PSYCHO-ONCOLOGY

Psycho-Oncology 8: 260–263 (1999)

BRIEF REPORT

COLLECTING QUALITY OF LIFE DATA INEORTC CLINICAL TRIALS—WHAT HAPPENS IN

PRACTICE?TERESA YOUNG1,*,2 and JANE MAHER2

Lynda Jackson Macmillan Centre for Cancer Support and Information, Mount Vernon Hospital, Northwood,Middlesex, HA6 2RN, UK

SUMMARY

Problems with poor compliance when collecting quality of life data (QoL) in randomised clinical trials haveprompted investigators to suggest measures to improve data collection. This study sought to look at the practicalproblems encountered by data managers and nurses in the cross-cultural setting of EORTC trials. A literaturesearch was followed by a poster workshop session at a meeting of the EORTC Study Group on Data Managementand finally a postal questionnaire. The key problems identified centred around the lack of interest from someclinicians, lack of resources, inadequate protocols and a desire for training in the rationale for collecting QoL datato aid discussions with patients. Despite these problems many data managers and nurses found it rewarding to beon the ‘frontline’. Since this study, the EORTC Data Center and in particular the Quality of Life Study Groupand the Quality of Life Unit have implemented a number of measures to improve compliance. These includewritten guidelines for EORTC trials and a training course planned for Autumn 1999. Copyright © 1999 JohnWiley & Sons, Ltd.

‘Evaluating the quality of life is a terrible bearing tosome people, the real truth to others. Some find itjust a fashion, others think it is the next step inoncological research. Without taking quality of lifeinto account we cannot pretend to understand themeaning of lengthening the life of the cancer patient.However many of us remain sceptical about ourability to assess a subjective element such as qualityof life. The data manager, in charge of administeringthe questionnaires, is in the middle of these doubts.’Belgian Data Manager (Van Vreckem and Parme-nier, 1994)

AIMS

Recent publications have examined the attitudesof physicians towards evaluating Quality of Life(QoL) (Taylor and MacDonald, 1996). This col-laborative European Organisation for Researchand Treatment of Cancer (EORTC) project be-tween the Study Group on Data Management(SGDM), the Quality of Life Study Group(QLSG) and the Quality of Life Unit (QLU)sought to identify the practical problems encoun-tered by the data managers (DMs) and nurseswho are often entrusted with the day to day issuesof collecting the data within EORTC clinicaltrials.

METHOD

A literature search identified a number of mea-sures that had been proposed to improve compli-ance. These recommendations were used to

* Correspondence to: Lynda Jackson Macmillan Centre forCancer Support and Information, Mount Vernon Hospital,Northwood, Middlesex, HA6 2RN, UK.1 On behalf of the EORTC Study Group on Data Manage-ment.2 A member of the EORTC Quality of Life Study Group.

CCC 1057–9249/99/030260–04$17.50Copyright © 1999 John Wiley & Sons, Ltd.

Recei6ed 29 December 1997Accepted 30 No6ember 1998

EO

RT

CC

LIN

ICA

LT

RIA

LS

261

Copyright

©1999

JohnW

iley&

Sons,L

td.P

sycho-O

ncology8:

260–

263(1999)

Table 1. A summarised version of the questionnaire and results

MissingYesQuestion No N/A or don’tknow

58 5 0Are you responsible for collecting clinical data and completing Case Report Forms for EORTC Trials? 0Do you participate in discussions with patients to explain the clinical trial, and that quality of life data will be 0025 38

collected, and seek their informed consent?Do you seek out the patients in person and give them a quality of life questionnaire then collect it from them after 29 34 0 0

a short while?If yes, do you ever have to help patients complete quality of life questionnaires? 23/29 6/29 34 0

234If yes, do patients have relatives or carers with them whilst completing questionnaires? 18/29 9/294/18 14/18If yes, does this ever cause you any problems? 43 2

21/29 8/29If yes, do you check for missing data immediately a questionnaire is completed? 34 0If yes and data is missing do you ask the patient to complete all the questions? 20/21 1/21 42 0

0Do you arrange for the questionnaires to be left in an appropriate place (e.g. patient’s notes) and ask someone else 42 21 0(e.g. a clinic nurse or the doctor) to give it out and return it to you?

04716 0Do you mail questionnaires to patients at home?8/16If yes, do you send out reminders if not returned? 8/16 47 0

Do patients complete questionnaires in a busy OP clinic? 22 0142718 31Do patients complete questionnaires in a private room? 14 013 36Do patients complete questionnaires whilst an IP on a ward? 14 0

Do patients complete questionnaires at home? 27 22 14 049 11Are you always told a patient is entered in a study before the baseline questionnaire is due? 0 3

373617Do you get sent reminders that questionnaires are due, either from the study co-ordinator or the data centre?23/36 13/36If no, would you find them helpful? 24 3

Do you get feedback about quality of life results for studies you are involved with? 12 49 0 249/49 0/49If no, would you find such information useful? 21259 1Do you have access to the trial protocol? 0 360 2If the protocol is not clear can you contact the clinical investigator for further information or support? 0 1

Are you specifically funded to collect quality of life data? 0 60 0 3

T. YOUNG AND J. MAHER262

facilitate a poster workshop at an SGDM meetingto identify other new problems. The workshopwas not audiotaped, however, extensive noteswere taken. Finally, a postal questionnaire wassent to EORTC DMs, nurses and principalinvestigators.

RESULTS

Workshop

The workshop was attended by 89 DMs fromBelgium (22), France (2), Italy (2), Netherlands(49), Spain (2) and the UK (12). The main areasof concern were:

� Collecting QoL data is rarely part of routineclinical care, it is usually an ‘extra’ task thatmust be remembered. If forgotten or missed,the data cannot be inferred or obtained at alater date.

� Job descriptions, working practices andavailability of DMs vary across Europe. Inmany countries DMs are unavailable whenQoL assessments are due.

� Most DMs have a sense of responsibility andcommitment to any trial with which they areinvolved. Where they are unable to be person-ally involved and have to rely on staff notnecessarily committed to the trial, this was feltto be detrimental.

� Some clinicians were thought to regard collect-ing QoL data as less important than collectingclinical data. Consequentially, there was a lackof resources and funding for the extra workinvolved.

� An initial personal explanation to the patient isrequired to ensure high quality data. DMs arenot always aware of the justification and ratio-nale for collecting QoL data, and providingsuch an explanation is not necessarily withintheir field of expertise.

� Some DMs are reluctant to have personalcontact with patients. They feel vulnerable toquestions from the patient about their diagnosisor prognosis. In particular, when a clinician’smotivation for collecting QoL data is thoughtto be poor and the patient’s prognosis worsens,selection bias may be introduced as not allpatients are approached on all occasions.

� Sometimes patients do not fully understand thequestionnaires and want additional informa-

tion, or are confused and reluctant to completea questionnaire. It is not always clear from theprotocol how much ‘help’ is acceptable or howquestions should be represented to the patientwithout biasing results.

� When patients are unable to complete theirtreatment according to the protocol (e.g. de-layed chemotherapy due to infection), guidanceis not always available on the procedure tofollow.

Postal questionnaire

Three hundred and forty questionnaires weremailed to SGDM members (but not all those onthe SGDM mailing list at that time were ‘actively’involved in data management for current EORTCtrials), Oncology Nursing Study Group membersand EORTC Principal Investigators. One hundredand twenty replies were received. This paper onlyrelates to the 63 respondents actively responsiblefor collecting QoL data for EORTC trials. Asummarised version of the questionnaire plus re-sults are given in Table 1. Asterisks denote addi-

Table 2. Key problems identified from workshop and ques-tionnaire

(1) Resources� Lack of clinician support� Lack of time� Lack of funding� Inability to be present when quality of life data duefor collection

(2) Inadequate protocols� Lack of information on collection procedure iftreatment schedule disrupted

(3) Missing data� Baseline questionnaires missed� No opportunity to check for missed questions� Missing questionnaires when responsible person not inclinic

(4) Communication skills� Lack of knowledge on justification and rationale forcollecting QoL� Reluctance to approach all patients on alloccasions—selection bias� No guidance for providing ‘help’

(5) Reminders and feedback

(6) Site of questionnaire completion� No private rooms� Questionnaires taken home

Copyright © 1999 John Wiley & Sons, Ltd. Psycho-Oncology 8: 260–263 (1999)

EORTC CLINICAL TRIALS 263

tional open ended questions where respondentscould add comments or details about specificlocal practices.

DISCUSSION

Table 2 lists the key problems identified duringthis project. If QoL data are to influence clinicalpractice and policy making about cancer treat-ment in Europe it is necessary to ensure that highquality data are collected from a large and repre-sentative sample of patients in well-designed QoLstudies within appropriate clinical trials. Moreresources and increased clinician support are onlylikely to be forthcoming when QoL data has beenshown to be the major endpoint in determiningthe outcome of important trials. Some advocatethat the current limited resources should be in-vested in a few trials where clinical endpoints,such as survival or recurrence, are unlikely to besignificantly different, but QoL results may be thedeciding outcome when making recommendationsfor future treatment. A demonstration of the im-pact of good QoL data in such trials wouldincrease the overall awareness of QoL issues andmaybe lead to realistic assessments of the re-sources and finances required.

The content requirements of a clinical trial pro-tocol are constantly increasing, as efforts aremade to improve compliance and reduce missingdata of all types. Within the EORTC, the QLUand members of the QLSG are able to offeradvice on incorporating QoL assessment intoEORTC clinical trials and a manual will shortlybe available (Young et al., 1999). One particularrecommendation is that completion of a baselineQoL questionnaire should be an eligibility crite-rion. However, collecting follow-up QoL datamay require significant changes in local practicenecessitating training and additional funding.

QoL compliance data is also prepared on a regu-lar basis for feedback to clinicians at EORTCStudy Group meetings. It is then the clinicians’responsibility to arrange local feedback to theirstaff.

QoL issues are regularly included in the pro-grammes of courses run by the EORTC and bydata management groups in the UK and theNetherlands. The QLU organised a workshop todiscuss a standard training programme with rep-resentatives from 14 European countries, and atraining manual has been prepared (Kiebert,1997). A specific QoL course is planned for Au-tumn 1999. Topics to be covered will include thejustification and rationale for collecting QoLdata, identifying trials where it is beneficial toassess QoL outcomes, selecting appropriate in-struments, the design and conduct of the trial andcross-cultural issues so that reliable, valid andclinically relevant data are collected.

Despite these problems some DMs find it re-warding to be on the ‘frontline’ where collectingQoL data is concerned and are keen to be in-volved in policies to improve compliance.

REFERENCES

Kiebert, W. (on behalf of EORTC Quality of Life Unit)(1997) Manual for a Standard Training Course forQuality of Life Assessment in Clinical Trials.Brussels.

Taylor, K.M. and MacDonald, K.G. (1996) Physicians’perspective on quality of life: An exploratory studyof oncologists. Qual. Life Res. 5(1), 5–14.

Van Vreckem, A. and Parmenier, N. (1994) Bias ofQuality of Life data. EORTC SGDM Annual Meet-ing Proceedings, Maastricht.

Young, T., de Haes, H., Curran, D., Fayers, P. andBrandberg, Y. (on behalf of EORTC Quality of LifeStudy Group) (1999) Guidelines for Assessing Qualityof Life in EORTC Clinical Trials. Brussels: EORTC.

Copyright © 1999 John Wiley & Sons, Ltd. Psycho-Oncology 8: 260–263 (1999)