RESEARCH ARTICLE Open Access

Processes of consent in research for adultswith impaired mental capacity nearing theend of life: systematic review andtransparent expert consultation(MORECare_Capacity statement)C. J. Evans1,2* , E. Yorganci1, P. Lewis3, J. Koffman1, K. Stone1, I. Tunnard1, B. Wee4, W. Bernal5, M. Hotopf6,I. J. Higginson1 and on behalf of MORECare_Capacity

Abstract

Background: Involving adults lacking capacity (ALC) in research on end of life care (EoLC) or serious illness isimportant, but often omitted. We aimed to develop evidence-based guidance on how best to include individualswith impaired capacity nearing the end of life in research, by identifying the challenges and solutions for processesof consent across the capacity spectrum.

Methods: Methods Of Researching End of Life Care_Capacity (MORECare_C) furthers the MORECare statement onresearch evaluating EoLC. We used simultaneous methods of systematic review and transparent expert consultation(TEC). The systematic review involved four electronic databases searches. The eligibility criteria identified studiesinvolving adults with serious illness and impaired capacity, and methods for recruitment in research, implementingthe research methods, and exploring public attitudes. The TEC involved stakeholder consultation to discuss andgenerate recommendations, and a Delphi survey and an expert ‘think-tank’ to explore consensus. We narrativelysynthesised the literature mapping processes of consent with recruitment outcomes, solutions, and challenges. Weexplored recommendation consensus using descriptive statistics. Synthesis of all the findings informed theguidance statement.

(Continued on next page)

© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate ifchanges were made. The images or other third party material in this article are included in the article's Creative Commonslicence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commonslicence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtainpermission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to thedata made available in this article, unless otherwise stated in a credit line to the data.

* Correspondence: catherine.evans@kcl.ac.uk1Cicely Saunders Institute of Palliative Care, Policy & Rehabilitation, FlorenceNightingale Faculty of Nursing, Midwifery & Palliative Care, King’s CollegeLondon, Bessemer Road, London SE5 9PJ, UK2Sussex Community NHS Foundation Trust, Brighton General Hospital,Brighton, UKFull list of author information is available at the end of the article

Evans et al. BMC Medicine (2020) 18:221 https://doi.org/10.1186/s12916-020-01654-2

(Continued from previous page)

Results: Of the 5539 articles identified, 91 met eligibility. The studies encompassed people with dementia (27%)and in palliative care (18%). Seventy-five percent used observational designs. Studies on research methods (37studies) focused on processes of proxy decision-making, advance consent, and deferred consent. Studiesimplementing research methods (30 studies) demonstrated the role of family members as both proxy decision-makers and supporting decision-making for the person with impaired capacity. The TEC involved 43 participantswho generated 29 recommendations, with consensus that indicated. Key areas were the timeliness of the consentprocess and maximising an individual’s decisional capacity. The think-tank (n = 19) refined equivocalrecommendations including supporting proxy decision-makers, training practitioners, and incorporating legislativeframeworks.

Conclusions: The MORECare_C statement details 20 solutions to recruit ALC nearing the EoL in research. Thestatement provides much needed guidance to enrol individuals with serious illness in research. Key is involvingfamily members early and designing study procedures to accommodate variable and changeable levels of capacity.The statement demonstrates the ethical imperative and processes of recruiting adults across the capacity spectrumin varying populations and settings.

Keywords: Palliative care, Terminal care, Decision-making, Consent, Methods, Ethics, Systematic review, Consensus

BackgroundThere is an urgent need for evidence on best practice inpalliative care. The projected increases in global serioushealth-related suffering demand immediate action. By2060, an estimated 48 million people will die globallywith serious related-suffering, representing an 87% in-crease from the 26 million in 2016 [1]. Failing to re-spond will see 80% of people globally with little or noaccess to palliative care services and treatment [2]. Amajor barrier in research on palliative care is ethicalconcerns about the perceived vulnerability of adults withserious illness and including them in research, particu-larly if the person also has impaired mental capacity [3].Exclusion of adults with impaired capacity to consentfor themselves impedes evidence-based care and treat-ment that is applicable across the illness trajectory andend of life (EoL) [4]. New interventions require robustevaluation to examine benefit and potential of harm forthe population intending to benefit [5, 6]. Studies, espe-cially clinical trials in palliative care, are often compro-mised by insufficient sample size to detect change [7–13], and impaired understanding of legislation governingresearch involving adults with impaired capacity [14].The ethical challenges of recruiting individuals with im-paired capacity are examined across fields involvingadults with serious illness including palliative care [15],dementia [16–18], mental health [19], and intensive care[20]. Systematic reviews have considered consent pro-cesses in specific conditions (e.g. dementia [21], schizo-phrenia [22]) and aspects of involving adults lackingcapacity in research (e.g. capacity assessment [22], en-hancing informed consent with older people [23, 24],and strategies for designing research studies [25] and in-creasing the recruitment rate in palliative care [26]). But,

in palliative care, intervention studies are few and oftenexclude adults lacking capacity, for example in the dyingphase [27]. There is literature from both within and out-side the field of palliative care that could inform muchneeded guidance on best practice on processes of con-sent across the capacity spectrum in serious illness. Thisstudy aimed to determine how best to include individ-uals with impaired capacity in research on EoLC byidentifying challenges for and solutions to processes ofconsent across the capacity spectrum. This paper reportsthe integrated results from a systematic review andtransparent expert consultation to form the MORECare_Capacity statement on processes of consent in researchon EoLC. This furthers the Methods Of ResearchingEnd of Life Care (MORECare) statement on evaluatingcomplex circumstances in EoLC [28] by giving detailedconsideration on processes of consent for adults withserious illness across the capacity trajectory. The MORE-Care statement omitted this area, focusing on outcomemeasurement, response shift and attrition, integratingmixed methods and economic evaluation.In this paper, ‘capacity’ refers to mental capacity to

make an informed decision regarding research participa-tion. ‘The spectrum of capacity’ of individuals rangesfrom potentially impaired, and anticipated to have im-paired capacity, to lacking capacity. The legislation gov-erning involvement of adults lacking capacity in researchand terminology is jurisdiction specific. In this paper,the term ‘consultee’ (someone who has capacity) is usedto encompass the different terms used in respective ju-risdictions including but not limited to proxy-decisionmaker, personal consultee and nominated consultee. Adistinction is made between a personal consultee (e.g.family member) and a nominated consultee (e.g. health

Evans et al. BMC Medicine (2020) 18:221 Page 2 of 55

professional) [29]. ‘The process of consent’ refers to thesteps taken to ensure that an eligible research participantis sufficiently informed about the purposes, content, af-filiations of the study, and their right to withdraw fromthe study at any point, enabling them or their consulteeto decide freely about research participation [30].

MethodsStudy designWe used a parallel iterative research design detailed inFig. 1. We used methods of systematic literature review toidentify and map the challenges and solutions for pro-cesses of consent for adults with impaired capacity, andthe MORECare transparent expert consultation (TEC) todebate key areas [28] of uncertainty/contention. The TECinvolved expert stakeholder consultations using consensusmethods of modified nominal group technique to generaterecommendations [31], and then presenting the recom-mendations in an online Delphi survey to explore levels ofagreement [32]. We held a final expert think-tank to ex-plore areas of contention/uncertainty and synthesise thefindings to develop the statement. King’s College LondonResearch Ethics Committee approved the TEC compo-nent (ref no. BDM/10/11-90).

Systematic reviewDesignWe used systematic review method of narrative synthesisto systematically identify, appraise, and synthesise

quantitative and qualitative literature [33]. Methods ofanalysis and inclusion criteria were pre-defined in thestudy protocol. Reporting followed the PRISMA guid-ance [34] (see Additional file 1: Table S1).

Eligibility criteria

Population Adults (≥ 18 years old) with impaired cap-acity encompassing declining capacity (e.g. mild to mod-erate dementia), fluctuating capacity (e.g. delirium), andlack of capacity (e.g. dying, advanced dementia) areincluded.

Context The scoping of the literature identified areasrecruiting adults lacking capacity with serious illness inresearch. We included studies from palliative care, men-tal health (delirium, dementia, learning disabilities), oremergency medicine/critical care.

Interest Studies discussing consent in its various forms(e.g. informed, advanced, proxy) and impaired mentalcapacity are included. We did not restrict by health orbehaviour outcome. We focused on research studies in-vestigating either of the following: (i) methods for in-volving adults with impaired capacity in research, (ii)implementing research methods to enable recruitment,or (iii) exploring public attitudes and ethical issues oninvolvement in research.

Fig. 1 Overview of the study design

Evans et al. BMC Medicine (2020) 18:221 Page 3 of 55

Design Randomised controlled trials (RCTs) and quasi-experimental, mixed method, or observational qualitativeor quantitative designs are included. We included pub-lished study protocols that reported the methods of con-sent to enhance understanding for the main study. Weexcluded systematic reviews but used relevant systematicreviews for reference chaining by screening the citedpublications for eligibility. We excluded opinion piecesand commentaries and non-English language papers.Studies concerning treatment/clinical decision-makingor bioethics were out of scope.

Search strategyWe developed search strategies for each of the popula-tion groups—palliative care, mental health (including de-mentia), and emergency medicine/critical care. MeSHterms for the palliative care group included ‘Terminallyill’ or ‘Palliative Care’ AND ‘Decision making’ or ‘MentalCompetency’ AND ‘Informed consent’ or ‘Third-partyconsent’. Key search terms were used as free-text, andwith use of truncation symbol to retrieve variations inthe terminology. Search terms were piloted pre-studyand mapped to assess their relevance and specificity, andrefined working with a specialist librarian (see Add-itional file 2: Table S2 - electronic search terms). Wesearched four electronic databases: MEDLINE (1966–Present), EMBASE (1947–Present), PsychINFO (1887–Present), and CINAHL (1937–Present), and supple-mented with referencing chain, grey literature electronicsurvey, and expert recommendations. The last searchwas run on 30 October 2018.

Quality appraisalStudy risk of bias was assessed using the validated Qual-Syst review tool suitable for quantitative and qualitativestudies [35] by one reviewer (EY, IT, CJE), and a random10% sample checked by a second author (CJE). Scoresthat diverged by > 10% were discussed within the re-search team. A single quality appraisal was undertakenfor studies reported in multiple publications (e.g. proto-col paper and a main trial results paper). The QualSystassessment criteria include 14 items for quantitativestudies and 10 items for qualitative studies. Each item isscored from 0 to 2 (0, not present; 1, partial; 2, yes; ornot applicable). The percentage of the total possible scoreindicates the quality grade: < 50%, low; > 50 and < 70%,medium; or > 70%, high. Study designs were categorisedusing the Cochrane Effective Practice and Organisation ofCare (EPOC) grade system [36] (see Additional file 2:Table S3).

Data screening, extraction, and analysisReferencing software (Endnote version x8) [37] was usedto manage a database of search findings and remove

duplicates. Title and abstract screen by one reviewer(EY/KS) and second independent review of 20% to testthe application of the eligibility criteria (CJE). Titles/ab-stracts that met the review criteria, or if insufficientinformation to determine eligibility, were subject to full-text screening. Full-text articles were single-screened bythree reviewers (EY/KS/CJE). Full-text papers with un-certain eligibility were reviewed by two reviewers andeligibility agreed (EY/KS with CJE). A standardised dataextraction form was developed and piloted based on theCochrane Consumers and Communication RevieGroup’s data extraction template (see Additional file 2:Table S4). Data included the study design and aim, thepopulation and context, method(s) of consent, recruit-ment rate, and challenges and solutions. Study data wereextracted by one reviewer (CJE, EY, KS) and checked bya second reviewer (CJE, EY, KS). We contacted two au-thors to check availability of the publications in English.Using narrative synthesis [33], textual descriptions fromextracted data for all studies were mapped to formmatrices for studies innovating research method, studiesusing innovative methods, or considered ethics, legisla-tion, or public attitudes. Each matrix was analysed andcoded in Microsoft Excel using thematic analysis to ex-plore prominent themes. Higher quality studies were val-ued with a greater strength in the final synthesis.

Transparent expert consultationThe TEC aimed to enhance the systematic review find-ings by exploring the application of the findings in re-search studies and areas little considered or uncertain inthe evidence. The TEC explored researchers’ and serviceusers’ perspectives on recruiting individuals with im-paired capacity in research on EoLC. The TEC sought togenerate recommendations on processes of consent toenable recruitment and explore the level of consensus.

Setting and participantsParticipants were purposively sampled based on their ex-pertise in conducting research involving adults with im-paired capacity (including ethicists), caring for patientswith advanced disease, or a service user/carer (e.g. pallia-tive care services), or a voluntary sector representative(e.g. Alzheimer’s Society). Participants were invited tothe workshops held in the Cicely Saunders Institute,King’s College London. Eligible participants includedmembers of the project’s expert panel (project appli-cants), Project Advisory Group (invited experts in, forexample, ethics, and PPI and voluntary sector represen-tatives) [see the “Acknowledgements” section], respon-dents in the systematic review grey literature survey andinvited ethicists, clinicians, commissioners, researchers,members of ethical committees, policymakers, and ser-vice user and lay voluntary sector representatives.

Evans et al. BMC Medicine (2020) 18:221 Page 4 of 55

Identified professional participants received email invita-tions for the workshop. Service user and lay voluntarysector representatives were recruited via voluntary sectorgroups including, for example, Alzheimer’s Society andIndependent Cancer Patients’ Voice. The respectiveorganisations circulated the invitation letter to theirmembers targeting those known to have an interest/ex-perience of either a carer for an adult with impairedcapacity, being a patient with progressive illness, or sup-porting research involving adults with impaired capacity.The TEC used four stages:Stage I: Identifying critical issues. The initial workshop

focused on critical issues identified from the systematicreview preliminary findings and expert opinion (e.g.areas with limited empirical evidence and relevance inthe processes of consent for adults across the capacityspectrum).Stage II: Stakeholder workshop. Participants received a

pre-workshop briefing pack detailing the aim, critical is-sues, and workshop format. The workshop comprisedpresentations on the critical areas overviewing findingsfrom the systematic review followed by structured groupdiscussion involving 10–14 participants focusing on oneof the critical areas. Group discussions were digitally re-corded. We used a structured nominal group process fa-cilitated by a member of the research team. Thefacilitator guided participants through a structuredprocess of (1) a brief discussion, (2) individual writing ofrecommendations and ranking, and (3) participants inturn stating their highest ranked recommendations untilindividual lists were exhausted (or time exceeded) [31].Scribes wrote the recommendations and ranking on aflipchart, and each small group discussed and agreed onthe final priority order, then presented and discussedwith the whole group. Participants individually listedand ranked recommendations from one to five (highestto lowest) on structured A4 sheets detailing the respect-ive group question, ranking scale and boxes to list rec-ommendations, rank, and detail rationale.Stage III: Delphi online consensus exercise. This is a

two-round online consensus exercise [32]. Recommen-dations generated in the workshop were posted online tothe workshop participants, members of the expert paneland Project Advisory Group, and respondents to thegrey literature survey. Participants received a persona-lised email invitation and reminder after 2 weeks. Theonline participants anonymously ranked, from one tonine (strongly disagree to strongly agree), the extent theyagreed with a recommendation and used free-text spacesto comment on each recommendation. Findings fromround 1 informed requirements to revise recommenda-tions where comments suggested ambiguity. Round 2 re-presented the revised recommendations and the medianscore for each recommendation from round 1.

Participants again indicated their level of agreementranked from one to nine and provided free-text com-mentary on, for example, rationale for ranking score.Stage IV: Expert ‘think-tank’. The expert ‘think tank’

workshop aimed to aid data synthesis and inform thesolutions and recommendations in the statement by crit-ically considering areas of contention/uncertainty identi-fied in the consensus exercise and systematic reviewfindings. The think-tank aimed to understand the de-bates surrounding these areas, the strengths and limita-tions of the evidence, and the solutions for practice.Participants were purposively selected from the work-shop participants based on expertise, e.g. ethicists, layvoluntary sector representative, researcher, and clinician.Think-tank participations received a briefing report thatsummarised for the respective area the systematic reviewfindings on the challenges and solutions identified in theevidence base, and the recommendations and level ofagreement from the consensus exercise. The think-tankused a format of presentations and debate, drawing onstructured nominal group process to facilitate partici-pant agreement on the top two or three key solutionsfor each area, and commentary on their thinking.Participants discussed and debated these areas in groupsof 6–7. Discussions were digitally recorded, and scribesrecorded on flipchart the key debates.

Data analysisIndividual recommendations from the workshop andtheir ranking were entered in Excel spreadsheets withassigned participant identification numbers. Two re-searchers (CJE, KS) coded and arranged recommenda-tions by themes, duplicates were combined, andrecommendations arranged by priority ranking (1 high-est to 5 lowest). Free-text comments were collated.Digital recordings were reviewed to inform understand-ing on the recommendations and debates presented,with key points noted on the Excel spreadsheet for therespective recommendation. The recommendationsretained participants’ original language where possiblewith amendments guided by the expert panel to enhanceclarity and avoid repetition. The final recommendationswere those ranked the highest (≤ 3) and reviewed andagreed by the expert panel and piloted (e.g. for clarity),and then posted on the online consensus survey. Ana-lysis of the consensus survey-scaled data used descriptivestatistics (frequencies and medians) and plots (box andwhisker plots) of interquartile ranges to analyse and in-terpret levels of agreement. We used a conventional cat-egorisation to interpret agreement (indicated, equivocal,or not indicated) and strength of agreement (strict orbroad) used in previous consensus studies [28]. Table 1details the categories by the respective median regionand IQR [38]. Narrative comments were collated by

Evans et al. BMC Medicine (2020) 18:221 Page 5 of 55

recommendation, and themes identified to understandthe issues raised and provide illustrative examples [32].

ResultsSystematic review search resultsThe electronic database searches identified 5539 ab-stracts after removal of duplicates with a further 179publications identified from other sources (see Fig. 2).Ninety-one publications met the eligibility criteria,reporting 89 studies. Two studies included a protocoland main results papers [39–42]. Studies were con-ducted mainly in dementia (n = 23), palliative care (n =16), and intensive care (n = 15) (Table 2). Publicationsincreased over time with the majority published after2010 (n = 54). Studies were conducted mainly in theUSA (n = 35), UK (n = 29), or Canada (n = 9) (see Table 2and Table 3, and Additional file 3: Table S8). The stud-ies formed three main areas of (1) innovating researchmethods to recruit adults across the capacity spectrum,(2) applying consent processes across the capacityspectrum in studies on serious illness, and (3) public at-titudes on involving adults lacking capacity in research.

Quality appraisalOverall, the quality of the included articles was mediumto high. Most quantitative (95.8%, n = 71) and qualitativestudies (88.9%, n = 17) were assessed as medium or highquality (see Additional file 2: Table S5 quantitative stud-ies and Table S6 qualitative studies). The proportion ofhigh-quality studies included was consistent across thethree main areas (56.8% ‘innovating research methods’,56.6% ‘applying consent processes’, and 59.1% ‘public at-titudes’). However, in the area of ‘public attitudes on in-volving adults lacking capacity’, 9% were assessed as lowquality, compared with 0% in ‘innovating researchmethods’ and 2% in ‘applying consent processes’. Thisreflected in part the methodological nature of the studiesand poorer fit with the Qualsyst item criteria. The in-cluded studies were mainly descriptive (n = 36) cate-gorised as ‘non-experimental, longitudinal, cohort,matched pairs, or cross-sectional, sound qualitative, oranalytical studies’, with few experimental (n = 20) or

quasi-experimental designs (n = 3) (see Additional file 3:Table S7).

Innovating research methods to recruit adults across thecapacity spectrumThirty-seven studies were categorised as innovating re-search methods (Table 3). Studies focused on participa-tion in research involving individuals with cancer/receiving palliative care (n = 6), dementia (n = 13), geriat-ric care (all settings) (n = 2), delirium and mental healthservices (n = 7), or intensive care (n = 5). While numer-ous studies used standardised capacity assessment tools,existing tools were often regarded as time-consuming,and administration of a formal capacity assessment re-duced recruitment, for example in an observationalstudy involving patients with delirium [99]. Formal cap-acity assessment was considered of little value unlessaligned to the decisional requirements for study partici-pation, notably the risks and the potential direct orindirect benefits of participation. Overall, studies incor-porated multiple components of the processes of con-sent. These were tailored to individuals’ level of capacityfrom mild to moderate impairment with a focus on en-hancing informed consent, through to lacking capacityrequiring involvement of a consultee. For example, inpopulations such as psychiatric or stroke patients, whereparticipants experienced varying levels of impaired cap-acity, studies incorporated processes of enhanced in-formed consent and consultee involvement [87, 90]. Inboth studies, consultee advice was sought for a third ofparticipants (30.6% and 35.7%, respectively). The innova-tions broadly mapped onto two sub-categories of ‘maxi-mise individuals’ autonomy and decisional capacity inthe consent process’ and ‘processes of consent to enableadults across the capacity spectrum to participate inresearch’.

Maximise individuals’ autonomy and decisional capacity inthe consent processFifteen studies aimed to examine ways to enhance theinformed consent process to maximise decisional cap-acity for adults with mild/moderate capacity or fluctuat-ing capacity (see Table 3). Clinical trials requiredapplicable methods to facilitate understanding of com-plex procedures, e.g. randomisation and clinical equi-poise. For populations with mild/moderate dementia,and other neuropsychological disorders such asParkinson’s disease, studies aimed to enhance variousfacets of cognition. This involved in the informed con-sent, for example, enhancing decisional capacity, under-standing [90], reasoning, comprehension, and recall ofinformation. Across patient groups, key challenges to en-abling participation in the informed consent processwere addressing concerns about causing distress for the

Table 1 Levels of consensus and agreement by median regionsand IQR [38]

Median regions and IQR Interpretation

7–9 Recommendation indicated

4–6 Recommendation equivocal

1–3 Recommendation not indicated

IQR in one region Strict agreement for recommendation

IQR in any three-point region Broad agreement for recommendation

IQR interquartile range

Evans et al. BMC Medicine (2020) 18:221 Page 6 of 55

person from receiving and considering study information[43], and enabling understanding and recall of the infor-mation provided [59, 103]. Moreover, in ensuring an in-formed consent rather than participation based on trustin the clinician and the study proposal considered a rec-ommendation to take part [103]. A study observing con-sent process for people with dementia (n = 46) and theirsurrogate decision-maker (e.g. spouse) across six trials(drug and non-drug) showed approaches to facilitatingthe consent process and wide variation [60]. Importantly,

the consent process involved the person with dementiaand their surrogate, with the surrogate playing a majorrole in facilitating the consent process. Use of a dualconsent was common, but assent from the person wasoften implicitly implied not explicitly asked. The studyrevealed wide variability in the conduct of the consentprocess with higher risk drug trials generally more com-prehensive in giving study information and assessing un-derstanding compared to the non-drug trials. Theauthors make recommendations on using a dual consent

Fig. 2 PRISMA flow diagram

Table 2 Population of interest categorised by the study focus

Patient population ofinterest

Study focus on processes of consent

Innovating research methods (n = 37) Applying research methods(n = 30)

Attitudes and ethical considerations(n = 22)

Total studies(n = 89)*

Palliative care/cancer 6 [43–48] 8 [49–56] 2 [57, 58] 16

Dementia 13 [59–71] 10 [40, 42, 72–79] 1 [80] 24

Geriatric care 2 [81, 82] 3 [83–85] 0 5

Parkinson’s disease 1 [86] 0 0 1

Cerebral ischaemic stroke 2 [87, 88] 0 1 [89] 3

Mental health 6 [90–95] 2 [96, 97] 1 [98] 9

Delirium 2 [99, 100] 2 [101, 102] 0 4

Intensive care 5 [103–107] 5 [108–112] 4 [113–116] 14

General population 0 0 13 [117–129] 13

*89 studies are reported from 91 publications (includes two study protocols [40, 42] reported with the main study papers [39, 41]

Evans et al. BMC Medicine (2020) 18:221 Page 7 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

Enhanced

inform

edconsen

tprocesses

Dob

ratzet

al.

[43],U

SA,B3

2003

Retrospe

ctiveresearch

stud

ywith

avulnerablepo

pulatio

nto

describ

eissues

and

dilemmas

relatedto

non-

participation,attrition

,and

need

forassistance

inresearch

with

vulnerable

homeho

spiceparticipants.

Hom

eor

preferred

settingof

homeho

spice

agen

cyrecipien

tsfro

mtw

ometropo

litan

settings

Palliativecare:h

ome

hospiceagen

cyrecipien

tsInform

edco

nsen

t:stud

ywas

explaine

dto

all

participantsover

theph

one.

Participantsprovided

‘teleph

oneconsen

t’.During

thestud

yvisit,participants

provided

inform

edconsen

t.Adults

withdeclin

ing

capacity

97/113

(86%

)Key

findings:fivepe

ople

who

agreed

totake

part

unableto

provideinform

edconsen

tdu

eto

distress;five

othe

rparticipantswith

cogn

itive

impairm

ent

preclude

dinform

edconsen

t.So

lution

s:participants

requ

ireregu

larmon

itorin

gof

theirph

ysicaland

psycho

logicalsym

ptom

s,oversamplingto

anticipate

andplan

forwith

draw

als,and

cogn

itive

assessmen

ttool

for

allp

oten

tialp

articipants

regardless

ofdiagno

sis(e.g.

brainmetastases),and

careful

screen

ingof

psycho

logical

behaviou

rsto

redu

cedistress.

Participantpo

sitivefeed

back

abou

tthestud

y.

Siminoffet

al.

[71],the

Nethe

rland

s,C1

2004

Qualitativestud

yexam

ining

thefactorsthat

were

impo

rtantin

makingresearch

participationde

cision

sam

ong

patientswith

Alzhe

imer’s

disease(AD),cancer,critically

illchildren.

Unclear

Dem

entia:

patientswith

Alzhe

imer’sdiseasewith

cogn

itive

impairm

ent

Inform

edco

nsen

t:po

tential

subjectsweregiven

inform

ationabou

tthestud

yby

phon

e.Duringthe

conversatio

n,ifpatientsmet

initialeligibility

criteria

and

expressedcontinuedinterest

inparticipation,attend

edthe

clinicforform

alinform

edconsen

tproced

ure.The

consen

tform

was

read

tothe

subjectsandqu

estio

nswere

addressedthroug

hout

andat

theen

dof

review

ingthe

writtenconsen

tform

.Adults

withdeclin

ing

capacity

46ADpatients,meanage72,

91.3%

white,61%

female,

78.3%

married,

63.6%

more

than

ahigh

scho

oled

ucation,65.6%

income>

$25,000

Key

findings:keyelem

ents

ofinform

edconsen

t.Inform

ation—

theAD(m

ean

=1.78)grou

preceived

more

inform

ationabou

tvoluntary

participationthan

the

paed

iatric(m

ean=1.27)

grou

p(F=4.1;p<.05).

Con

fiden

tiality—

weremost

likelydiscussedwith

theAD

grou

p(87%

)versusthe

cancer

(22%

)and

paed

iatric

(9.2%)grou

ps(χ2=85.24;p<

.001).Discussionof

‘no

treatm

ent’as

aviableop

tion

occurred

mostoftenin

the

ADgrou

p(58.7%

)versus

the

adultcancer

(41.6%

)and

paed

iatric(3.9%)subjects(χ2

=46.38;p<.001).AD

subjectsreceived

themost

inform

ationabou

tvoluntary

participationand

confiden

tiality,andno

Evans et al. BMC Medicine (2020) 18:221 Page 8 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

treatm

entop

tion.

Cha

lleng

e:lack

ofdiscussion

sabou

tconfiden

tialityandno

treatm

entas

anop

tion.

Buckleset

al.

[59],U

SA,B3

2003

Long

itudinalstudy

ofhe

althy

ageing

andde

men

tiato

evaluate

unde

rstand

ingof

inform

edconsen

tby

olde

rparticipantsacross

arang

eof

demen

tiaseverityusinga

briefteston

theelem

entsof

inform

edconsen

tforalow-

riskstud

y.

Not

stated

Dem

entia:

415participants,

165with

outde

men

tia,250

with

demen

tia

Inform

edco

nsen

tan

dassessmen

tdem

entia

seve

rity

with

Clinical

Dem

entia

Ratin

g(CDR),

MMSE

Adults

withdeclin

ing

capacity

415

Key

findings:afteradjusting

fored

ucation,pe

rform

ance

onthetestvariedwith

demen

tiaseverityin

mean

differences

andby

correlation.Allno

n-de

men

ted

andvery

mildlyde

men

ted

participantsand92%

ofmildlyde

men

tedparticipants

provided

correctansw

erfor

atleast8of

10true-false

items,whe

reas

only67%

ofthemod

eratelyde

men

ted

participantsachieved

this

levelo

faccuracy.

Solution

s:by

mod

erate

demen

tiastage,involvem

ent

ofarespon

siblecaregiverin

consen

tprocessshou

ldbe

mandatory.

Mangset

etal.

[103],Norway,C

12008

Qualitativestud

yto

explore

critically

illpatients’

expe

riencewith

theprinciple

ofinform

edconsen

tin

aclinicaltrialand

theirability

togive

valid

inform

edconsen

t.

2Norweg

ian

hospitals

Stroke

:strokepatientswho

wereinvitedto

take

partin

aninternationalstroketrial

Inform

edco

nsen

tAdults

withdeclin

ing

capacity

11Key

findings:theresults

challeng

ethevalidity

ofinform

edconsen

tforan

expe

rimen

taltrialo

btaine

dfro

mcritically

illpatients,and

theconcep

tof

theconsen

tas

acontract

obtained

ona

ratio

nalb

asisbe

tweeneq

ual

andautono

mou

sparties.

Cha

lleng

es:m

ostpatients

didno

tun

derstand

the

purposeof

rand

omisationor

theconcep

tof

clinical

equipo

ise.Themainreason

sforconsen

twerepatients’

depe

nden

cyon

thedo

ctor,

theirne

edforatrustfu

lrelatio

nship,

andseeing

stud

yinform

ationas

arecommen

datio

n.

Cho

uliara

etal.

[44],U

K,C1

2004

Individu

alinterviewson

some

ethicaland

metho

dological

challeng

esinvolved

in

Careof

the

elde

rlywards

andacancer

Can

cer:pe

oplewith

cancer

>65

yearsold

Inform

edco

nsen

t:ob

tained

afterexplaining

theproject

verballyandwritten.Family

37/50(74%

)Key

findings:meanage74.3

(SD=7.5),M

MSE

24.0(4.1)—

mild

demen

tia.Involvemen

t

Evans et al. BMC Medicine (2020) 18:221 Page 9 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

cond

uctin

gresearch

with

olde

rpe

oplewith

cancer

byreferringto

researchers’

expe

riences

inan

on-going

research

project.

centre

mem

bersinvolved

toassist

patient

decision

-making.

Ininstanceswhe

rethey

feltlike

thepatient

was

not

unde

rstand

ing,

they

also

involved

.Adults

withdeclin

ing

capacity

ofvulnerable,elderly

individu

alswas

considered

ethicalinalow-risk

qualitativestud

y.Cha

lleng

es:Fluctuatin

gcapacity,fatigue,frailty,

physicalandcogn

itive

limitatio

ns.

Solution

s:asemi-structured

interview

sche

duleallowed

patientsto

talkfre

ely.A

rigorou

sproced

ureto

obtain

valid

consen

t,includ

ingthe

view

pointsof

allthe

parties

involved

.

Mittalet

al.

[61],U

SA,B3

2007

RCTcomparin

gtw

oprocesses

ofconsen

tto

evaluate

the

feasibility,accep

tability,and

prelim

inaryefficacyof

two

enhanced

consen

tproced

ures

forpatientswith

Alzhe

imer’s

diseaseor

mild

cogn

itive

impairm

entthat

used

either

aPo

werPo

intpresen

tatio

nor

printedconsen

tform

.

2med

ical

centres

Dem

entia:

referred

patientswith

possibleor

prob

ableAlzhe

imer’s

disease(M

MSE

≥19),mild

cogn

itive

impairm

ent

Enha

nced

inform

edco

nsen

t.Po

werPo

int

slidesho

wpresentation(no

written

consen

tform

)(SSP

)or

enha

nced

written

consen

tprocedure

(researcher

readsthe

inform

ationalou

d,whilethe

potentialp

articipantcan

follow

from

theirow

ncopy

with

largefonts).Followed

with

MacCAT-CRassessmen

tof

capacity,the

nrepe

atof

either

oftheconsen

tprocesses.MMSE

≥19

Adults

withim

paired

capacity

35Key

findings:participants

improved

their

unde

rstand

ingscores

after

verbalre-explanatio

nof

consen

tinform

ation.There

wereno

sign

ificant

differences

inlevelo

fun

derstand

ingam

ongthose

intheSSPversus

theEW

CP

cond

ition

sat

either

trial,bu

twefoun

dtheSSPtook

sign

ificantlyless

timeto

administer.

Rubright

etal.

[63],U

SA,A

32010

RCTtestingwhe

ther

amem

oryandorganisatio

nal

aidim

proves

ADpatient

perfo

rmance

onmeasuresof

capacity

andcompe

tencyto

give

inform

edconsen

t.

Alzhe

imer’s

diseasecentre

Dem

entia:

patientswith

Alzhe

imer’sdisease

Enha

nced

inform

edco

nsen

twithmem

oryan

dorgan

isationa

laid

orstan

dardised

inform

edco

nsen

t:interven

tiongrou

preceived

theadditio

nalaid

which

summarised

thekey

elem

entsin

thedrug

Z-298

inform

edconsen

tform

.It

presen

tedinform

ationin

the

samesequ

ence

andhe

ader

titlesas

presen

tedin

the

inform

edconsen

tform

.The

text

simplified

impo

rtant

pointsfro

mtheconsen

tform

usinglang

uage

atasixth

80/112

(71%

)ofp

oten

tialA

Dpatients,30/33(91%

)of

cogn

itivelyno

rmalolde

radults

Key

findings:the

interven

tiongrou

pwas

more

likelyto

bejudg

edcompe

tent

than

thecontrol

grou

p(χ2=8.2,df

=1,p=

0.004)

andhadhigh

erscores

onMacCAT-CRmeasure

ofun

derstand

ing(z=2.86,p

=0.004).ThisRC

Tshow

sthat

amem

oryandorganisatio

nal

aidtailoredto

thedistinctive

cogn

itive

patterns

ofAD

patientscanim

provethe

ability

ofpatientswith

very

mild

toearly

mod

erateADto

providetheirow

ninform

ed

Evans et al. BMC Medicine (2020) 18:221 Page 10 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

gradereadinglevel.All

participantswen

tthroug

ha

capacity

assessmen

t.

consen

tto

enrolinan

early-

phaseclinicaltrial.

Cha

lleng

e:thistool

was

design

edspecifically

forAD

patientsandmay

not

gene

ralisableto

othe

rpo

pulatio

ns.U

seof

form

alcapacity

assessmen

tsshou

ldno

tbe

treatedwith

strictcut-

offm

easuremen

ts.M

ore

challeng

ingforhigh

-risktrials.

Solution

s:forpatientsat

anearly

phaseof

AD,capacity

canbe

enhanced

.

Ford

etal.[86],

USA

,B2

2008

RCTto

evaluate

theeffectsof

socialsupp

orton

compreh

ension

andrecallof

consen

tform

inform

ationin

astud

yof

Parkinson’sdisease

patientsandtheircaregivers.

Med

ical

centre

Parkinson’sdisea

se:

Parkinson’sdiseasepatients

(meanage71

(SD8.6)

years)andtheircaregivers

Enha

nced

inform

edco

nsen

t:in

thesocial

supp

ortgrou

p,patient-

caregiverpairwas

askedto

completetheconsen

tform

inthesameroom

compared

tothecontrolg

roup

who

completed

theform

sin

separate

room

s.Adults

withim

paired

capacity

136/143(95%

)Key

findings:1-weekfollow-

up,nosign

ificant

differences

inQualityof

Inform

edCon

sent

(QuIC)scores

betw

eenparticipants

receivingthesocialsupp

ort

interven

tionandthecontrol

grou

p.Regardless

ofthe

grou

pallocatio

n,participants

scored

approxim

ately50%

oftheQuICqu

estio

ns.But

the

finding

sshow

edthat

compreh

ension

ofconsen

tform

inform

ationwas

increasedthroug

hthesocial

supp

ortinterven

tionin

a‘real-w

orld’clinicalsetting.

Cha

lleng

e:initial

compreh

ension

was

low

and

remaine

drelativelyconsistent

with

inthe1-mon

thpe

riod.

Solution

s:inform

ational

supp

ortprovided

byfamily

caregivers.C

areg

iverswho

scored

high

oncorrectQuIC

wereassociated

with

patient

participantswho

with

high

QuICscores,e.g.

unde

rstand

inginform

ation.

Cam

pbelletal.

[94],Sou

thAfrica,

B3

2017

Case-controlstudy

exploring

ifusingiterativelearning

improves

participants’

unde

rstand

ingof

theresearch

stud

yandpred

ictorsof

better

Psychiatric

hospitalsand

clinics

Men

talh

ealth:

patients

with

psycho

sis/

schizoph

renia

Enha

nced

inform

edco

nsen

t(with

iterative

learning

):following

explanation,University

ofCalifornia,San

Diego

Brief

1056

participants—528

matched

casesandcontrols.

(targe

twas

181pairs)

Key

findings:be

fore

iterative

learning

,55%

ofcasesand

33%

ofcontrolswerescoring

lower

than

thecut-offpo

int

forstud

yparticipation.Afte

r

Evans et al. BMC Medicine (2020) 18:221 Page 11 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

unde

rstand

ingof

thestud

yat

theinitialscreen

ing.

Assessm

entof

Capacity

toCon

sent

Questionn

aire

(UBA

CC)isadministrated

,if

thepe

rson

achieves

>14.5

demon

stratin

gcapacity

toconsen

t;inform

edconsen

tis

provided

.Adults

withim

paired

capacity

iterativelearning

,only7%

ofcasesand3%

ofthecontrols

wereun

ableto

consen

tto

participate.Iterativelearning

processim

proved

decision

alcapacity

andun

derstand

ing

ofthestud

yin

both

cases

andcontrols.Thisprocessis

repe

ated

afteriterative

learning

.So

lution

s:thestud

yrecruitersplay

asign

ificant

rolein

managingthequ

ality

oftheinform

edconsen

tprocess.

Palm

eret

al.

[62],U

SA,A

32018

RCTto

evaluate

theefficacyof

amultim

edia-aided

enhanced

consen

tprocessincorporating

correctivefeed

back,

comparedwith

routine

consen

t,am

ongindividu

als

with

mild-to-mod

erate

Alzhe

imer’sdiseaseandno

n-ne

urop

sychiatriccomparison

(NC)sub

jects.

Unclear

Dem

entia:

individu

alswith

Alzhe

imer’sdisease(m

ildto

mod

erate)

Enha

nced

inform

edco

nsen

t:theen

hanced

consen

tproced

ureexpand

edon

routineconsen

tby

adding

amorestructured

,iterativeprocessandby

incorporatingmultim

edia

toolsinto

theconsen

tpresen

tatio

n.Participants

wererand

omised

toroutine

consen

tanden

hanced

consen

tproced

ures

for

participatingin

ahypo

thetical

RCTlower

risk(FDA-app

roved

med

ication)

orahigh

risk

(phase

2im

mun

othe

rapy).

Assessm

entof

capacity

using

MacCAT-CR.

Adults

withim

paired

capacity

248—

134control,114

Alzhe

imer’sdisease

Key

findings:regardless

ofwhe

ther

rand

omised

tothe

lower

orhigh

erriskprotocol

type

,participantswho

received

theen

hanced

consen

tproced

uredidno

tde

mon

strate

sign

ificantly

better

decision

alcapacity

scores

comparedwith

those

who

received

theroutine

consen

tproced

ure.Find

ings

couldbe

dueto

rapid

forgettin

g.

Moser

etal.

[93],U

SA,B2

2006

Quasi-experim

ental,pre-po

ststud

yto

determ

inewhe

ther

abriefinterven

tioncould

improvede

cision

alcapacity

inpe

oplewith

schizoph

renia.

Unclear

Men

talh

ealth:

individu

als

with

schizoph

renia

Enha

nced

inform

edco

nsen

t:avery

brief(less

than

30min),semi-

individu

alised

interven

tion

consistin

gof

acompu

terised

presen

tatio

nof

the

hypo

theticalstud

yinform

ationin

abu

lleted,

simplified

form

at,w

ithon

ekeypo

intpe

rslide.

Participantsview

edthe

presen

tatio

nandread

alon

gas

theexam

iner

read

alou

d

30individu

alswith

schizoph

reniaand30

healthy

comparison

participants

Key

findings:at

follow-up,

theschizoph

reniagrou

phad

improved

sign

ificantlyon

unde

rstand

ing(t[27]

=2.85,

p=.008)andwas

nolong

ersign

ificantlydifferent

from

thecomparison

grou

pon

anyof

thefour

dimen

sion

sof

decision

alcapacity

onthe

MacCART-CRscale(p

=.13–.33).Follow-upanalyses

also

show

edasign

ificant

effect

oftheinterven

tionon

Evans et al. BMC Medicine (2020) 18:221 Page 12 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

each

slide.Followingthe

educationalrem

ediatio

n,the

exam

iner

review

edwith

the

participantallM

acCAT-

CR

‘und

erstanding

’itemsfor

which

theparticipantdidno

treceivemaxim

umcred

it.Participantsreceived

both

astandardised

interven

tion,

individu

alised

discussion

,and

correctivefeed

back

regarding

theaspe

ctsof

theresearch

protocol

that

they

foun

dconfusing.

TheMacCAT-CR

interview

was

then

repe

ated

toassess

participants’

decision

alcapacity

following

theed

ucationalrem

ediatio

n.Finally,the

exam

iner

abriefer

structured

interview

design

edto

assess

theadeq

uacy

ofparticipants’und

erstanding

ofthehypo

theticalstud

y.Adults

withflu

ctua

ting

capacity

asub-setof

the

schizoph

reniagrou

pwho

hadpe

rform

edmostpo

orly

atbaseline,fro

mabaseline

meanof

18.0(SD=4.7)

to20.6(SD=4.9;t[7]=2.59,p

=.029).Theeffect

size

forthis

change

ismod

eratein

size

(Coh

en’sd=0.6).Participants

with

schizoph

reniaearned

sign

ificantlylower

scores

than

thosein

thecomparison

grou

pacross

multip

lene

urop

sycholog

icaldo

mains.

Cha

lleng

e:un

ableto

determ

inewhataspe

ctof

the

interven

tionused

was

most

helpful(activeingred

ients),as

allp

articipantsin

the

schizoph

reniagrou

preceived

both

thestandardised

compu

terpresen

tatio

nand

theindividu

alised

corrective

feed

back

compo

nents.Had

healthycomparison

s,no

tothe

rschizoph

reniapatients.

Solution

s:thoseindividu

als

who

initiallylacked

decision

alcapacity

may

bene

fitsign

ificantlyfro

men

hanced

consen

tproced

ures.Further

research

isne

eded

toun

pick

which

compo

nentsof

the

interven

tionarethose

causingtheim

provem

ents.

Jesteet

al.[92],

USA

,A3

2009

RCTto

evaluate

the

effectiven

essof

amultim

edia

versus

routineconsen

tproced

ure(aug

men

tedwith

a10-m

incontrolvideo

presen

tatio

n)to

enhance

unde

rstand

ingam

ongadults

with

schizoph

reniaand

healthycomparison

subjects.

Outpatient

clinicsof

anolde

rmen

tal

health

service

Men

talh

ealth:

olde

rpatientswith

schizoph

renia

Enha

nced

inform

edco

nsen

t:multim

edia

consen

t;theresearch

assistant(RA)p

rovide

dparticipantswith

theprinted

consen

tform

.Sub

jects

watched

aDVD

that

explaine

dtheprotocol.

Multip

lerepresen

tatio

nand

contiguity

principles

were

presen

tthroug

hout

theDVD

bypresen

tingconsen

t-relevant

inform

ationthroug

hanarrator

explaining

key

128middle-aged

andolde

rpe

rson

swith

schizoph

renia

and60

healthycomparison

subjects

Key

findings:ou

tpatients

with

schizoph

reniaprovided

with

amultim

edia-aided

consen

tproced

ure

demon

stratedbe

tter

compreh

ension

ofaresearch

protocol

andweremore

likelyto

becatego

rised

asbe

ingcapableof

consen

tun

derthreedifferent

standardsexam

ined

,comparedwith

those

presen

tedwith

anen

hanced

routineconsen

tproced

ure.

Evans et al. BMC Medicine (2020) 18:221 Page 13 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

points,w

ithsimultane

ous

visualpresen

tatio

nusing

graphics,p

ictures,anim

ations,

andsummarytext

(bullet-

pointed).Sub

jectswere

encouraged

toasktheRA

tostop

theDVD

andrepe

atany

segm

entsthat

wereun

clear.

Participantswereen

couraged

todiscussandclarify

issues

with

theRA

s.Such

discussion

isim

portantformultim

edia

consen

taids

toaidpe

rson

-to-personinteractionno

tto

substitute.

Adults

withim

paired

capacity

MacCAT-CRun

derstand

ing

subscaleforou

tpatientswith

schizoph

reniatrial1

(d=

0.6384,p

=0.0055,95%

CI

0.54,0.74),trial2

(d=0.6108,

p=0.0237,95%

CI0.52,0.71),

trial3

(d=0.6117,p

=0.0169,

95%

CI0.52,0.70),UBA

CC

total(d=0.6795,p

=0.0003,

95%

CI0.59,0.77).Therewere

few

differences

betw

eenthe

two(ro

utineandmultim

edia)

consen

tcond

ition

sam

ong

thehe

althycontrols.

Cha

lleng

e:compreh

ension

canbe

improved

with

simple

proced

ures

such

ascorrective

feed

back/iterativelearning

.Hen

ce,con

side

ringthe

additio

nalresou

rces,w

hether

afull-multim

edia

presen

tatio

nisne

eded

isqu

estio

nable.

Solution

s:multim

edia

consen

tproced

ures

may

bea

valuableconsen

taidthat

shou

ldbe

considered

foruse

whe

nen

rolling

participantsat

riskforim

pairedde

cision

alcapacity,p

articularlyfor

complex

and/or

high

-risk

research

protocols.

Harmelletal.

[91],U

SA,A

32012

RCTto

evaluate

the

prelim

inaryfeasibility

and

potentialeffectiven

essof

aweb

-med

iaapproach

toconsen

t,i.e.tode

term

ine

whe

ther

developm

entof

such

web

-med

iabasedtools

warrantsfurthe

rpu

rsuit.

Unclear

Men

talh

ealth:

patients

with

schizoph

renia

Enha

nced

inform

edco

nsen

t:participants

allocatedto

web

-med

iaconsen

treview

edthestud

yinform

ationon

aweb

-med

iaprototype,which

involved

vide

oclips,staticim

ages/

graphics,and

bullet-po

inted

text

toexplaintheconsen

tform

.The

printedconsen

tform

was

presen

tedon

the

screen

insections

coverin

g:e.g.

stud

yintrod

uctio

n,a

timelinewith

stud

yvisits,

descrip

tionof

proced

ures,

risks/discomforts,po

ssible

bene

fits,actio

nifinjured,

and

19patientswith

schizoph

reniaand16

norm

alcomparison

Key

findings:relativeto

thosereceivingtheroutine

consen

tproced

ure,those

receivingtheweb

-med

iaconsen

teviden

cedbe

tter

UCSD

BriefAssessm

entof

Capacity

toCon

sent

(UBA

CC)

scores,U

=19,z

=−2.15,p

=0.03

(d=0.94;‘large’effect

size).Participantsratedthe

quality

oftheen

hanced

consen

tproced

ureas

‘better’,

andno

participantrepo

rted

worse

expe

rience.

Cha

lleng

e:increasedleng

thof

administrationandthat

compu

ter-basedapproaches

Evans et al. BMC Medicine (2020) 18:221 Page 14 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

voluntaryparticipation.The

tool

includ

edqu

estio

nswith

correctivefeed

back

after

each

sectionto

check

unde

rstand

ing.

Participants

couldaskqu

estio

nsat

any

pointandreplay

the

presen

tatio

n.Adults

withim

paired

capacity

may

intim

idatepe

oplewith

less

compu

terliteracy.

Solution

s:incorporating

audio-visualmaterialson

acompu

ter/web

platform

toen

ableamoreinteractive

andflexiblepresen

tatio

nis

feasibleandmoreacceptable

than

presen

tatio

non

aDVD

.Such

presen

tatio

nmay

enableresearchersto

capitaliseon

thebe

nefitsof

audio-visuallearning

,while

circum

ventingthelim

itatio

nsof

aDVD

presen

tatio

n.

Sudo

reet

al.

[82],U

SA,B3

2006

Observatio

naln

estedin

atrial

oftw

oadvancedirectives

tode

scrib

eamod

ified

research

consen

tprocess,and

determ

inewhe

ther

literacy

andde

mog

raph

iccharacteristicsareassociated

with

unde

rstand

ingconsen

tinform

ation.

Hospital

Geriatriccare/older

patients:ethn

icallydiverse

subjects,age

d≥50,

consen

tingforatrialto

improvetheform

sused

for

advancedirectives

Enha

nced

inform

edco

nsen

t:amod

ified

consen

tprocess—

consen

tform

(writtenat

asixthgradelevel)

read

toparticipants,

combine

dwith

7compreh

ension

questio

nsandtargeted

education,

repe

ated

until

compreh

ension

achieved

(teach-to-go

al).

329po

tentialp

articipants.

Twen

typarticipantsrefused

toparticipate,39

exclud

eddu

eto

sche

dulingissues,61

didno

tmeettheeligibility

criteria,and

data

were

missing

for1participant,

leaving208participants.208/

329(63.2%

)

Key

findings:de

spite

sign

ificant

consen

tmod

ificatio

ns(im

proving

readability

oftheconsen

tform

,havingbilingu

alresearch

assistantsread

the

consen

tform

toparticipants,

andallowingtim

efor

discussion

),few

participants

(28%

)had

complete

compreh

ension

andrequ

ired

only1pass

throug

hthe

consen

tprocess.How

ever,

furthe

ruseof

ateach-to-goal

strategy

was

successful

inachievingcomplete

compreh

ension

in98%

ofall

participantswho

engage

din

theconsen

tprocess,

includ

ingthosewith

literacy

orlang

uage

barriers.

Cha

lleng

e:the

compreh

ension

statem

ents

couldhave

addressedallo

ftherequ

iredelem

entsof

inform

edconsen

t,making

ourresults

more

gene

ralisable.

Solution

s:forthemajority

oftheseparticipants,little

additio

naledu

catio

nwas

requ

ired.

Evans et al. BMC Medicine (2020) 18:221 Page 15 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

Rikkertet

al.

[81],N

ethe

rland

s,C1

1997

Pre-

andpo

st-teststud

yof

astep

-wiseconsen

tprocessto

determ

inetheeffectsof

research

expe

rienceon

the

capacity

toconsen

t.

Hospital

Geriatriccare:g

eriatric

patients

Step

-wiseco

nsen

t:(1)

Eligibility

screen

ing.(2)

Research

experienc

ewas

givenby

atry-ou

tpe

riodof

aweek.Verbalandwritten

inform

ationabou

tthetry-ou

twas

givento

alleligible

subjects(n

=78).For40%

ofpo

tentialsub

jects,thefamily

mem

berswerewillingto

accompany

them

whe

nthey

received

inform

ation.70

subjects(90%

)provide

dve

rbal

consen

tto

participate

inthetry-ou

t.(3)After

the

try-ou

t,written

inform

edco

nsen

twas

reque

sted

.The

verbalinform

ationwas

repe

ated

.Assessm

entof

the

capacity

toconsen

twas

cond

uctedbe

fore

andafter

expe

riencingresearch

bytestingcompreh

ension

and

ability

toweigh

risks

and

inconven

iences.

Adults

withim

paired

capacity

53/78(68%

)Key

findings:thetry-ou

teffect

onthecompreh

ension

scores

couldbe

tested

in53

subjectswho

provided

writteninform

edconsen

t.Initially,the

subjects

answ

ered

only5.0of

theten

questio

nscorrectly,this

numbe

rincreasedto

7.0after

thetry-ou

t.Thisstep

-wise

consen

tproced

ureresulted

inaparticipationrate

of68%

(53/78)o

falleligiblesubjects.

Duringthetry-ou

t,seven

subjectswith

drew

consen

t.Tensubjectsrefused

inform

edconsen

tto

continue

research

following

thetry-ou

t.So

lution

s:research

expe

rienceseem

sto

improve

thecapacity

toconsen

tin

demen

tedandde

pressed

subjectsas

wellasin

subjects

with

outpsycho

geriatric

illne

sses.

Processesto

enableadultslackingcapacity

toparticipatein

research

Advance

andprocessconsen

t

Olazaranet

al.[64],Spain,C1

2012

Research

protocol

toprovide

anoverview

oftheclinical

research

protocol

ofthe

ACRSF(Alzhe

imer

Cen

ter

ReinaSofia

Foun

datio

n),to

analysetheadeq

uacy

ofthe

assessmen

tinstrumen

ts,and

torepo

rton

change

sto

the

protocol.

ACRSF

research

centre

Dem

entia:

patientswith

Alzhe

imer’sor

othe

rde

men

tiaandtheirrelatives

who

agreed

toreceive

treatm

entfro

mtheACRSF

(MMSE

6.7(6.1)—

nursing

hometo

9.1(7.6)—

day-care

centre)

Advanc

eco

nsen

tup

onarrivaltothefacility;

lega

lrepresentativeto

provide

inform

edco

nsen

t:up

onadmission

,one

ortw

oACRSF

research

physicians

introd

uced

thepatient

and

family

caregiversto

the

ACRSFresearch

prog

ramme

andinvitedthelegal

represen

tativeto

sign

consen

tto

participate.The

consen

tform

includ

edseparate

boxesforagreem

ent

tospecifically

participatein

theclinical,b

iochem

ical,

gene

tic,M

RI,and

neurop

atho

logical

prog

rammes.

180(80%

ofthetotal)

Inform

edconsen

twas

obtained

from

180patients.

Thosepatientsrepresen

ted

approxim

ately80%

ofall

patientsadmitted

atthe

ACRSFdu

ringthat

perio

d.Dem

ograph

ics:thetw

ogrou

psof

patientsstud

ied

wereold(outpatients)or

very

old(inpatients),had

very

low

educationalachievemen

t,andwerepred

ominantly

wom

en.

Solution

s:multid

isciplinary

actio

n.

Evans et al. BMC Medicine (2020) 18:221 Page 16 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

Adults

withim

pairedor

dim

inishe

dcapacity

Gyselset

al.

[45],U

K,C1

2013

Con

sultatio

nworksho

pand

TECto

presen

ttheprocesses

andou

tcom

esof

aworksho

pandconsen

susexercise

onagreed

bestpracticeto

accommod

ateethicalissues

inresearch

onpalliativecare.

N/A

Palliativecare

Advanc

eco

nsen

t,process

consen

tAdults

withim

pairedor

dim

inishe

dcapacity

28Key

findings:16

recommen

datio

nsge

nerated.

Therecommen

datio

nson

obtainingandmaintaining

consen

tfro

mpatientsand

families

werethemost

conten

tious.

Cha

lleng

e:fluctuatin

gcapacity,tim

e,risks

involved

inparticipating.

Solution

s:inform

ingallthe

patients/relatives

onadmission

that

thefacility

cond

uctsresearch,m

inim

ises

gatekeep

ing,

andiden

tifies

peop

leinterested

inresearch

participation.Thelevelo

fde

tailon

theinform

ation

sheetsshou

ldbe

prop

ortio

naltobu

rden

and

risks.A

dvance

consen

t(early

inform

edconsen

twhe

nthe

patient

stillhascapacity)for

allresearch,no

tjustCTIMPs.

Con

tempo

rane

ousassent

shou

ldalso

beob

tained

for

alltrials.C

onsent

shou

ldbe

acontinuo

usprocess.Con

sent

process<24

hafter

approach

with

clear

justificatio

nto

avoidcoercion

.

Cow

dellet

al.[70],UK,B3

2008

Ethn

ograph

icstud

yexploring

strategies

that

wereused

toen

ableolde

rpe

oplewith

demen

tiato

becomeactively

engage

din

research

with

them

.

Hospital

Dem

entia:

inpatients≥65,

with

ade

men

tiadiagno

sis

atan

advanced

stageof

theillne

ss

Proc

essco

nsen

t:verbalor

behaviou

ralcon

sent

was

takenfro

mparticipantsat

the

beginn

ingof

everype

riodof

observationto

ensure

they

werewillingto

continue.This

consen

twas

nego

tiated

betw

eenthepe

rson

with

demen

tia,the

researcher,the

staffon

duty,and

onoccasion

thene

xtof

kin.

Participantswereob

served

foranysign

sthat

they

might

125hof

observations

and

interviews(n

unclear)in

aninpatient

setting

Key

findings:actively

engaging

olde

rpe

oplewith

demen

tiaeven

atadvanced

stages

(e.g.instead

ofhaving

apre-de

fined

MMSE

cut-off)

inresearch

ispo

ssible.

Researchersne

edto

apply

ethicalp

rinciples

andrules

sensitivelyandflexibly.

Cha

lleng

e:en

surin

gthe

inform

ednature

ofthe

consen

t,with

outhaving

aform

alcapacity

assessmen

t.

Evans et al. BMC Medicine (2020) 18:221 Page 17 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

wishto

with

draw

.For

the

interview

partof

thestud

y,participantswereaskedto

sign

aconsen

tform

.

Dun

ning

etal.[46],Australia,

B3

2012

Individu

alsemi-structured

interviews,field

notes,

philosoph

icalframew

orkto

discusstheethicaland

metho

dologicalissues

encoun

teredwhe

nun

dertakingresearch

tode

velopgu

idelines

for

managingdiabetes

atthe

EoL.

Participants’

homes

Palliativecare:sem

i-structured

interviewswith

14men

andwom

enwith

diabetes

and10

spou

sesof

the14

participants

Proc

essco

nsen

t:involved

asking

participantswhe

ther

they

wantedto

continue

the

conversatio

ndu

ringthe

interview.Process

consen

twas

used

whe

naparticipant

becameph

ysicallyor

emotionally

distressed

(inadditio

nto

inform

edco

nsen

t).

Adults

withim

paired

capacity

Not

know

nKey

findings:qu

alitative

research

deem

edto

bethe

mosteffectivedata

collection

metho

d.So

lution

s:attentionto

protectin

gparticipants’

privacy,en

surin

gthey

can

give

inform

edconsen

t,be

ing

awareof

theirph

ysicaland

men

talstate,and

perio

dically

checking

theirwillingn

essto

continue

participatingdu

ring

interviewsandfocusgrou

ps,

isessential.

Hug

heset

al.[65],UK,C1

2015

Qualitativeconsultatio

naimingto

developan

approach

with

inthegu

idance

oftheMen

talC

apacity

Act

(2005)

tomeaning

fully

includ

epe

oplediagno

sed

with

demen

tiain

research

ende

avou

rs.

Integrated

demen

tiaday

care

services

Dem

entia:

peop

lewith

ade

men

tiadiagno

sis

residing

atreside

ntialcare

homes

Proc

essco

nsen

tAdults

withim

paired

capacity

8/9(one

declined

toparticipatedu

eto

unexpe

cted

housingissues)

Con

sent

process:first

consen

tof

participating

serviceleadsto

allow

iden

tificationof

eligible

patients.Then

,the

tested

consen

tprocesswas

implem

ented,

andpatients’

consen

twas

assessed

and

recorded

.Con

sent

reliability—

nothaving

aon

e-offcon

sent

andrene

wing

consen

tat

everyen

coun

ter.

Cha

lleng

e:participantslost

trackof

thepu

rposeof

the

research.Impo

rtantto

attend

tono

n-verbalcues.

Gatekeeping

from

relatives

andsw

itchof

decision

-maker

from

patient

torelativebased

onthecontextof

the

decision

.So

lution

s:researchersshou

ldbe

traine

din

reflexive

assessmen

tof

consen

tusing

verbalandbe

haviou

ralcues.

Initialapproach

toparticipate

couldbe

byatraine

dservice

user

consultant

tobalance

thepo

wer

dynamics.

Researchersshou

ldinvolveall

Evans et al. BMC Medicine (2020) 18:221 Page 18 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

partiesno

tlim

itto,for

exam

ple,family

caregiver.

Carey

etal.

[95],Ireland

,B3

2017

Qualitativestud

yto

explore

theo

reticalun

derpinning

sof

intellectuald

isability

research,

andto

discusstheethicaland

metho

dological

considerations

inrecruitin

gandob

taininginform

edconsen

tfro

madultswith

intellectuald

isabilities.

Intellectual

disability

service/

participants’

preferred

placeand

time

Men

talh

ealth:

adultswith

intellectuald

isability

Proc

essco

nsen

tAdults

withim

paired

capacity

12/14

Con

sent

process:o

ngoing

inform

edconsen

t—de

tail

inform

ationsession,op

tionto

have

asupp

ortpe

rson

presen

t.Writtenconsen

t.Six

participantsalso

wantedtheir

consen

tto

berecorded

.Non

-verbalcues

wereconsidered

whileassessingon

going

consen

t.So

lution

:the

structureof

thesemeetin

gsfacilitated

discussion

sabou

tthenature

ofthestud

y.Key

findings:making

reason

ableaccommod

ations

tosupp

ortde

cision

-making,

makingspaceforthe

developm

entof

empathic

relatio

nships

with

both

the

potentialp

articipantsand

with

thestructures

and

servicesupp

orts.

Deferredconsen

t

Adamiset

al.[100],U

K,B3

2010

Prospe

ctivecoho

rtto

assess

serum

IGF-Iinpatientswith

delirium

andtheway

inwhich

results

alteredwhe

ninclud

ingpatientswith

delirium

who

lacked

capacity.

Elde

rlycare

unit

Delirium:p

atients70

years

ormorewith

thepresen

ceof

delirium

usingCon

fusion

Assessm

entMetho

d–Fluctuatingcapacity

Deferredco

nsen

t:those

who

lacked

capacity

were

entered(deferred)to

stud

yandtheircapacity

werere-

assessed

toseeifthey

gaine

dcapacityor

proxy

assent

was

obtained

.Adults

withdim

inishe

dcapacity

164/233(70%

).13/23

recruitedlacked

capacity

and

151/210recruitedwith

capacity.

Key

findings:theinclusion

ofthemoreincapacitated

subjectsallowed

asign

ificant

finding

(lower

serum

IGF-Iin

prevalen

tde

lirium

cases).

Streng

then

edtheeviden

cethat

IGF-Ih

asaroleto

play

inthepathop

hysiolog

yof

delirium.

Solution

:informalapproach

tocapacity

may

allow

for

morerepresen

tativeresults

ofthestud

ypo

pulatio

n.

Hon

armand

etal.[104],

Canada,B3

2018

Prospe

ctive,pilotstud

yto

describ

ethefeasibility

ofthe

deferred

consen

tmod

elin

alow-risk,ob

servationalstudy

ofcritically

illpatients

(Progn

ostic

Valueof

Elevated

Trop

oninsin

CriticalIllne

ssStud

y[PRO

-TRO

PICS])andto

Intensivecare

units

atthree

stud

ysites

across

Canada

Intensivecare:criticallyill

patients

Deferredco

nsen

t:patients

areen

rolledto

thestud

yand

then

them

selves

ortheir

surrog

atede

cision

-maker

isapproached

forconsen

t.Con

sent

canbe

provided

toon

goingstud

yparticipation,

useof

data

collected

sofar,

214/267(80%

)Key

findings:de

ferred

consen

tmod

elwas

feasible

with

80%

consen

trate.O

f53

person

sde

cliningconsen

t,37

(70%

)agreedto

theuseof

thedata

collected

tothat

point.One

patient

with

drew

consen

tafteritwas

provided

Evans et al. BMC Medicine (2020) 18:221 Page 19 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

determ

inethefactors

associated

with

consen

tprocurem

ent.

orno

consen

tfordata

tobe

used

.Adults

withdim

inishe

dcapacity

byaproxyde

cision

-maker.

But,patientsun

likelyto

recoverwereexclud

ed.

Con

sent

rate

didno

tdiffer

basedon

who

(patient/

surrog

ate)

was

consen

ting.

Cha

lleng

e:exclusionof

patientswho

might

not

recover/dieandexclusionof

patientswho

dieearly

orcann

otprovideconsen

twith

inthestud

ytim

eframe

lead

toselectionbias,

redu

cedstatisticalpo

wer,and

decreasedexternalvalidity.

Con

sulteeadvice

Blacket

al.

[60],U

SA,C

12007

Metho

dologicalp

aper

focusing

onthreeaspe

ctsof

theconsen

tprocessfor

demen

tiaresearch:(1)

providinginform

ation,(2)

assessingun

derstand

ingand

capacity

toconsen

t,and(3)

obtainingassent

and

inform

edconsen

t.Foreach

aspe

ct,the

differences

betw

eendrug

andno

n-drug

stud

iesin

CDRS

exam

ined

.

Sixparent

demen

tiastud

ies

Dem

entia

Inform

edco

nsen

tan

d/or

persona

lcon

sultee

advice

(dua

lcon

sent)

Researchers(n

=11),patients,

andtheirpe

rson

alconsultees

from

sixde

men

tiastud

ies—

46consen

tprocess

observations

Key

findings:stud

yrevealed

widevariabilityin

how

inform

edconsen

twas

obtained

.(1)

Con

sentsform

swereprovided

tothepatients

andpe

rson

alconsultees

prior

toen

rolm

entvisitsandoften

served

asagu

ideforconsen

tdiscussion

s;(2)consen

tdiscussion

sweremore

consistent

and

compreh

ensive

fordrug

stud

iesthan

non-drug

stud

ies;(3)stud

yproced

ure

explanations

dominated

the

discussion

s,whe

reas

the

rightsof

research

subjects

weremen

tione

dless

frequ

ently;(4)

assessmen

tsof

affected

individu

als’

unde

rstand

ingandcapacity

toconsen

toccurred

ina

minority

ofcasesbu

twere

morelikelyto

occuron

drug

stud

ies;(5)pe

rson

alconsulteeadvice

was

soug

htmoreoftenusingan

implicit

rather

than

anexplicit

approach;(6)

dualconsen

tby

both

theaffected

individu

alandsurrog

atede

cision

-maker

was

mostcommon

onbo

th

Evans et al. BMC Medicine (2020) 18:221 Page 20 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

drug

andno

n-drug

stud

ies;

(7)pe

rson

alconsultees

often

played

amajor

rolein

facilitatingtheconsen

tprocess.

Solution

s:de

scrib

ingthe

purposeof

thestud

y;discuss

theindividu

al’srig

htsin

detail;involvingthepe

rson

alconsultees;explicitlyaskthe

potentialp

articipantsfortheir

involvem

ent;usea

standardised

way

ofassessing

capacity;explain

why

ape

rson

alconsulteeadvice

isne

eded

.

Agarw

alet

al.[69],UK,C1

1996

Observatio

nalstudy

exam

iningtherelevanceof

theLaw

Com

mission

recommen

datio

nsin

accessinginform

edconsen

tfro

mearly

demen

tiapatients

andtheircarerssubjectedto

ado

uble-blind,

placeb

o-controlledtrialo

fapo

tentially

therapeutic

agen

t.

Unclear

Dem

entia:

patientsand

carers

Person

alco

nsulteead

vice:

twoqu

estio

nnaires(fo

rpatientsandtheircarers)

werede

sign

edto

exam

ine

whe

ther

subjectsfulfilledthe

criteria

fora‘cog

nitive’or

‘functio

n’testof

capacity

toconsentto

participatein

aresearch

study.Thiswas

anattempt

toestablishwhe

ther

consen

twas

a‘true

choice’.

Adults

withim

paired

capacity

15patientsandcarers

Key

findings:asing

lelegal

‘test’,with

strin

gent

criteria,

appliedacross

thebo

ardfor

alltreatmen

tandresearch

cond

ition

s,may

impe

defuture

research

activity

asno

neof

thesubjectsfulfilled

thecriteria

forde

term

ining

whe

ther

participationwas

atrue

choice.

Cha

lleng

e:im

pliedconsen

t(opt-out)cou

ldlead

toexploitatio

nof

vulnerable

patients.

Solution

s:theroleand

involvem

entof

carersin

the

decision

-makingprocess

need

tobe

considered

.Provided

that

they

areactin

gin

thepatient’sbe

stinterests,

that

thepatient

hasno

tactivelyexpressedade

sire

notto

participate,andthat

theresearch

ispo

tentially

therapeutic,w

iththeresearch

drug

having

negligibleside

effects,thisisun

likelyto

violatehisfund

amen

tal

rights.

Gaino

ttiet

al.[68],Italy,A

3(m

etho

dspape

r

2010

Metho

dologicalp

aper

hypo

thesisingthat

the

requ

iremen

tthat

inform

ed

Outpatient

clinic

Dem

entia:

outpatients

seekingmed

icaladvice

for

cogn

itive

complaints

Legally

appointed

consulteead

vice:the

proced

ureto

obtain

inform

ed

78/172

(46.2%

)req

uiredlegal

consulteeappo

intm

ent,55/

78(70.5%

)received

Key

findings:the

requ

iremen

tthat

thelegal

represen

tativebe

appo

inted

Evans et al. BMC Medicine (2020) 18:221 Page 21 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

repo

rtingRC

T[79])

consen

tforan

incapacitated

subject’s

participationto

research

begivenby

alegal

represen

tativeappo

intedby

thecourtsslow

sdo

wnthe

recruitm

entprocessin

research

thus

complicating

thecond

uctio

nof

demen

tiaresearch

inItaly.

consen

tin

thestud

ywas

quite

elaborated

.First,

subject’s

compe

tencewas

evaluatedby

means

ofthe

MMSE.Ifthesubject’s

score

was

≥20,the

nhe

/she

unde

rwen

tfour

additio

nal

neurop

sycholog

icaltests.If

thesubject’s

scoreto

thefour

testswas

high

erthan

the

establishe

dcut-offs,the

subjectwas

deem

edableto

give

inform

edconsen

t.Ifthe

subject’s

MMSE

scorewas

<20,adjustedforageand

education,or

ifthesubject’s

scoreto

theothe

rfour

tests

was

lower

than

the

establishe

dcut-offs,the

subjectwas

deem

edun

able

togive

inform

edconsen

tand

consen

thadto

begivenby

alegally

authorised

represen

tative.

Adults

withim

paired

capacity

appo

intm

ent

bythecourtsmay

impe

dea

subject’s

participationin

research.Itmay

cause

embarrassm

entandconflicts

amon

gfamily

mem

bers,it

may

have

been

received

asa

bureaucraticand

burden

sometask,and

relatives

may

bereluctantto

goto

courtdu

eto

stigma.

Thisresults

inon

lya

privilege

dselectionof

patientsbe

inginvolved

inthestud

ies.

Solution

s:Removalof

legal

proced

ureforthe

involvem

entof

consultees

orfasten

ingthelegalp

rocesses

andredu

cing

burden

.

Adamiset

al.[99],UK,A3

2005

RCTto

investigatewhe

ther

different

metho

dsof

obtaininginform

edconsen

taffected

recruitm

entto

astud

yof

delirium

inolde

r,med

icallyillho

spital

inpatients.

Acute

med

ical

servicefor

olde

rpe

ople

ataho

spital

Delirium:p

atients70

years

orolde

radmitted

tothe

unitwith

in3days

ofho

spitaladm

ission

Inform

edco

nsen

tor

proxy

assent:b

othgrou

psof

patientsweregivenroutine

sameinform

ation(verbally

andwritten).A

fter

aform

alcapacity

assessmen

t,assent

was

soug

htfro

maproxy(if

available)

ifpatient

lacked

capacity

ingrou

pA,w

hereas

ingrou

pB,an

inform

alcapacity

assessmen

ttook

placeandthoseindividu

als

who

deem

edto

lack

capacity

wereexclud

ed.

57assessed

ingrou

pA

(43.8%

),25/57(43.9%

)en

teredthestud

y.73

assessed

ingrou

pB(56.2%

),54/73(74%

)entered

the

stud

y.20

patientsin

each

grou

pwererecorded

as‘case

note

delirium’.

Key

findings:im

plem

entin

gbe

stethicalp

racticeby

aform

alassessmen

tof

capacity

toconsen

tto

aresearch

projectin

anacutemed

ical

wardwilllead

toa

considerableredu

ctionin

the

prop

ortio

nen

terin

gthe

stud

y.Astrin

gent

assessmen

tof

capacity

may

lead

toredu

cedge

neralisability

ofthestud

yfinding

s.In

turn,

thisun

derm

ines

theethical

justificatio

nof

thestud

y.Of

the20

patientsin

each

ofthe

initialrand

omised

grou

ps(A

andB)

with

case

note

delirium,7

(35%

)ingrou

pA

and16

(80%

)ingrou

pB(p

=0.004)

enteredthestud

y(χ2=

8.29,d

f=1;p=0.004).

Cha

lleng

e:researcher

assessingthecapacity

was

Evans et al. BMC Medicine (2020) 18:221 Page 22 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

notblinde

dto

grou

pallocatio

n.Manypo

tential

participantswith

delirium

dono

thave

form

alcapacity

toconsen

t.In

thisstud

y,includ

ingalmostall

prospe

ctivepatientsadmitted

toan

elde

rlycare

unit,40%

lacked

capacity

togive

consen

tto

thisresearch

whe

njudg

edform

ally.The

processof

form

altestingof

capacity

might

have

resulted

inbias

byindu

cing

high

erratesof

decliningto

give

consen

t.So

lution

s:theconsen

trate

may

begreaterifastep

-wise

approach

toconsen

tdu

ring

participationisused

.Inthis

approach,called‘experienced

consen

t’,verbalconsen

tis

accepted

initially,and

after

thesubjecthasexpe

rienced

theproject,writtenconsen

tis

soug

ht.

Morán-

Sánche

zet

al.[90],

Spain,C1

2016

Cross-sectio

nalsurveyto

evaluate

theassociation

betw

eencapacity

toconsen

tto

research

andthemore

prevalen

tpsychiatric

disorders,andto

characterise

factorsassociated

with

impairm

entsin

capacity

across

diagno

sticgrou

ps.

Men

talh

ealth

care

Men

talh

ealth:

psychiatric

patients

Inform

edco

nsen

tan

dlegal

gua

rdianco

nsen

t:capacity

was

assessed

usingMMSE

andMacCAT-CR.

Adults

withim

paired

capacity

139/235(59%

)Con

sent

process:informed

consen

tpatientswith

capacity

orfro

mlegal

guardian

iflacked

capacity.

Cap

acity:

thelevelo

fun

derstand

ingne

eded

toprovidemeaning

fulcon

sent

toparticipatein

thisminim

al-

riskprotocol

was

muchlower

than

that

requ

iredfora

complex

orhigh

er-riskclinical

trial,such

asthat

describ

edin

thehypo

theticalprotocol

used

toevaluate

capacity

inthestud

y.MacCAT-CRused

toassess

capacity.

Key

findings:no

subjectwas

exclud

edbe

causeof

alack

ofcapacity.31%

ofthe

participantslacked

decision

alcapacity

toprovideinform

edconsen

t.Thoselacking

capacity

weremorelikelyto

Evans et al. BMC Medicine (2020) 18:221 Page 23 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

beolde

r,with

severe

illne

ssover

along

ertim

e.The

numbe

rof

psychiatric

admission

swas

high

erin

the

incapacitatedgrou

p.They

weremorelikelyto

have

apsycho

ticor

moo

ddisorder

andto

scorelower

onthe

MMSE.

Solution

s:cogn

ition

mustbe

considered

incapacity

assessmen

t.Und

erstanding

canbe

improved

throug

hen

hanced

consen

tproced

ures.

Thom

allaet

al.[87],Germany,

A3

2017

RCT(baselinedata

only).Aim

tode

term

ineifthemanne

rof

consen

t,i.e.informed

consen

tby

theparticipantor

byproxy

decision

-maker,affected

clinicalcharacteristicsof

samples

ofacutestroke

patientsen

rolledin

clinical

trials.

Hospital

Stroke

:strokepatients

Inform

edco

nsen

t(w

ritten

ororal)bypatient,

persona

lorlega

lproxy,

consen

susbetwee

nthe

inve

stigator

andan

indep

enden

tclinician:

six

optio

nsgive

includ

ing

writtenor

oralconsen

tby

thepatient,leg

algu

ardian

consen

t,NoK

consen

t,investigator’sde

cision

(followed

with

consen

tfro

mNoK

assoon

aspo

ssible).

Adults

withdim

inishe

dcapacity

1005/1039(ong

oing

trial)

Key

findings:in

646(64%

)patients,inform

edconsen

twas

givenby

thepatients;in

359(36%

),consen

twas

bya

proxy.Therelativefre

quen

cyof

theinform

edconsen

ttype

used

variedam

ongcoun

tries

(p<0.001).Inthisanalysisof

baselinedata

ofthefirst1005

patientsen

rolledin

the

WAKE-UPtrial,abou

t1in

3patientswereen

rolledby

proxyconsen

t.In

thesecases,

consen

twas

provided

bythe

legalg

uardian,by

next

ofkin,

byan

inde

pend

ent

consultant,orby

the

investigator

basedon

anem

erge

ncyclause.

Cha

lleng

e:lim

itedgu

idance

arou

ndregu

latio

nof

clinical

research

inpatientslacking

capacity

togive

inform

edconsen

t,andthe

conseq

uenceof

different

approaches

forinform

edconsen

tused

indifferent

stroke

trials,amon

gcoun

tries

ortrialsites.

Kim

etal.

[66],U

SA,C

12011a

Toassess

theextent

towhich

person

swith

Alzhe

imer’s

disease(AD)retain

their

capacity

toappo

intaresearch

Interview

stud

yDem

entia:

peop

lewith

Alzhe

imer’sdisease(M

MSE

18–23)

Prox

yco

nsen

tAdults

withdim

inishe

dor

impairedcapacity

700

Key

findings:successful

recruitm

enthadthehigh

est

prop

ortio

n(46.3%

)of

participantswith

MMSE

score

Evans et al. BMC Medicine (2020) 18:221 Page 24 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

proxy.

rang

eof

18–23.Reliabilityof

thejudg

es’d

eterminationof

capacity

was

high

.61.7%

ofparticipantshadthecapacity

toappo

intaresearch

proxy,

41.4%

hadcapacity

toconsen

tto

thedrug

RCT,and

15.6%

hadthecapacity

toconsen

tto

thene

urosurgical

RCT.Asubstantialp

ropo

rtion

ofADsubjectsthou

ght

incapableof

consen

tingto

lower

orto

high

erriskstud

ies

hadcapacity

toappo

inta

research

proxy.

Solution

s:providingforan

appo

intedproxyeven

after

theon

setof

ADmay

help

addresskeyethical

challeng

esto

ADresearch.

App

ointingaproxyis

advocatedearly

inthe

diseasetrajectory.

Warrenet

al.

[47],U

SA,C

11986

Qualitativeinterviewsto

exam

inethebaseson

which

theproxiesmadetheir

decision

,toiden

tify

characteristicsthat

distingu

ishe

dproxieswho

refusedconsen

tfro

mthose

who

gave

consen

t,andto

determ

inereason

sforrefusal.

Nursing

home

Palliativecare:n

ursing

homereside

nts’(tho

sewho

were>65

yearsold)

proxies

Prox

y/surrog

ateco

nsen

tAdults

withdim

inishe

dcapacity

90/168

(54%

)Key

findings:54%

(n=90)

ofproxiesapproached

consen

tedto

patients’

participation.60%

ofproxies

consultedothe

rpe

opleabou

ttheirde

cision

;27%

consulted

aclinicianabou

tadvising

participation(orno

t).N

osign

ificant

differencein

the

frequ

ency

ofconsen

tam

ong

thosewho

decide

dalon

e,thosewho

consultedothe

rs,

andthosewho

consulteda

clinician.Mostproxieswere

notop

posedto

research

inge

neral,bu

tas

beliefsand

percep

tions

abou

tresearch

becamemorerelevant

totheirow

nfamily

mem

bers,

theirsupp

ortforresearch,

even

intheabstract,

declined

.96%

thou

ght

research

inge

neralw

asim

portantformed

icalcare;

87%

agreed

forresearch

tobe

unde

rtaken

inho

spital;

Evans et al. BMC Medicine (2020) 18:221 Page 25 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

83%

thou

ghtthat

elde

rlype

opleshou

ldparticipatein

research;only66%

thou

ght

that

research

shou

ldbe

cond

uctedin

nursingho

mes.

Solution

s:broaded

ucational

effortto

increase

awaren

ess

oftherelevanceof

research

tothehe

alth

ofolde

rpe

ople

across

care

settings.To

discussparticipationin

research

with

patientswhile

they

arecompe

tent

andto

includ

epo

tentialp

roxies

intheseearly

discussion

s.

Karlawishet

al.[67],USA

,C1

2008

Com

panion

stud

yto

anRC

Tplaceb

ocontrolledof

apo

tentialA

lzhe

imer’sdisease

treatm

ent(drug)

stud

yto

exam

inetheview

sof

Alzhe

imer’sdiseasepatients

andtheirstud

ypartne

rson

theethics

ofproxyconsen

tforclinicalresearch.

Universities

Dem

entia:

patientswith

mild-to-mod

erateAD

(MMSE

12to

16),andtheir

stud

ypartne

rs(spo

useor

adultchild)

Prox

yco

nsen

tAdults

withdim

inishe

dcapacity

59/73(81%

)patients,60/75

(80%

)study

partne

rsKey

findings:stud

ypartne

rsof

patientsjudg

edincapable

ofgiving

inform

edconsen

trepo

rted

thesamede

gree

ofpatient

involvem

entin

the

decision

toen

rolasthestud

ypartne

rsof

patientscapable

ofgiving

inform

edconsen

t.Moststud

ypartne

rsand

patientssupp

ortedproxy

consen

tfortheclinicaltrial,

andne

arlyallp

atientschose

theirstud

ypartne

randtheir

proxy.Stud

ypartne

rsge

nerally

maderesearch

enrolm

entaligne

dwith

maxim

isingthepatient’swell-

being.

Solution

:pursueaprocessof

shared

decision

-making

betw

eenpatient

andstud

ypartne

rto

recruitpatients

with

impairedcapacity.

Smith

etal.

[107],Canada,C1

2013

Qualitativestud

ywith

interactivefocusgrou

psto

presen

tstrategies

that

may

optim

isetheprocessof

obtaininginform

edconsen

tfro

msubstitutede

cision

-makersforparticipationof

critically

illpatientsin

trials.

Intensivecare

unit

Intensivecare:research

coordinatorsworking

with

critically

illpatients

Surrog

ateco

nsen

t:inform

edco

nsen

tfrom

substitutedecision-makers

ofcritically

illpatients

Adults

withdim

inishe

dcapacity

71Key

findings/solution

s:(1)

brandthetrialw

ithkey

message

s,(2)trainthelocal

person

nel,(3)prom

otea

cultu

reof

research,(4)

befamiliar

with

patient

family

dynamics,(5)involvebe

dside

staff—

makethem

awarethat

youareinterested

inrecruitin

gtheirpatient,(6)

Evans et al. BMC Medicine (2020) 18:221 Page 26 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

introd

ucetheidea

ofresearch

inaprofession

almanne

r—explainwhy

you

areapproachingthe

surrog

atede

cision

-maker,(7)

presen

tthefactsabou

tthe

research

prob

lem

and

outcom

es,(8)

convey

risks

andbe

nefitstransparen

tly,(9)

describ

ealternatives

toparticipationandsupp

ortthe

consen

tde

cision

,(10)explain

allresearch-relatedactivities,

(11)

Docum

enttheconsen

tprocess,(12)

providethanks

andon

goingstud

yup

dates

toallstakeho

lders,and(13)

follow-upwith

thepatient

toen

sure

ongo

ingconsen

t.Strategies

reinforce

requ

iremen

tsou

tline

din

existin

glegislations

and

additio

nalp

rocess

toen

hancetheintegrity

ofthe

consen

tprocess.

Bolcic-

Jankovicet

al.

[105],USA

,C1

2014

Cross-sectio

nalq

uantitative

survey

on(a)the

determ

inantsof

confiden

cein

asurrog

ate’sability

tomakea

decision

forthepatient,(b)

thedifferencebe

tween

surrog

ates’and

patients’

confiden

ce,(c)ifgreater

confiden

ceincreases

agreem

entbe

tweenthe

surrog

ate’sandpatient’s

respon

se.

Intensivecare

unit

Intensivecare:p

atients

who

requ

iredICUandwho

hadpo

tentialtoregain

capacity

afterrecovery,and

surrog

atede

cision

-makers.

Surrog

ateco

nsen

tAdults

withdim

inishe

dcapacity

445surrog

ates,214

patients

Con

sent

process:surrogate

consen

tob

tained

durin

gthe

patient’sICUadmission

.Key

findings:theresearch

fund

er,kno

wledg

eabou

tthe

stud

y,anddiscussing

with

atrustedpe

rson

were

associated

with

surrog

ates’

confiden

cein

advocatin

gparticipationandattitud

estowards

research.Patients’

confiden

cein

their

surrog

ates’d

ecisionwas

associated

with

aprevious

discussion

abou

tresearch

participation(p

<.001).

Con

fiden

tsurrog

ates

respon

dedin

agreem

entwith

patients’wishe

s(80%

).Most

surrog

ates

wantedto

represen

tthepe

rson

’swishe

s.So

lution

:Earlydiscussion

sbe

tweentheproxyand

patient

abou

tresearch

Evans et al. BMC Medicine (2020) 18:221 Page 27 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

participation.

Fowelletal.

[48],U

K,A3

2006

Rand

omised

crossovertrialto

explorethefeasibility

oftw

ode

sign

sof

consen

tfordying

patients:rand

omised

consen

t(aka

Zelen’sde

sign

)and

clusterconsen

tto

seewhich

design

ismoreeffectivefor

trialsin

palliativecare.

One

oncology

andon

epalliativecare

unit

Palliativecare/can

cer:

patientswith

aterm

inal

cancer

diagno

siswho

were

onan

integrated

care

pathway

forthedying

Cluster

consen

t:consen

tat

unitlevelfor

agrou

pfro

ma

‘cluster

guardian’,and‘cluster

gatekeep

er’respo

nsiblefor

individu

alpatient

approach.

Both

guardian

and

gatekeep

ermustgive

written

agreem

entfortheirclusterto

participatein

thetrial.

Rand

omised

consen

t(Zelen

’sdesign):seeks

inform

edconsen

tafter

rand

omisationbu

ton

lyifthe

patient

isto

receivethe

expe

rimen

taltreatmen

t.Adults

withdeclin

ing,

impaired,o

rdim

inishe

dcapacity

20/50(60%

)Key

findings:theinitial

requ

estto

abstract

data

was

iden

ticalin

both

design

sand

sign

ificantlyfewer

Zelen

patientsin

thelarger

unit

consen

tedto

this.Zelen

’sde

sign

redu

cesthebu

rden

ofseekingconsen

tfor

treatm

entallocatio

nbu

tdid

notim

proverecruitm

ent.

Solution

s:cluster

rand

omisationruns

inthe

backgrou

nd,red

ucingbu

rden

onthepatient,carer,and

clinician.Con

sent

totreatm

entallocatio

nisat

the

unitlevelw

ithindividu

alpatient

consen

tforaccess

toconfiden

tialm

edicaldata.

Thisstud

yillustrates

how

clusterrand

omisationexploits

thesenaturaladvantage

s,particularlywith

dying

patients.

Levine

etal.

[106],USA

,C1

2017

Expe

rtconsultatio

n—electron

icsurvey

followed

with

Delph

irou

ndsaimingto

establishabroade

rconsen

sus

onthebarriersto

emerge

ncy

care

research

glob

allyand

prop

oses

acompreh

ensive

arrayof

new

recommen

datio

nsto

overcomethesebarriers.

Global

emerge

ncy

med

icine

coverin

g

Intensivecare/emergen

cymed

icine:

expe

rtsin

glob

alem

erge

ncymed

icine

Com

mun

ityco

nsen

tAdults

withdim

inishe

dcapacity

80So

lution

s/suggestion

s:stream

linedata

collection,

iden

tifying

alternatives

tolocalIRB

approvalandthe

useof

commun

ityconsen

twhe

napprop

riate

whe

rethe

individu

alscanchoo

seto

opt-ou

tof

thestud

ylateron

.Keyfinding

sweredivide

dinto

four

catego

ries.(1)

Limitedavailabilityof

research

training

.(2)

Logisticalissues

andlack

ofdata

collection

standardisation.(3)Ethical

barriersregardingcond

uctin

gresearch

inlow-in

come

coun

tries.(4)Dearthof

fund

ingforglob

alem

erge

ncyresearch.

Key

findings:ne

edfor

ethicalcurriculum

sinclud

ing

impo

rtanttopics

relatedto

Evans et al. BMC Medicine (2020) 18:221 Page 28 of 55

Table

3Stud

iesinno

vatin

gresearch

metho

dsto

recruitadultsacross

thecapacity

trajectory

(group

edby

solutio

ns)(n

=37

stud

ies)(Con

tinued)

Autho

rs,cou

ntry,

EPOCgrade

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadults

across

thecapacity

spectrum

No.patients/eligible(%)

Keyfinding

s,challeng

es,and

solutio

ns

theethics

ofacutecare

research

internationally

such

asconsen

t,loss

tofollow-up,

coercion

,and

undu

einfluen

ceforen

rolm

entwas

high

lighted

.

Boxallet

al.

[88],U

K,B3

2016

Qualitativefocusgrou

psand

interviewsexploringthe

barriersto

recruitin

gstroke

patientsto

clinicaltrialsfro

mtheview

pointof

expe

rienced

nurseresearchers.Second

ary

aimsinclud

edexploringthe

factorsaffectingthe

recruitm

entof

stroke

patients,

explorethemainthem

esthat

influen

cerecruitm

ent,and

determ

ineifstroke

research

facesun

ique

recruitm

ent

issues.

Hospitals

Stroke

:strokeresearch

nurses

Surrog

ateco

nsen

t,param

edic/early-on-scen

eco

nsen

t,exception

/delayed

consen

tAdults

withdim

inishe

dcapacity

12Cha

lleng

es:restrictive

inclusion/exclusioncriteria,

physicianen

dorsem

ent,and

lack

ofclinicaleq

uipo

ise.

Impairm

entsaffecting

capacity

toconsen

t—lack

ofvalidated

toolsto

help

assess

capacity.A

cute

timeframeto

recruit,paternalism

of(especially

less-experienced

)nu

rseresearchers.

Solution

s:en

gaging

caregiversand,

ifpo

ssible,

usingsurrog

ateconsen

t.Find

ingabalancebe

tween

giving

patientsthe

oppo

rtun

ityyetno

tcoercing

them

.Con

sent

process

(sug

gestion

s):

■Tradition

al,in-ho

spital

consen

twith

aclinicianand

writteninform

ation.

■Paramed

ics/early

on-scene

consen

t.■Exceptionor

delayed

consen

t.■Surrog

ate(re

lative,legal

represen

tative,or

inde

pend

entph

ysician)

consen

t.■Shortor

abbreviated

writteninform

ation.

■Use

ofpictorialinformation

sheetsor

vide

osto

explaina

stud

y■Teleph

oneor

vide

oconsen

t.

Evans et al. BMC Medicine (2020) 18:221 Page 29 of 55

process and standardising the information provided toencompass both study procedures and participants’rights to, for example, withdrawal.Successful enhanced informed consent processes often

included providing information in more than one format(generally verbal and written) [61], improving under-standing by using novel techniques (e.g. simplified story-book, video) [91–93], and tailoring the process of studyapproach to a person’s psychological and physical healthstatus [43]. In studies involving patients with schizophre-nia [91, 93, 94] or dementia [62, 63], using a combin-ation of multimedia techniques and flexibility to repeataspects of the study information significantly improvedindividuals’ understanding about the study, and to givean informed consent. Studies using enhanced informedconsent processes reported generally high consent ratesfor eligible participants ranging from 68.0% (n = 53/78)in an elderly patient population [81] (mean age 80.1years) to 95.1% (n = 136/143) in patients with Parkin-son’s disease (mean age 71 years) [86]. The addition ofsocial support from family members enhanced older in-dividuals’ decision-making capacity. For example, a studywith hospitalised older cancer patients (MMSE mean24.0, SD 4.1) achieved 74% (37/50) recruitment rate byformal involvement of family members including pres-ence during the consent process, asking their views, en-couraging patients to discuss with the family members,and taking their concerns into account [44]. In an RCT(n = 136) on the effect of social support on the consentprocess for patients with Parkinson’s disease (mean age71 years, SD 8.6), the presence of a family member com-pared to the patient alone showed effect on enhancingcomprehension and recall of the study information at1 week (p = 0.012) and 1 month (p = 0.040) [86].However, the evidence for using multiple or iterative

techniques to enhance individuals’ understanding andreasoning was mixed. Variation was related to the con-text and the processes used. Using techniques of itera-tive learning by, for example, quizzing potentialparticipants’ understanding of the study and enhancinghow information was provided [86, 92] enabled studyengagement for individuals with lower levels of healthliteracy from, for example, lower socioeconomic back-grounds [82] or low- to middle-income countries(LMICs) [94]. However, findings were mixed for studiesrecruiting adults with dementia. RCTs showed enhancedunderstanding (F [1, 29] = 7.17, p = 0.012) using a Power-Point presentation, combined with verbal consent andverbal re-explanation for participants with mild Alzhei-mer’s disease (MMSE ≥ 19) (n = 53) [61], and using asimplified memory and organisational aid to improveunderstanding (χ2 = 8.2, df = 1, p = 0.004) (n = 110) [63].Conversely, a RCT (n = 114) using multimedia and itera-tive learning for participants with mild to moderate

Alzheimer’s disease (MMSE 20.9 (SD 3.9) and 22.5 (SD3.4), by treatment arm) did not detect a significant effecton decisional capacity (χ2(3) = 2.63, p = 0.453) [62].

Processes of consent to enable recruitment of adults acrossthe capacity spectrumTwenty-one studies explored research methods in theconsent process for adults across the capacity spectrum.Advance consent was advocated in two studies for par-ticipants with the anticipated loss of capacity associatedwith a progressive condition, e.g. dementia [45, 64]. Ad-vance consent involved early informed consent when theperson had capacity, for example, on admission to a clin-ical facility, with consent upheld at the point of loss ofcapacity [64]. An area of contention was the requirement(or not) in all research studies (Clinical Trial of an Inves-tigational Medicinal Product (CTIMP) and non-CTIMP)for the person to nominate a consultee for contempor-aneous advice on continued participation in the study[45]. Best practice was conducting the advance consentprocess with the person and the consultee (e.g. a familymember), to discuss the person’s preference for contin-ued participation should they lose capacity, and the roleof the consultee on contemporaneous advice aligned tothe person’s wishes and the context [45, 64]. This strat-egy may minimise potential gatekeeping about continuedparticipation with disease progression [45]. Process con-sent was proposed for studies with multiple time pointsthat involved individuals able to understand and appraiseinformation in the moment, but with difficulty retainingand recalling study-related information in the future, forexample, individuals with mild/moderate dementia, pal-liative care, and mental health populations [46, 65, 95].Process consent involved continuous monitoring of thevalidity of the informed consent provided at the begin-ning of the study, for example, verbal confirmation ofconsent at each time point with the formality of theprocess consent proportionate to the risks involved inparticipation.Two studies reported the use of deferred consent in

circumstances of fluctuating capacity with participantsanticipated to regain capacity with reversibility of theunderlying cause, for example, an infection [100, 104].Patients who were unlikely to recover or die were typic-ally excluded. Patients entered the study without theirprior consent, with consent deferred until they regainedcapacity and/or a consultee was approached. These typesof studies typically involved low-risk observational pro-cedures conducted with, for example, older patientswith delirium [100] or in the ICU [104]. Using de-ferred consent enabled the inclusion of participants atthe acute point of their illness spectrum. Deferredconsent appeared acceptable in the ICU study with an80.1% (n = 214/267) consent rate [104], and sensitivity

Evans et al. BMC Medicine (2020) 18:221 Page 30 of 55

analysis of the delirium study indicated that excludingALC would have compromised the detection of statis-tically significant findings [100].The main process of enabling adults lacking capacity

to participate in research was seeking advice from a con-sultee about study participation. Most studies (n = 11)explored the involvement of consultees across researchstudies in geriatric/stroke care (all settings) (n = 7), ICU(n = 2), or mental health services (n = 1) (Table 3). Stud-ies conferred that the role of a consultee was not to pro-vide substituted judgement, but rather to give advicealigned with the patient’s wishes and well-being. This re-quired researchers to consult patients as much as pos-sible about participation and encourage them to identifya consultee to advise on their behalf if they lost capacity[66]. Key to this process was discussion and shareddecision-making between the consultee and the patientwhen they had capacity [67], and including potentialconsultees in early discussions with the person, whilecapacitous about future participation in research [47].An observational study in ICUs involving patients (n =214) considered likely to regain capacity, and familymembers (n = 445) as surrogate decision-makers, re-ported that the only factor significantly associated withpatients’ confidence in their consultee’s consent decisionwas having a previous conversation about research par-ticipation [105]. Consultees also reported reduced stressfrom their role when they could align their advice withunderstanding of the person’s priorities [67]. However, astudy involving nursing home residents reported no as-sociations with patients’ characteristics and consulteedecision to advise enrolment (54%, n = 90) (or not) [47].While consultees were generally family members (in-

cluding close friends), consent processes also allowedprofessional consultees or legal representatives to adviseon behalf of an individual lacking capacity. However,guidance and regulation around the participation ofincapacitated adults in research and involvement ofconsultees varied by jurisdiction. An international studyon stroke (n = 1005, across six European countries) [87]reported that one in three patients lacked capacity toconsent. The study found considerable variation in thejurisdictions’ respective enrolment requirements, detail-ing four different processes for adults lacking capacity asto who to approach as the proxy decision-maker (e.g. alegal guardian, next of kin, or independent physicianconsulting with the next of kin on the patient’s pre-sumed will). In the USA, institutional review boards(IRBs) were reported to differ in their rates of allowingresearch involving incapacitated adults regardless of therisks and benefits, and in who could act as a consulteewithin the studies [117]. For instance, 15% of the IRBsdisallowed participation in research without direct bene-fit regardless of risks, while 22% of IRBs accepted only

an authorised proxy, spouse, or parent as surrogates, ex-cluding adult children and other family members. Stud-ies conducted with patients with dementia [68, 69]showed that requiring a legally appointed consultee ledto declining participation from family members due tothe bureaucracy and the time involved in the legal ap-pointment. Such legal requirements could halt accumu-lation of evidence required for enhancing care ofpatients with impaired capacity.Cluster consent was advocated for trials involving

adults with impaired capacity in palliative care [48] andemergency medicine [106]. Cluster consent was under-taken by a ‘cluster guardian’ giving written agreementfor treatment allocation of the defined cluster at thelevel of a unit, e.g. a hospital ward. This enabled the allo-cation of treatment at the cluster level. However, indi-vidual consent from the person, or consultee advice, wasrequired for access to personal medical data for researchpurposes. Recommendations in emergency medicine re-search included enhancing recruitment using commu-nity consent as a similar process with ‘group consent’ forthe study and individual ‘opt-out’ of the study at a laterstage, for example, once capacity was regained using asimilar process to deferred consent [106].

Applying consent processes across the capacity spectrumin studies on serious illnessThirty-two publications reported 30 original studies in-volving adults with serious illness and impaired mentalcapacity (n = 9046) (see Table 4). These studies wereconducted across clinical settings (e.g. hospices, hospi-tals) and populations (e.g. elderly, dementia, delirium).Fourteen (46.6%) studies were RCTs (including feasibil-ity), and the remaining observational including prospect-ive and cross-sectional designs. The studies enabledanalysis on the use of different consent processes andthe outcome of the recruitment rate for adults acrossthe capacity spectrum in clinical studies.Most studies were conducted in palliative care (n = 8)

or dementia/geriatric care (n = 10). The studies used abreadth of consent processes tailored to the respectivepopulation. Methods included personal and/or profes-sional consultee advice, advance and process consent,enhanced informed consent, and deferred consent. Pro-cesses showed variation in recruitment of eligible partici-pants (range 23.9% [72] to 78.8% [42]). Use of a personalconsultee alone (e.g. a family member) was uncommonand showed the greatest variability in the recruitmentrate. Four studies used personal consultee advice onlyfor adults lacking capacity [49, 50, 83, 108]. The recruit-ment rates varied by study population and study design.In a cross-sectional survey exploring patients’ palliativecare needs in hospital, the overall recruitment rate was48.1% (654/1359) [49]. While most eligible patients

Evans et al. BMC Medicine (2020) 18:221 Page 31 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

Palliativecare

Abe

rnethy

etal.[51],

Australia,A

3

2006

Cluster

RCT2×2×2to

test

GPed

ucationalo

utreach

visitin

gandcase

conferen

cing

toim

provepatient

outcom

es,

e.g.

pain

managem

ent.

Com

mun

itysettings

Palliativecare

Assen

tpersona

lcon

sultee

,or

nominated

consultee,e.g.

GP

capacity

assessmen

tMMSE

≤24

indicate

requ

ireassent.

n=461/607(76%

)Sampleno

tdifferentiatedby

capacity/lack

capacity

Key

findings:strategies

used

enabledrecruitm

entof

largest

commun

itystud

yin

palliative

care.

Solution

s:(1)stud

yde

sign

pragmatic2×2×2cluster

RCTandmetho

ds,b

road

inclusioncriteria,d

efined

recruitm

entplan,d

etailed

interven

tion;(2)minim

ise

patient

burden

,and

clinician

burden

,e.g.researchnu

rses

collect

data.

Cha

lleng

es:(1)

timeand

resource

complex

triald

esign.

(2)7patientsineligibleno

caregiveravailable/no

pain.

Gardine

ret

al.[49],UK,B3

2013

Cross-sectio

nalsurveyto

explorepalliativecare

need

inho

spitaland

agreem

ent

betw

eeninform

ants.

Allho

spital

inpatient

adult

ward

Palliativecare

Assen

tpersona

lcon

sultee

ALC

(n=38

).Cap

acity

assessmen

tMCAcriteria:

HCP,family

ifavailable.

n=654/1359

(48%

)(meanage

78years)

Key

findings:36%

patients’

palliativecare

need

s.Low

iden

tificationmed

ical(16%

)andnu

rsing(17%

)staff.In

23cases,consultees

completed

questio

nnaireson

behalfof

patientswho

lacked

capacity

toconsen

t,andrespon

sesgiven

viaconsulteemay

notbe

accurate.

Cha

lleng

es:n

on-con

senting

patients(n

=582)—(1)patient/

consulteede

clined

(e.g.too

ill)

(n=407);(2)

consulteeno

tcontactable(n

=109).

Rees

etal.

[54],U

K,A3

2003

Feasibility

RCTon

anti-

muscarin

icmed

ication

(hyoscineversus

glycop

yrronium

)for‘death

rattle’tode

velopaprocessof

advanceconsen

ten

abling

research

tobe

unde

rtaken

intheterm

inalph

ase.

Palliativecare

ward

Palliativecare

Advanc

ean

dprocess

consen

t:patientsiden

tifiedon

admission

,asked

ifprep

ared

toen

terstud

yifde

velop

secretions,informed

consen

tdo

cumen

tedmed

icalno

tes,

consen

treconfirm

edat

readmission

s.

n=58/107

(54%

)(n=15

develope

dde

athrattle

rand

omised

)

Key

finding:A

dvance

consen

tisaviableandacceptable

metho

dto

consen

tfortrialsin

dyingph

ase.Patient

accrual

ratesto

date

arelower

than

need

edto

recruitadeq

uate

numbe

rsin

thetim

eallotted

toansw

ertheresearch

questio

n.Cha

lleng

es:(1)

complexities

ofrecruitin

gpatients,e.g.

too

unwell(n=15),died

elsewhe

re(n

=16),died

pre-

rand

omisation(n

=15);(2)

resource

intense—

estim

ate

Evans et al. BMC Medicine (2020) 18:221 Page 32 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

takes3yearsto

meetsample

size;(3)

consen

tprocesstim

e-consum

ingandem

otionally

draining

.So

lution

s:(1)stud

yacross

care

settings,e.g.hospices;(2)

involveallH

CPdisciplines;(3)

patient

acceptability

trialsdying

phase(16patientsde

clined

).

Whe

lanet

al.

[52],U

K,A3

2013

RCTexploringtheim

pact

oftherequ

iremen

tfor‘proxy

assent’onrecruitm

entin

atrial

ofantib

odyrespon

seto

influen

zavaccinationanduse

ofabo

osterdo

sewhe

nindicatedversus

usualcarein

care

homes

(FEVER

Trial).

Allcare

homes

inthreeLond

onbo

roug

hs

Older

peop

lein

care

homes

Person

alco

nsulteead

vice

(n=82,14%

)ifun

available

professiona

lcon

sultee

care

homestaff(n

=40,13%

).Cap

acityassessmen

tinform

al(study

pre-datesMCA2005).

Lack

capacity

n=122/557

(22%

)Capacity

n=155/411(38%

)Overall=277/968(29%

)

Key

findings:difficulties

attainingconsulteeadvise

causeof

recruitm

entbias

with

lower

recruitm

ent,e.g.

olde

rparticipants.C

areho

mestaffas

consulteerarelyused

,e.g.

reluctance

ofrespon

sibility.

Furthe

rresearch

requ

ired

inde

pend

entrisk/be

nefit

expe

rtpane

ls,e.g.Ind

epen

dent

Men

talC

apacity

Advocates

(MCA2005).

Cha

lleng

es:(1)

high

lack

ofcapacity

toconsen

t(62%

,n=

602);(2)

nocontactable

consultee(n

=304,55%);(3)

relativede

clined

perm

ission

enrolm

ent(n

=27,5%);(4)

researcher

considered

patient

likelyto

resistproced

ures

(n=

146,26%).

Hen

woo

det

al.[56],

Australia,B3

2014

Cross-sectio

nalstudy

with

rand

omsampling.

Aim

toestablishtheprevalen

ceand

riskfactorsto

sarcop

enia

amon

golde

radultswith

comprom

ised

well-b

eing

residing

inreside

ntialage

dcare

(RAC).

Reside

ntialage

dcare

Very

old(m

eanage

84.5years)reside

nts

ofthecare

facility

Allparticipantswererequ

ired

togive

inform

edco

nsen

tdirectlyor

bythesubstitute

decision-maker,o

rbythe

serviceman

ager

ordirector

ofnu

rsingfollowing

discussion

with

thesubstitute

decision

-maker.

102/273(37%

)—91

provided

consen

tand11

were

consen

tedby

proxy

Key

findings/ch

alleng

es:

high

estreason

ofno

n-recruitm

entwas

notwantin

gto

participate(79%

),followed

by‘MyGPdo

esno

twantmeto’

(7%),change

sin

health

status,

andde

ath.

Solution

s:To

facilitate

recruitm

ent,requ

ires

involvem

entof

keystaff

mem

bersin

planning

and

executionof

theresearch

stud

yto

supp

ortthestud

yand

unde

rstand

ingof

thestud

yprotocol.

Evans et al. BMC Medicine (2020) 18:221 Page 33 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

Myerset

al.

[50],U

SA,B3

2018

Prospe

ctivecoho

rtstud

yto

determ

inewhe

ther

unplanne

dho

spitaltransfercanbe

avoide

d.

One

largecoun

tywith

asing

lesystem

ofem

erge

ncy

med

icalservices

Assistedliving

reside

ntswho

fall

Each

patient’sprim

arycare

physicianinform

edthepatient

orhisor

herhe

althcare

Power

ofAttorne

yof

thestud

ydu

ring

usualcare.Pa

tien

tschoo

sing

toparticipateor

theirpow

ers

ofattorney

sign

edawritten

inform

edco

nsen

tdo

cumen

t.

953/1473

(65%

)Key

findings:of

the953

reside

ntsin

thestud

y,359had

840falls

durin

g43

mon

ths.The

protocol

recommen

dedno

n-transfer

after553falls.Elevenof

thesepatientshadatim

e-sensitive

cond

ition

.549

ofthe

553patients(99%

[CI,98

to100%

])with

aprotocol

recommen

datio

nforno

n-transfer

received

approp

riate

care.

Cha

lleng

e:participantaccrual

was

slow

erthan

anticipated

andtim

e-sensitive

cond

ition

swereless

prevalen

t.

Irwin

etal.

[55],U

SA,C

12008

Pilotstud

yusingaconven

ience

sampleto

assess

cogn

ition

and

evaluate

thepresen

ceof

cogn

itive

impairm

entin

alert

andaw

akeho

spicepatients

who

didno

thave

acurren

tor

pastdiagno

sisof

acogn

itive

disorder

orcogn

itive

impairm

ents.

Hospice

Hospice

inpatients/

palliativecare

patients

approachingthe

endof

life

(progn

osisof

<6

mon

ths)

Written(n

=14)o

roral

inform

edco

nsen

t(n

=16)

30Key

findings:cogn

itive

impairm

entsarecommon

amon

ginpatient

hospice

patients(12/30

patientswere

diagno

sedwith

demen

tia—

DSM

V).Finding

shigh

light

the

unde

r-recogn

ition

ofcogn

itive

impairm

entin

thispo

pulatio

n.So

lution

s:(1)psycho

education

forfamily

mem

bersand

caregivers,(2)

psycho

pharmacolog

ical

treatm

entforthepatient,and

(3)moretim

elyfinal

prep

arations

tomaxim

ise

functio

n,adeq

uate

unde

rstand

ingandcoping

mechanism

s,complete

person

alaffairs,e.g.w

ills,

advanced

directives,leg

acy

workandsaying

‘goo

d-byes’.

Davieset

al.

[53],U

K,A3

2018

Afeasibility

clusterRC

T.Aim

‘can

ade

finitive

(ade

quately

powered

)stud

ybe

done

?’Hypothe

siswas

that

adeq

uate

clinicallyassisted

hydration

durin

gthelastfew

days

oflife

wou

ldmaintainrenalp

erfusion

andpreven

thype

ractive

delirium

(‘terminalagitatio

n’).

Cancercentres

andho

spices

Cancerpatientsat

theen

dof

lifewith

delirium

Inform

edco

nsen

tfro

mthe

patient

whe

repo

ssible,ifno

tpersona

lcon

sultee

assent

from

arelativeor

afrien

dor

nominated

consulteeassent

from

thesite

Stud

yGuardian

(aninde

pend

entsenior

clinician).P

rocess

consen

t—if

patient

lostcapacity

durin

gthe

stud

y,pe

rson

al/nom

inated

200/219(91%

)Key

findings:thestud

yrecruited91.3%

ofeligible

patientsby

usingmultip

leconsen

tprocesses.Only13/219

(6%)de

clined

totake

partand

nowith

draw

als.Datacollection

burden

from

thepatient

and

family

mem

berswas

minim

ised

.Informed

consen

twas

received

from

patients(16,

Evans et al. BMC Medicine (2020) 18:221 Page 34 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

consulteewas

requ

iredto

confirm

continuedinvolvem

ent

inthestud

y.

8%),advice

from

person

alconsultees

(161,81%

),and

nominated

consultees

(23,

12%).

Cha

lleng

e:Unb

alancedtrial

armsdu

eto

differences

ineligibleparticipantsand

‘com

petitive’recruitm

ent

strategy.

Solution

s:Multip

leconsen

tprocessesandminim

isedata

collectionbu

rden

.

Dem

entia

andge

riatriccare

Baskin

etal.

[76],U

SA,A

31998

RCTto

iden

tifychalleng

esto

inform

edconsen

tin

research

involvingsubjectswith

advanced

demen

tiato

check

purposeof

bid.

Inpatient

hospital

Advancedde

men

tiaPe

rson

alco

nsulteead

vice

n=75/146

(51%

)Key

findings:thisstud

yisthe

firstto

exam

inebarriersto

research

inpatientswith

advanced

demen

tia.49%

(n=

71)eligibleparticipantscould

notbe

enrolledin

astud

yon

palliativeapproaches

tocare.

68be

causeof

aninability

toen

gage

theproxyin

the

consen

tprocess(22be

causeof

theabsenceof

asuitable

proxy),and

4be

causeproxy

declined

consen

t.Thefinding

sindicate

barriersin

clinical

research

inen

d-stage

demen

tia,and

implications

for

med

icalde

cision

-makingin

this

vulnerablepo

pulatio

n.Cha

lleng

es:(1)

relative

declined

perm

ission

enrolm

ent

(n=4);(2)

proxyun

contactable

(n=41/63);nosuitableproxy

iden

tified(n

=22).

Sampson

etal.[40]*,U

K,B3

2018

A9-mon

thprospe

ctivecoho

rtstud

yaimingto

describ

e(1)

physicalandpsycho

logical

symptom

s,(2)he

alth

andsocial

care

serviceutilisatio

n,and(3)

care

aten

dof

lifein

peop

lewith

advanced

demen

tia.

14nu

rsing

homes

intheUK

orparticipant’s

ownho

mes

Peop

lewith

advanced

demen

tia(Fun

ctional

Assessm

entStaging

Scale6e

and

above),age

d65

orover

Person

alco

nsulteeadvice,

consulteeun

contactable

approach

professiona

lco

nsultee

85/159

(54%

)Key

findings:recruitm

ent

target

notreache

dwith

30%

ofpatients/carersde

clining.

Of

157eligiblecare

home

reside

nts,80

peop

lerecruited

bycarerconsultee(62,1

patient

died

),or

profession

alconsultee(18).32de

clined

,28

uncontactableconsultee,and

17died

.42pe

opleresiding

atho

me,6recruitedby

carer

Evans et al. BMC Medicine (2020) 18:221 Page 35 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

consultee.16

consultees

declined

,19un

contactable,and

onedied

.

Jone

set

al.

[39]*,UK,B3

2012

Protocol

fortheabovecoho

rtstud

y.Careho

mes

and

ownho

me

Dem

entia

(early

stage)

Person

alco

nsulteead

vice,

consulteeno

tcontactable

approach

professiona

lco

nsultee

Stud

yprotocol

(recruitm

ent

target

n=100)

Person

alco

nsultee,consultee

notcontactableapproach

professiona

lcon

sultee

,e.g.

staffno

tinvolved

instud

y.

Blacket

al.

[73],U

SA,B3

2006

Coh

ortstud

yto

describ

ehe

alth

prob

lemsandcomorbiditiesin

nursingho

mereside

nts,and

correlations

with

staffiden

tified

pain.

Nursing

homes

Advancedde

men

tiaPe

rson

alor

legally

appointed

representative

Capacity

assessmen

tinform

alHCP

n=126/289(44%

)Key

findings:ph

ysicianhigh

recogn

ition

symptom

sat

EoL,

butde

tectionde

creaseswith

increasing

cogn

itive

impairm

entindicate

sample

bias.

Cha

lleng

es:(1)

consultee

denied

enrolm

entpe

rmission

(18%

);(2)no

respon

seconsultee(5%);(3)patient

died

(23%

);(4)ph

ysiciande

nied

recruitm

entpe

rmission

(5%).

Goo

dman

etal.[74],UK,B3

2011

Coh

ortmixed

metho

dde

sign

toun

derstand

EoLC

need

sand

supp

ortforpe

oplewith

demen

tiain

care

homes.

Careho

mes

(reside

ntial,6

homes)

Advancedde

men

tiaCon

tinu

ousco

nsen

tadults

ableconsen

tin

mom

ent(n

=65);persona

lcon

sultee

advice

adultslackingcapacity

(n=68).

Cap

acityassessmen

tMCA

criteria

researcher

andcare

staff.

n=133/215(62%

)Key

findings:65

(74%

)peo

ple

with

demen

tiawho

had

capacity

toun

derstand

the

stud

yagreed

toparticipate,

andthoselackingcapacity

person

alconsultees

advised

participation(68/126,54%).

Solution

s:greatestim

pact

onrecruitm

entcare

homecultu

re,

e.g.

working

practices.

Cha

lleng

es:(1)

consultees

declined

perm

ission

participant

enrolm

ent(n

=9);(2)

consultee

uncontactable(n

=49);(3)

resource—5mon

thsfull-tim

eresearcher

recruit133.

Solution

s:(1)en

gage

men

tfamilies

andcare

staff;(2)

accessibility

stud

yinform

ation,

e.g.

A4summarysheet.

Hansonet

al.

[75],U

SA,A

32010

Cluster

RCTof

ade

cision

aidto

inform

andsupp

ortthechoice

betw

eentube

feed

ingand

assisted

oralfeed

ingin

demen

tia.Tode

scrib

erecruitm

entstrategies

innu

rsingho

mes

andethical

Nursing

homes

(n=24)

Advancedde

men

tiaPe

rson

alco

nsultee,e.g.

son/

daug

hter

(68%

).Cap

acity

assessmen

t:Cog

nitive

Perfo

rmance

ScaleandGlobal

Deterioratio

nScale.

n=256/425(60%

)(paired

reside

ntandpe

rson

alconsultee)

Key

findings:clinicaltrials

involvingadultslacking

capacity

requ

iremultip

lestrategies

toen

gage

consultees

andrecruitin

nursingho

mes.

Recruitm

entapproaches

ensuredadeq

uate

timeto

Evans et al. BMC Medicine (2020) 18:221 Page 36 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

approaches

tominim

iseharm

andmaxim

isebe

nefits.

addressconcerns

offamily

surrog

ates,and

compe

nsation

fortheirtim

ecommitm

entto

thestud

yinterviews.

Cha

lleng

es:(1)

care

home

cultu

rerecruitm

entrate

varied

bynu

rsingho

mes

(3to

36,30–

94%).

Solution

s:(1)stud

yde

sign

clearpo

tentialb

enefits

for

reside

nts;(2)en

gage

consultees,e.g.m

onetary

reim

bursem

entfortim

e;researcher

frequ

entcalling

tobo

okinterview

andflexibility—

interviewsconven

ient

timeand

locatio

n/by

teleph

one,

repe

ated

provisionaccessible

stud

yinform

ation;(3)en

gage

care

homestaff,e.g.

prior

contacts,tim

eandresource,

minim

albu

rden

forcare

staff.

Sampson

etal.[42]*,U

K,B3

2015

Coh

ortstud

yto

definethe

prevalen

ceof

pain

usingself-

ratedandob

servationalp

ain

scales

inpe

oplewith

demen

tiain

thege

neralh

ospital,andto

exam

inede

mog

raph

icand

clinicalfactorsassociated

with

pain

andto

explorea

hypo

thesised

association

betw

eenpain

andbe

haviou

ral

andpsychiatric

symptom

sof

demen

tia(BPSD).

Twolargeacute

gene

ralh

ospitals

inLond

on

Patientswith

demen

tiawho

were

admitted

toge

neral

med

icalwards

oftw

olargeacute

hospitals

Con

sent

process:patient

togive

writteninform

edconsen

tor

with

aninform

alcareror

‘professionalcon

sultee’.The

consen

tproced

urefollowed

theEnglandandWales

MCA

2005.N

oform

alcapacity

assessmen

t.Ape

rson

alconsulteewas

iden

tifiedto

give

agreem

enteither

verbally

over

theteleph

one,andpo

sted

anagreem

entform

tosign

and

return.Ifform

swereno

treturned

,participants’d

ata

werede

stroyed.

Ifape

rson

alconsulteecouldno

tbe

reache

din

48hof

screen

ing,

aprofession

alconsulteeadvice

(e.g.g

eriatricianno

tinvolved

inthestud

yor

patient

care).

230/292(79%

)Cha

lleng

e:62

exclud

edbe

causethey

didno

tfulfilthe

inclusioncriteria

orbe

cause

carerswho

gave

teleph

one

agreem

entdidno

treturn

sign

edform

s.So

lution

s:consen

tprocess

practicalandfeasiblein

the

acutesetting.

Scottet

al.

[41]*,UK,B3

2011

Protocol

fortheabovecoho

rtstud

y.Hospitalinp

atient

Dem

entia

and

unplanne

dho

spital

admission

Assen

tpersona

lcon

sultee

,if

noconsulteeiden

tifiedin

48h,

nominated

‘professiona

l’co

nsultee—

senior

mem

ber

clinicalteam

notinvolved

Stud

yprotocol

(recruitm

ent

target

250)

Key

finding:p

rotocol

develope

dMCAgu

idance,i.e.

assess

thepe

rson

using

structured

assessmen

tfor

capacity,seekassent

from

a

Evans et al. BMC Medicine (2020) 18:221 Page 37 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

stud

y.Cap

acityassessed

MCA

criteria

iden

tifyconsen

t/assent.

carerforpe

rson

lostcapacity,

anduseprofession

alconsultees

whe

ncarerno

tcontactable.

Warne

ret

al.

[72],U

K,A3

2007

RCTof

demen

tiatreatm

ent,

Ginkgobiloba,toassess

the

prop

ortio

nof

peop

lewith

capacity

toconsen

t,de

scrib

euseof

theMCA

2005

toassess

capacity

inaresearch

setting,

andiden

tifyvariables

associated

with

thepresen

ceof

capacity.

Com

mun

itysettings

Dem

entia

Inform

edco

nsen

tor

persona

lcon

sultee

advice

for

adultslackingcapacity.

Cap

acityassessed

MCA

criteria

andMMSE

.

n=42/176

(24%

)Key

findings:cogn

itive

testing

alon

eisinsufficien

tto

assess

thepresen

ceof

capacity.

Recommen

dresearchersrecord

outcom

ecapacity

assessmen

tandbasisforthefinding

;cogn

itive

measuresconsidered

poor

proxyforjudg

ing

capacity.Researchmay

beun

dertaken

with

incapacitous

participantswho

neverthe

less

appe

arto

agreeto

participation,bu

tison

lype

rmissibleifthereissome

potentiald

irect

orindirect

bene

fitto

thepatient

andthe

research

cann

otbe

carriedou

ton

patientswho

have

capacity.

Cha

lleng

es:(1)

high

cogn

itive

impairm

ent,76%

participants

assessed

lacked

capacity

for

inform

edconsen

t;(2)MMSE

didno

tpred

ictthepresen

ceof

capacity.

Ellis-Smith

etal.[78],UK,B2

2018

Prospe

ctive,mixed

metho

dsfeasibility

stud

yto

explorethe

mechanism

sof

actio

n,feasibility,accep

tability,and

implem

entatio

nrequ

iremen

tsof

ameasure,the

Integrated

Palliativecare

OutcomeScale

(IPOS-Dem

),used

inroutine

care

tosupp

ortcompreh

ensive

assessmen

tof

symptom

sand

concerns

ofcare

home

reside

ntswith

demen

tiaand

theirfamily

mem

bers.

Threereside

ntial

care

homes

registered

toprovidecare

for

peop

leaged

65andover

ina

Lond

onbo

roug

h,UK

Palliative/reside

ntial

care

homereside

nts

with

demen

tia

Theresearch

team

met

with

reside

ntsto

ascertain

willingn

essto

participateand

assess

men

talcapacity

toconsen

tforthem

selves.Tho

sereside

ntsthat

hadcapacity

gave

writteninform

edconsen

t.Adviceof

consultees

was

soug

htifALC

.The

care

home

thereforesent

aletter

onbe

halfof

theresearch

team

toaclosefrien

dor

family

mem

berto

invite

them

toadvise

onwhe

ther

thereside

ntshou

ldparticipatein

research

(personalcon

sultee).Two

lettersweresent.Ifno

respon

sewas

received

after1

weekof

thesecond

letter

beingsent,a

nominated

10family

mem

bers

approached

.22profession

als

approached

.47reside

nts

approached

.32baselineand30

finaltim

epo

intpatient

data.20

profession

alsinvolved

infocus

grou

psandserialinterview

s.Six

family

mem

bersinvolved

ina

focusgrou

pandinterviews

Key

findings:of

47participantsapproached

,one

patient

provided

inform

edconsen

t,45

wereassessed

asno

thaving

capacity

toconsen

t.Con

sulteeapproached

forn=

39.Personalcon

sultees

provided

advisedforn=15,

andprofession

alconsultees

for

n=24.O

nereside

ntde

clined

,four

NoK

sun

contactable.Not

recruitedon

consulteeadvice

(n=3),advised

against

participation(n

=2),

profession

alconsultee(external

tothecare

homeandstud

y)advisedNoK

involvem

ent,bu

tNoK

didno

trespon

d(n

=1).

Find

ings

show

edthat

possible

tointrod

uceameasure

into

Evans et al. BMC Medicine (2020) 18:221 Page 38 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

consulteewas

askedto

advise

onreside

ntparticipation.The

nominated

consulteewas

inde

pend

entfro

mtheresearch

stud

yandused

allavailable

inform

ation(includ

ingmeetin

gwith

thereside

nt,reviewing

case

notes,andspeaking

tocare

homestaff)in

orde

rto

give

advice

onreside

ntparticipation.

routinecare

ofreside

ntsand

change

care

processesto

improvereside

ntandfamily

outcom

es.

Solution

s:mutlipleproced

ures

ofapproachingtheNoK

,and

ifun

contactable,aprofession

alconsulteeisfeasiblein

care

homes.The

staffmem

bers

know

thereside

ntsandcould

inform

theresearchersabou

tpreferen

cesforresearch

participation(orno

t).

Mitche

llet

al.[77],USA

,B3

2006

Coh

ortstud

yto

exam

ineEoLC

inadvanced

demen

tiaandto

describ

eho

wresearch

challeng

esweremet

(CASC

ADE

stud

y).

Nursing

homes

(n=15)

Reside

ntswith

advanced

demen

tiaandtheirhe

althcare

proxies—

Cog

nitive

Perfo

rmance

Scale

score5/6

Profession

alco

nsultee

(health

care

staff)involving

consen

tforthem

selves

and

providingproxyconsen

tfor

reside

nts(datacollectionchart

review

,briefclinical

exam

ination,nu

rsinginterview).

Cap

acityassessmen

t:Cog

nitivePerfo

rmance

Score5

or6.

n=189/343(55%

)(patient/HCP

dyad)

Key

findings:involvingALC

inresearch

onpalliativecare,

includ

ingthedyingph

ase,

acceptableto

staffsurrog

ate

decision

-makersin

nursing

homes.Finding

ssugg

estthat

theem

otionald

istressof

families

does

notde

clinewhe

npatientswith

demen

tiaare

institu

tionalised

.Cha

lleng

es:clinicians’refusal

(n=154),e.g.too

burden

some

(n=30),lack

ofinterest(n

=53).

Solution

s:(1)nu

rsingho

me

cultu

re,e.g.track

record

cond

uctin

gresearch;(2)

engage

care

homestaff,e.g.

researcher

traine

dge

riatric

nurse,participationlow

burden

;(3)research

team

cred

ibility.

Davieset

al.

[84],U

K,B3

2010

Coh

ortstud

yto

exam

inehe

alth

trajectoriesandou

tcom

esof

peop

leaged

85+coho

rtand

associations

with

biolog

ical,

med

ical,and

socialfactors.

Usualplaceof

reside

nce,e.g.

atho

me,care

home

Older

peop

leaged

>85

years

Advanc

eco

nsen

t—consen

tdo

cumen

tparticipant

nominateapersona

lco

nsulteeshou

ldthey

lose

capacity.U

sedcontinuo

usconsen

tprocess,fluctuatin

gcapacity

awaitedre-gain

capacity.C

apacity

assessmen

t.MCAcriteria

research

er.

n=1042/1453(72%

)Key

findings:72%

recruitm

ent

rate

usingconsultees,sep

arate

consen

tprotocolsfor

participantsin

care

homes

and

flexibilityregardingthosewith

fluctuatin

gcapacity.C

ognitive

impairm

entiscommon

requ

iring

considerationof

the

relevant

ethicaland

legal

issues.

Evans et al. BMC Medicine (2020) 18:221 Page 39 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

Cha

lleng

es:(1)

consultee

declined

recruitm

ent(n

=5);(2)

consulteeun

contactable(n

=9).

Solution

s:(1)compreh

ensive

protocolsforconsen

tand

capacity

assessmen

tthat

anticipated

high

cogn

itive

impairm

entgu

ided

bylegislation;(2)training

researcherson

consen

tand

capacity

processes,sensitively

hand

lingcomplex

situations;(3)

timeandresource

toen

gage

family

andcare

homestaff;(4)

follow-upteleph

onecall/visit1

weekafterreceiptstud

yinform

ation;(5)de

sign

stud

yinform

ationforolde

rpe

rson

,e.g.

font

size,langu

age.

Mason

etal.

[83],U

K,A3

2006

RCTto

determ

ineeffectiven

ess

andcost-effectiven

essof

pressure

relieving

mattresses

onpressure

ulcers.

10research

centresinvolving

inpatient

units

Patientswith

pressure

ulcersaged

>55

yearswith

orwith

outcogn

itive

impairm

ent

Person

alco

nsulteead

vice

for

ALC

.Lack

capacity

n=87/2445(4%)

Capacity

n=1972/2395(82%

)Key

findings:ahigh

erprop

ortio

nof

relatives

declined

enrolm

ent,than

patients

them

selves.Proxy

consen

tallowed

onlyasm

allincreasein

trialrecruitm

ent.

Recruitm

entch

alleng

es:(1)

norelativeavailable(n

=2286);

(2)relatives

declined

perm

ission

toen

rol(n=72/

159,45%).

Botker

etal.

[85],D

enmark,

A3

2018

RCT.Aim

toexam

ineifthe

additio

nof

brainnatriuretic

peptidemeasuremen

tto

the

routinediagno

sticwork-up

bypreh

ospitalcriticalcare

team

physicians

improved

triage

inpatientswith

severe

dyspno

ea.

Preh

ospital

criticalcareun

itsat

Cen

tral

Den

markRegion

hospitals

Patientswith

severe

dyspno

eaDeferredco

nsen

t:includ

edpatientspriorto

inform

edconsen

t.Subseq

uent

oraland

writteninform

edconsen

twas

obtained

from

allp

articipants

orne

xtof

kinby

oneof

10stud

yinvestigators.

711/747(95%

)Key

findings:28

(4%)

participantswith

draw

nconsen

tbasedon

patient

orproxy

requ

est.7(0.9%)patientswere

exclud

edas

consen

tcouldno

tbe

obtained

.The

routine

additio

nof

preh

ospitalN

T-proB

NPmeasuremen

tdidno

tim

provethetriage

ofpatients

with

dyspno

eaof

cardiaccause

directlyto

thede

partmen

tof

cardiology

anddidno

tsign

ificantlyim

provetreatm

ent

orpatient

outcom

es.

Evans et al. BMC Medicine (2020) 18:221 Page 40 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

Galeo

ttietal.

[79],Italy,A

32012

RCT.TheADCarestud

yaimed

toevaluate

thelong

-term

safety

andefficacyprofilesof

threeatypicalantip

sychotic

drug

sandon

econven

tional

antip

sychoticdrug

intreatin

gpsycho

sis,aggression

,and

agitatio

nin

outpatientswith

AD.

Theaim

ofthispape

risto

repo

rttheADCarestud

yexpe

rienceandto

analysein

depththepo

ssiblereason

sfor

thelow

accrual.

N/A

Alzhe

imer’sdisease

patients

Inform

edco

nsen

t83/~

800eligiblepatientsfro

m19

clinicalcentres

Key

findings:~800eligible

patientsin

1year

were

iden

tifiedin

theclinicalcentres.

Alth

ough

initiallywillingto

participate,mostof

these

patients/family

mem

bers

declined

becausealegalcou

rtno

minationwas

requ

iredto

participatein

thetrial.Cou

rtno

minationwas

perceivedby

patientsandfamily

mem

bersas

intrusiveandpo

tentially

creatin

gconflictsin

families.

Cha

lleng

es:o

nly9clinical

centresparticipatingin

the

stud

yen

rolledpatientswith

theinvolvem

entof

alegal

proxy,even

with

initiatives

todissem

inateinform

ationon

legalage

ncyforresearch

participants.The

mainob

stacles

werecaregivers’reluctanceto

design

atealegalp

roxy

and

courtde

lays.

Intensivecare

Day

etal.

[109],UK,B3

2015

PilotRC

Tsing

lecentre.A

imto

repo

rtchalleng

esto

unde

rtake

theRC

Tandfeasibility

ofde

liveringandevaluatin

ga

complex

interven

tionin

acriticalcareun

it.

Twomixed

criticalcareun

itsof

alargeinne

rLond

onho

spital

Mixed

med

ical,

surgical,and

trauma

patient

popu

latio

nrequ

iring

either

level3

(intensive)

orlevel2

(high

depe

nden

cy)care

Researchersno

tinvolved

indata

collectionwere

respon

siblefortheconsen

tprocess.In

thisstud

y,patients

werewillingto

consen

tve

rbally

but

deferredto

their

relative

toprovidewritten

consen

t.IfALC

,the

irrelative

provided

advice,the

researchersthen

ensuredthat

whe

npatientsregained

capacity

andthey

requ

estedan

inform

edconsen

t.

158/221(72%

)Key

findings:rushingthe

decision

-makingprocesswas

aconcern.Somepatients

approached

feltun

ableto

say

no,b

utkept

saying

comeback

later.Thismay

indicate

concern

that

theircare

may

beaffected

ifde

cline.View

inginform

edconsen

tas

aprocessand

revisitin

gitthroug

hout

thetrial

perio

dto

ensure

participants

canrecallprovidingconsen

t.Cha

lleng

es:M

ostof

those

approached

felttootired

orlacked

theconcen

trationto

read

thestud

yinform

ation.

Patientswith

noapprop

riate

consulteewereexclud

ed.

Solution

s:(1)avoidconsen

tbias,riskof

coercion

andallow

moretim

eforde

cision

-making.

(2)Participantinform

ation

Evans et al. BMC Medicine (2020) 18:221 Page 41 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

sheetclearandconcise.(3)To

review

capacity

inpatientswith

mem

oryloss.

Benchet

al.

[110],UK,A3

2015

PilotRC

T.Aim

(1)providingan

initialevaluatio

nof

auser-

centredcriticalcaredischarge

inform

ationpack

(UCC

DIP),(2)

inform

decision

sabou

tits

furthe

rde

velopm

entand

evaluatio

n,and(3)estim

atethe

samplesize

forafulltrial.

TwoICUswith

inasing

leteaching

hospitalin

centralLon

don

Mixed

med

ical,

surgical,and

trauma

patient

popu

latio

nrequ

iring

either

level3

(intensive)

orlevel2

(high

depe

nden

cy)care

Inform

edco

nsen

tfro

mthe

patient

was

then

obtained

prior

todata

collectionon

theward.

ForALC

,informed

written

consen

tat

thepoint

ofICU

disch

arge(deferred),

persona

lcon

sultee

declaration

s,usually

from

the

patient’sne

xtof

kin,were

soug

ht.The

relatives

ofall

recruitedpatientsweregiven

stud

yinform

ationwhe

nthey

visitedtheICUor

teleph

oned

andinvitedto

participate.

Writtenconsen

twas

obtained

from

relatives

durin

gtheirne

xtho

spitalvisit.

158/221(36de

clined

toparticipate)

(72%

)Key

findings/ch

alleng

es:The

apriorien

rolm

entgo

alwas

not

reache

d,andattrition

was

high

leadingto

insufficien

tstatistical

power

tode

term

ineou

tcom

ebe

nefit.101

(64%

)patient

participantsprovided

prim

ary

outcom

edata

attim

epo

int1.

Atotalo

f48(60%

)patients’

relatives

provided

atleaston

esetof

outcom

edata.Twen

ty-

seven(17%

)patientsand32

(40%

)relatives

werelostto

any

follow-up.

Higginson

etal.[111],U

K,B3

2016

Prospe

ctiveethn

ograph

icstud

y.Aim

toexplorethe

nature

andpatterns

ofde

cision

-makingprocesses

durin

gICUadmission

,including

sourcesof

conflictand

resolutio

n.

TwoICUsin

aninne

rcity

hospitalserving

anethn

icallyand

sociallydiverse

popu

latio

n

ICUpatientswhe

reclinicians

had

potentialend

oflife

concerns,

discussion

sor

ahigh

riskof

dying

durin

gthecurren

tadmission

Whe

repo

ssible,informed

consen

tfro

mpatients.For

patientswho

lacked

capacity,

followingtheMen

talC

apacity

Act

2005,firstly

capacity

was

assumed

tobe

presen

t,un

less

proven

tobe

absent.C

linicians

assessed

capacity.Iflacked

capacity,app

roachforadvice

apersona

lcon

sultee

(e.g.fam

ilymem

ber),

ifno

tiden

tifieda

nominated

consultee(e.g.

clinician)

16patientsand19

relatives

Solution

s:theresearchers

looked

forinstanceswhe

recapacity

was

presen

tand

discussedwith

theclinicians.

Theresearch

team

ensuredthat

theno

minated

consulteewas

notpartof

theresearch

team

.Relatives

also

provided

inform

edconsen

t/assent

for

interviewsandob

servations.

Rouzéet

al.

[112],France,

A3

2017

RCTto

testthehypo

thesisthat

theuseof

analgo

rithm

based

onfung

albiom

arkerswou

ldincrease

thepe

rcen

tage

ofearly

discon

tinuatio

nof

empiricalantifun

galtreatmen

tam

ongcritically

illpatients.

Mixed

50-bed

ICUde

partmen

tof

theUniversity

Hospitalo

fLille,

France

ICUpatientswith

Cand

idainfection

Aninform

edwritten

consen

twas

obtained

before

rand

omisationfro

mthe

patientsor

theirproxies.

110/387(28%

)Key

findings:8/510patients

who

wereassessed

for

eligibility

refused.

One

patient

with

drew

consen

tafter

receivingthebiom

arker

strategy.The

useof

abiom

arker-basedstrategy

increasedthepe

rcen

tage

ofearly

discon

tinuatio

nof

empiricalantifun

galtreatmen

tam

ongcritically

illpatientswith

suspectedinvasive

Cand

ida

infection.

Evans et al. BMC Medicine (2020) 18:221 Page 42 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

Men

talh

ealth

Hoet

al.[97],

Australia,B3

2018

Metho

dologicalp

aper

aiming

tode

scrib

ean

inform

edconsen

tprocessused

whe

nrecruitin

gpe

rson

swith

intellectuald

isability

forastud

ywhich

iscurren

tlyinvestigating

falls

amon

gpe

oplewith

intellectuald

isability,and

toreflect

onthemetho

dsof

inform

edconsen

tused

.

Com

mun

ity—

participant’s

home,sm

all

grou

pho

mes

with

upto

two

tofour

co-

inhabitantswith

paid

supp

ort

Peop

lewith

intellectuald

isability

Inform

edco

nsen

tan

dproxy

advice

from

theparticipants

them

selves

whe

repo

ssible.The

processwas

design

edto

provideacollectivepe

rspe

ctive

ofthecapacity

ofthepe

rson

with

intellectuald

isability

toconsen

t.Theresearcher

uses

repe

ated

observations

toestablishifthepe

rson

has

capacity

toprovideconsen

t,andconsultwith

thecaregiver.

Use

ofade

cision

alqu

estio

nnaire

involvingscore

of>3ou

tof

sixsugg

est

capacity

toprovideinform

edconsen

t.Th

ene

xtof

kin,

family

mem

ber,o

rcaregive

risaske

dto

bepresent

during

thisprocess

toprovidea

supportive

,com

fortab

leen

vironm

entan

dto

provide

oversight

tothediscu

ssion.

[84]

40/68(59%

)Key

findings:on

ly3ou

tof

40participantswereableto

provideinform

edconsen

t.22

participantswereableto

have

adiscussion

abou

ttheir

involvem

entin

thestud

ywith

thesupp

ortof

theircaregiver.

15participantswereun

ableto

engage

intheprocess.The

decision

alqu

estio

nnaire

gave

anaccurate

represen

tatio

nof

participants’d

ecision-making

capacity.

Cha

lleng

es:m

orethan

40%

caregiversde

clined

taking

part

inthestud

yon

behalfof

the

person

.Thisconsen

tprocess

was

time-consum

ing.

Solution

s:Training

and

buildingrelatio

nships

with

patientsandtheircaregivers.

Ramerman

etal.[96],

Holland

,B3

2018

Expe

rimen

talstudy

toexam

ine

challeng

ingbe

haviou

r,ph

ysical

symptom

s,andqu

ality

oflife

associated

with

antip

sychotic

drug

.

Hospital

intellectual

disabilitymen

tal

healthcare

clinics

Adu

ltswith

intellectual

disabilities

Writteninform

edconsen

twas

obtained

from

participants

and/or

theirlega

lrepresentatives.

159(nodata

onnu

mbe

rof

peop

leapproached

)Key

findings:meanagewas

46.2(SD17.7).Health

quality

oflife(HQoL)was

negatively

associated

with

both

symptom

sof

challeng

ingbe

haviou

rand

physicalsymptom

sassociated

with

antip

sychoticdrug

s.No

data

ondifferentiatio

nproxies

andpatientsin

inform

edconsen

tprocess.

Delirium

Marcanton

ioet

al.[101],U

SA,

A3

2010

RCTto

determ

inewhe

ther

ade

lirium

abatem

ent

prog

ramme(DAP)

canshorten

duratio

nof

delirium

inne

wadmission

sto

postacutecare

(PAC).

Hospital

Patientswith

delirium

olde

rthan

65(m

eanage=84);

MMSE

andCAM

used

toassess

capacity

Inthosewho

hadde

lirium,

family

caregivers,actingas

proxies,p

rovide

dinform

edco

nsen

tfortrialp

articipation.

457/667(69%

)Key

findings:thisstud

ywas

thelargestcoho

rtof

patients

with

delirium

ever

enrolledin

aresearch

stud

y.Facilities

received

asm

allincen

tive

basedon

theirpe

rform

ance.

Clusteringeffect

widen

edthe

confiden

ceintervals.Po

ssibleto

ascertainadhe

renceand

outcom

eson

lyin

trial

participantswho

seproxies

Evans et al. BMC Medicine (2020) 18:221 Page 43 of 55

Table

4Stud

iesapplying

consen

tprocessesacross

thecapacity

trajectory

inserio

usillne

ss(n

=30

stud

iesrepo

rted

in32

publications)(Con

tinued)

Autho

rs,

coun

try,EPOC

Year

Stud

yde

sign

andaim

Setting

Samplede

scrip

tion

Con

sent

processforadultswith

declining,

impaired,

orlacking

capacity

No.patients/eligible

(%recruited)

Keyfinding

s,challeng

es,and

solutio

ns

provided

inform

edconsen

tto

allow

med

icalrecordsto

bereview

ed.

Cha

lleng

es:O

f667patients

who

wereeligible,138

relatives

refused(21%

),56

relatives

uncontactable(8%).

Coleet

al.

[102],Canada,

A3

2002

RCTto

determ

ineifsystem

atic

detectionandmultid

isciplinary

care

ofde

lirium

inolde

rpatientscouldredu

cetim

eto

improvem

entin

cogn

itive

status.

Hospital

Older

patients65

yearsoldor

more

admitted

toa

gene

ralh

ospital

med

icalservice

Inform

edco

nsen

twas

obtained

from

thepatient

orsubstitutedecision-maker.

Con

fusion

assessmen

tmetho

dwas

used

toassess

capacity.

227/299

(76%

)Key

findings:72

were

exclud

edas

they

couldno

tprovideconsen

t.Stud

ywas

not

powered

tode

tect

any

statisticallysign

ificant

differences.

Emerge

ncymed

icine

Offerm

anet

al.[108],U

SA,

B3

2013

Observatio

nalp

rospective

multicen

trestud

yto

describ

etherate

ofsuccessful

consen

tusingan

altered(deferred

teleph

one)

consen

tprocessin

emerge

ncyde

partmen

t(ED)

patients.

Emerge

ncy

departmen

tAdu

ltswho

had

attend

edED

with

blun

the

adtrauma

Inform

edco

nsen

twas

obtained

from

thepatient

orsubstitutedecision-maker.

506

Key

findings:follow-up

teleph

onecontactwas

successfullyaccomplishe

din

501of

the506subjects(99.0%

;95%

CI=

97.7to

99.7%).

Con

sent

forstud

yinclusionand

cond

uctof

theteleph

one

survey

was

obtained

in500of

501subjectsat

timeof

the

follow-upcall(99.8%

;95%

CI=

98.9to

100.0%

).Surrog

ate

consen

twas

obtained

in199of

the501subjects(39.7%

;95%

CI

=35.4to

44.2%)

Evans et al. BMC Medicine (2020) 18:221 Page 44 of 55

lacking capacity had an identifiable personal consulteeable to advise, 8.1% of eligible patients (n = 109 /1359)were not approached as no personal consultee was avail-able. In contrast, in an RCT [83] on reducing pressureulcers, only 3.6% (n = 87/2445) of eligible patients wholacked capacity were recruited compared to 82.3% (n =1972/2395) of patients able to giveinformed consent.The main reason for non-recruitment was no availablepersonal consultee (93.5%, 2286/2445), and when avail-able, 45.3% declined to give advice (n = 72/159).Three studies used both a personal and a professional

consultee to provide advice with an initial approach to apersonal consultee, and if unavailable an approach to aprofessional consultee [51–53]. Abernethy and colleagues[51] achieved a high recruitment rate (n = 461, 75.9%)using this joint approach in a low-risk non-invasive clusterRCT on palliative care in community settings. A MMSEscore of ≤ 24 indicated the need for consultee advice. Onlyseven patients were ineligible due to consultee unavailabil-ity. The recruitment process used a resource-intensive de-fined recruitment plan that sought to facilitaterecruitment by minimising patient and clinician burden.Similarly, a feasibility cluster RCT of clinically assisted hy-dration in cancer patients in the last days of life used mul-tiple processes of informed consent for patients withcapacity (n = 16, 8%), and advice from a personal (n = 161,80.5%) or professional consultee (n = 23, 11.5%) for adultslacking capacity, and process consent throughout thestudy [53]. The multiple consent processes enabled a91.3% recruitment rate (n = 200/219) of eligible patientsfrom four cancer centres and eight hospices. Only 5.9%(n = 13/219) of the eligible participants declined to partici-pate, and none withdrew. Using clinical observations fordata collection minimised burden for patients and familymembers. However, a competitive recruitment strategycaused variation in recruitment rate between clusters andimbalance in the trial arms (n = 73 treatment arm and n =127 control arm). Conversely, an RCT of antibody re-sponse to influenza vaccination for older people in carehomes incorporated multiple processes of consent [52].But care home staff were hesitant to act as a professionalconsultee in the absence of a personal consultee. A totalof 304 (54.5%) eligible individuals were excluded for thisreason, and recruitment of only 8.9% of eligible residentswho required a professional consultee (n = 40/448). How-ever, the study acceptable was mixed with 75.2% of per-sonal consultee advising enrolment (n = 82/109), but only37.7% of residents with capacity consented to recruitment(n = 155/411). Greater clarity on the responsibility of theprofessional consultee was required to enable care staff toact in this role, particularly in a trial involving invasiveprocedures.Seeking advice from a consultee on enrolment was

prominent in studies involving patients with dementia or

delirium. Depending on the context of the study, recruit-ment rates ranged from 23.9% [72] to 78.8% [42]. Moststudies in dementia took place in nursing homes/resi-dential care homes [40, 42, 72–75], with only two RCTsconducted in hospital both on delirium [101, 102]. Com-mon challenges to participation were relatives advisingagainst patient enrolment [40, 73, 101], unavailability ofa personal consultee [74, 76, 101], and care home staffacting as professional consultees declining to give advicefor 44.9% of the eligible participants [77]. To overcomepotential hesitancy from consultees required active en-gagement with consultees, keeping in touch and beingflexible, highlighting potential benefits, and loweringburden for the consultee [75]. To minimise exclusion ofeligible patients due to unavailability of personal consul-tees, studies used processes of a professional consulteesand a clear recruitment strategy detailing the procedures[42, 78]. A cohort study involving people with dementiafrom six care homes achieved 62.1% patient recruitmentrate (n = 133/215) using a process of informed consentwhen possible (n = 65/89, 73.8%) and personal consultees(n = 68/126, 54%) [74]. The study reported the import-ance of engaging with patients and families by using ashort summary sheet to enable understanding and par-ticipation. However, 49 eligible patients had no availablepersonal consultee and were excluded [74].Two studies [54, 84] in palliative care employed advance

consent, followed by process consent (or personal con-sultee) processes. In a feasibility RCT of two medicationsfor the treatment of death rattle, 54.2% (58/107) of eligiblepatients with capacity provided advance consent. However,due to the complexities of the patient population and un-certain prognosis, only 25.9% (15/58) of patients were ran-domised [54]. The consent process was resource intensiveto recruit the target sample size and emotionally drainingfor patients as it required conversations about an eventwhich may not occur. Most patients accepted the advanceconsent process, with only 16 of 58 declining. A prospectivecohort study examining health trajectories and outcomes inpatients over 85 years old in usual place of residenceachieved a 72.0% (1042/1453) recruitment rate [84]. Indi-viduals who provided advance consent were also asked tonominate a personal consultee. Throughout the study, cap-acity of the recruited participants was assessed, and thenominated personal consultee was contacted if needed.Only five consultees declined, and nine were unavailable.The recruitment process was facilitated in multiple ways in-cluding using different consent and capacity assessmentprotocols for respective settings, tailoring the study infor-mation to the target population, flexibility towards thosewith fluctuating capacity, identifying personal consultees,training researchers, sensitively handling complex situa-tions, and allocating time and resource to engage and keepin touch with family and care home staff.

Evans et al. BMC Medicine (2020) 18:221 Page 45 of 55

Four studies that took place in ICUs used informedconsent and consultee advice, and deferred consent pro-cesses [109–112]. Pilot RCTs of complex interventions(non-invasive) used deferred consent processes [109,110]. Patients agreed verbally to study enrolment, andtheir personal consultees gave written advice. Once pa-tients recovered, they were approached for an informedconsent [109]. The study achieved 71.5% recruitmentrate. However, the possibility of coercion was raised withsome patients indicating they felt unable to decline ver-bal enrolment, and consent bias with the exclusion ofpatients with no personal consultee. In another pilotRCT, the next of kin of 83.6% of patients who lackedcapacity in the moment of data collection provided de-ferred advice at the point of discharge. Although thisstudy had a 71.5% recruitment rate, 16.3% declined toparticipate. A further RCT involving both adults withcapacity to provide informed consent and personalconsultee advice for incapacitated patients showed only8/510 eligible individuals declined participation [112].Similarly, an RCT involving a blood test for patientswith severe dyspnoea used deferred consent [85] andshowed high acceptability with only 3.8% (28/747) with-drawing consent once regained capacity or on consulteeadvice. Only 0.9% (7/747) were excluded as unable toobtain deferred consent or consultee advice. An ethno-graphic observational study in ICU used informedconsent and personal or professional consultee advicesuccessfully to recruit 16 patients and 19 relatives witha clear consent protocol aligned to minimal risk ofparticipation [111].Enhanced consent processes were limited to studies

involving adults with intellectual disabilities [96, 97] andin palliative care [55]. A study investigating falls inpeople with intellectual disability aimed to maximise anindividual’s ability to provide informed consent forthemselves achieved a recruitment rate of 58.8% (n = 40/68). The study enhanced consent processes by involvingfamily members and healthcare professionals in theprocess, using a questionnaire to assess capacity, andconducting multiple observations [97]. A personal con-sultee was sought for adults lacking capacity. However,over 40% of relatives advised non-participation in thestudy on behalf of the person as they considered theconsent process too time-consuming and burdensome.A cross-sectional study on use of antipsychotic medica-tion for people with intellectual disabilities recruited 159eligible participants using informed consent andprofessional consultee comprising legal representative[96]. However, no denominator was stated. Finally, apilot study of patients admitted to a hospice exploringlevels of cognitive impairment sought to minimise theburden of informed by allowing written or oral informedconsent [55].

Public attitudes on enrolling adults lacking capacity inresearchStudies (n = 22) on attitudes towards enrolling adults lack-ing capacity in research consistently reported the accept-ability of involving consultees to enable recruitment.However, the level of acceptability varied associated withthe nature and purpose of the study (see Additional file 3:Table S8 reporting the studies’ key findings, and suppl. 9 –additional reporting [130–133]). Most studies (59.1%) wereconducted with public members (including researchers,healthy populations, IRBs) and used observational designs(e.g. survey) to explore attitudes about research with critic-ally ill patients (e.g. conducted in intensive or emergencycare settings), or focused on Alzheimer’s disease, stroke,cancer, intellectual disabilities, overall incapacitated indi-viduals, or proxy decision-making for study enrolment.Overall, stakeholders considered it acceptable to use substi-tute decision-making to enrol adults lacking capacity in re-search, especially for low-risk studies. Acceptability focusedon prioritising a person’s wishes and the potential benefitsfor the person, rather than the burden of acting as a proxydecision-maker.

Transparent expert consultationThe systematic review findings identified three criticalareas debated in the TEC including:

1) Time and design of the consent process: ‘How canwe enhance the timeliness of the consent processesfor adults with fluctuating or deteriorating capacityin research on palliative and EoLC?’

2) Enhancing consultee and supportive decision-making: ‘How can we enhance proxy decision-makers’ role in the process of consent for adultswho lack capacity?’

3) Ethics, resources, and expertise: ‘What are the keyconsiderations in planning the consent process instudies involving adults lacking capacity andcommunicating this to a research ethicscommittee?’

The TEC stakeholder workshop involved 39 partici-pants of the 83 invitees (47.0%). The participants repre-sented service users/lay voluntary sector representatives(n = 13, 33.3%), researchers (n = 15, 38.5% including ethi-cists), clinical academics (n = 9, 23.1%), and clinicians(n = 2 5.1%). Voluntary sector representatives includedlay members who were carers and/or patients living witha progressive condition recruited from charities forpeople with dementia (Age UK and Alzheimer’s Society)and cancer/palliative care (Brainstrust, Marie Curie,Independent Cancer Patients’ Voice, National CancerResearch Institute, and the then National Council forPalliative Care). The group generated 184

Evans et al. BMC Medicine (2020) 18:221 Page 46 of 55

recommendations on the three areas (area 1, n = 60 rec-ommendations; area 2, n = 72 recommendations; andarea 3, n = 53 recommendations). Following data ana-lysis, 29 recommendations were presented in the onlineDelphi Survey round 1 (see Additional file 4: Table S10 -Delphi participant characteristics). The recommenda-tions pertained to timeliness and design of the consentprocess (area 1, n = 7 recommendations), enhancing con-sultee and supportive decision-making (area 2, n = 10recommendations), and ethics resources and expertise(area 3, n = 12 recommendations) (see Additional file 4:Table S11 – Delphi round 1 recommendations and Fig.S12 – round 1 box and whisker plots). The Delphi sur-vey round 1 involved the workshop participants (n = 39)and individuals unable to attend the workshop (n = 4).51.8% participated (n = 43). Round 2 involved the round1 respondents only (83.7% response rate, n = 36). Find-ings from round 1 analysing 454 free-text comments in-formed revisions to 11 recommendations to reduceambiguity. This mainly concerned the use of the term‘consultee’ in relation to legislation. In round 2, consen-sus was apparent for 24 recommendations with strict/broad agreement and five considered equivocal (seeAdditional file 4: Table S13 - Delphi round 2 recommen-dations and Fig. S14 – round 2 box and whisker plots).No recommendations were not indicated. The top rec-ommendations by respective areas were as follows:

� Area 1. ‘Information about a research study iscomprehensible, short and written in accessiblelanguage’ (R2, median 9, IQR 9–9) and ‘Althoughthe legal significance of advance consent will varydepending on the relevant legal framework, it isgood practice for researchers to seek an advanceconsent while individuals have capacity to consentfor themselves, for example, shortly after a diagnosisof a progressive illness’ (R10, median 8, IQR 7–9)

� Area 2. ‘The consent process is tailored to theindividual’s needs, capabilities and values withresearchers observing for non-verbal and verbal cuesthat may indicate an individual may wish to with-draw’ (R9, median 9, IQR9–9) and ‘Advance careplanning to include discussing and recording in aStatement of Wishes document an individual’s nom-inated or personal consultee whose opinion on par-ticipating in a research study is sought if the patientloses capacity’ (R21, median 8, IQR 7–8)

� Area 3. ‘Researchers to demonstrate to Research EthicsCommittees a clear process of consent for potentialparticipants with compromised capacity that detailshow the researchers will proceed to tailor the consentprocess to maximise individuals’ ability to consent forthemselves and when and how they will seek anopinion from a consultee’ (R29, median 9, IQR9–9)

and ‘Health and social care practitioners to recogniseresearch as a core clinical activity in a similar way asteaching and training’ (R23, median 9, IQR 9–9)

Three main areas were considered equivocal and weredebated in the expert ‘think-tank’, and solutions pro-posed by consensus (see Additional file 4: Box S15 –main areas and top solutions). Nineteen experts attendedthe think-tank, representing researchers/clinical aca-demics (n = 12), clinicians (n = 2), lay voluntary sectorrepresentatives (n = 3), and PPI member (n = 1). Thethree equivocal areas comprised the following:

� Area 1. Involving and supporting consultees in thedecision-making process with uncertainty on how bestto support consultees to engage in the research process,with four recommendations indicating equivocal broadagreements (recommendations 6, 8, 15, 22, 27)

� Area 2. Practitioner training and education withagreement on the requirement to increase training andsupport to researchers and clinicians, but areas ofuncertainty concerned for example how to provide anddisseminate training and guidance, and how to fund (R5)

� Area 3. Legislative frameworks and theincorporation into research practice

A prominent equivocal area surrounded the Mental Cap-acity Act 2005 [29] legislation in England and Wales. Thekey solutions proposed concerned amendments in the sup-porting guidance for the Act to enable greater flexibility inthe role of the ‘professional’ consultee in, for example, howthey are identified. The Expert Panel concurred that theMCA process of review of an advance consent being upheld(or not) by a consultee if a participant lost capacity to con-sent should be applied to all clinical trials and not limitedto trials of non-pharmacological interventions. This wouldinclude for example pharmaceutical trials where advanceconsent is upheld without requirement for consultee reviewif a person loses capacity in European legislation [134].

DiscussionThis synthesis of evidence from systematic review andTEC identified challenges and solutions to including indi-viduals across the capacity spectrum in research on EoLC.Our findings produce the MORECare_Capacity statementdetailing 20 best practice solutions and implementationrequirements to maximise study participation across thecapacity spectrum (Table 5 and Fig. 3). The statementprovides much needed guidance to maximise opportun-ities for adults across the capacity spectrum to participatein research. It is relevant for researchers, members of re-search ethical committees, individuals overseeing researchgovernance, clinicians, the public, service users, and vol-untary sector representatives. The statement details

Evans et al. BMC Medicine (2020) 18:221 Page 47 of 55

solutions for study designs at the participant level to maxi-mise individual autonomy, enhance the contribution ofproxy decision-makers, and ensure time and resources toenable participation and recruitment strategies that antici-pate and plan for varying and changeable levels of cap-acity. At the structural level, key activities are engaging

research ethical committees in how to include adults lack-ing capacity in research by auditing ethical approvals toreview decision-making and inform research practice, andpromoting the importance of research in health and socialcare to improve clinical care and raise awareness onrecruiting adults from across the capacity spectrum to

Table 5 MORECare_Capacity statement solutions on recruiting adults with impaired mental capacity at the end of life in research

Solutions

Ethics 1. Researchers should design all aspects of the study in the context of potential risk, burden, and benefit ofstudy participation.

2. Institutional review boards and research ethical committees should have transparent decision-making pro-cesses to ensure consistency on ethical approvals for studies from various health specialties involving adultsacross the mental capacity trajectory.

3. The individuals (e.g. relative, friend, formal carer, nursing home staff, healthcare staff) who can act as aconsultee to advise on whether the individual would have wanted to participate in the research study hadthey had capacity should depend on the nature of the study, rather than legal restrictions.

4. Researchers should be able to demonstrate a clear process of consent for potential participants withcompromised capacity that details how the researchers will tailor the consent process to maximise anindividual’s ability to consent for themselves and when and how they will seek an opinion from a consultee.

Maximising individual autonomy 5. Clinicians should engage in research participation conversations with patients at the early stages of illness,discussing varying levels of risk, burden, and benefit, and document the person’s preferences and wishes in forexample an advance directive.

6. Individuals who are likely to lose capacity should be asked to designate a consultee whose opinion ontheir participation in a research study will be sought if the individual loses capacity.

7. For individuals who have capacity to consent in the moment, but overtime may not remember thediscussion, process consent should be adopted whereby researchers re-confirm the individual’s wish to partici-pate at each data collection time point.

8. Researchers should check for non-verbal (e.g. agitation) and verbal cues (e.g. ‘I’m unsure why you are ask-ing me’) that may indicate a wish to withdraw during the study.

Involving consultees 9. Personal consultees should be present in research participation conversations with patients at the earlystages of illness.

10. Where possible, to improve social support, personal consultees/family members should be engaged inthe (enhanced) informed consent process for adults with impaired capacity.

11. The nature and extent of the responsibility of a consultee acting on behalf of a patient should beclarified.

12. Establish a national body to provide support and information to family members/informal carers acting aspersonal consultees.

Tailoring recruitment process toneed

13. Researchers should incorporate in the study design research participation information and (whereapplicable) data collection tools in multiple formats (e.g. verbal, written, electronic).

14. Studies where potential participants are expected to be adults across the mental capacity trajectoryshould incorporate multiple consent processes (e.g. personal and professional consultees, informed andprocess consent).

Time 15. When possible, potential participants should be allowed time for further discussions regarding theirresearch participation decision.

16. The study design should allow time proportionate to the risk, burden, and benefit of participation forconsent or consultee decision (before or after research participation if consent is deferred).

17. Researchers should build in the time required to engage with and train clinical staff who will be involved,and those who might act as professional consultees in the study.

Enhancing the research culture andinfrastructure

18. Health and social care practitioners should recognise research as a core clinical activity in a similar way asteaching and training.

19. Clinicians should be supported and provided with training to ensure they are confident in their skills todiscuss research studies with patients (and/or family members) during routine clinical contact.

20. Health and social care practitioners should support adults across the mental capacity trajectory at allstages of the research study by considering the person’s best interests and individual wishes and preferencesto uphold individual autonomy and minimise the risk of harm, and enable family members to act as apersonal consultee by ensuring sufficient information and understanding about the role.

Evans et al. BMC Medicine (2020) 18:221 Page 48 of 55

enhance the applicability of research findings for serviceusers. The statement is intended as a framework for ‘best’research practice. A statement cannot always anticipatechange in an individual’s circumstances or wishes. On-going review is always required by the consultee and re-searchers of an individual’s best interests.Our findings demonstrate that despite ethical chal-

lenges and legislative requirements, recruiting adultswith impaired capacity in research is possible with care-ful consideration of the context, study design, and theresources required to enable participation. The main so-lutions for maximising research with incapacitated adultscentred around early involvement of the person whilethey had capacity to indicate their preferences for re-search participation, and nomination and engagementwith consultee(s) in these discussions. It is important torecognise that acting as either a personal (e.g. familymember) or nominated (e.g. clinician) consultee can beburdensome, with burden increasing in studies whichcarry risks or little benefit the patient directly. A priorityfor consultees is to have sufficient time and informationabout their role and to have an awareness of the person’swishes through earlier discussions and documenting aperson’s wishes about research participation in, for ex-ample, an advance directive [21] or study specific in anadvance consent. Figure 3 summarises the different pro-cesses across the capacity spectrum and sets out key

factors to anticipate and plan for in a research study. In-corporation of our solutions in research studies on EoLCcould increase the potential to recruit a representativesample of those intending to benefit from palliative careservices and treatments. Involving patient groups withimpaired capacity would increase generalisability of re-search findings by not limiting the evidence base tothose with capacity and less advanced disease, and sus-tainability of the tested interventions in clinical care.Our solutions are informed by methodological and ori-ginal research studies, and expert consensus. Import-antly, surveys of the public and stakeholders consistentlyendorse the involvement of adults with impaired cap-acity in research and specifically in palliative care [135].Ethically, research studies must endeavour to ask ques-

tions important for care and treatment, be designed toenable participation and minimise potential for harm,and demonstrate to research ethical committees a clearrationale for the study design and processes of consentfor adults across the capacity spectrum (R29). In turn,ethical research committees must ensure transparencyand consistency on ethical approvals for adults lackingcapacity (R28). Working in this way seeks to uphold in-dividual rights of, for example, autonomy to participate(or not), beneficence, and justice [136]. Beneficence inresearch on EoLC is important. Individuals with ad-vanced disease are unlikely to benefit from the research

Fig. 3 How ‘best’ to include adults nearing the end of life in research

Evans et al. BMC Medicine (2020) 18:221 Page 49 of 55

findings seeking to improve care and treatment, but par-ticipation can enable a sense of contribution to society,of their ‘voice’ being valued and foster a sense of altru-ism [137]. Excluding individuals because of perceptionsof vulnerability associated with advanced disease and im-paired capacity may be considered unjust [3]. It deniesequal opportunity for individuals living with progressiveconditions to contribute to care and treatment in pallia-tive care across the illness spectrum. To facilitate partici-pation in research, our findings identify two key areas:‘enabling and empowering consultees’ and ‘a carefullyconstructed and resourced recruitment strategy’.

Enabling and empowering consulteesWhile techniques for enhancing autonomy are imperativeto maximise individual decision-making, for adults withimpaired capacity in EoLC studies, proxy decision-makerswere crucial to enable recruitment. Consultees as proxydecision-makers sought to represent patients’ wishes andbest interests [105]. Their role was vital in enabling par-ticipation in research for adults across the capacityspectrum. While an array of consent processes can be im-plemented depending on the condition and level of cap-acity of the individual, successful participation ofindividuals relied on empowering and educating all thoseinvolved (patients, families, health and care staff, re-searchers, ethical committee members). Introducing pa-tients to research studies while they were competent, andtheir family members, empowered patients and their fam-ilies. Careful consideration is required to identity the rightproxy decision-maker. Consultees sought to consider thebenefits, risks, and burden to the person, and align theiradvice with understanding of the person’s wishes. Earlierconversations enabled family members to align advice onparticipation with understanding on the person’s wishes,particularly in comparatively high-risk studies [47, 67].Similarly, West et al. in their systematic review on ethicalchallenges in dementia research reported differences inconsultees advising enrolment with a higher proportionwilling to provide consent for non-invasive studies com-pared to invasive studies, even with no potential directbenefit for the patient [21]. Our findings show that thepresence of family members during enhanced consentprocesses enabled study participation by increasing oppor-tunity and support for the person to discuss their prior-ities [44, 86]. Although personal consultees were crucial,studies reported higher recruitment rates when they useda nominated/ professional consultee to advise if no per-sonal consultee was available (compared to designs with apersonal consultee alone). In contexts where being a pro-fessional consultee was unfamiliar, support and trainingfor professional consultees were required to increase con-fidence and understanding of their role, for example, in acare home with little/no embedded research culture.

A carefully constructed and resourced recruitmentstrategyRecruitment strategies in research on EoLC must align withthe target population and anticipate and plan for varied andchangeable levels of capacity associated with progressivedisease. Even with strict eligibility criteria, potential partici-pants within one study will likely present with varying andchangeable levels of capacity. Our findings identify the ne-cessity to have multiple consent processes in place alignedto the target population, context, and study aim to recruitindividuals across the capacity spectrum [51, 53]. Studieson EoLC demonstrate the feasibility of using multiple pro-cesses in, for example, trials in the dying phase [54] and ob-servational cohort studies [40, 77]. The studies demonstratethe resource intensity and careful planning required to im-plement complex processes of advance consent when theperson had capacity, and if/when capacity is lost, the in-volvement of proxy decision-makers (e.g. family or profes-sional). Boland et al. echo these findings in their systematicreview on recruitment strategies in trials involving peoplewith serious illness [26]. Our findings show that where pos-sible, researchers sought to maximise individual’s auton-omy, through consent processes such as enhanced,advanced, deferred, and process consent, and when lackedcapacity incorporating a personal and/or professional con-sultee declaration that sought to align with understandingof the person’s wishes and preferences. Enhanced consentwas feasible and acceptable to augment understanding andreasoning by using methods of multiple formats and involv-ing family members [61]. Similarly, Hostiuc et al. reportedfrom their meta-regression that using multimedia topresent study information significantly improved under-standing, reasoning, and appreciation and enabled informedconsent for adults with schizophrenia in clinical trials [22].There is clearly opportunity to incorporate advances fromother specialities, but with consideration if a successfulstrategy in one population is transferable to another. Differ-ent conditions impair different components of cognition,with variance in transience and progression (e.g. betweenschizophrenia and dementia), and prospect of recovery ornot (e.g. between emergency medicine and palliative care).Further process evaluation of multimedia techniques isneeded to determine the content and duration as the activeingredients of the techniques which leads to enhanced deci-sional capacity in studies for specific populations. Toolkitsare useful resources to inform consent processes thatencompass the heterogeneity of for example palliative carepopulations and the legislative framework for therespective jurisdiction and context [138] (See page 96in for example [139]).

LimitationsOur incorporation of systematic review and expert consult-ation and consensus enabled consideration of a breadth of

Evans et al. BMC Medicine (2020) 18:221 Page 50 of 55

evidence and the application in study designs to enable par-ticipation across the capacity spectrum. The search strategyenabled consideration of a breadth of research methodsemployed in different clinical populations and contexts re-quired to underpin the MORECare_Capacity statement.But the breadth of our strategy may have limited the identi-fication of all relevant original research studies involvingadults with serious illness and impaired capacity. While wepresent evidence from an array of original research studiesincluding qualitative work, future work is required to sys-tematically review recruitment processes for clinical trialsin palliative care. The use of search terms to identify eligiblestudies was limited by poor reporting of recruitment pro-cesses and details regarding mental capacity of eligible par-ticipants [55]. To address this, we used supplementaryprocesses of reference chaining and seeking recommenda-tions from experts of key studies. In line with internationalreporting guidelines [28, 140], researchers should providedetailed information about recruitment processes and dif-ferentiate between participants who consented for them-selves and those requiring a proxy decision-maker. Thiswould improve assessment of the success of the recruit-ment methods used to inform research practice [55]. Whilerecommended research methods can be generalisable, theseneed to be tailored to the respective setting, context, andculture. Our findings were limited to predominantly West-ern contexts. This reflects the concentration of palliativecare research in, for example, the USA and Europe [141].Finally, the study included lay representatives in the work-shops and Delphi Survey, but PPI in the data synthesis andvoluntary sector representation was limited to membershipof the Project Advisory Group. Greater exploration is re-quired in areas directly impacting on patients and families,for example, the use of advance directives to record a per-son’s preferences for research participation (R21), and es-tablishing a national body to provide support andinformation to family members acting as consultees (state-ment solution 12).

ConclusionsTo meet the increasing need for palliative care requiresgreater provision of evidence-based services and treatments,which is informed by research that includes adults fromacross the capacity spectrum. Our findings show that con-ducting research involving adults with impaired capacity isfeasible and acceptable, and it is ethically unjust to excludethem. Inclusion requires careful planning of processes ofrecruitment that are aligned to varying and changeablelevels of capacity, and the nature and intent of the researchstudy to minimise risk of harm. Studies must ask researchquestions important for patients and families in theprovision of services, care, and treatment in serious health-related illness. This review and TEC present innovative re-search methods, solutions on enhancing use, and critical

consideration on implementation in research studies in dif-ferent populations, clinical settings, and research designs.The MORECare_Capacity statement provides solutionsunderpinned by carefully considered evidence on involvingand supporting consultees, consent processes across thecapacity spectrum, and decision-making with adaptationsfor the respective care setting. The solutions have applic-ability for vulnerable patient populations in palliative careand beyond such as mental health and emergency medi-cine. The application of the solutions stated requires con-sideration within respective jurisdiction’s legislativeframework. Future research is required on the applicabilityof the MORECare_C statement for non-Western culturesand low- to middle-income countries.

Supplementary informationSupplementary information accompanies this paper at https://doi.org/10.1186/s12916-020-01654-2.

Additional file 1: Table S1. PRISMA statement and checklist of items.

Additional file 2: Additional methods [Tables S2-S4]. Table S2:Systematic review electronic search terms for the respective database.Table S3: Study design categories using the Cochrane Effective Practiceand Organisation of Care taxonomy. Table S4: Systematic review dataextraction template.

Additional file 3: Additional results for the systematic review[Tables S5-S8; suppl. 9]. Table S5: Qualsyst quality assessment of theincluded quantitative studies. Table S6: Qualsyst quality assessment of theincluded qualitative studies. Table S7: Reported study designs categorisedby the Cochrane Effective Practice and Organisation of Care taxonomy.Table S8: Public attitudes and ethical issues in recruiting adults across thecapacity spectrum. Suppl. 9: Additional reporting on public attitudes andethical issues.

Additional file 4: Additional results for the Transparent ExpertConsultation [S10-S15]. Table S10: Delphi Survey participants in roundone and round two. Table S11: Delphi Survey round onerecommendations and level of consensus. Fig. S12: Delphi Survey boxand whisker plots round one recommendations. Table S13: Delphi Surveyround two recommendations and level of consensus. Fig. S14: DelphiSurvey box and whisker plots round two recommendations. Box S15:Expert ‘think-tank’ - equivocal areas and the priority solutions.

AbbreviationsAD: Alzheimer’s disease; A&E: Accident and emergency; AMI: Acutemyocardial infarction; CTIMP: Clinical Trial of an Investigational MedicinalProduct; DMC: Decision-making capacity; EIC: Enhanced informed consent;EoLC: End of life care; EPOC: The Effective Practice and Organisation of Care;GP: General practitioner; HCP: Healthcare professional; ICU: Intensive careunit; IPOS-Dem: Integrated Palliative care Outcome Scale–Dementia;IRB: Institutional review board; LMICs: Low- to middle-income countries; Mac-CAT-CR: MacArthur competence assessment tool for clinical research;MCA: Mental Capacity Act 2005; MRC: Medical Research Council; MMSE: Mini-Mental State Examination; NoK: Next of kin; Non-CTIMP: Clinical Trial of anInvestigational Non-medicinal Product; PRISMA: Preferred Reporting Items forSystematic Reviews and Meta-Analyses; QoIC: Quality of informed consent;RCT: Randomised controlled trial; SD: Standard deviation; TEC: Transparentexpert consultation

AcknowledgementsWe would like to thank our Project Advisory Group members: DeborahTanner, Claire Henry, Gunn Grande, Steve Dewar, Gareth Owen, RachelBurman, Dimitrios Adamis, Michael Dunn, Scott Kim, and Simon Woods forthe oversight of the study and contributions, and Rowena Vohora for heradministrative support in the initial stages of the study.

Evans et al. BMC Medicine (2020) 18:221 Page 51 of 55

Authors’ contributionsCJE, PL, JK, BW, WB, MH, and IJH were responsible for the conception anddesign of the study. CJE, EY, and KS were responsible for conducting thesearches, data extraction, and analysis. CJE and KS were responsible forconducting and analysing the TEC, and EY in the reporting. CJE, EY, PL, JK,KS, IT, BW, WB, MH, and IJH wrote and revised the manuscript. All authorsread and approved the final manuscript.

FundingMORECare_Capacity was funded by Marie Curie (MCCC-RP-11-A12544).MORECare_Capacity aimed to determine how best to include individualsnear to death in research on EoLC by identifying solutions and developingbest practice guidance on processes of consent for people with impairedmental capacity. CJE is funded by HEE/NIHR Senior Clinical Lectureship (ICA-SCL-2015-01-001). The views expressed are those of the author(s) and notnecessarily those of the NIHR, the Department of Health and Social Care. Thefunders had no role in the design of the study and collection, analysis, andinterpretation of data and in writing the manuscript.

Availability of data and materialsAll data analysed for this review are included in this published article and itsadditional files.

Ethics approval and consent to participateThe transparent expert consultation received ethical review and approvalfrom King’s College London Research Ethics Committee (ref no. BDM/10/11-90). All participants who took part provided informed consent. Thesystematic review included previously published studies. We could notinfluence the design of the prior studies upon which this work is based andcannot comment on individual ethics approval or consent.

Consent for publicationNot applicable.

Competing interestsThe authors declare that they have no competing interests.

Author details1Cicely Saunders Institute of Palliative Care, Policy & Rehabilitation, FlorenceNightingale Faculty of Nursing, Midwifery & Palliative Care, King’s CollegeLondon, Bessemer Road, London SE5 9PJ, UK. 2Sussex Community NHSFoundation Trust, Brighton General Hospital, Brighton, UK. 3Centre of MedicalLaw and Ethics, The Dickson Poon School of Law, King’s College London,London, UK. 4Oxford University Hospitals NHS Foundation Trust and HarrisManchester College, University of Oxford, Oxford, UK. 5King’s CollegeHospital, London, UK. 6Psychological Medicine, Institute of Psychiatry,Psychology & Neuroscience, King’s College London, London, UK.

Received: 12 December 2019 Accepted: 3 June 2020

References1. Sleeman KE, de Brito M, Etkind S, Nkhoma K, Guo P, Higginson IJ, Gomes B,

Harding R. The escalating global burden of serious health-related suffering:projections to 2060 by world regions, age groups, and health conditions.Lancet Glob Health. 2019;7(7):e883–92.

2. Knaul FM, Farmer PE, Krakauer EL, De Lima L, Bhadelia A, Jiang Kwete X,Arreola-Ornelas H, Gomez-Dantes O, Rodriguez NM, Alleyne GAO, et al.Alleviating the access abyss in palliative care and pain relief-an imperativeof universal health coverage: the Lancet Commission report. Lancet. 2017;391(10128):1391–454.

3. Koffman J, Morgan M, Edmonds P, Speck P, Higginson IJ. Vulnerability inpalliative care research: findings from a qualitative study of black Caribbean andwhite British patients with advanced cancer. J Med Ethics. 2009;35(7):440–4.

4. Department of Health. End of Life Care Strategy - promoting high qualitycare for adults at the end of life. London: Crown; 2008.

5. Neuberger J, Guthrie C, Aaronovitch D. More care, less pathway: a review of theLiverpool Care Pathway. London: Department of Health and Social Care; 2013.

6. Koffman J, Yorganci E, Yi D, Gao W, Murtagh F, Pickles A, Barclay S, JohnsonH, Wilson R, Sampson L, et al. Managing uncertain recovery for patients

nearing the end of life in hospital: a mixed-methods feasibility clusterrandomised controlled trial of the AMBER care bundle. Trials. 2019;20(1):506.

7. Goodwin DM, Higginson IJ, Edwards AG, Finlay IG, Cook AM, Hood K,Douglas HR, Normand CE. An evaluation of systematic reviews of palliativecare services. J Palliat Care. 2002;18(2):77–83.

8. Lorenz KA, Lynn J, Dy SM, Shugarman LR, Wilkinson A, Mularski RA, MortonSC, Hughes RG, Hilton LK, Maglione M, et al. Evidence for improvingpalliative care at the end of life: a systematic review. Ann Intern Med. 2008;148(2):147–59.

9. Zimmermann C, Riechelmann R, Krzyzanowska M, Rodin G, Tannock I.Effectiveness of specialized palliative care: a systematic review. JAMA. 2008;299(14):1698–709.

10. Gysels M, Higginson IJ. Improving supportive and palliative care for adultswith cancer: research evidence. London: NICE; 2004.

11. Finlay IG, Higginson IJ, Goodwin DM, Cook AM, Edwards AG, Hood K,Douglas HR, Normand CE. Palliative care in hospital, hospice, at home:results from a systematic review. AnnOncol. 2002;13(Suppl 4):257–64.

12. Chan RWJ. End-of-life care pathways for improving outcomes in caring forthe dying. Chichester: Cochrane Database of Systematic Reviews: Reviews2010 Issue 1 John Wiley & Sons; 2010.

13. Higginson IJ, Finlay IG, Goodwin DM, Hood K, Edwards AG, Cook A, DouglasHR, Normand CE. Is there evidence that palliative care teams alter end-of-life experiences of patients and their caregivers? J Pain Symptom Manag.2003;25(2):150–68.

14. Shepherd V, Griffith R, Sheehan M, Wood F, Hood K. Healthcareprofessionals’ understanding of the legislation governing research involvingadults lacking mental capacity in England and Wales: a national survey. JMed Ethics. 2018;44(9):632–7.

15. Casarett DJ, Karlawish JH. Are special ethical guidelines needed for palliativecare research? J Pain Symptom Manag. 2000;20(2):130–9.

16. Dewing J. From ritual to relationship: a person centred approach to consentin qualitative research with older people who have dementia. Int J Soc ResPract. 2002;1(2):156–71.

17. Dewing J. Participatory research: a method for process consent withpersons who have dementia. Dementia. 2007;6(1):11–25.

18. Nuffield Council on Bioethics. Dementia ethical issues. London: NuffieldCouncil on Bioethics; 2009.

19. Owen GS, David AS, Richardson G, Szmukler G, Hayward P, Hotopf M.Mental capacity, diagnosis and insight in psychiatric in-patients: a cross-sectional study. Psychol Med. 2009;39(8):1389–98.

20. Lemaire F, Blanch L, Cohen SL, Sprung C. Informed consent for researchpurposes in intensive care patients in Europe--part II. An official statementof the European Society of Intensive Care Medicine. Working group onethics. Intensive Care Med. 1997;23(4):435–9.

21. West E, Stuckelberger A, Pautex S, Staaks J, Gysels M. Operationalisingethical challenges in dementia research--a systematic review of currentevidence. Age Ageing. 2017;46(4):678–87.

22. Hostiuc S, Rusu MC, Negoi I, Drima E. Testing decision-making competencyof schizophrenia participants in clinical trials. A meta-analysis and meta-regression. BMC Psychiatry. 2018;18:2.

23. Mukherjee A, Livinski AA, Millum J, Chamut S, Boroumand S, Iafolla TJ,Adesanya MR, Dye BA. Informed consent in dental care and research for theolder adult population. J Am Dental Assoc. 2017;148(4):211–20.

24. Sugarman J, McCrory DC, Hubal RC. Getting meaningful informed consentfrom older adults: a structured literature review of empirical research. J AmGeriatr Soc. 1998;46(4):517–24.

25. Wohleber AM, McKitrick DS, Davis SE. Designing research with hospice andpalliative care populations. Am J Hosp Palliat Med. 2012;29(5):335–45.

26. Boland J, Currow DC, Wilcock A, Tieman J, Hussain JA, Pitsillides C, AbernethyAP, Johnson MJ. A systematic review of strategies used to increase recruitmentof people with cancer or organ failure into clinical trials: implications forpalliative care research. J Pain Symptom Manag. 2015;49(4):762–72 e765.

27. Bruera E, Hui D, Dalal S, Torres-Vigil I, Trumble J, Roosth J, Krauter S,Strickland C, Unger K, Palmer JL. Parenteral hydration in patients withadvanced cancer: a multicenter, double-blind, placebo-controlledrandomized trial. J Clin Oncol. 2013;31(1):111.

28. Higginson IJ, Evans CJ, Grande G, Preston Nancy J, Morgan M, McCrone P,Lewis P, Fayers P, Harding R, Hotopf M, et al. Evaluating complex interventionsin End of Life Care: the MORECare Statement on good practice generated by asynthesis of transparent expert consultations and systematic reviews. BMCMed. 2013;11(1):111 https://doi.org/10.1186/1741-7015-11-111(111).

Evans et al. BMC Medicine (2020) 18:221 Page 52 of 55

29. Department of Health. Mental Capacity Act. London: Department of Health;2005.

30. Association WM. World Medical Association Declaration of Helsinki. Ethicalprinciples for medical research involving human subjects. Bull World HealthOrgan. 2001;79(4):373.

31. Fink A, Kosecoff J, Chassin M, Brook R. Consensus methods: characteristicsand guidelines for use. Am J Public Health. 1984;74(9):979–83.

32. Sinha IP, Smyth RL, Williamson PR. Using the Delphi technique to determinewhich outcomes to measure in clinical trials: recommendations for thefuture based on a systematic review of existing studies. PLoS Med. 2011;8(1):e1000393.

33. Pope C, Mays N, Popay J. Synthesizing qualitative and quantitative healthevidence: a guide to methods. Maidenhead and New York: Open UniversityPress, McGraw Hill Education; 2007.

34. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P,Stewart LA. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1.

35. Kmet LM, Lee RC, Cook LS. Standard quality assessment criteria forevaluating primary research papers from a variety of fields: Alberta HeritageFoundation for Medical Research; 2004.

36. EPOC. Cochrane effective practice and organisation of care review group.Data collection checklist. In. Ottowa: Cochrane Effective Practice andOrganisation of Care; 2002.

37. Endnote X8. http://endnote.com/.Accessed 2 Nov 2016.38. Jones J, Hunter D. Consensus methods for medical and health services

research. In: Mays N, Pope C, editors. Qualitative research in health care. 2nded. London: British Medical Journal Publishing Group; 1999.

39. Jones L, Harrington J, Scott S, Davis S, Lord K, Vickerstaff V, Round J, CandyB, Sampson EL. CoMPASs: IOn programme (Care Of Memory Problems inAdvanced Stages of dementia: Improving Our Knowledge): protocol for amixed methods study. BMJ Open. 2012;2(6).

40. Sampson EL, Candy B, Davis S, Gola AB, Harrington J, King M, Kupeli N,Leavey G, Moore K, Nazareth I. Living and dying with advanced dementia: aprospective cohort study of symptoms, service use and care at the end oflife. Palliat Med. 2018;32(3):668–81.

41. Scott S, Jones L, Blanchard MR, Sampson EL. Study protocol: the behaviourand pain in dementia study (BePAID). BMC Geriatr. 2011;11:61.

42. Sampson EL, White N, Lord K, Leurent B, Vickerstaff V, Scott S, Jones L. Pain,agitation, and behavioural problems in people with dementia admitted togeneral hospital wards: a longitudinal cohort study. Pain. 2015;156(4):675–83.

43. Dobratz MC. Issues and dilemmas in conducting research with vulnerablehome hospice participants. J Nurs Scholarsh. 2003;35(4):371–6.

44. Chouliara Z, Kearney N, Worth A, Stott D. Challenges in conducting researchwith hospitalized older people with cancer: drawing from the experience ofan ongoing interview-based project. Eur J Cancer Care. 2004;13(5):409–15.

45. Gysels M, Evans CJ, Lewis P, Speck P, Benalia H, Preston NJ, Grande GE,Short V, Owen-Jones E, Todd CJ. MORECare research methods guidancedevelopment: recommendations for ethical issues in palliative and end-of-life care research. Palliat Med. 2013;27(10):908–17.

46. Dunning T, Duggan N, Savage S, Martin P. Diabetes and end of life: ethicaland methodological issues in gathering evidence to guide care. Scand JCaring Sci. 2013;27(1):203–11.

47. Warren JW, Sobal J, Tenney JH, Hoopes JM, Damron D, Levenson S,DeForge BR, Muncie HL Jr. Informed consent by proxy. N Engl J Med. 1986;315(18):1124–8.

48. Fowell A, Johnstone R, Finlay IG, Russell D, Russell IT. Design of trials withdying patients: a feasibility study of cluster randomisation versusrandomised consent. Palliat Med. 2006;20(8):799–804.

49. Gardiner C, Gott M, Ingleton C, Seymour J, Cobb M, Noble B, Bennett M,Ryan T. Extent of palliative care need in the acute hospital setting: a surveyof two acute hospitals in the UK. Palliat Med. 2013;27(1):76–83.

50. Myers JB. Improving decisions about transport to the emergency departmentfor assisted living residents who fall. Ann Intern Med. 2018;168(3):179–86.

51. Abernethy AP, Currow DC, Hunt R, Williams H, Roder-Allen G, Rowett D,Shelby-James T, Esterman A, May F, Phillips PA. A pragmatic 2 x 2 x 2factorial cluster randomized controlled trial of educational outreach visitingand case conferencing in palliative care-methodology of the Palliative CareTrial [ISRCTN 81117481]. Contemp Intern Med. 2006;27(1):83–100.

52. Whelan PJ, Walwyn R, Gaughran F, Macdonald A. Impact of the demand for‘proxy assent’ on recruitment to a randomised controlled trial of vaccinationtesting in care homes. J Med Ethics. 2013;39(1):36–40.

53. Davies AN, Waghorn M, Webber K, Johnsen S, Mendis J, Boyle J. A clusterrandomised feasibility trial of clinically assisted hydration in cancer patientsin the last days of life. Palliat Med. 2018;32(4):733–43.

54. Rees E, Hardy J. Novel consent process for research in dying patients unableto give consent. BMJ. 2003;327(7408):198.

55. Irwin S, Zurhellen C, Diamond L, Dunn L, Palmer B, Jeste D, Twamley E.Unrecognised cognitive impairment in hospice patients: a pilot study. PalliatMed. 2008;22(7):842–7.

56. Henwood TR, Keogh JW, Reid N, Jordan W, Senior HE. Assessing sarcopenicprevalence and risk factors in residential aged care: methodology andfeasibility. J Cachexia Sarcopenia Muscle. 2014;5(3):229–36.

57. Wood F, Prout H, Bayer A, Duncan D, Nuttall J, Hood K, Butler CC. Consent,including advanced consent, of older adults to research in care homes: aqualitative study of stakeholders’ views in South Wales. Trials. 2013;14(1):247.

58. Kleiderman E, Avard D, Black L, Diaz Z, Rousseau C, Knoppers BM. Recruitingterminally ill patients into non-therapeutic oncology studies: views of healthprofessionals. BMC Med Ethics. 2012;13:33.

59. Buckles V, Powlishta K, Palmer J, Coats M, Hosto T, Buckley A, Morris J.Understanding of informed consent by demented individuals. Neurology.2003;61(12):1662–6.

60. Black BS, Kass NE, Fogarty LA, Rabins PV. Informed consent for dementiaresearch: the study enrollment encounter. IRB Ethics Hum Res. 2007;29(4):7–14.

61. Mittal D, Palmer BW, Dunn LB, Landes R, Ghormley C, Beck C, Golshan S,Blevins D, Jeste DV. Comparison of two enhanced consent procedures forpatients with mild Alzheimer disease or mild cognitive impairment. Am JGeriatr Psychiatry. 2007;15(2):163–7.

62. Palmer BW, Harmell AL, Dunn LB, Kim SY, Pinto LL, Golshan S, Jeste DV.Multimedia aided consent for Alzheimer’s disease research. Clin Gerontol.2018;41(1):20–32.

63. Rubright J, Casarett DJ, Gur R, Xie SX, Karlawish J, Sankar P. A memory andorganizational aid improves Alzheimer disease research consent capacity:results of a randomized, controlled trial. Am J Geriatr Psychiatry. 2010;18(12):1124–32.

64. Olazarán J, Agüera-Ortiz L, Osorio RS, León-Salas B, Dobato JL, Cruz-OrduñaI, González B, Valentí M, Gil-Ruiz N, Frades B. Promoting research inadvanced dementia: early clinical results of the Alzheimer Center ReinaSofía Foundation. J Alzheimers Dis. 2012;28(1):211–22.

65. Hughes T, Castro RM. A processural consent methodology with peoplediagnosed with dementia. Qual Ageing Older Adults. 2015;16(4):222–34.

66. Kim SY, Karlawish JH, Kim HM, Wall IF, Bozoki AC, Appelbaum PS.Preservation of the capacity to appoint a proxy decision maker: implicationsfor dementia research. Arch Gen Psychiatry. 2011;68(2):214–9.

67. Karlawish J, Kim SY, Knopman D, Van Dyck CH, James BD, Marson D. The viewsof Alzheimer disease patients and their study partners on proxy consent forclinical trial enrollment. Am J Geriatr Psychiatry. 2008;16(3):240–7.

68. Gainotti S, Imperatori SF, Spila-Alegiani S, Maggiore L, Galeotti F, VanacoreN, Petrini C, Raschetti R, Mariani C, Clerici F. How are the interests ofincapacitated research participants protected through legislation? An Italianstudy on legal agency for dementia patients. PLoS One. 2010;5(6).

69. Agarwal MR, Ferran J, Ost K, WILSON KC. Ethics of ‘informed consent’in dementiaresearch—the debate continues. Int J Geriatric Psychiatry. 1996;11(9):801–6.

70. Cowdell F. Engaging older people with dementia in research: myth orpossibility. Int J Older People Nursing. 2008;3(1):29–34.

71. Siminoff LA, Caputo M, Burant C. The promise of empirical research in thestudy of informed consent theory and practice. In: Hec Forum: 2004:Springer Science & Business Media; 2004. p. 53.

72. Warner J, McCarney R, Griffin M, Hill K, Fisher P. Participation in dementiaresearch: rates and correlates of capacity to give informed consent. J MedEthics. 2007;34(3):167–70.

73. Black BS, Finucane T, Baker A, Loreck D, Blass D, Fogarty L, Phillips H,Hovanec L, Steele C, Rabins PV. Health problems and correlates of pain innursing home residents with advanced dementia. Alzheimer Dis AssocDisord. 2006;20(4):283–90.

74. Goodman C, Baron N, Machen I, Stevenson E, Evans C, Davies S, Illiffe S.Culture, consent, sosts and care homes: enabling older people withdementai to participate in research. Aging Ment Health. 2011;15(4):6.

75. Hanson LC, Gilliam R, Tae JL. Successful clinical trial research in nursinghomes: the improving decision-making study. Clin Trials. 2010;7(6):735–43.

76. Baskin SA, Morris J, Ahronheim JC, Meier DE, Morrison RS. Barriers toobtaining consent in dementia research: implications for surrogate decision-making. J Am Geriatr Soc. 1998;46(3):287–90.

Evans et al. BMC Medicine (2020) 18:221 Page 53 of 55

77. Mitchell SL, Kiely DK, Jones RN, Prigerson H, Volicer L, Teno JM. Advanceddementia research in the nursing home: the CASCADE study. Alzheimer DisAssoc Disord. 2006;20(3):166–75.

78. Ellis-Smith C, Higginson IJ, Daveson BA, Henson LA, Evans CJ. BuildCare: Howcan a measure improve assessment and management of symptoms andconcerns for people with dementia in care homes? A mixed-methodsfeasibility and process evaluation of IPOS-Dem. PLoS One. 2018;13(7):e0200240.

79. Galeotti F, Vanacore N, Gainotti S, Izzicupo F, Menniti-Ippolito F, Petrini C,Chiarotti F, Chattat R, Raschetti R. How legislation on decisional capacity cannegatively affect the feasibility of clinical trials in patients with dementia.Drugs Aging. 2012;29(8):607–14.

80. Kim SY, Kim HM, McCallum C, Tariot PN. What do people at risk forAlzheimer disease think about surrogate consent for research? Neurology.2005;65(9):1395–401.

81. Rikkert M, Van Den Bercken J, Ten Have H, Hoefnagels W. Experiencedconsent in geriatrics research: a new method to optimize the capacity toconsent in frail elderly subjects. J Med Ethics. 1997;23(5):271–6.

82. Sudore RL, Landefeld CS, Williams BA, Barnes DE, Lindquist K, Schillinger D.Use of a modified informed consent process among vulnerable patients: adescriptive study. J Gen Intern Med. 2006;21(8):867–73.

83. Mason S, Barrow H, Phillips A, Eddison G, Nelson A, Cullum N, Nixon J. Briefreport on the experience of using proxy consent for incapacitated adults. JMed Ethics. 2006;32(1):61–2.

84. Davies K, Collerton JC, Jagger C, Bond J, Barker SA, Edwards J, Hughes J,Hunt JM, Robinson L. Engaging the oldest old in research: lessons from theNewcastle 85+ study. BMC Geriatr. 2010;10:64.

85. Botker MT, Jorgensen MT, Stengaard C, Seidenfaden SC, Tarpgaard M,Granfeldt A, Mortensen TO, Grofte T, Friesgaard KD, Maerkedahl R, et al.Prehospital triage of patients suffering severe dyspnoea using N-terminalpro-brain natriuretic peptide, the PreBNP trial: a randomised controlledclinical trial. Eur Heart J Acute Cardiovasc Care. 2018;7(4):302–10.

86. Ford M, Kallen M, Richardson P, Matthiesen E, Cox V, Teng E, Cook K,Petersen N. Effect of social support on informed consent in older adultswith Parkinson disease and their caregivers. J Med Ethics. 2008;34(1):41–7.

87. Thomalla G, Boutitie F, Fiebach JB, Simonsen CZ, Nighoghossian N, PedrazaS, Lemmens R, Roy P, Muir KW, Heesen C, et al. Effect of informed consenton patient characteristics in a stroke thrombolysis trial. Neurology. 2017;89(13):1400–7.

88. Boxall L, Hemsley A, White N. Exploring recruitment issues in stroke research: aqualitative study of nurse researchers’ experiences. Nurs Res. 2016;23(5):8–14.

89. Ali K, Roffe C, Crome P. What patients want: consumer involvement in thedesign of a randomized controlled trial of routine oxygen supplementationafter acute stroke. Stroke. 2006;37(3):865–71.

90. Moran-Sanchez I, Luna A, Perez-Carceles MD. Assessment of capacity toconsent to research among psychiatric outpatients: prevalence andassociated factors. Psychiatry Q. 2016;87(1):89–105.

91. Harmell AL, Palmer BW, Jeste DV. Preliminary study of a web-based tool forenhancing the informed consent process in schizophrenia research.Schizophr Res. 2012;141(2–3):247–50.

92. Jeste DV, Palmer BW, Golshan S, Eyler LT, Dunn LB, Meeks T, Glorioso D,Fellows I, Kraemer H, Appelbaum PS. Multimedia consent for research inpeople with schizophrenia and normal subjects: a randomized controlledtrial. Schizophr Bull. 2009;35(4):719–29.

93. Moser DJ, Reese RL, Hey CT, Schultz SK, Arndt S, Beglinger LJ, Duff KM,Andreasen NC. Using a brief intervention to improve decisional capacity inschizophrenia research; 2006.

94. Campbell MM, Susser E, Mall S, Mqulwana SG, Mndini MM, Ntola OA,Nagdee M, Zingela Z, Van Wyk S, Stein DJ. Using iterative learning toimprove understanding during the informed consent process in a SouthAfrican psychiatric genomics study. PLoS ONE. 2017;12(11):e0188466.

95. Carey E, Griffiths C. Recruitment and consent of adults with intellectualdisabilities in a classic grounded theory research study: ethical andmethodological considerations. Disabil Soc. 2017;32(2):193–212.

96. Ramerman L, Hoekstra PJ, de Kuijper G. Health-related quality of life inpeople with intellectual disability who use long-term antipsychotic drugsfor challenging behaviour. Res Dev Disabil. 2018;75:49–58.

97. Ho P, Downs J, Bulsara C, Patman S, Hill AM. Addressing challenges ingaining informed consent for a research study investigating falls in peoplewith intellectual disability. Br J Learn Disabil. 2018;46(2):92–100.

98. Hamilton J, Ingham B, McKinnon I, Parr JR, Tam LYC, Le Couteur A. Mentalcapacity to consent to research? Experiences of consenting adults with

intellectual disabilities and/or autism to research. Br J Learn Disabil. 2017;45(4):230–7.

99. Adamis D, Martin F, Treloar A, Macdonald A. Capacity, consent, andselection bias in a study of delirium. J Med Ethics. 2005;31(3):137–43.

100. Adamis D, Treloar A, Martin FC, Macdonald AJ. Ethical research in delirium:arguments for including decisionally incapacitated subjects. Sci Eng Ethics.2010;16(1):169–74.

101. Marcantonio ER, Bergmann MA, Kiely DK, Orav EJ, Jones RN. Randomizedtrial of a delirium abatement program for postacute skilled nursing facilities.J Am Geriatr Soc. 2010;58(6):1019–26.

102. Cole MG, McCusker J, Bellavance F, Primeau FJ, Bailey RF, Bonnycastle MJ,Laplante J. Systematic detection and multidisciplinary care of delirium in oldermedical inpatients: a randomized trial. Can Med Assoc J. 2002;167(7):753–9.

103. Mangset M, Forde R, Nessa J, Berge E, Wyller TB. I don’t like that, it’s trickingpeople too much...: acute informed consent to participation in a trial ofthrombolysis for stroke. J Med Ethics. 2008;34(10):751–6.

104. Honarmand K, Belley-Cote EP, Ulic D, Khalifa A, Gibson A, McClure G, SavijaN, Alshamsi F, D’Aragon F, Rochwerg B, et al. The deferred consent modelin a prospective observational study evaluating myocardial injury in theintensive care unit. J Intensive Care Med. 2018;33(8):475–80.

105. Bolcic-Jankovic D, Clarridge BR, LeBlanc JL, Mahmood RS, Roman AM, FreemanBD. Exploring determinants of surrogate decision-maker confidence: anexample from the ICU. J Empir Res Hum Res Ethics. 2014;9(4):76–85.

106. Levine AC, Barry MA, Agrawal P, Duber HC, Chang MP, Mackey JM, HansotiB. Global Emergency Medicine Think Tank Clinical Research Working G:Global health and emergency care: overcoming clinical research barriers.Acad Emerg Med. 2017;24(4):484–93.

107. Smith OM, McDonald E, Zytaruk N, Foster D, Matte A, Clarke F, Fleury S,Krause K, McArdle T, Skrobik Y, et al. Enhancing the informed consentprocess for critical care research: strategies from a thromboprophylaxis trial.Intensive Crit Care Nurs. 2013;29(6):300–9.

108. Offerman SR, Nishijima DK, Ballard DW, Chetipally UK, Vinson DR, Holmes JF.The use of delayed telephone informed consent for observationalemergency medicine research is ethical and effective. Acad Emerg Med.2013;20(4):403–7.

109. Day TL, Bench SD, Griffiths PD. The role of pilot testing for a randomisedcontrol trial of a complex intervention in critical care. J Res Nurs. 2015;20(3):167–78.

110. Bench S, Day T, Heelas K, Hopkins P, White C, Griffiths P. Evaluating thefeasibility and effectiveness of a critical care discharge information pack forpatients and their families: a pilot cluster randomised controlled trial. BMJOpen. 2015;5(11):e006852.

111. Higginson IJ, Rumble C, Shipman C, Koffman J, Sleeman KE, Morgan M,Hopkins P, Noble J, Bernal W, Leonard S, et al. The value of uncertainty incritical illness? An ethnographic study of patterns and conflicts in care anddecision-making trajectories. BMC Anesthesiol. 2016;16:11.

112. Rouze A, Loridant S, Poissy J, Cornu M, Nseir S, Sendid B. Impact of the useof biomarkers on early discontinuation of empirical antifungal therapy incritically ill patients: a randomised controlled study. Med Mycol. 2018;56:S49.

113. Buckley JM, Irving AD, Goodacre S. How do patients feel about taking partin clinical trials in emergency care? Emerg Med J. 2016;33(6):376–80.

114. van Beinum A, Hornby L, Dhanani S, Ward R, Chambers-Evans J, Menon K.Feasibility of conducting prospective observational research on critically ill,dying patients in the intensive care unit. J Med Ethics. 2017;43(1):47–51.

115. Kamarainen A, Silfvast T, Saarinen S, Virta J, Virkkunen I. Conduct of emergencyresearch in patients unable to give consent--experiences and perceptions ofpatients, their consent providing next of kin, and treating physicians followinga prehospital resuscitation trial. Resuscitation. 2012;83(1):81–5.

116. Barrett KA, Ferguson ND, Athaide V, Cook DJ, Friedrich JO, McDonald E,Pinto R, Smith OM, Stevenson J, Scales DC. Surrogate decision makers'attitudes towards research decision making for critically ill patients.Intensive Care Med. 2012;38(10):1616–23.

117. Gong MN, Winkel G, Rhodes R, Richardson LD, Silverstein JH. Surrogateconsent for research involving adults with impaired decision making: surveyof Institutional Review Board practices. Crit Care Med. 2010;38(11):2146–54.

118. Burns KE, Magyarody NM, Duffett M, Nisenbaum R, Cook DJ. Attitudes ofthe general public toward alternative consent models. Am J Crit Care. 2011;20(1):75–83.

119. Kim SY, Kim HM, Knopman DS, De Vries R, Damschroder L, Appelbaum PS.Effect of public deliberation on attitudes toward surrogate consent fordementia research. Neurology. 2011;77(24):2097–104.

Evans et al. BMC Medicine (2020) 18:221 Page 54 of 55

120. Karlawish J, Rubright J, Casarett D, Cary M, Ten Have T, Sankar P. Olderadults’ attitudes toward enrollment of non-competent subjects participatingin Alzheimer’s research. Am J Psychiatry. 2009;166(2):182–8.

121. Bravo G, SY HK, Dubois M-F, Cohen CA, Wildeman SM, Graham JE. Surrogateconsent for dementia research: factors influencing five stakeholder groupsfrom the SCORES study. IRB Ethics Hum Res. 2013;35(4):1–11.

122. Bravo G, Duguet AM, Dubois MF, Delpierre C, Vellas B. Substitute consentfor research involving the elderly: a comparison between Quebec andFrance. J Cross Cult Gerontol. 2008;23(3):239–53.

123. Farber NJ, Aboff BM, DeJoseph MR, Castellano J, Weiner J, Boyer EG. Physicians’understanding of consent requirements for phase I clinical trials in cognitivelyimpaired or highly vulnerable populations. Account Res. 2004;11(1):63–78.

124. Kim SY, Kim HM, Langa KM, Karlawish JH, Knopman DS, Appelbaum PS.Surrogate consent for dementia research: a national survey of olderAmericans. Neurology. 2009;72(2):149–55.

125. De Vries R, Ryan KA, Stanczyk A, Appelbaum PS, Damschroder L, KnopmanDS, Kim SY. Public’s approach to surrogate consent for dementia research:cautious pragmatism. Am J Geriatr Psychiatry. 2013;21(4):364–72.

126. Karlawish JH, Knopman D, Clark CM, Morris JC, Marson D, Whitehouse PJ,Kawas CH. Informed consent for Alzheimer’s disease clinical trials: a surveyof clinical investigators. IRB. 2002;24(5):1–5.

127. Cahill M, Wichman A. Research involving persons with cognitiveimpairments: results of a survey of Alzheimer disease research centers in theUnited States. Alzheimer Dis Assoc Disord. 2000;14(1):20–7.

128. van der Vorm A, Vernooij-Dassen MJ, Kehoe PG, Olde Rikkert MG, vanLeeuwen E, Dekkers WJ. Ethical aspects of research into Alzheimer disease.A European Delphi Study focused on genetic and non-genetic research. JMed Ethics. 2009;35(2):140–4.

129. Peterson G, Wallin A. Alzheimer disease ethics--informed consent and relatedissues in clinical trials: results of a survey among the members of the ResearchEthics Committees in Sweden. Int Psychogeriatr. 2003;15(2):157–70.

130. Garcia J, Evans J, Reshaw M. “Is there anything else you would like to tellus”- methodological issues int he use of free-text comments form postalsurveys. Qual Quant. 2004;38:113–25.

131. Dickert NW, Hendershot KA, Speight CD, Fehr AE. Patients’ views of consentin clinical trials for acute myocardial infarction: impact of trial design. J MedEthics. 2017;43(8):524–9.

132. Olazaran J, Aguera-Ortiz L, Osorio RS, Leon-Salas B, Dobato JL, Cruz-OrdunaI, Gonzalez B, Valenti M, Gil-Ruiz N, Frades B, et al. Promoting research inadvanced dementia: early clinical results of the Alzheimer Center ReinaSofia Foundation. J Alzheimers Dis. 2012;28(1):211–22.

133. Gibbins J, Reid CM, Bloor S, Burcombe M, McCoubrie R, Forbes K.Overcoming barriers to recruitment in care of the dying research inhospitals. J Pain Symptom Manag. 2013;45(5):859–67.

134. European Parliament and of the Council of the European Union. Regulation(EU) No 536/2014 on clinical trials on medicinal products for human use, andrepealing Directive 2001/20/EC. Off J Eur Union. 2014;57:L58-51-5820176.

135. Gysels MH, Evans C, Higginson IJ. Patient, caregiver, health professional andresearcher views and experiences of participating in research at the end oflife: a critical interpretive synthesis of the literature. BMC Med Res Methodol.2012;12(1):123.

136. Foster C. The ethics of medical research on humans. Cambridge: CambridgeUniversity Press; 2001.

137. Koffman J, Stone K, Murtagh F. Ethics in palliative care research. In: BrueraEHI, von Gunten C, Morita T, editors. Textbook of Palliative Medicine. 2nded. London: CRC Press; 2015. p. 211–20.

138. University of Leicester UoB. Adults lacking capacity – on-line toolkit: NRES; 2010.139. Koffman J, Yorganci E, Murtagh F, Yi D, Gao W, Barclay S, Pickles A,

Higginson I, Johnson H, Wilson R. The AMBER care bundle for hospitalinpatients with uncertain recovery nearing the end of life: the ImproveCarefeasibility cluster RCT. Health Technol Assess. 2019;23(55):1.

140. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, VandenbrouckeJP. The Strengthening the Reporting of Observational Studies inEpidemiology (STROBE) statement: guidelines for reporting observationalstudies. Ann Intern Med. 2007;147(8):573–7.

141. Kaasa S, Dale O. Building up research in palliative care: an historicalperspective and a case for the future. Clin Geriatr Med. 2005;21(1):81–92.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Evans et al. BMC Medicine (2020) 18:221 Page 55 of 55