headley court trust report headley court trust proj ect · project report – january 2013 ... the...
TRANSCRIPT
1
Headley Court
Trust Proj ect
Headley Court Trust Report
Project Report – January 2013 20132012November 2012
2
THE HEADLEY COURT TRUST PROJECT
The identification of factors which conduce to and prevent rehabilitation and
adjustment in military personnel following combat-related injury: a pragmatic
evaluation of patients and their partners.
Emeritus Professor David A Alexander,
Professor Susan Klein,
Major Kevin Forbes
January 2013
Aberdeen Centre for Trauma Research,
The Robert Gordon University, ABERDEEN
3
Table of Contents
Table of Contents ................................................................................................................ 3
List of Tables ....................................................................................................................... 4
List of Figures ...................................................................................................................... 5
Acknowledgements .............................................................................................................. 6
Introduction ......................................................................................................................... 7
Literature review ................................................................................................................. 8
Mental health and combat-related disorders ..................................................................... 8
Post-traumatic psychopathology ..................................................................................... 11
The ‘ripple effect’ of trauma ........................................................................................... 11
Study aims and research questions .................................................................................. 13
Study design ....................................................................................................................... 14
Population and samples .................................................................................................. 14
Selection criteria ............................................................................................................. 16
Recruitment..................................................................................................................... 17
Data collection ................................................................................................................ 19
Participant assessments ................................................................................................... 19
Clinical outcome data ..................................................................................................... 20
Self-report data ............................................................................................................... 20
Research governance ...................................................................................................... 24
Data analysis ................................................................................................................... 26
Results ................................................................................................................................ 27
Recruitment..................................................................................................................... 27
Response Rates ............................................................................................................... 29
Demographics ................................................................................................................. 30
Primary outcome variables – ‘independent samples’ ..................................................... 46
Primary outcome variables – ‘Paired samples’ ............................................................... 56
Health and lifestyle variables .......................................................................................... 62
Qualitative findings (Partner study) ................................................................................ 67
Conclusions ........................................................................................................................ 72
Recruitment..................................................................................................................... 72
Response rates ................................................................................................................ 73
Demographics ................................................................................................................. 74
Primary outcome variables ............................................................................................. 77
Health and lifestyle variables .......................................................................................... 80
Qualitative findings (Partner study) ................................................................................ 82
Recommendations ............................................................................................................. 83
References .......................................................................................................................... 84
4
List of Tables
Table 1: 36-Month study period .............................................................................................. 15
Table 2: Study groups .............................................................................................................. 16
Table 3: Summary of selection criteria (all groups) ................................................................ 17
Table 4: Participant assessments ............................................................................................. 19
Table 5: Participant assessment schedule and summary of self-report measures.................... 24
Table 6: Response rates (all groups) ....................................................................................... 30
Table 7: Ethnicity of participants (01PG) and (01RG) ............................................................ 31
Table 8: Ethnicity of participants (02PG) and (02RG) ............................................................ 31
Table 9: Gender and incident type (01PG) and (01RG) .......................................................... 32
Table 10: Age at first admission to the DMRC (01PG) and (01RG) ...................................... 32
Table 11: Relationship status (01PG) and (01RG) (n=226) .................................................... 33
Table 12: Financially dependent children (01PG) and (01RG) ............................................... 35
Table 13: Highest level of educational achievement (01PG) and (01RG) .............................. 36
Table 14: Highest level of educational attainment (02PG) ..................................................... 37
Table 15: Employment status (02PG) and (02RG) ................................................................. 38
Table 16: Military service by type ........................................................................................... 39
Table 17: Branch of service ..................................................................................................... 40
Table 18: Current rank ............................................................................................................ 41
Table 19: Length of service ..................................................................................................... 42
Table 20: Type of incident (01PG) and (01RG) ...................................................................... 43
Table 21: Nature of incident (01PG) and (01RG) ................................................................... 45
Table 22: PCL-C – Baseline, six months and 12 months (01PG) ........................................... 47
Table 23: PCL-C - Baseline, six months and 12 months (02PG) ............................................ 48
Table 24: GHQ-12 - Baseline, six months and 12 months (01PG) ......................................... 52
Table 25: GHQ-12 - Baseline, six-months and 12-months (02PG)......................................... 54
Table 26 CDRIS-10 - Baseline, six-months and 12-months (01PG)....................................... 55
Table 27: Paired samples statistics (PCL-C) ........................................................................... 60
Table 28: Paired samples tests (PCL-C) .................................................................................. 60
Table 29: Paired samples statistics (CDRIS-10) ..................................................................... 61
Table 30: Paired samples tests (CDRIS-10) ............................................................................ 61
Table 31: Paired samples statistics (GHQ-12) ........................................................................ 61
Table 32: Paired samples tests (GHQ-12) ............................................................................... 62
Table 33: AUDIT scores (01PG) ............................................................................................. 63
Table 34: Mean outcome measures for Abstainees compared with non-abstainees................ 63
Table 35: Mean outcome measures for abstainees and non-abstainees (non-parametric tests)64
Table 36: Regular social activities outside of work (01PG) .................................................... 64
Table 37: Were most of the people you socialised with involved with
the military? (Prior to injury) ........................................................................................ 65
Table 38: Social activity prior to injury - going to the gym with friends (Baseline)............... 65
Table 39: Marital/relationship satisfaction for panel (n=11) ................................................... 66
Table 40: SSI - Baseline, six months and 12 months for
strict panel (n=11) (01PG) ............................................................................................ 67
5
List of Figures
Figure 1: Summary of clinical outcome data........................................................................... 20
Figure 3: Age at first admission to DMRC (01PG) and (01RG) ............................................. 33
Figure 4: Relationship status (01PG) and (01RG) (n=125) ..................................................... 34
Figure 5: Highest level of educational achievement (01PG) and (01RG) ............................... 35
Figure 6: Highest level of educational achievement (02PG) ................................................... 37
Figure 7: Military service by type ........................................................................................... 39
Figure 8: Branch of service ..................................................................................................... 40
Figure 9: Current rank (or rank at completion of service) ....................................................... 41
Figure 10: Length of service (01PG) and (01RG) - Regular and FTRS .................................. 42
Figure 11: Length of service (01PG) and (01RG) - Reservists ............................................... 43
Figure 12: Type of incident (01PG) and (01RG) .................................................................... 44
Figure 13: Nature of incident (01PG) and (01RG) .................................................................. 45
Figure 14: Injury category ....................................................................................................... 46
Figure 15: PCL-C - Baseline (01PG) ...................................................................................... 48
Figure 16: PCL-C - Six months (01PG) .................................................................................. 48
Figure 17: PCL-C - 12 months (01PG) ................................................................................... 48
Figure 18: PCL-C - Baseline (02PG) ...................................................................................... 49
Figure 19: PCL-C - Six months (02PG) .................................................................................. 49
Figure 20: PCL-C - 12 months (02PG) ................................................................................... 49
Figure 21: GHQ-12 - Baseline using Likert and GHQ scoring methods (01PG) .................... 50
Figure 22: GHQ-12 - Six-months using Likert and GHQ scoring methods (01PG) ............... 51
Figure 23: GHQ-12 - 12-months using Likert and GHQ scoring methods (01PG) ................ 52
Figure 24: GHQ-12 - Baseline using Likert and GHQ scoring methods (02PG) .................... 53
Figure 25: GHQ-12 - Six-months using Likert and GHQ scoring methods (02PG) ............... 53
Figure 26: GHQ-12 - 12-months using Likert and GHQ scoring methods (02PG) ................ 54
Figure 27: CDRIS-10 - Baseline, six-months and 12-months (01PG) .................................... 56
Figure 28: PCL-C and CDRIS-10 - Means for panel sample (n=12) ...................................... 57
Figure 29: PCL-C and CDRIS-10 - Medians for panel sample (n=12) ................................... 57
Figure 30: PCL-C and GHQ-12 Medians for panel sample (n=12) ........................................ 58
Figure 31: PCL-C and GHQ-12 Means for panel sample (n=12) ........................................... 59
6
Acknowledgements
No study of this size could be successfully completed without the support, advice and
goodwill of a great many people. Although it is not possible to mention them all
personally, we are especially grateful to the following individuals whose personal
contribution greatly facilitated the progress of the study: Colonel John Etherington
(Director of Defence Rehabilitation); Flight Lieutenant Stuart Haggerty (Former
Officer Commanding, Mental Health Flight at the Centre for Mental and Cognitive
Health, DMRC), Mrs Anne Brannagan (Complex Trauma Team Manager) and Mr Kit
Malia (Project Manager, Mild Traumatic Brain Injury).
Thanks must also go to Colonel Jeremy Tuck (former Commanding Officer at the
DMRC), Major Helen Rickard, Mrs Lucia Piruch (Manager, Officers Mess) and to all
of the clinical staff for making us welcome at the DMRC and for assisting in the day
to day management of the study.
To the Board of the Headley Court Trust for their financial support we extend our
thanks. Special thanks are due to Commodore Tobin Elliott for his encouragement
and support throughout the study.
Finally, this project would have achieved little without the commitment and co-
operation of all of those who took time to reply to our questionnaires, patient and
partner alike. We made many demands on them, all of which were met with
enthusiasm and a strong desire to contribute to the care of those service personnel and
their families affected by combat-injury. We sincerely hope that our own efforts and
the outcomes of this study will go some way towards recompensing them.
Emeritus Professor David A Alexander
Professor Susan Klein
Major Kevin J Forbes
7
THE HEADLEY COURT TRUST PROJECT
The identification of factors which conduce to and prevent rehabilitation and
adjustment in military personnel following combat-related injury: a pragmatic
evaluation of patients and their partners.
Introduction
In 2009 the Defence Medical Rehabilitation Centre (DMRC) based at Headley Court
received considerable praise from the Healthcare Commission for its “excellent”
quality of care and treatment of UK service personnel in the first-ever independent
review of the Defence Medical Services (DMS) (Healthcare Commission, 2009). The
National Audit Office (NAO) further recognised the contribution of the DMRC in
2010, describing the service provided there as being ‘unique’ within the United
Kingdom (NAO, 2010). While these reports highlight the high standard of care
provided at Headley Court, they also highlight that, due to increased numbers of
operational casualties, additional demands placed upon the DMRC represent a
significant challenge for the rehabilitation services there and for the longer-term
management of those sustaining combat injury.
Between the 7th October 2001 and the 31st October 2012, 394 military and
civilian personnel were killed in action or died of wounds sustained on military
operations in Afghanistan (OP HERRICK)1. Over the same period, 587 individuals
were classified as being either ‘Very Seriously Injured or Wounded’ (VSI) (n=289) or
‘Seriously Injured or Wounded’ (SI) (n=298)2 and a closer examination of year on
year figures reveals a steady increase in casualty numbers from the beginning of OP
HERRICK until 2009 (DASA, 2012). The upward trend in operational casualties
sustained on OP HERRICK has had a corresponding effect on inpatient admission at
the DMRC. The NAO report Treating injury and illness arising on military operation
1 The total number of UK military and civilian fatalities in Afghanistan as reported by the Defence Analytical Services and Advice (DASA) agency totals 437 individuals. This total includes 43 ‘other’ deaths occurring as a result of accidental or violent
causes while deployed on operations and deaths due to disease related causes during the deployment. This total also includes
fratricide or so called ‘blue-on-blue’ or ‘green-on-blue’ incidents.
2 The Notification of Casualty or 'NOTICAS' system is used to categorise and report casualty status within the UK Armed
Forces. The classification 'Very Seriously Injured or Wounded' (VSI) is the definition used where the illness or injury is of such severity that life or reason is imminently endangered. The classification 'Seriously Injured or Wounded' (SI) is the definition
used where the patient's condition is of such severity that there is cause for immediate concern, but there is no imminent danger to life or reason.
8
identifies that this upward trend represents a doubling of complex trauma inpatients at
the DMRC between 2006 and 2009 (NAO, 2010).
While 2009 saw annual casualty rates from OP HERRICK peak and subsequently
reduce to pre-2009 levels; the DMRC has continued to face the legacy of an increase
in inpatient admission. In order to address the demands of this increase in inpatient
numbers the DMRC has seen a rise of 23% in its staffing establishment, 83% in its
ward bed provision and an expansion of service both at Headley Court itself and
through the inception of DMRC-led rehabilitation throughout the United Kingdom
and Germany at the Regional Rehabilitation Units (RRU’s). While these measures, in
no small part, attempt to address capacity issues and ensure the continuation of a high
standard of inpatient care, significant challenges remain in respect of the longer term
management of those with combat injuries post rehabilitation and, often, post military
service. A recent report on the efforts of Combat Stress presents the argument that
many of the positive gains achieved by rehabilitation services; military and civilian
alike are frequently short term (Fletcher, 2007). There is therefore, a clear need to
find ways of sustaining any initial positive effect of rehabilitation and to identify
factors which either compromise, or facilitate the durability of such effects in the
longer term.
Literature review
A review of the literature has highlighted three initial areas of concern in respect of
the potential compromise, or facilitated durability, of the positive effects of
rehabilitation and the effective management of those with combat injuries in the
longer term. These areas of concern are: the long-term nature of mental health
problems and of combat-related disorders; the continued pre-occupation with Post-
Traumatic Stress Disorder (PTSD) as a single diagnosis following trauma; and the
nature of observed family and marital relationship problems following operational
deployment.
Mental health and combat-related disorders
Post-traumatic reactions have been observed throughout recorded human history but it
is only with the introduction of PTSD into the Diagnostic and Statistical Manual of
Mental Disorders-III (DSM-III) (American Psychiatric Association, 1980), that a
more widespread acceptance that major catastrophes and military combat could result
in identifiable psychopathology and problems of psychosocial adjustment (Alexander,
9
1996). In terms of contemporary conflicts, Hotopf et al (2006) estimate a 4%
incidence of PTSD among UK troops returning from Iraq (without combat-injury);
Hoge et al (2004) however described a figure of just under 13% for returning troops,
and Engelhard et al (2007) report rates varying from 4%-21% among Dutch troops
returning from Iraq. Higher rates of PTSD were associated with higher levels of
direct combat exposure and minor wounds or injury, and rates of PTSD among
soldiers returning from war increased over time post-deployment as the majority of
soldiers with PTSD or depression at seven months had not met the criteria for either
condition one month following injury. These findings highlight the clinical
challenges of providing early psychiatric care for combat exposed personnel and the
difficulties associated with the management of these conditions in the longer term.
Whilst the evidence suggests that exposure to combat may be a potent source
of such sequelae, it is also recognized that no single traumatic event is guaranteed to
cause psychopathological reactions in all those exposed to it (Alexander, 1996). Why
considerable individual differences exist remains a challenge to clinical investigators
(Klein et al, 2002). Meta-analyses have identified several risk factors for PTSD in
both the military (Iversen et al, 2008) and the general population (Brewin et al, 2000;
Ozer et al, 2003) in terms of pre-trauma, peri-trauma, and post-trauma factors. Over
the past twenty years there has been a particular interest in the psychological
consequences of physical injury caused by a traumatic event among survivors of a
variety of traumatic events including road traffic incidents, terrorism, criminal assault,
and burn injuries (O’Donnell et al, 2003). While a few studies have shown that risk
for PTSD is associated with severity of injury, other studies have failed to replicate
these results in civilian populations.
Even although the research literature on traumatic injury is not yet sufficiently
advanced to differentiate whether some types of injury are more likely to elicit
emotional traumatic responses than others (O’Donnell et al, 2008), epidemiological
data suggest that certain traumatic events such as those involving interpersonal
violence, result in higher rates of PTSD (Klein & Alexander, 2006). Further studies
are required to investigate in depth whether there are potentially key mechanisms
inherent in certain physical injuries that increase the risk of poor psychological
outcomes. In one of the few studies which have attempted to identify predictors of
PTSD following combat-related injury, Koren et al (2005) directly compared, in a
matched case-control design, injured and non injured Israeli soldiers who experienced
10
the same combat events to estimate the unique contribution of physical injury over
and above that of the trauma itself to the subsequent development of PTSD.
Consistent with the outcome of some earlier studies (e.g., Kulka et al, 1990; Michaels
et al, 1999), their findings unequivocally indicated that bodily injury is a risk factor –
rather than a protective one – for PTSD. Approximately fifteen months post-injury,
16.7% of injured soldiers had PTSD compared to 2.5% of non-injured soldiers with
similar combat experiences. Moreover, the data also suggested that the odds of
developing PTSD following traumatic injury are approximately eight times higher
than following injury-free trauma. Interpretation of the validity of these figures
requires careful analysis of such factors as the intensity of combat; whether the data
were from reservists or full time troops; whether the data derived from self-report or
from structured clinical interviews; whether the psychopathology was genuinely
combat-related, and how long after deployment the surveys were conducted.
Moreover, objective measures of physical injury are often not related to PTSD
outcome. High levels of physical problems, chronic pain, and seeking of medical care
are often seen in patients diagnosed with PTSD. However, the relationship of pain and
other physical problems shortly after injury to PTSD outcome has not been well
studied. Thus, it is not clear what factors contribute to low initial rates of PTSD. For
example, it may be due to high levels of social and medical support, pain or sleep
medications routinely administered during the immediate post-injury period and a
focus on recovery. In addition, the literature does not support a predictable trend of
recovery from PTSD over time.
Given that a delayed onset of symptoms has been reported in several studies
(e.g., Hoge et al, 2004), further research is required to understand the longitudinal
course of post-traumatic reactions and to assess the factors which may affect the
reporting of symptoms at follow-up such as persisting physical problems, anticipation
of returning home and starting work, concern over health care, and pending disability.
This is particularly pertinent in light of the increasing numbers of injured combat
troops returning from Afghanistan with blast related poly-trauma (Sayer et al, 2008)
and associated traumatic brain injury [TBI] (French & Parkinson, 2008). Although the
acute clinical management of poly-trauma has been well described, the post-acute
clinical management of such cases in rehabilitation and primary care settings has not
been adequately addressed in the literature.
11
Post-traumatic psychopathology
It is now well established that PTSD is not the only psychiatric condition that may
develop in the aftermath of trauma (Klein & Alexander, 2006). Traditionally,
however, PTSD has tended to adopt centre stage thereby dominating the focus of
military research (Wessely, 2005). Comorbidity and PTSD however is the norm
rather than the exception (O’Donnell et al, 2004). In large community samples, for
example, between 80% and 85% of individuals with a diagnosis of PTSD also meet
criteria for at least one other psychiatric condition (Creamer et al, 2001). A meta-
analysis of psychopathology following civilian trauma suggested that increased rates
of generalised anxiety disorder, substance abuse, phobias, and major depressive
disorder occur following exposure to traumatic events (Brown et al, 2000). The
military literature confirms that symptoms from which soldiers may be at risk include
depression, anxiety, anger, sleep disturbance, somatisation, substance misuse,
dissociation and sexual problems (Goff et al, 2007). Currently, the literature indicates
that these additional symptoms (particularly depression and anxiety symptoms) are
higher in soldiers with PTSD than those without. Research conducted by Hoge et al
(2004) and Hoge (2005) reported an increase in post-deployment rates of PTSD
symptoms and anxiety in soldiers after deployment from Iraq and Afghanistan.
Studies which report symptoms, rather than diagnoses, imply a much higher level of
psychopathology. Lapierre et al (2007), having followed up US troops from Iraq and
Afghanistan, reported that about 44% of soldiers had symptoms of depression, and
post-traumatic stress. The available literature on the prevalence of these other
psychiatric conditions following combat-related injury however is very limited. Few
studies report prevalence and, of those that do, many do not clarify whether the
diagnosis is the sole condition or comorbid with PTSD or another diagnosis.
Moreover, quality of life, perception of recovery, and return to work are generally not
systematically indexed by researchers. Yet, the validity of findings based solely on
psychopathology is questionable in that they fail to address other major parameters
relevant to this population (O’Donnell et al, 2003).
The ‘ripple effect’ of trauma
Military personnel returning from Afghanistan face psychological challenges that can
exert profound effects on families and couples (Erbes et al, 2008). Particular
symptoms of Post-Traumatic Stress Disorder (PTSD), such as feelings of detachment,
12
or estrangement from others, restricted affect, irritability, and outbursts of anger, may
adversely affect the interpersonal relationships of service members with their spouses/
partners. In dealing with such problems, family members may experience a burden of
care, and may themselves develop psychological symptoms. Thus, the impact of a
traumatic event is not limited to the traumatized person but can also impact on their
spouse/partner or family member; a concept which has been referred to as the “Ripple
Effect” of trauma (Klein & Alexander, 2005) or secondary traumatisation (Figley,
1998). Whilst evidence suggests that an accumulation of secondary traumatic stress
can lead to emotional exhaustion and emotional burnout, little is known about the
factors that bring about secondary traumatisation (Dirkzwager et al, 2005). There is
some evidence to suggest that combat or other military-related traumatic experiences
may be particularly detrimental to marriage. Cook and colleagues (2004) reported that
31% of World War II prisoners of war with PTSD reported marital distress compared
with 11% of those without PTSD. Research focussing on the perspective of the
spouse/partner of war veterans also reports higher levels of impairment in marital,
family and social relations (e.g., Dirkzwager et al, 2005). A major limitation of this
research however is the undue emphasis placed on the diagnosis of PTSD to the
detriment of identifying problems associated with other post-traumatic
psychopathology such as substance abuse, depression, relationship conflict, and
occupational dysfunction that are more common and potentially more troublesome to
service members and their families (Batten & Pollack, 2008).
In a subsequent investigation of reasons for the excess of ill health in
reservists, Browne et al (2007) included problems at home (readjustment or
relationship difficulties) and marital satisfaction. Following deployment, reservists
reported experiencing significantly more major problems and readjustment difficulties
than regulars. Because couple and family relationships may serve as either a vital
support or challenging obstacle to recovery from combat-related pathology, including
recovery from physical injury, research is required to identify factors associated with
adjustment and recovery. Collins and Kennedy (2008) further extend this particular
approach to the case of service members experiencing multiple and severe injuries,
requiring treatment in the poly-trauma system of care, due to the substantial burden
which may be placed on family members in fulfilling an unexpected caretaking role.
In summary therefore, there is a need to adopt a dual focus on the individual needs of
the service member in his or her rehabilitation and on the support and functioning of
13
the care-giving spouse/ partner (Sammons & Batten, 2008). It represents a highly
topical issue as it concerns an “invisible healthcare system” which constitutes the core
long-term care in the community. Despite increasing recognition by policy makers of
the need to consider such effects (e.g., National Service Framework for Mental Health
[Department of Health 1999]), relatively few studies have systematically pursued a
focus on this important aspect of trauma care, and none has specifically examined this
issue with regards to combat-injured personnel despite the likelihood that the welfare
of partners or immediate family members will impact on the rehabilitation and
adjustment of the injured.
Study aims and research questions
The primary aims of this unique research study were to identify in combat-injured
personnel the:
1. durability of psychiatric and psychosocial gains from rehabilitation following
combat-related injury, and
2. impact of combat-related injury on the partner of military personnel in terms
of mental health, psychosocial adjustment, and relationships.
In fulfilling the first of these aims the study addressed the following research
questions with respect to military personnel admitted to the DMRC.
a) What are the long-term effects of rehabilitation of combat-injured personnel
following admission to the Defence Medical Rehabilitation Centre, Headley
Court?
b) What factors either compromise or facilitate the durability of therapeutic gain?
In fulfilling the second aim of this study, we addressed the following research
question with respect to partners of military personnel admitted to the DMRC.
c) What is the association between the durability of therapeutic gain and the
‘Ripple Effect’ of trauma?
14
Study design
There are a number of well-established difficulties and common methodological
problems associated with the study of military cohorts (Hotopf & Wessely, 2005),
(O’Donnell et al, 2003). Difficulties associated with low response rates, recall bias,
self-selection of participants and lack of comparable ‘non-combat’ control groups
consistently affect the validity and reliability of such studies. This study sought to
avoid these difficulties and common methodological problems by seeking to: facilitate
high participation rates; minimise selection bias; enhance the interpretation of the
relationship between combat-related injury and outcome; minimise recall bias through
the inclusion of a prospective cohort; employ multimodal assessments, which include
objective measures, and provide a suitable ‘homogeneous’ control group for
comparison.
In order to effectively answer all three of the research questions the study
design also incorporated the linked and parallel study of military personnel admitted
to the DMRC with combat-related injuries (patient study) and their partners (partner
study).
Population and samples
In order to achieve the first of the research aims, the patient study consisted of a 36-
month case-control design to facilitate the effective comparison of complex trauma
patients with combat-related injury (study group) and an exact age- and gender-
matched control group of musculoskeletal patients with non-combat-related injury
(control group). In order to ensure a sufficient sample and adequate numbers for
follow-up this study utilised both prospective and retrospective cohorts.
Within the patient study, the prospective cohort consisted of individuals
admitted to the DMRC for the first time as in-patients during the designated 18-month
recruitment period3. In contrast, the retrospective cohort consisted of individuals
discharged from their final cycle of treatment in the 6 to 24 months preceding the
prospective recruitment period4. The utilisation of both prospective and retrospective
cohorts was intended to effectively extend the period studied from initial in-patient
admission to 36 months post-admission (see Table 1: 36-Month study period). The
inclusion of both prospective and retrospective cohorts within the study was intended
3 The eighteen month recruitment period from the 3rd January 2010 and the 30th June 2011 was agreed by Col. Jeremy Tuck, the
then Commanding Officer of the DMRC, Col. John Etherington, Director of Defence Rehabilitation and Prof. David A. Alexander and Prof. Susan Klein, the Principal Investigators. 4 1st January 2008 to 31st June 2009
15
to facilitate higher response rates (by maximising participation) and to minimise the
effects of recall bias. Initial estimates of participant numbers, based on previous in-
patient admission to the DMRC, predicted that up to 340 patient participants across
both cohorts (prospective and retrospective) and both groups (study and control)
could be expected during the 18-month period.
Table 1: 36-Month study period
Prospective Group
(Days/Months post admission to
DMRC)
Retrospective Group
(Months post discharge from DMRC)
Recruitment
Within 7 days of first admission 6 12 18 24 Baseline
Assessment
Six-Month
Assessment 6 12 18 24 30
12-Month
Assessment 12 18 24 30 36
In order to achieve the second of the study aims a parallel 36-month case-control
study of partners was incorporated within the design. Once again, in order to ensure a
sufficient sample and adequate follow-up of partner participants the study made use of
both prospective and retrospective cohorts. In order to ensure sufficient numbers for
analysis following the anticipated attrition of partners due to unsuitability, refusal to
participate or loss at follow-up, initial estimates suggested that up to 300 partner
participants would be required across both cohorts (prospective and retrospective) and
both groups (study and control).
As the study design incorporated the linked and parallel study of military
personnel admitted to the DMRC and their partners (patient study and partner study)
and utilised both study and control groups as well as prospective and retrospective
cohorts, eight distinct study groups may be described. For the purpose of the report,
however, discussion of the study groups will be limited to the four main groupings:
Prospective Patient Group (01PG); Retrospective Patient Group (01RG); Prospective
Partner Group (02PG) and the Retrospective Partner Group (02RG) (see Table 2:
Study groups).
16
Table 2: Study groups
Patient Study (01) Partner Study (02)
Prospective
Cohort (PG)
Retrospective
Cohort (RG)
Prospective
Cohort (PG)
Retrospective
Cohort (RG)
Study
Group(S) 01PG 01RG 02PG 02RG
Control
Group(C)
Selection criteria
Within the patient study, participants were required to be a minimum of 18 years of
age at recruitment and able to give full informed consent. Study group patient
participants had to have sustained an operational (combat-related) injury and this
injury must have been the cause of their admission to the DMRC. Alternatively,
members of the control group were required to have been the victims of a non-
combat-related injury sustained out-with the operational environment, occurring
either on or off military duty (e.g., road traffic incident, sporting injury or other
accident). This injury also must have been the cause of their admission. Recruitment
of both study and control groups from a core population of DMRC in-patients was
intended to ensure the provision of a suitable ‘homogeneous’ control group. A
summary of selection criteria is presented in Table 3: Summary of selection criteria
(all groups).
Potential patient participants meeting the selection criteria were invited to
participate in the study unless otherwise presenting with moderate or severe
Traumatic Brain Injury (TBI)5 or other identifiable cerebral impediment likely to
compromise their understanding of, or ability to respond to, verbal and written
assessments. Moderate or severe traumatic brain injury has previously been identified
as a confounding factor in the assessment of post-injury mental health (Hill, Mobo, &
Cullen, 2009). Those individuals presenting with traumatic brain injury often exhibit
symptoms consistent with PTSD and considerable overlap in presentation and
sequelae serves only to complicate their diagnosis and management (Vasterling et al,
2009).
Within the partner study, participants were also required to be: able to give full
informed consent and; a minimum of 18 years of age at time of recruitment. Potential
5 The criteria utilised for the identification of traumatic brain injury in participants was that adopted by the World Health
Organisation (WHO) and employed within the Centre for Mental and Cognitive Health at DMRC. For the purpose of this study,
therefore, mild TBI (mTBI) was defined as injury resulting from, or being represented by, a Glasgow Coma Scale (GCS) score of >12 within 30 minutes of injury; a loss of consciousness of <30 minutes, and posttraumatic amnesia <24 hours. This
definition includes complex mTBI where focal lesions may be evident.
17
partner participants were those identified as being in what they themselves regarded
as being a ‘committed relationship’ with an index patient enrolled on the patient study
(in either study or control group) and that index patient must have given prior consent
for the partner to be approached. Potential partner participants who were non-
English-speaking, or presented with impediments that may have limited their
understanding of the purpose and nature of the study were excluded from
participation. Where potential partner participants were involved in legal proceedings
against the patient participant or were the co-victim of the index trauma (e.g. road
traffic incident) they were also excluded from participation.
Table 3: Summary of selection criteria (all groups)
Study Group/Cohort Inclusion Criteria Exclusion Criteria
General Criteria (All Groups)
Must be able to give informed consent.
Must be aged eighteen years and above at
time of recruitment to the study.
Patient Participants
Study Group
Both Prospective and
Retrospective Cohorts
Victim of an operational
combat-related injury Moderate or severe TBI
Cerebral impediment
Discharged from DMRC outwith the
previous six to twenty-four months Patient Participants
Control Group
Both Prospective and
Retrospective Cohorts
Victim of non-combat-related injury
Partner Participants
Prospective and Retrospective
Cohorts
In a committed relationship
with an index patient
Non-English speaking
Impediments to understanding the
purpose of the project
Involved in legal proceedings
against the patient
Co-victim of the index trauma
Recruitment
Patient Group (Prospective cohort)
Potential prospective patient participants were identified in collaboration with the
complex trauma team at the DMRC and their initial eligibility was determined by the
clinical summary of injuries presented in the weekly complex-trauma admission list.
Medical and psychiatric opinions were available regarding the suitability of potential
participants in the event of any uncertainty or disagreement regarding their
competence or their ability to participate. During the agreed 18-month prospective
recruitment period (3rd January 2010 to 30th June 2011) every complex trauma
patient admitted to the DMRC during their first cycle of inpatient care who met the
selection criteria was approached by the Headley Court Fellow (HCF) in order to
further assess their eligibility. It is a considerable strength of the study design that
every patient admitted to the complex trauma patient group at the DMRC was
18
assessed and, if meeting the selection criteria, invited to participate. The study design
incorporated face-to-face recruitment of consecutive admissions in order to facilitate
maximum participation, maximum response, and to minimise problems associated
with selection bias.
Eligible potential participants were approached within the first week of their
initial admission to the DMRC and the purpose and nature of the research project was
explained to them. During this initial approach, potential participants were provided
with a patient-information leaflet in a ‘commonly asked questions’ format in support
of this initial briefing and were provided with the opportunity to ask questions.
Eligible potential participants were subsequently left for a minimum of 24 hours to
consider their participation before being approached a second time. Those eligible
potential participants willing to participate at second contact with the HCF were taken
through the consent process, enrolled into the study and provided with the baseline
assessment document.
Patient Group (Retrospective cohort)
Potential retrospective participants were initially identified through the Defence
Patient Tracking System operated by the Defence Analytical Services Agency
(DASA) and Joint Medical Command (JMC). The contact information for potential
retrospective participants meeting the study selection criteria was then requested from
the Service Personnel and Veterans Agency (SPVA) and each potential participant
was subsequently written to by the HCF, informed about the purpose of the study and
invited to participate. Letters of invitation were accompanied by patient information
leaflets, consent forms and self-addressed envelopes to allow the return of the
appropriate documentation. Retrospective patient participants returning completed
consent forms and patient information documents were then sent the baseline
assessment document by post.
Partner Group (Prospective and retrospective cohorts)
Potential partner participants in both prospective and retrospective cohorts were
identified through their existing relationship with index patient participants. During
the consent process, patient participants were asked to identify whether or not they
were in a ‘committed relationship’ and whether or not they would consent to their
partner being approached by the HCF in order to invite them to participate in the
partner study. For prospective patient participants the request was made during the
19
face-to-face consent process, and for retrospective patient participants the request was
made by post. Where patient participants consented to their partner being
approached, contact details were obtained and letters of invitation accompanied by
partner information leaflets, consent forms and self-addressed envelopes, to allow the
return of the appropriate documentation, were sent to potential partner participants.
Data collection
Following their identification, consent and recruitment, a three staged follow up of
participants was undertaken using a comprehensive multimodal participant
assessment previously piloted by the HCF at the DMRC6. Each patient and partner
participant was assessed initially at baseline and then again at six months and 12
months using one of 12 standardised questionnaires (see Table 4: Participant
assessments). Each of the participant questionnaires consisted of a number of
validated objective measures and supporting questions and, with the exception of the
baseline assessment of the prospective patient group, was administered as a postal
self-report questionnaire. In order to minimise participant burden and to avoid
causing unnecessary distress only those measures deemed essential in order to fulfil
the aims of the study were included.
Table 4: Participant assessments
Patient Study (01) Partner Study (02)
Prospective
Cohort (PG)
Retrospective
Cohort (RG)
Prospective Cohort
(PG)
Retrospective
Cohort (RG)
Baseline
Assessment (BA) 01-PG-BA 01-RG-BA 02-PG-BA 02-RG-BA
Six-month
Assessment (6M) 01-PG-6M 01-RG-6M 02-PG-6M 02-RG-6M
12-month
Assessment (12M) 01-PG-12M 01-RG-12M 02-PG-12M 02-RG-12M
Participant assessments
In order to comprehensively answer the stated research questions the study design
made use of a combination of clinical outcome and standardised self-report data.
Clinical data obtained through outcome measure assessments conducted at the DMRC
during day to day practice were utilised in order to assess spinal injury, mobility, pain,
6 The baseline assessment document was piloted at the DMRC in December 2009 with a representative sample of twenty-five
complex-trauma patients with both combat-related and musculoskeletal injury. The pilot study consisted of individuals from all
three Services, male and female participants, and UK and Foreign and Commonwealth personnel. The response rate of the pilot study was 60%, and a number of revisions to the proposed assessment documents were made in response to comments from
participants and clinicians and the analysis of completed pilot assessment documents.
20
balance, anxiety and depression, quality of life and functional capability in patient
participants. Data regarding injury severity and neurological status of participants
following their injury were obtained through the Academic Department of Military
Emergency Medicine (ADMEM). Self-report data, collected during the 12-month
follow-up phase, provided information on participants’ socio-demographic
background, military experience and exposure to traumatic events as well as self-
reported physical health, resilience, perceived social support, general trauma-related
psychopathology and quality of life.
Clinical outcome data
The systematic review of medical, nursing, occupational therapy and physiotherapy
case notes in addition to the interrogation of a number of electronic sources, i.e. the
Defence Medical Information Capability Programme (DMICP) and the UK Joint
Theatre Trauma Registry (JTTR) facilitated the collection of a range of clinical
outcome data. A summary of the clinical outcome measures is to be found in Figure
1: Summary of clinical outcome data. The clinical outcome data was collected at
baseline, six months and at 12 months, or during the nearest admissions to those
dates.
Figure 1: Summary of clinical outcome data
DASA/JTTR Data
o Injury severity (ISS, NISS)
o Head injury (GCS)
DMRC Clinical Outcome Data
o ASIA Score (Spinal cord injury)
o Mobility Scores
SIGAM mobility grades,
Six-minute timed walk,
Stair Assessment,
AMPQ
o Pain Score
o Balance
Fregley & Graybiel balance, Y balance test
o Mental health
Hospital Anxiety and Depression Scale (HADS)
o Quality of life
SF-36, Independence in Activities of Daily Living;
o Functional performance
FIT-HaNSA,
Return to driving,
Return to work
o Patient satisfaction
21
Self-report data
Self-report data were collected through the administration of the participant
questionnaires described above. The objective measures used within the standardised
questionnaires were intended to provide a comprehensive assessment of participant
health and well-being and facilitated the collection of: socio-demographic data;
physical health data; resilience and social support data as well as data regarding
general and trauma-related psychopathology, quality of life and caregiver strain. A
summary of these measures and an example of their use is contained in Table 5:
Participant assessment schedule and summary of self-report measures (Patient
Group).
Socio-demographic variables
The collection of socio-demographic data was achieved through the administration of
a modified version of the Accident Assessment Interview Schedule (AAIS), (Alexander
et al, 1995) and the Partner Semi-Structured Interview (PSSI), (Klein & Alexander,
2005). These measures were incorporated into the baseline assessment of both patient
and partner participants, and were intended to provide information on: the socio-
demographic characteristics of participants; the specific details of the incident and of
the injury sustained, as well as the reactions of participants to their injury and specific
predictors of general psychopathology.
As patient participants are, or have been, members of a military cohort,
permission was obtained from the King’s Centre for Military Health Research
(KCMHR)7 to incorporate sections of the Health & wellbeing survey of serving & ex-
serving members of the UK Armed Forces: Phase 2 within the participant
assessments. Utilisation of elements of the ‘Phase 2’ study, including those in respect
of participants’ previous deployment history, experience of operations and exposure
to traumatic events, provided a sound ‘military context’ for this study and helped to
ensure the comparability of results with the existing KCMHR database.
Physical health variables
The Physical Symptom Checklist (PSC), (Unwin et al, 1999) and the Chalder Fatigue
Scale (CFS), (Chalder et al, 1993) were included in all participant assessments in
order to assess somatisation and to measure fatigue symptoms across both studies and
all groups of participants.
7 Approval has been obtained from Professor Simon Wessely (Director, KCMHR) to incorporate elements of the Phase 2 study
questionnaire.
22
Resilience variables
‘Resilience’ refers to positive adaptation in the face of stress or trauma (Luthar et al,
2000), and two standardised measures thereof have been employed within the study
design to assess participant resilience.
Dispositional Resilience Scale-15 (DRS-15), (Bartone, 2007).
This brief 15-item scale was developed in order to measure the three
personality traits of hardiness, i.e. commitment, control and challenge, and has
been included in the baseline assessment of patient participants.
Connor-Davidson Resilience Scale-10 (CDRIS-10), (Connor & Davidson,
2003). This 10-item scale was developed from the original 25-item self-report
scale that measures the ability to cope with adversity and is used within each
of the participant assessments.
Social support variables
In order to measure perceived social support in both patient and partner groups the
Social Support Inventory (SSI), (Holeva et al, 2001) has been included within patient
and partner participant assessments. The social support inventory is a 5-item measure
of perceived social support based on an expanded version of the Perceived Criticism
Scale (Hooley & Teasdale, 1989).
General and trauma-related psychopathology variables
The collection of data regarding participants’ general mental health and, more
specifically, any symptoms of PTSD was achieved through the administration of the
General Health Questionnaire (GHQ-12), (Goldberg & Williams, 1988) and the
PTSD Checklist-Civilian Version (PCL-C), (Blanchard et al, 1996).
General Health Questionnaire (GHQ-12; Goldberg & Williams, 1988)
This 12-item version has been widely used to measure symptoms of common
mental disorder among individuals in community settings and in non-
psychiatric settings.
PTSD Checklist-Civilian Version (PCL-C; Blanchard et al, 1996)
The PCL-C measure is a 17-item checklist that can be used to screen
individuals for PTSD, as a diagnostic tool, or to monitor symptom change
during and after treatment. While a dedicated military version of the PCL
exists, the civilian version (PCL-C) has been utilised within this study to
23
facilitate the direct comparison of symptoms of PTSD in both patient and
partner groups and to ensure the compatibility of collected data with the
existing KMHCR database.
Quality of life and lifestyle variables
The Short Form-36 health survey measure (SF-36), (Ware et al, 1993) constitutes a
generic self-report measure of quality of life that has been widely used to provide an
assessment that is not specific to age, disease or treatment group (Fayers & Machin,
2001). As a measure, it places emphasis upon physical, social, and emotional
functioning.
While the primary measure of quality of life variables remains the Short Form-
36 measure (SF-36), the study design makes use of a number of other lifestyle
variables including a brief measure of alcohol consumption to assess the extent and
pattern of alcohol misuse (Alcohol Use Disorders Identification Test (AUDIT; Babor
et al, 2001)).
Caregiver burden variables
Caregiver Strain Index-Modified (CSI-M), (Robinson, 1983) is a 16-item index used
to identify the level of strain experienced by partners in relation to the provision of
care for the patient in respect of the, emotional, financial, physical, social and time
‘costs’ and strains. Each item is framed to ensure that any change is attributable to the
patient’s injury.
24
Table 5: Participant assessment schedule and summary of self-report measures (Patient Group)
Baseline Assessment Six-month Assessment 12-month Assessment
PG/RG Prospective
Cohort
Retrospective
Cohort
Prospective
Cohort
Retrospective
Cohort
Prospective
Cohort
Retrospective
Cohort
Clinical Variables DMRC DMRC
Socio-demographic
variables
AAIS/KCMHR
(Phase 2)
AAIS/KCMHR
(Phase 2)
Physical health
variables
PSC
CFS
PSC
CFS
PSC
CFS
PSC
CFS
PSC
CFS
PSC
CFS
Resilience
variables
DRS-15
C-DRIS-10
DRS-15
C-DRIS-10 C-DRIS-10 C-DRIS-10 C-DRIS-10 C-DRIS-10
Social support
variables SSI SSI SSI SSI SSI SSI
General and trauma-
related
psychopathology
variables
GHQ-12
PCL-C
GHQ-12
PCL-C
GHQ-12
PCL-C
GHQ-12
PCL-C
GHQ-12
PCL-C
GHQ-12
PCL-C
Quality of Life
variables SF-36* SF-36 SF-36 SF-36 SF-36 SF-36
*SF-36 included with baseline assessment of retrospective cohort only
Research governance
By virtue of the fact that this study involved participants who may have suffered
traumatic injury, and investigated topics about which some participants may have
been sensitive, embarrassed or upset, it is recognised that participation in this project
may be harmful. At each stage, measures were taken to minimise the potential risk to
participants and the study was conducted in full compliance with the Research Ethics
Policy of The Robert Gordon University (RGU). The study was reviewed by the
Surgeon General’s Research Strategy Group and has been granted full ethical
approval by both the Ministry of Defence Research Ethics Committee (MODREC)8
and the National Research Ethics Service – North of Scotland (NRES) Committee9.
8 Ethical approval from the Ministry of Defence Research Ethics Committee (MoDREC) was granted on the 1st October 2009
(MoDREC references, 62/Gen/09 (Patient study) and 63/Gen/09 (Partner study)). 9 A favourable ethical opinion was granted by the National Research Ethics Service Committee – North of Scotland (NRES) on
the 29th July 2009 (NRES reference number 09/S0801/47).
25
Participant consent
Informed written consent was obtained from each participant prior to their enrolment
in the study. For the purpose of the study, written consent included permission from
the participant to access their medical and resettlement records and, where
appropriate, to contact their partner. During the induction process, participants were
provided with a verbal briefing, informed as to the nature, aims and risks of the study
and provided with a detailed patient information sheet. Participants were reminded
that they were entitled to ask questions on any aspect of the study that may occur to
them at any time during their participation.
Potential pressures
As potential patient participants were currently, or had previously been, in receipt of
treatment from the DMRC and because they may continue to serve in the armed
forces, it was feasible that some individuals might feel compelled to participate in the
study. Participants were informed, therefore, that although the study was being
conducted in collaboration with the DMRC and was being funded and supported by
the Headley Court Trust, the researchers remained independent of these bodies and
that participation in the study was entirely voluntary. Participants were reminded also
that they would retain the right to withdraw their consent at any time during the study
without prejudicing either their medical care or their military career (or medical care
or military career of their partner).
Confidentiality, Anonymity and Data Storage
Participants were assured that their responses would remain confidential throughout
and that on completion of the study; no personally identifying information would be
released into the public domain. Participants were informed that no individual would
be identified in reports or publications and that all findings would be reported on the
basis of aggregated data.
As participants were asked to consent to the storage of personal information it
was also made clear that any data collected would be used only for the purpose of the
current study and that their information would be handled in accordance with the Data
Protection Act 1998.
Computerised data (which may be militarily sensitive and/or clinically
confidential) have been stored on an encrypted hard-disk that remains accessible only
by authorised members of the research team. As MoDREC requires the retention of
the original hard copy of participant assessments for a period of not less than 100
26
years, hard copy information is stored in a locked cupboard within the secure ACTR
site which is accessible only by designated RGU staff.
Data analysis
The statistical analysis of data collected through the participant assessments was
carried out using the software package SPSS for Windows (Version 17). The primary
outcome (dependent) variables for both patient and partner study were the GHQ-12,
PCL-C, and the SF-36, and the analysis of these measures fulfilled the primary
descriptive aim of the study (the estimate of the prevalence of general and trauma-
related psychopathology and quality of life and its variation by group, socio-
demographic characteristics, clinical, psychosocial and resilience status).
A series of univariate analyses (such as the paired t-test or non parametric
equivalents if deemed appropriate) were conducted to identify the variables most
closely associated with outcome in terms of group and postulated risk factors.
27
Results
Recruitment
Patient Group (Prospective cohort)
Between the 3rd of January 2010 and the 30th of June 2011 details of 234 potential in-
patient admissions to the DMRC were published by the complex-trauma team.
During the recruitment phase 224 individuals were admitted to the complex-trauma
group. Non-admission (n=10) was as a result of: improvement in patient condition
following the initial review by the DMRC consultant; referral to a regional
rehabilitation centre, or to a facility other than the DMRC, or death at the Queen
Elizabeth Hospital in Birmingham during acute management.
An initial assessment, by the HCF, of those presenting as in-patients at the
DMRC resulted in the exclusion of a further ten potential participants due to their
unsuitability or failure to meet the selection criteria (e.g. those presenting at the
DMRC following elective procedures, as medical admissions, or those admitted to an
rehabilitation group other than to complex-trauma or by virtue of their being non UK
nationals or civilians).
Two-hundred and fourteen potential patient participants were approached by
the HCF and invited to participate in the study. One-hundred and ninety-nine
individuals (92.99%) consented to participate (see Figure 2: Participant recruitment).
Patient Group (Retrospective cohort)
The Defence Patient Tracking System (DPTS) provided details of 126 individuals,
previously treated at the DMRC, who met the study selection criteria. The Service
Personnel and Veterans Agency (SPVA) provided contact information for each of
those 126 individuals, and they were contacted by post by the HCF and invited to
participate in the study. From the initial tranche of 126 potential retrospective patient
participants, 20 (15.87%) responded to the invitation either on initial contact or
following a subsequent reminder letter, and were enrolled on to the study.
Partner Group (Prospective and retrospective cohorts)
Contact information for potential partner participants was obtained through index
patient participants enrolled in both the prospective or retrospective cohorts. Within
the prospective patient cohort (n=199), 139 (69.85%) individuals identified
themselves as being in a ‘committed relationship’ and 107 (76.98%) of them
28
consented to having their partner approached10. Within the retrospective patient
cohort, of those responding to the invitation to participate in the study, 14 (70.00%)
consented to having their partner approached.
Following the identification of potential prospective and retrospective partner
participants, 121 invitations to participate in the study were posted by the HCF.
Following either the initial approach or a subsequent reminder letter, 30 (24.79%)
potential partner participants consented to participate and were enrolled onto the
study. A summary of the participant recruitment for all four main study groups is
presented in Figure 2: Participant recruitment.
10 Sixty of the 199 enrolled prospective participants identified themselves as being single (30.15%) when they were initially
approached by the HCF and 32 (16.08%) did not consent to the approach of their partner.
Potential prospectivepatient participants
(n=234)
Non-admissions to Complex-Trauma
Group
(n=10)
Admitted to Complex-Trauma Group
(n=224)
Excluded
(n=10)
Approached
(n=214)
01PG
Consented
(n=199)
Not in a 'committed relationship'
(n=60)
In a 'committed relationship'
(n=139)
Refused consent to approach partner
(n=32)
Consented to partner being approached
(n=107)
Potential prospective partner participants
approached
(n=107)
02PG
Consented
(n=27)
Did not repond to invitation
(n=80)
Refused
(n=15)
Potential retrospectivepatient participants
(n=126)
Potential retrospective patient participants
approached
(n=126)
Responded to invitation
(n=20)
01RG
Consented
(n=20)
Not in a 'committed relationship'
(n=6)
In a 'committed relationship'
(n=14)
Consented to partner being approached
(n=14)
Potential retrospective partner participants
approached
(n=14)
02RG
Consented
(n=3)
Did not respond to invitation
(n=11)
Refused consent to approach partner
(n=0)
Did not respond to invitation
(n=106)
Retrospective Cohort Prospective Cohort
Figure 2: Participant recruitment
29
Response Rates
During the 12-month follow-up phase, participant assessments in both patient and
partner studies were administered at baseline (BA), at six months (6M) and again at
12 months (12M). With the exception of the prospective patient participant baseline
assessment (01PGBA), which was administered directly at the DMRC, the participant
assessments were administered by post. A summary of response rates for all groups is
presented in Table 6: Response rates (all groups).
Patient Group (Prospective cohort)
The response rate on the prospective patient baseline assessment (01PGBA) was
good, and achieved a return of 56.28% (n=112). The postal administration of the six-
month (01PG6M) and 12-month (01PG12M) prospective patient assessments,
however, achieved a significantly lower response. Of the six-month assessments
administered (n=199) only 14.57% (n=29) were returned. Expressing the response
rate of the 01PG6M assessments as a percentage of number of those completing the
original 01PGBA assessment (n=112), improves the rate to 25.89%. The response
rate for the 01PG12M assessment, expressed as a percentage of the total number of
12-month assessments administered (n=199), is 11.56% (n=23) and as a percentage of
those completing the 01PGBA assessment (n=112), 20.54%.
Patient Group (Retrospective cohort)
The response rate on the retrospective patient baseline assessment (01RGBA) was
also good, achieving a return of 65.00% (n=13). In contrast to the prospective patient
baseline assessment (01PGBA) the baseline assessment for the retrospective group
(01RGBA) was administered by post in the same way as the six-month and 12-month
follow-up assessments for that group. The administration of the six-month
(01RG6M) and 12-month (01RG12M) retrospective patient assessments also achieved
a better response rate than those of the prospective patient study, albeit in much lower
numbers. Of the 20 six-month assessments administered by post eight were returned.
The response rate for the 01RG12M assessment was 25.00% (n=5).
Partner Group (Prospective and retrospective cohorts)
The response rate to the partner baseline assessments in both prospective (02PGBA)
and retrospective cohorts (02RGBA) was 70.37% (n=19) in prospective partners and
66.66% (n=2) in retrospective partners. The administration of the six-month
(02RG6M) and 12-month (02RG12M) prospective partner assessments was 44.44%
30
(n=12) and 37.04% (n=10) respectively. Within the retrospective partner group there
were no six-month (02RG6M) or 12-month (02RG12M) returns.
Table 6: Response rates (all groups)
Patient Study (01) Partner Study (02)
Prospective Group
(PG)
Retrospective
Group (RG)
Prospective Group
(PG)
Retrospective
Group (RG)
Baseline
Assessment (BA)
Administered 199* 20 27 3
Returned (Rate) 112
(56.28%)
13
(65.00%)
19
(70.37%)
2
(66.66%)
Returns as % of
previous N/A N/A N/A N/A
Six-month
Assessment (6M)
Administered 199 20 27 3
Returned (Rate) 29
(14.57%)
8
(40.00%)
12
(44.44%)
0
(0.00%)
Returns as % of
previous 25.89% 61.53% 63.16% 0%
12-month
Assessment
(12M)
Administered 199 20 27 3
Returned (Rate) 23
(11.56%)
5
(25.0%)
10
(37.04%)
0
(0.00%)
Returns as % of
previous 79.31% 62.50% 83.33% 0%
*Baseline Assessment (BA) for the Patient Study Prospective Group was administered at the DMRC
Demographics
Ethnicity
The majority of patient participants enrolled in the patient study (both prospective and
retrospective cohorts) identified themselves as being ‘White: British’ (89.9%, n=197).
The remainder of the participants identified themselves as being either ‘White: Irish’
(2.5%, n=5), ‘White: Other’ (1.5%, n=3,) or as being either ‘Asian or Asian British’
(3.0%, n=6) or ‘Black or Black: British: African’ (3.6%, n=8). The distribution of
ethnic origin was roughly in line with that reported by the DASA (2012) in respect of
UK Regular Forces at 1st April 2011 where, for the NATO ranks of OF3 and below,
91.36% (n=164,980) were identified as being ‘White’ and 6.71% were reported as
being ‘Black and Minority Ethnic’. Those participants reporting their ethnicity as
‘White: Other’ were exclusively South African and, while the use of UK Census
categories for ethnicity is not particularly useful in identifying Nepalese participants,
the majority of those identifying themselves as ‘Asian or Asian British: Other’ were
31
Ghurkha participants (2.28%, n=5). A summary of patient participant ethnicity is
presented in Table 7: Ethnicity of participants (01PG) and (01RG).
Table 7: Ethnicity of participants (01PG) and (01RG)
Frequency Percent Valid Percent Cumulative Percent
White: British 197 89.9 89.9 89.9
White: Irish 5 2.3 2.3 92.2
White: Other 3 1.4 1.4 93.6
Asian or Asian British: Other 6 2.7 2.7 96.4
Black or Black British: African 8 3.6 3.6 100.0
Total 219 100.0 100.0
Within the partner study, 86.6% (n=26) of participants identified themselves as
‘White: British’. One partner participant identifying herself as being ‘Asian or Asian
British: Other’ and was Nepalese in origin and the partner of a Ghurkha patient
participant. A summary of partner participant ethnicity is presented in Table 8.
Table 8: Ethnicity of participants (02PG) and (02RG)
Frequency Percent Valid Percent Cumulative Percent
White: British 26 86.6 86.6 86.60
White: Irish 1 3.33 3.33 89.93
White: Other 1 3.33 3.33 93.26
Asian or Asian British: Other 1 3.33 3.33 96.59
Black or Black British: African 1 3.33 3.33 100.0
Total 30 100.0 100.0
Gender
The vast majority of participants within the patient study were male (98.6%, n=216).
Only three female participants took part in the patient study (1.4%). This distribution
of gender is not representative of the overall distribution of gender in UK Regular
Forces11 but is more likely to be representative of the gender distribution of those
undertaking combat duties (see Table 9: Gender and incident type). Only one of the
female participants in the patient study sustained injuries on military operations and
those injuries were sustained during a mortar attack on Shaibah Logistics Base (SLB)
rather than on foot patrol or while deployed in forward areas (Patrol Base (PB) or
Forward Operating Base (FOB)).
11 At 1 April 2011 there were 168,500 (90.4 per cent) male and 17,850 (9.6 per cent) female UK Regular Forces personnel.
32
Table 9: Gender and incident type (01PG) and (01RG)
CDS3 - Type of incident
Total Combat related Non-combat-related
CDS1 - Gender Male 179 30 209
Female 1 2 3
Total 180 32 212
Participants within the partner study were exclusively female. While there were three
female participants in the patient study, only one of them consented to the HCF
contacting their partner and, in this case, the male partner did not respond to the
invitation letter or subsequent follow-up.
Age
Patient participants ranged in age between 19 and 56 years on first admission to the
DMRC (range=37 years, minimum age 19 years and maximum age 56 years.). The
mean age of patients at first admission to the DMRC was 26.33 years (see Table 10:
Age at first admission to the DMRC (01PG) and (01RG)). At 1st April 2011, the
mean age of those serving in the UK Regular Forces as ‘Other Ranks’ was 29 years
and as ‘Officers’, 37 years of age. The mean age of patient participants is noticeably
lower therefore than that of the general UK Regular Forces figure, but this is most
likely a reflection of the ‘junior’ nature (in both age and rank) of those deployed in
forward areas on operations. A summary of age and age distribution is presented in
Figure 3: Age at first admission to DMRC (01PG) and (01RG).
Table 10: Age at first admission to the DMRC (01PG) and (01RG)
N Valid 210
Missing 9
Mean 26.33
Range 37
Minimum 19
Maximum 56
Within the partner study (02PG) and (02RG) the mean age of participants was 28.4
years (range=25 years, minimum age 19, maximum age 44).
33
Figure 3: Age at first admission to DMRC (01PG) and (01RG)
Relationship Status
Overall, within both prospective and retrospective patient cohorts 125 participants
(64.8%) regard themselves as being in ‘committed relationship’ at baseline and either
describe themselves as being ‘married’ (27.2%, n=34), ‘living with a partner’ (10.4%,
n=13) or ‘in a long-term relationship’ (27.2%, n=34). The remaining 35.2% of
participants (n=44) report being ‘Single’, ‘Separated’ or ‘Divorced’. A summary of
the relationship status of patient participants is presented in Table 11: Relationship
status (01PG) and (01RG) (n=226) and the distribution of relationship status of
participants (less those failing to report the relationship status) is presented in Figure
4: Relationship status (01PG) and (01RG) (n=125).
Table 11: Relationship status (01PG) and (01RG) (n=226)
Relationship Status Frequency Percent Valid Percent Cumulative Percent
Valid
Married 34 15.0 27.2 27.2
Living with a partner (cohabiting) 13 5.8 10.4 37.6
In a long term relationship 34 15.0 27.2 64.8
Single (never married) 36 15.9 28.8 93.6
Separated 4 1.8 3.2 96.8
Divorced 4 1.8 3.2 100.0
Total 125 55.3 100.0
Missing System 101 44.7
Total 226 100.0
34
Figure 4: Relationship status (01PG) and (01RG) (n=125)
Financially dependent children
The definition of ‘financially dependent children’ is difficult and does not rely solely
upon residential status. Participants were asked, firstly, whether or not they regarded
themselves as having any financially dependent children. If they answered ‘yes’ they
were then asked: how many financially dependent children they had; their ages; their
gender and whether or not the children lived with them. The majority of patient
participants reported that they did not have financially dependent children (74%,
n=91). The numbers of children reported by the remaining 26% (n=32) of
participants varied from one child to six children (mean number of children per
family=1.81). The ages of the reported children ranged from two weeks to 23 years
of age, with a mean age of 6.5 years. A summary of ‘family size’ and distribution of
financially dependent children is presented in Table 12: Financially dependent
children (01PG) and (01RG).
35
Table 12: Financially dependent children (01PG) and (01RG)
Frequency Percent Valid Percent Cumulative Percent
Valid
0 91 41.6 74.0 74.0
1 15 6.8 12.2 86.2
2 12 5.4 9.8 96.0
3 3 1.4 2.4 98.4
4 1 .5 .8 99.2
6 1 .5 .8 100.0
Total 123 56.2 100.0
Missing System 96 43.8
Highest level of education
Within the patient study, participants were asked to report their highest level of
education attainment. The majority of participants (70.2%, n=87) reported that the
completion of secondary school was their highest level of education. Of those
reporting the completion of secondary school as their highest level of educational
achievement, almost one quarter reported leaving school with A-Level or equivalent
qualifications (24.1%, n=21).
Figure 5: Highest level of educational achievement (01PG) and (01RG)
Overall, 29.8% of patient participants reported completing post-secondary education,
18.5% (n=23) completed further education gaining a college certificate, diploma or
equivalent qualification, and 10.5% (n=13) completed a programme of higher
education to certificate (0.8%, n=1), degree (8.9%, n=11) or to higher degree (0.8%,
n=1) level. A significant percentage of participants (13.7%, n=17) left secondary
school prior to sixteen years of age with no formal qualification. A summary of these
36
results can be found in Figure 5: Highest level of educational achievement (01PG)
and (01RG) and Table 13: Highest level of educational achievement (01PG) and
(01RG).
Table 13: Highest level of educational achievement (01PG) and (01RG)
Frequency Percent Valid Percent Cumulative Percent
Valid Left school before age 16 (No
formal qualifications)
17 7.8 13.7 13.7
Left school with GCSE/CSE/O-
Level or equivalent
49 22.4 39.5 53.2
Left school with A-Level or
equivalent qualification(s)
21 9.6 16.9 70.2
College certificate or diploma 22 10.0 17.7 87.9
University degree 11 5.0 8.9 96.8
Higher university degree (e.g. MSc,
PhD)
1 .5 .8 97.6
City and Guilds Certificate 1 .5 .8 98.4
Overseas Qualification (Nepal) 1 .5 .8 99.2
Certificate of Higher Education 1 .5 .8 100.0
Total 124 56.6 100.0
Missing System 95 43.4
Total 219 100.0
Within the partner study, 23.8% (n=5) of participants reported the completion of
secondary school as their highest level of educational achievement. Of those partners
who reported completion of secondary school as their highest level of educational
achievement only one participant reported having left school with A-Level or
equivalent qualifications (n=1, 4.8%).
37
Figure 6: Highest level of educational achievement (02PG)
In contrast to the patient participants, a higher percentage of partner
participants (51.8%, n=14) reported further education or higher education as their
highest level of educational achievement. Within this group, a significantly higher
number of participants identified completion of higher education to degree level
(38.1%, n=8) or to post-graduate certificate level (4.8%, n=1). While none of the
partner participants reported having left secondary school without any formal
qualification, one participant reported the completion of GCSE/CSE/O-Level or
equivalent in Further Education.
Table 14: Highest level of educational attainment (02PG)
Frequency Percent Valid Percent Cumulative Percent
Valid Left school with GCSE/CSE/O-
Level or equivalent
4 14.8 19.0 19.0
Left school with A-Level or
equivalent qualification(s)
1 3.7 4.8 23.8
College certificate or diploma 5 18.5 23.8 47.6
University degree 8 29.6 38.1 85.7
Overseas Qualification (Nepal) 1 3.7 4.8 90.5
GCSE/CSE/O-Level or equivalent
in Further Education
1 3.7 4.8 95.2
Postgraduate Certificate 1 3.7 4.8 100.0
Total 21 77.8 100.0
Missing System 6 22.2
Total 27 100.0
38
Employment status (02PG) and (02RG)
Partner participants were asked about their current occupation, employment status and
whether or not they worked full- or part-time (Table 15: Employment status (02PG)
and (02RG)). The majority of partner participants indicated that they were employed
full-time (67%, n=14) and a significant proportion of those partners further reported
working in a professional or managerial role (64%, n=9). Four partner participants
identified themselves as being in full-time education (19%) while three reported
working in a semi-skilled role. The remaining number of partner participants reported
being either unemployed (n=2) or being employed as a house-person.
Table 15: Employment status (02PG) and (02RG)
Do you work (or study)?
Total Full-Time Part-Time Not Applicable
What is your employment
status?
Professional/Managerial 9 1 0 10
Semi-Skilled 1 2 0 3
Student 4 0 0 4
House-person 0 0 2 2
Unemployed 0 0 2 2
Total 14 3 4 21
Military Service
Both prospective and retrospective patient participants were asked if they were
currently serving in the Armed Forces and, if they were, to which branch of the
Armed Forces they belonged (see Table 16: Military service by type). Within the
patient study, those in the prospective cohort, as first time in-patient admissions to the
DMRC, were all currently serving in the Armed Forces. The retrospective cohort, by
contrast, having been recruited six to 24 months after discharge from the DMRC,
consisted a number individuals who were ‘ex-forces’ (2.4%, n=3) in addition to those
individuals who where currently still in the Armed Forces. Of those participants still
serving in the Armed Forces, 91.2% (n=114) were either in regular service or on full-
time reserve service (FTRS). Members of the Reserve Forces, i.e. Territorial Army
(TA), Royal Auxiliary Air Force (RAuxAF), Royal Naval Reserve (RNR) or Royal
Marine Reserve (RMR) formed 6.4% (n=8) of the overall patient participant sample.
Those patient participants identifying themselves as members of the Reserve Forces
were all members of the Territorial Army.
39
Figure 7: Military service by type
Table 16: Military service by type
Frequency Percent Valid Percent Cumulative
Percent
Valid A regular or on full-time reserve
service (FTRS)
114 50.4 91.2 91.2
A volunteer reserve (e.g. TA,
RAuxAF, RNR or RMR (mobilized
or not))
8 3.5 6.4 97.6
Ex forces 3 1.3 2.4 100.0
Total 125 55.3 100.0
Missing System 101 44.7
Total 226 100.0
Branch of service
All branches of the Armed Forces were represented in the study (see Table 17: Branch
of service). For the purposes of the report, and to ensure comparability with the
existing KMHCR ‘Phase 2’ dataset, the Army and the Royal Marines were regarded
as one group and constituted the majority of participants with complex trauma injuries
(96%, n=120). Members of the Royal Navy represented 1.6% (n=2) of those with
complex-trauma injuries (all non-combat-related). Royal Air Force personnel
belonging to the Royal Air Force Regiment presented with a combination of combat-
related and non-combat-related injury and constituted 2.4% of the sample (n=3).
40
Figure 8: Branch of service
Table 17: Branch of service
Frequency Percent Valid Percent Cumulative Percent
Valid Royal Navy 2 .9 1.6 1.6
Army and Royal Marines 120 54.8 96.0 97.6
Royal Air Force 3 1.4 2.4 100.0
Total 125 57.1 100.0
Missing System 94 42.9
Total 219 100.0
Current rank
The rank range of those serving in the Army and Royal Marines was Private (Pte),
(NATO rank OR1) to Captain (Capt), (NATO rank OF2). While, as is traditional,
Private Soldiers of different cap badges and regiments (as well as of different Corps,
trades or occupations) report their rank differently depending on their Regimental or
Corps affiliations. The rank of Private Soldier was used to describe individuals
holding rank equivalent of NATO rank OR1, e.g. Gunner, Drummer, Ranger, Marine,
Rifleman etc. Within the Army and Royal Marine grouping 44.0% (n=55) of
participants reported their rank as being equivalent to Private Soldier. Junior Non-
Commissioned Officers (JNCOs’, i.e. Lance Corporals and Corporals) constituted
34.4% (n=43) of the Army and Royal Marine group and Senior Non-Commissioned
Officers (SNCOs’, Sergeant to Warrant Officer Class 1) constituted 10.4% (n=13).
41
Junior Officers (Second Lieutenant (2Lt) to Captain (Capt)) made up 8.8% (n=11) of
the sample (see Table 18: Current rank).
Royal Naval participants (n=2) were ranked Petty Officer (PO) and Warrant
Officer Class 1 (WO1) and comprised 1.6% of the sample and the Royal Air Force
participants reporting their rank between Senior Aircraftman (SAC) and Lance
Corporal (LCpl).
Figure 9: Current rank (or rank at completion of service)
Table 18: Current rank
Frequency Percent Valid Percent Cumulative Percent
Valid Pte 55 24.3 44.0 44.0
AC/LAC/SAC/JT 1 .4 .8 44.8
L/Cpl to Cpl 43 19.0 34.4 79.2
Sgt to WO1 13 5.8 10.4 89.6
2Lt to Capt 11 4.9 8.8 98.4
PO to WO1 2 .9 1.6 100.0
Total 125 55.3 100.0 100.0
Missing System 101 44.7
Total 226 100.0
Length of Service
Those patient participants reporting that they currently served as regular members of
the Armed Forces or that they were undertaking Full-time Reserve Service (FTRS)
had a mean length of service of 6.88 years (range from one to 23 years of service).
Those patient participants serving in the Reserve Forces had periods of service
42
ranging from two to 35 years of service and the mean length of service was 11.71
years.
The length of service for regular members of the Armed Forces and for those serving
on FTRS was biased toward the lower end of the range and 47.5% of patient
participants had served in the Armed Forces for five years or fewer and 20.3% had
served for two years or fewer.
Table 19: Length of service
Length of Service
(Regular/FTRS)
(YEARS)
Length of Service
(Reservist)
(YEARS)
N Valid 118 7
Missing 101 212
Mean 6.8814 11.7143
Range 22.00 33.00
Minimum 1.00 2.00
Maximum 23.00 35.00
Figure 10: Length of service (01PG) and (01RG) - Regular and FTRS
43
Figure 11: Length of service (01PG) and (01RG) - Reservists
Type and nature of incident
The majority of patient participants within the study were admitted to the DMRC as a
result of a combat-related injury (84.9%, n=180) and non-combat-related injury
accounted for the remaining 15.1% (n=32) of admissions where the type of incident is
known. Within the retrospective patient participant group (01RG) there were a
number of individuals who failed to report the type of incident causing their injuries
and where the non-availability of medical records (from which this information could
have been recovered) has resulted in unknown or ‘missing’ values (n=7, 3.2%). Table
20: Type of incident (01PG) and (01RG) and Figure 12: Type of incident (01PG) and
(01RG)) present the distribution of combat-related and non-combat-related injury.
Table 20: Type of incident (01PG) and (01RG)
Frequency Percent Valid Percent Cumulative Percent
Valid Combat related 180 82.2 84.9 84.9
Non-combat-related 32 14.6 15.1 100.0
Total 212 96.8 100.0
Missing System 7 3.2
Total 219 100.0
44
Figure 12: Type of incident (01PG) and (01RG)
Closer examination of the nature of the incident indicates that 68.4% (n=145) of
patient participants (both 01PG and 01RG) sustained combat-related injuries resulting
from explosion/blast causes, i.e. Improvised Explosive Device (IED), Rocket
Propelled Grenade (RPG), Anti-Personnel Mine (APM) or through Grenade strike.
Thirty five patient participants (16.5%) report having been injured as a result of a
Gunshot Wound (GSW). Within the non-combat-injury category incidents involving
vehicles (predominantly motorcycles) were the most common cause of injury (6.6%,
n=14). A summary of the reported causes of injury in the patient participant study is
presented in Table 21: Nature of incident (01PG) and (01RG) and in Figure 13:
Nature of incident (01PG) and (01RG).
45
Table 21: Nature of incident (01PG) and (01RG)
Frequency Percent Valid Percent Cumulative Percent
Valid Explosion / Blast (IED, RPG, APM,
Grenade)
145 66.2 68.4 68.4
Fragment 1 .5 .5 68.9
Bullet 35 16.0 16.5 85.4
Fall 5 2.3 2.4 87.7
Vehicle 14 6.4 6.6 94.3
Sport 7 3.2 3.3 97.6
Other: Alcohol Related 1 .5 .5 98.1
Other 4 1.8 1.9 100.0
Total 212 96.8 100.0
Missing System 7 3.2
Total 219 100.0
Figure 13: Nature of incident (01PG) and (01RG)
Range of injuries
The range of injuries sustained by patient participants admitted to the DMRC is broad
and diverse. In order to best describe the range of injuries for the purpose of this
report, participant injuries are identified using following categories: ‘Gunshot Wound’
(GSW); ‘Spinal Injury’ (SI); ‘Blast/Fragmentation Injuries’ (BFI); ‘Single Traumatic
Amputation’, ‘Double Traumatic Amputation’ and ‘Triple Traumatic Amputation’
(STA, DTA and TTA); ‘Other’ injuries, and ‘Unknown’.
The three largest categories of injury are single traumatic amputation
(44.22%), blast and fragmentation injuries (39.2%) and gunshot wounds (36.12%).
46
Double traumatic amputation accounts for 34.17% of the injuries sustained by the
study group and triple traumatic amputation accounts for a further 9.4% of injuries.
Spinal injuries and ‘Other’ injuries account for 33.62% of admissions to the DMRC
during the study period. The category ‘Other’ includes sporting injuries; injuries
related to Road Traffic Accident (RTA), burns, falls and crush injuries.
Figure 14: Injury category
Primary outcome variables – ‘independent samples’
The primary outcome variables identified for both patient and partner studies were the
General Health Questionnaire (GHQ-12), the civilian version of the Post-Traumatic
Checklist (PCL-C) and the Short-Form Health Survey (SF-36). While the SF-36
measure has been designed to capture data regarding a range of health concepts (both
physical and mental), it was identified that, in respect of the military rehabilitation
population, it lacked the sensitivity and specificity required to provide meaningful
results and the DMRC discontinued its clinical use during the early stages of this
research study. Consequently, the partial SF-36 dataset will be excluded from this
analysis.
The analysis of the remaining measures is intended to fulfil the primary
descriptive aims of the study and to provide an estimate of the prevalence of general
and trauma-related psychopathology and quality of life. The resilience status of
participants is to be estimated using the Dispositional Resilience Scale (DRS-15) and
the Connor-Davidson Resilience Scale (CDRIS-10).
36, 18%
44, 22%
34, 17%
9, 4%
39, 20%
10, 5%
23, 12%
4, 2%
Gunshot Wound
Single Traumatic Amputation
Double Traumatic Amputation
Triple Traumatic Amputation
Blast and Fragmentation Injuries
Spinal Injuries
Other
Unknown
47
PCL-C
The civilian version of the Post-Traumatic Checklist (PCL-C) was used at baseline, at
six months and again at 12 months to estimate the prevalence of PTSD within each of
the study groups. Within this study the PCL-C was used as a screening, rather than as
a diagnostic tool, and the total severity score for each participant was calculated and
compared against a predetermined normative threshold. For the purpose of this report
a cut-off score of 50 was used as the normative threshold and those participants
scoring 50 or more were taken to have met the screening criteria for PTSD. A cut-off
score of 50 was used to ensure consistence with the data reported by KMHCR.
Considering the baseline, six-month and 12-month assessments of the PCL-C as
independent samples provides an effective estimate of the prevalence of PTSD at each
stage.
Patient participants (01PG)
The baseline assessment of PCL-C for the prospective patient group (01PG) returned
a mean score of 31.27 (range=58, scores from 17 to 75). Of the total number of
participants completing the PCL-C assessment at baseline (n=108) 11 participants
achieved scores greater than, or equal to, 50 providing an estimated prevalence of
PTSD within the group of 10.18% (see Figure 15: PCL-C - Baseline (01PG)). While
the six-month assessment of PCL-C returned a higher mean score of 35.28 (range=65,
scores from 17 to 82) it did so for a considerably smaller number of participants
(n=29). Six patient participants achieved scores greater than, or equal to, 50 yielding
an estimated prevalence of 20.69% at six months (see Figure 16: PCL-C - Six months
(01PG)). At 12 months the total number of participants returning completed PCL-C
questionnaires was 23 and the mean PCL-C score was 37.82 (range=64, scores from
18 to 82). Four of the patient participants achieved scores greater than, or equal to, 50
providing an estimated PTSD prevalence of 18.18% at this stage (See Figure 17:
PCL-C - 12 months (01PG)).
Table 22: PCL-C – Baseline, six months and 12 months (01PG)
Baseline Six Months 12 Months
N Valid (Meets criteria)* 108 (11) 29 (6) 22 (4)
Missing 91 170 177
Mean 31.27 35.28 37.82
Range 58 65 64
Minimum 17 17 18
Maximum 75 82 82
*Screening criteria for PTSD defined as a score on PCL-L of ≥ 50
48
Figure 15: PCL-C - Baseline
(01PG)
Figure 16: PCL-C - Six months
(01PG)
Figure 17: PCL-C - 12 months
(01PG)
Partner participants (02PG)
The PCL-C was also used to estimate the prevalence of PTSD within the partner
group. The baseline assessment of PCL-C for the prospective partner group (02PG)
returned a mean score of 34.07 (range=33, scores from 21 to 54). Of the total number
of participants completing the PCL-C assessment in the partner group at this stage
(n=14) only one achieved a score greater than, or equal to, 50 providing an estimated
prevalence of PTSD of 7.14% (see Figure 18: PCL-C - Baseline (02PG)). The six-
month assessment of PCL-C in partners returned a mean score of 37.17 (range=41,
scores from 17 to 58). At six months however, the number of partner participants
meeting the screening criteria for PTSD increased to four, and yields an estimated
prevalence of 33.33% (Figure 19: PCL-C - Six months (02PG)). At 12 months the
mean PCL-C score in partner participants was 30.20 (range=41, scores from 17 to 58)
and the number of those meeting the screening criteria for PTSD had returned to one,
providing an estimated PTSD prevalence of 10.00% (Figure 20: PCL-C - 12 months
(02PG)). A summary for partner participants is presented in Table 23: PCL-C -
Baseline, six months and 12 months (02PG).
Table 23: PCL-C - Baseline, six months and 12 months (02PG)
Baseline Six Months 12 Months
N Valid (Meets criteria)* 14 (1) 12 (4) 10 (1)
Missing 12 14 16
Mean 34.07 37.17 30.20
Range 33 41 41
Minimum 21 17 17
Maximum 54 58 58
*Screening criteria for PTSD defined as a score on PCL-L of ≥ 50
49
Figure 18: PCL-C - Baseline
(02PG)
Figure 19: PCL-C - Six months
(02PG)
Figure 20: PCL-C - 12 months
(02PG)
GHQ-12
The 12-item General Health Questionnaire (GHQ-12) is used to identify psychiatric
‘caseness’ in participants and is sensitive to short-term mental health disorder rather
than to the long-standing attributes of the respondent. The GHQ-12 is scored by
calculating the sum of the scores for each individual scale item and then comparing
the total score achieved against a pre-determined threshold12. There are two methods
of scoring the GHQ-12, the Likert method (where individual items are scored 0-1-2-3)
and the GHQ (binary) method (where items are scored 0-0-1-1). While the test
authors (Goldberg et al, 1997) advocate the use of the GHQ (0-0-1-1) scoring method
for identifying caseness, both methods have been included within this report. As with
the PCL-C, the GHQ-12 was administered at baseline, at six months and at 12 months
in each of the four study groups in order to produce an estimate of the prevalence of
common mental health disorder. While the GHQ-12 is not sensitive enough to
identify the nature of the common mental health disorder it is able to identify caseness
in participants, and this allows an estimate of prevalence to be made.
The recommended threshold level for GHQ-12 is 11/12 (max score 36) using
the Likert method and 1/2 (max score 12) using the GHQ method. For the purpose of
this report, those participants scoring ≥ 12 using the Likert method or ≥ 2 using the
GHQ method were deemed to have met the criteria for caseness. Once again, the
baseline, six-month and 12-month assessments of the GHQ-12 will be considered as
independent samples to provide an effective estimate of prevalence of common
mental health disorder at each stage.
12 Within GHQ-12 each of the 12 scale items have a 4 point scoring system that ranges from a 'better/healthier than normal'
option, through a 'same as usual' and a 'worse/more than usual' to a 'much worse/more than usual' option. The exact wording will
depend upon the particular nature of the item.
50
Patient participants (01PG)
At baseline, the prospective patient group (01PG) returned mean GHQ-12 scores of
13.17 (Likert) and 3.62 (GHQ). Within the group (n=110) 57 individuals met the
criteria for caseness using the Likert method and 73 met the criteria using the
recommended GHQ scoring method (see Figure 21: GHQ-12 - Baseline using Likert
and GHQ scoring methods (01PG)) producing an estimate of the prevalence of
common mental health disorders within the prospective patient group of 51.81%
(Likert) or 66.36% (GHQ).
Figure 21: GHQ-12 - Baseline using Likert and GHQ scoring methods (01PG)
At six months the assessment of GHQ-12 returned mean scores of 13.61 (Likert) and
3.64 (GHQ) within the group (n=28) and again the majority of participants met the
criteria for caseness (n=18 using Likert method and n=16 using the GHQ method)
producing an estimate of prevalence of common mental health disorder at six months
of 64.29% (Likert) and 57.14% (GHQ).
51
Figure 22: GHQ-12 - Six-months using Likert and GHQ scoring methods (01PG)
At 12 months the calculated mean scores for GHQ-12 remained high 13.00
(Likert) and 2.74 (GHQ) and 12 patient participants met the criteria set for caseness
from an overall group of 23. The estimate of prevalence of common mental health
disorders at 12 months was the same for both Likert-scoring and GHQ scoring
methods (52.17%).
Table 24: GHQ-12 - Baseline, six months and 12 months (01PG)) provides a
summary of GHQ-12 scores at each stage.
52
Figure 23: GHQ-12 - 12-months using Likert and GHQ scoring methods (01PG)
Table 24: GHQ-12 - Baseline, six months and 12 months (01PG)
Baseline Six Months 12 Months
GHQ12
TOTAL
(Likert)
GHQ12
TOTAL
(GHQ)
GHQ12
TOTAL
(Likert)
GHQ12
TOTAL
(GHQ)
GHQ12
TOTAL
(Likert)
GHQ12
TOTAL
(GHQ)
N Valid (Meets criteria)* 110 (57) 110 (73) 28 (18) 28 (16) 23 (12) 23 (12)
GHQ Caseness 51.81% 66.36% 64.29% 57.14% 52.17% 52.17%
Missing 89 89 171 171 176 176
Mean 13.17 3.62 13.61 3.64 13.00 2.74
Range 34 12 27 12 31 12
Minimum 2 0 4 0 5 0
Maximum 36 12 31 12 36 12
*Criteria for GHQ-12 caseness defined as scores of 1/2 (max score 12) using the GHQ method (0-0-1-1) and 11/12 (max score 36) using the Likert method (1-2-3-4).
Partner participants (02PG)
The prospective partner group (02PG) returned mean GHQ-12 scores of 16.00
(Likert) and 4.35 (GHQ) at baseline and within the group (n=20) 17 individuals met
the criteria for caseness using the Likert method and 13 met the criteria using the
recommended GHQ scoring method (see Figure 24: GHQ-12 - Baseline using Likert
and GHQ scoring methods (02PG)) producing an estimate of prevalence of common
mental health disorder of 85.00% (Likert) or 65.00% (GHQ).
53
Figure 24: GHQ-12 - Baseline using Likert and GHQ scoring methods (02PG)
At six months the assessment of GHQ-12 returned mean scores of 15.25 (Likert) and
4.42 (GHQ) within the group (n=12) and again the majority of participants met the
criteria for caseness (n=10 using Likert method and n=8 using the GHQ method)
producing an estimated prevalence of common mental health disorder at six months of
83.33% (Likert) and 66.66% (GHQ).
Figure 25: GHQ-12 - Six-months using Likert and GHQ scoring methods (02PG)
At 12 months the calculated mean scores for GHQ-12 within the partner group
were 15.00 (Likert) and 4.11 (GHQ). Four patient participants met the caseness
criterion using the Likert scoring method and five met the criterion using the GHQ
scoring method from an overall group of nine participants. Estimated prevalence of
common mental health disorders at 12-months was therefore 44.44% (Likert) and
55.55% (GHQ). Table 25: GHQ-12 - Baseline, six-months and 12-months (02PG)
provides a summary of GHQ-12 scores at each stage.
54
Figure 26: GHQ-12 - 12-months using Likert and GHQ scoring methods (02PG)
Table 25: GHQ-12 - Baseline, six-months and 12-months (02PG)
Baseline Six Months 12 Months
GHQ12
TOTAL
(Likert)
GHQ12
TOTAL
(GHQ)
GHQ12
TOTAL
(Likert)
GHQ12
TOTAL
(GHQ)
GHQ12
TOTAL
(Likert)
GHQ12
TOTAL
(GHQ)
N Valid (Meets criteria)* 20 (17) 20 (13) 12 (10) 12 (8) 9 (4) 9 (5)
GHQ Caseness 85.00% 65.00% 83.33% 66.66% 44.44% 55.55%
Missing 6 6 14 14 17 17
Mean 16.00 4.35 15.25 4.42 15.00 4.11
Range 20 12 21 11 20 11
Minimum 6 0 5 0 7 0
Maximum 26 12 26 11 27 11
*Criteria for GHQ-12 caseness defined as scores of 1/2 (max score 12) using the GHQ method (0-0-1-1)
and 11/12 (max score 36) using the likert method (1-2-3-4).
DRS-15
The Dispositional Resilience Scale (DRS-15) is a brief 15-item scale used to measure
the three personality traits of hardiness, i.e. commitment, control and challenge in
both patient participant groups (01PG) and (01RG). The DRS-15 measure was
administered in order to provide a baseline measure of the hardiness of participants.
The baseline assessment of prospective patient participants (n=105) returned a mean
DRS-15 score of 26.24 with a SD of 4.389 (Range=13 to 35).
55
CDRIS-10
The Connor-Davidson Resilience Scale-10 (CDRIS-10) is a 10-item scale that
measures the ability of participants to cope with adversity. The CDRIS-10 scale was
administered at each stage of the participant patient assessments (01PG) and (01RG)
and the mean scores at each stage calculated. Comparison of mean CDRIS-10 scores
at baseline, six months and 12 months will provide insight into the role and effect of
resilience as well as the effect of traumatic injury upon resilience.
At baseline assessment the mean CDRIS-10 score for the prospective patient
group (01PG) was 31.33 (SD= 7.220). At six months the mean resilience of the group
had decreased to 28.72 (SD=8.799) and at 12 months had increased slightly to 28.87
(SD=8.915). Once again, the number of participants returning completed CDRIS-10
scales decreased significantly over time (n=109 at baseline, n=29 at six months and
n=23 at 12 months). Table 26 CDRIS-10 - Baseline, six-months and 12-months
(01PG)and Figure 27: CDRIS-10 - Baseline, six-months and 12-months (01PG)
provide a summary of CDRIS-10 scores.
Table 26 CDRIS-10 - Baseline, six-months and 12-months (01PG)
Baseline Six Months 12 Months
N Valid 109 29 23
Missing 90 170 176
Mean 31.33 28.72 28.87
Std. Deviation 7.220 8.799 8.915
Range 35 35 39
Minimum 5 5 1
Maximum 40 40 40
56
Figure 27: CDRIS-10 - Baseline, six-months and 12-months (01PG)
Primary outcome variables – ‘Paired samples’
In order to answer effectively the first two research questions regarding the long-term
effects of rehabilitation and the factors which either compromise or facilitate the
durability of therapeutic gain, a strict panel of only those participants who responded
to the questionnaires at all three stages (baseline, six months and 12 months) was
required. Due to poorer response rates at six months and at 12 months there were
only 12 such participants, and it should be noted that due to the small sample size it is
difficult to draw strong conclusions from the data.
Repeated Measures Analysis (RMA) was performed on PCL-C, CDRIS-10
and GHQ-12 scores for the panel of 12 participants. For all three measures, there was
no significant time effect observed. In other words, average levels of these measures
over the three time periods fail to show any significant pattern. Friedman’s non-
parametric tests were also performed since the sample size is small. Non-parametric
tests also indicate that there is no significant time effect for these measures.
PCL-C and CDRIS-10
Figure 28 and Figure 29 show the pattern of mean and median PCL-C scores along
with those of the Connor-Davidson Resilience scale (CDRIS-10) for the prospective
patient participant group (01PG). While higher scores on the PCL-C indicate higher
severity of PTSD, higher CDRIS-10 scores indicate greater resilience in participants.
57
Figure 28: PCL-C and CDRIS-10 - Means for panel sample (n=12)
Figure 29: PCL-C and CDRIS-10 - Medians for panel sample (n=12)
Amongst the 12 participants represented in this sample one participant shows a very
large PCL-C score (82). This partly explains the reason for the 12-month mean PCL-
C score being the highest (Figure 28: PCL-C and CDRIS-10 - Means for panel sample
(n=12)). The means indicate that there is a steady worsening of PTSD over the
28.33
33.0035.67
32.00
29.58 29.17
0
5
10
15
20
25
30
35
40
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
Baseline 6-month 12-month
Co
nn
or-D
av
idso
n R
esi
lien
ce S
cale
PC
L-C
PCL-C
CDRIS-10
29.5032.50
30.00
34.50
30.0031.00
0
5
10
15
20
25
30
35
40
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
Baseline 6-month 12-month
Co
nn
or-D
av
idso
n R
esi
lien
ce S
cale
PC
L-C
PCL-C
CDRIS-10
58
follow-up period; however, focusing on medians rather than means (Figure 29: PCL-
C and CDRIS-10 - Medians for panel sample (n=12)) shows that while there is a
worsening of PTSD at the six month point, during the next six months, the average
level of PTSD (as represented by the median PCL-C score) falls back to a level
similar to baseline. A similar explanation holds for the CDRIS-10 also in that there is
one participant with a very low score (1) at the 12-month assessment which accounts
for a further drop in the mean level at this point (Figure 28: PCL-C and CDRIS-10 -
Means for panel sample (n=12)) rather than a slight improvement as indicated by the
median level.
Following the profile depicted by medians, it may be reasonable to conclude
that over a six-month period there is a worsening of PTSD and deterioration in
resilience, but by another six months these outcome measures return to baseline
levels.
PTSD and GHQ
The following figures (Figure 30 and Figure 31) show the pattern of median and mean
GHQ scores overlaid on the corresponding PTSD scores.
Figure 30: PCL-C and GHQ-12 Medians for panel sample (n=12)
29.50 32.50 30.00
14.00
12.00
11.00
0
3
6
9
12
15
18
21
24
27
30
33
36
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
Baseline 6-month 12-month
GH
Q-1
2
PC
L-C
PCL-C
GHQ-12
59
Figure 31: PCL-C and GHQ-12 Means for panel sample (n=12)
Both the median (Figure 30: PCL-C and GHQ-12 Medians for panel sample (n=12))
and mean scores (Figure 31: PCL-C and GHQ-12 Means for panel sample (n=12)) of
GHQ indicate that over the 12-month period from baseline, there has been a steady
improvement in general health, although PTSD, at best, has not changed.
PTSD – Paired comparisons
There is tentative evidence to suggest that there may have been a significant
worsening of PTSD during the first six months of the study. The tables below (Table
27: Paired samples statistics and Table 28: Paired samples tests) show parametric
comparisons among all possible pairings of the three study periods. As can be seen,
there is a significant (p=0.015) increase in mean PTSD score from 25.38 at baseline to
33.57 at the six-month assessment. This is also confirmed by a non-parametric test
(Wilcoxon Signed Ranks test with 2-tailed asymptotic significance of 0.015). Given
that the baseline and 12-month comparison does not show a significant difference,
this may suggest that experience of PTSD worsens in the short term before levelling
off.
These observations must be treated with caution since only a few cases used within a
pairing appear consistently in all pairings. However, the median levels of PTSD as
shown in Figure 29: PCL-C and CDRIS-10 - Medians for panel sample (n=12) and
28.3333.00 35.67
15.17
12.7512.67
0
3
6
9
12
15
18
21
24
27
30
33
36
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
Baseline 6-month 12-month
GH
Q-1
2
PC
L-C
PCL-C
GHQ-12
60
Figure 30: PCL-C and GHQ-12 Medians for panel sample (n=12) appear to
substantiate this finding.
Table 27: Paired samples statistics (PCL-C)
PTSD Mean N Std. Deviation Std. Error Mean
Pair 1 Baseline 25.38 21 7.839 1.711
6-month 33.57 21 13.710 2.992
Pair 2 6-month 35.71 14 17.652 4.718
12-month 39.00 14 20.138 5.382
Pair 3 Baseline 29.00 19 11.609 2.663
12-month 35.05 19 16.067 3.686
Table 28: Paired samples tests (PCL-C)
Pair 1 Pair 2 Pair 3
PTSD Baseline – 6month 6month – 12month Baseline - 12month
Paired Differences
Mean -8.190 -3.286 -6.053
Std. Deviation 14.073 11.180 16.827
Std. Error Mean 3.071 2.988 3.860
95% Confidence
Interval of the
Difference
Lower -14.597 -9.741 -14.163
Upper -1.784 3.169 2.058
t -2.667 -1.100 -1.568
df 20 13 18
Sig. (2-tailed) .015 .291 .134
CDRIS-10 – Paired comparisons
Paired comparisons were also done on mean levels of the CDRIS-10 scores (Table 29:
Paired samples statistics (CDRIS-10) and Table 30: Paired samples tests (CDRIS-
10)). As in the case with PTSD, a significant worsening was observed at the six-
month assessment as measured by mean levels of the CDRIS-10 measure. This is
also confirmed by the non-parametric equivalent – the Wilcoxon Signed Ranks Test
with asymptotic significance of 0.036.
61
Table 29: Paired samples statistics (CDRIS-10)
CDRIS-10 Mean N Std. Deviation Std. Error Mean
Pair 1 Baseline 33.14 21 5.944 1.297
6-month 28.76 21 8.123 1.773
Pair 2 6-month 28.93 14 8.426 2.252
12-month 27.29 14 10.462 2.796
Pair 3 Baseline 31.60 20 6.541 1.463
12-month 30.00 20 8.473 1.895
Table 30: Paired samples tests (CDRIS-10)
Pair 1 Pair 2 Pair 3
CDR Baseline – 6month 6month – 12month Baseline - 12month
Paired
Differences
Mean 4.381 1.643 1.600
Std. Deviation 9.319 8.688 11.170
Std. Error Mean 2.034 2.322 2.498
95%
Confidence
Interval of the
Difference
Lower .139 -3.373 -3.628
Upper 8.623 6.659 6.828
t 2.154 .708 .641
df 20 13 19
Sig. (2-tailed) .044 .492 .529
GHQ-12 – Paired comparisons
For the sake of completeness, paired comparisons were also done for GHQ-12, but no
significant pairings were found (using both parametric and non-parametric tests).
Table 31 and Table 32 show the results of the paired comparisons.
Table 31: Paired samples statistics (GHQ-12)
GHQ Mean N Std. Deviation Std. Error Mean
Pair 1 Baseline 13.10 21 7.300 1.593
6-month 12.57 21 6.577 1.435
Pair 2 6-month 13.57 14 8.680 2.320
12-month 12.64 14 7.249 1.938
Pair 3 Baseline 13.85 20 6.722 1.503
12-month 13.05 20 7.430 1.661
62
Table 32: Paired samples tests (GHQ-12)
Pair 1 Pair 2 Pair 3
GHQ Baseline – 6month 6month – 12month Baseline - 12month
Paired
Differences
Mean .524 .929 .800
Std. Deviation 7.061 5.567 9.036
Std. Error Mean 1.541 1.488 2.020
95% Confidence
Interval of the
Difference
Lower -2.690 -2.286 -3.429
Upper 3.738 4.143 5.029
t .340 .624 .396
df 20 13 19
Sig. (2-tailed) .737 .543 .697
Health and lifestyle variables
In Section Seven of the participant questionnaires (Your health and lifestyle)
participants were asked a series of questions regarding their use of alcohol, smoking
habits and their social life and relationships. Following the analysis of the primary
outcome variables, the health and lifestyle variables were analysed in order to identify
any significant associations with those primary outcome variables that may indicate
factors which are conducive to, or preventative of, effective rehabilitation.
Alcohol consumption
The World Health Organisation (WHO) ‘Alcohol use disorders identification test’
(AUDIT) was used to assess alcohol consumption at baseline, at six months and at 12
months. The AUDIT measure consists of a ten-item scale designed to reflect
participant’s relative level of risk related to their alcohol use. An AUDIT score of
eight or more is recommended by the WHO as an indicator of hazardous and harmful
alcohol use, as well as of possible alcohol dependence (higher scores indicating
greater likelihood of hazardous and harmful drinking). While the WHO recommend
the use of a cutoff score of eight, they acknowledge also that a secondary cut-off score
of ten provides greater specificity in terms of hazardous and harmful drinking, albeit
at the expense of sensitivity (Babor et al, 2001). Table 33 provides a summary of
mean AUDIT scores achieved by the prospective patient participants (01PG) at
baseline, six months and 12 months.
63
Table 33: AUDIT scores (01PG)
Number of
cases Mean Median Std. Deviation Minimum Maximum
WHO - BA - AUDIT -
Totals
93 9.80 9.00 5.272 2 29
WHO - 6m - AUDIT
TOTALS
26 9.00 8.50 5.044 1 22
WHO - 12m - AUDIT
TOTALS
20 8.05 8.00 3.677 2 16
While the mean scores achieved by the prospective patient group (01PG) are greater
than the cutoff of eight recommended by the WHO, they are below the secondary
cutoff score of ten. In respect of the mean AUDIT scores achieved by participants,
there is evidence to suggest that over the follow-up period, rates of hazardous or
harmful alcohol use decrease (as evidenced by a fall in mean and median scores).
There is therefore also some evidence to suggest a general reduction in alcohol
consumption over the follow up period as evidenced by a decrease in the maximum
levels of intake.
In respect of the primary outcome measures (PCL-C, GHQ-12 and CDRIS-10)
there is also evidence to suggest that (at baseline at least) outcomes are worse for total
abstainees from alcohol compared with non-abstainees (see Table 34). Mean PCL-C
scores for abstainees at baseline were significantly higher (Mean=45.08) when
compared to non-abstainees (Mean=29.38). Mean GHQ-12 scores at baseline,
similarly, were significantly higher for abstainees (Mean=16.85 (Likert)) when
compared to non abstainees (Mean=12.68)).
Table 34: Mean outcome measures for Abstainees compared with non-abstainees
Abstainees vs. Non-Abstainees N Mean Std. Deviation Std. Error Mean
BA - PCL-C - TOTAL
SEVERITY SCORE
0 13 45.08 18.549 5.145
1 95 29.38 10.944 1.123
BA - CDRIS10 - TOTAL 0 13 27.15 9.245 2.564
1 96 31.90 6.764 .690
BA - DRS15 - TOTAL 0 13 23.69 5.266 1.461
1 92 26.60 4.159 .434
BA - GHQ12 - TOTAL
(Likert)
0 13 16.85 8.774 2.433
1 97 12.68 5.940 .603
64
Table 35: Mean outcome measures for abstainees and non-abstainees (non-parametric tests)
BA - PCL-C -
TOTAL SEVERITY
SCORE
BA - CDRIS10 -
TOTAL
BA - DRS15 -
TOTAL
BA - GHQ12 -
TOTAL (Likert)
Mann-Whitney U 316.500 418.000 407.000 435.500
Wilcoxon W 4876.500 509.000 498.000 5188.500
Z -2.845 -1.930 -1.864 -1.809
Asymp. Sig. (2-tailed) .004 .054 .062 .070
a. Grouping Variable: Abstainees vs Non-Abstainees
While there is strong evidence to suggest that PTSD is worse for abstainees, these
results remain inconclusive with regard to resilience even on the CDRIS-10 scale
which shows borderline significance (Table 35). While DRS-15 and GHQ-12 scores
at baseline indicate worse results for abstainees rather than for non-abstainees these
results are not statistically significant at a 5% criterion of rejection of null hypothesis.
Findings were based upon non-parametric tests rather than upon their
parametric equivalents suggest that consuming alcohol (which includes the associated
social aspects and benefits) may be contributing towards lower levels of PTSD and
raising levels of resilience at baseline.
Organisations and social activities before injury
Participants were asked, in regard to their social life and relationships, which
organisations they belonged to and in which social activities they participated
regularly outside of work (see Table 36).
Table 36: Regular social activities outside of work (01PG)
Number % of sample of 110
Going to pubs/clubs 81 73.6%
Visiting family/friends 80 72.7%
Going to gym with friends 71 64.5%
Team sports / outdoor pursuits 71 64.5%
Going to watch sporting events with friends 52 47.3%
Social or hobby-related clubs 23 20.9%
Further Education 18 16.4%
Religious gatherings / societies 7 6.4%
Voluntary services 3 2.7%
Social activities including ‘going to the pubs/clubs’ (73.6%, n=81) and ‘visiting
family and friends’ (72.7%, n=80) were the main regular social activities out of work.
Similarly ‘going to the gym with friends’ (64.5%, n=71) and ‘team sports/outdoor
pursuits’ (64.5%, n=71) featured significantly in responses. This is perhaps to be
expected in the population and is consistent with the common perception of ‘the
military lifestyle’. When asked about their social life prior to their injury 56.9%
65
(n=64) of those responding (n=109) indicated that most of the people that they
socialised with were involved with the military (see Table 37).
Table 37: Were most of the people you socialised with involved with the military? (Prior to injury)
Frequency Percent Valid Percent Cumulative Percent
Valid Yes 62 31.2 56.9 56.9
No 47 23.6 43.1 100.0
Total 109 54.8 100.0
Missing System 90 45.2
Total 199 100.0
Further analysis of social activities prior to injury suggested that there may have been
some significance in the (baseline) response ‘going to the gym with friends’13. An
initial analysis of the primary outcome measures (PCL-C, GHQ-12, DRS-15 and
CDRIS-10) as independent samples indicated a statistically significant effect (see
Table 38).
Table 38: Social activity prior to injury - going to the gym with friends (Baseline)
Going to the gym with friends Baseline 6-Month 12-Month
Mean (n) Mean (n) Mean (n)
PCL-C Yes 30.32 (69) 33.83 (12) 32.09 (11)
N/A 32.43 (37) 32.00 (8) 39.43 (7)
GHQ12 Yes 12.17 (70) 11.83 (12) 11.67 (12)
N/A 14.95 (38) 13.75 (8) 15.29 (7)
DRS15 Yes 27.13 (68)
N/A 24.51 (35)
CDRSI10 Yes 32.01 (71) 30.08 (12) 31.25 (12)
N/A 30.06 (36) 26.88 (8) 28.86(7)
Cells highlighted in red are statistically significant results on both parametric (independent samples t-tests), and non-parametric (independent samples Mann-Whitney) tests.
When taken as independent samples, those who did not report ‘going to the gym with
friends’ prior to their injury appeared to fare worse on all outcome measures at the 12-
month point when compared with baseline. Prospective patient participants who
reported ‘going to the gym with friends’ prior to their injury at baseline, appeared to
fare better in terms of their resilience (CDRIS-10) and mental health (GHQ-12). This
‘improvement’ was not observed in respect of participants symptoms of PTSD (PCL-
C).
13 The baseline response to the question “Before your injury, which of the following organisations did you belong to or which social activities did you regularly choose to do outside of work” was used in preference to the same question asked at six months
and at 12 months as it was felt that responses at baseline would be a better reflection of participants actual social activity.
66
While a statistically significant difference was observed, on the DRS-15 and
GHQ-12 measures, between those who reported ‘going to the gym with friends’ and
those who did not report ‘going to the gym with friends’ at baseline, these results are
not based on a consistent panel sample. As a result, the sample of individuals
responding to this item at each time-point is treated as representative of those
responding at baseline. As the subsequent analysis of this phenomenon using the
strict panel of participants who responded to the question at each stage was
inconclusive and failed to verify the finding, it is recommended that these results be
verified with a larger panel.
Relationships and social support
Within the prospective patient group (01PG), 66.1% (n=74) of participants regard
themselves as being in ‘committed relationship’ at baseline and either describe
themselves as being ‘married’ (26.8%, n=30), ‘living with a partner’ (10.7%, n=12) or
‘in a long-term relationship’ (28.6%, n=32). Participants were asked about their
satisfaction with their current relationship, prior to their injury, at baseline, at six
months and again at 12 months. Taking a strict panel of participants responding to
these questions at each stage (n=11) indicated that while there was some increase in
marital dissatisfaction in individual cases, the general level of satisfaction was
unchanged at 12 months (see Table 39).
Table 39: Marital/relationship satisfaction for panel (n=11)
Prior to injury Baseline (01PGBA) Six months (01PG6M) 12 months (01PG12M)
Frequency % Frequency % Frequency % Frequency %
Extremely
satisfied 5 45.5 4 36.4 3 27.3 5 45.5
Satisfied 2 18.2 4 27.3 5 45.5 3 27.3
Neither
satisfied nor
dissatisfied
Dissatisfied 1 9.1
Extremely
dissatisfied 1 9.1
N/A 4 36.4 4 36.4 2 18.2 2 18.2
In respect of social support offered by partners, the Social Support Inventory (SSI)
was used to assess: perceived and received criticism, social support and total social
support inventory score. Using the same strict panel of participants (n=11), mean
scores on the perceived and received criticism sub-scale suggested that participants
tended to perceive their relationships as becoming less critical over time.
67
Alternatively, mean scores achieved on the social support sub-scale suggest that
participants were less satisfied with the support they were receiving from their
partners at six months and 12 months. Table 40: SSI - Baseline, six months and 12
months for strict panel (n=11) (01PG) provides a summary of mean SSI scores.
Analysis of these findings in relation to the primary outcome measures (PCL-C,
GHQ-12, CDRIS-10) indicated that there were no correlations between those
outcomes and perceived social support. Again it is it is recommended that these
results be verified with a larger panel.
Table 40: SSI - Baseline, six months and 12 months for strict panel (n=11) (01PG)
Mean Baseline (01PGBA) Six Months (01PG6M) 12 Months (01PG12M)
Perceived and
received criticism 11.45 10.45 9.91
Social support 6.27 7.18 7.17
SSI Total Score 17.73 17.64 17.09
Qualitative findings (Partner study)
Initial direct comparison of prospective patient participants (01PG) and prospective
partner participants (02PG) mean PCL-C and GHQ-12 (GHQ) scores at baseline, six
months and 12 months demonstrates higher mean PCL-C scores at baseline and six
months in partner participants and consistently higher GHQ-12 scores (GHQ) than
patient participants throughout the follow up period. Despite the relatively small
sample of participants in the prospective partner study, analysis of the primary
outcome variables does indicate a clinically significant level of psychopathology in
partners. Further examination of the partner ‘experience’ is therefore warranted.
As part of the overall assessment process partner participants were asked a
series of open questions at each stage designed to provide some insight into their
experience. Partner participants were asked, in general terms, about their experience
of the DMRC, about changes in their relationship with their partner and about their
perception of their own ability to cope.
Experience of the DMRC
Partner participants were asked if they had been to visit their partners at the DMRC
and 71% (n=15) indicated that they had, while 28.6 (n=6) indicated that they had not.
When asked ‘in general terms how convenient has it been to visit your partner at
DMRC: Headley Court?’ 50% (n=10) indicated that it had been ‘Very convenient’ or
‘Convenient’. Six partner participants indicated that it had been ‘Inconvenient’ to
68
visit. When asked what had made it inconvenient to visit, partner participants
indicated that: time factors; difficulty in securing time away from work and distance
had been the main factors. Those partner participants indicating that it had been
inconvenient to visit reported a general feeling that it was easier to wait for their
partners to return home at the weekends from the DMRC than for them to travel there.
One partner participant indicated that she did not feel that the DMRC “was a very
warm and friendly environment” and reported finding it “quite clinical”. One partner
participant indicated that she had not visited because her partner did not like her to
visit him at the DMRC.
Attitude of staff
When asked, ‘in general terms, how would you describe the attitude of the staff at
DMRC: Headley Court towards you?’ 66.7% of partner participants reported that the
staff at the DMRC had been either ‘Very supportive’ or ‘Supportive’ (n=10). One
third of partner participants (33.3%) indicated that the staff at the DMRC had been
‘Unsupportive’ (n=1) or ‘Very Unsupportive’ (n=3) or that they had not spoken to any
staff at the DMRC at all (n=1).
Perception of care
When asked about their perception of the level of care provided at the DMRC: 80%
(n=12) rated the care as being either ‘Good’ or ‘Very good’; 13.3% indicated that they
though that the care had been ‘Average’ (n=2) and one participant indicated that they
though that the level of care had been ‘Poor’. When asked to what extent had the
level of care affected their ability to cope with their partner since his or her injury the
majority of partner participants (60%, n=9) suggested that it made coping with their
partner ‘A little easier’ or ‘Much easier’. While one participant felt that the level of
care provided at the DMRC made no difference to their ability to provide care for
their partner at home, five participants (33.3%) suggested that it had made caring for
their partner at home ‘A little harder’ or ‘Much harder’.
Emotional support (Positive)
Partner participants were asked ‘Overall, in terms of the emotional support that you
may have received from staff at DMRC, what has been most helpful?’ Fifteen
partner participants provided responses to this question, and a number of themes were
identifiable within those responses. There were six generally negative responses to
69
this question; seven generally positive responses, and two responses that were neither
positive nor negative.
Staff
While one partner participant reports never having been to the DMRC, a number of
the others who had, reported having no contact with staff at all during the time that
their partner was an inpatient there (n=4). Four of the responses state that the staff at
the DMRC provided no emotional support and one partner participant suggests that,
as a girlfriend rather than as a wife or live-in partner, the support offered by the staff
at the DMRC “did not extend as far as me”. One partner participant indicated that
while she had no contact with staff when her partner was an inpatient at the DMRC
she had had contact with them subsequently, and that her partner’s Occupational
Therapist had been ‘the greatest support’.
One partner participant expressed the view that she felt ‘a lot easier knowing
that there was at least one friendly face’ to whom her partner could turn at Headley
Court. Those partner participants who reported positive experiences of support
indicated that “the willingness of certain staff” to “include and explain” was the most
helpful thing. In general terms, the staff identified as being most helpful were the
Social Work staff (n=5), Occupational Therapy staff (n=2), Prosthetics Department
(n=2) and the Physiotherapy staff (n=1). Neither Medical nor Nursing Staff were
mentioned in responses to this question. In their responses to this question, partner
participants did not make any distinction between Military and Civilian Staff.
Significant Moments
A number of partner participants indicated that the provision of information and an
orientation tour of the DMRC had been helpful in providing emotional support, and a
number of participants reported ‘significant moments’ like this or ‘milestones in care’
as being important in terms of their emotional support. Examples of ‘significant
moments’ identified by partner participants such as the witnessing of the fitting of
prosthetics and seeing their partner walk again, along with realising that their partner
‘wasn’t the odd one out’ were most helpful.
Emotional support (Negative)
Partner participants were asked ‘Overall in terms of emotional support that you may
have received from staff at DMRC, what has been least helpful?’ While eight partner
participants provided responses to this question, three of the responses were the same
70
as those they provided in the previous question. Once again, a number of themes
emerged.
Staff
While a number of members of staff were identified as being helpful in respect of the
provision of emotional support, concerns were raised by partner participants in
relation to the staff at the DMRC and their willingness to involve partner participants
in care. The lack of interaction also appeared to be a consistent concern for partner
participants in regard of staff, and one participant indicated that despite initial support,
promised regular contact was not forthcoming.
Slow Progress
While a number of partner participants indicated that ‘significant moments’ or
‘milestones in care’ were helpful in respect of their emotional support one partner
identified that, “The slow progress of my partner’s treatment and rehab, does nothing
for his self esteem and also for his anger issues which in turn does not help me
emotionally”, and it would appear that problems associated with frustration and anger
on the part of the patient participants over time constitutes a significant problem.
Provision of information
Partner participants were asked a series of specific questions in respect of the
provision of information by staff at the DMRC. These questions related to the overall
management of care; their partners’ future recovery; the psychological aspects of their
partners’ injury, and the potential sources of help available (see Figure 32). While
52.6% (n=10) of partner participants acknowledge that it was easy to obtain
information from staff when they requested it, eleven of the fourteen responses to this
question served to highlight problems or shortcomings associated with the provision
of information. A consistent view expressed by partner participants was that there
was a general lack of information provided by staff at the DMRC at all three stages of
the follow up process (see Figure 32: Provision of information to partners (02PG)).
Quality, quantity and timing of information
As a follow-up to the questions regarding the provision of information, those partners
who indicated that they had been provided with information were asked to comment
on the quality, quantity and timing of that information. On balance, those partner
participants responding indicated that the information provided in respect of the
overall management of their partners care, their partners’ future recovery and the
71
psychological impact of their partners’ injuries had been of poor quality and had often
been provided ‘too late’. In contrast, the majority of partner participants commenting
on the quality, quantity and timing of information regarding possible sources of help
and support (n=6) characterised the quality, quantity and timing of this information to
be ‘about right’ (66.7% - 88.3%, n=4 and n=5). It should be noted that the numbers
of participants responding to the questions regarding quality, quantity and timing of
information were very small and that it is difficult to draw general conclusions from
such a small number.
In terms of your
partners overall
management at DMRC, have you received any
information?
Did you receive any
information from the
staff regarding your partners future in terms
of recovery?
Did you receive any
information from the
staff with regard to how your partner might
react psychologically to
the injury?
Did you receive any
information from the
staff in terms of possible help available
to you or to your
partner?
Yes (Baseline) 4.80% 4.80% 9.50% 23.80%
No (Baseline) 95.20% 95.20% 90.50% 76.20%
Yes (Six months) 8.30% 8.30% 8.30% 8.30%
No (Six months) 91.70% 91.70% 91.70% 91.70%
Yes (Twelve months) 40.00% 10.00% 10.00% 30.00%
No (Twelve months) 60.00% 90.00% 90.00% 70.00%
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
90.00%
100.00%
Pro
vid
ed
wit
h i
nfo
rma
tio
n?
(%)
Figure 32: Provision of information to partners (02PG)
72
Conclusions
Recruitment
Within the prospective patient participant group (01PG) recruitment rates were
encouraging, and it seems likely that the high recruitment rate in that group (92.99%,
n=199) was due, in the most part, to the face-to-face contact between potential
participants and the HCF at the DMRC. Generally speaking, potential participants
demonstrated considerable willingness to participate in the research study when they
were first contacted and expressed a strong desire ‘to help anyone else who is injured
in the future’. It could also be argued that the willingness of those individuals to
participate in the research study could be explained by the fact that the HCF
approached each of the potential participants early in their first admission to the
DMRC, and as potential patient participants were approached in uniform (Combat 95)
this further facilitated the formation of a ‘credible military relationship’. While it
could also be argued that approaching potential participants in uniform may be
regarded as being coercive (as it served to highlight a significant rank differential
between the HCF (Major Forbes) and the majority of potential participants) it was felt
that the wearing of uniform on the part of the HCF would facilitate improved
participation and a more efficient integration within the DMRC.
As the invitation to participate in the study was extended to the retrospective
patient participant group (01RG) and to both partner groups (02PG) and (02RG) by
post, the recruitment rates within these groups are consequently considerably lower
(01RG=15.87%, 02PG=25.23% and 02RG=21.43%) than that of the prospective
patient participant group.
Observations
1. Future funded research within the DMRC setting should include provision for
a ‘military research fellow’ as face-to-face contact between a credible military
researcher and potential participants appears to facilitate effective recruitment
and higher levels of participation.
2. In order to establish a credible military relationship with participants, the
military research fellow should conduct research, wherever practicable, in
uniform. The wearing of military uniform also serves to facilitate access to
the clinical area and effective interaction with clinical staff. The military
research fellow must be of sufficiently high rank and of sufficient clinical
73
standing to allow effective interaction with military medical consultants and
other senior staff but not to limit interaction with potential participants.
Response rates
The response rates achieved within the prospective patient group (01PG) were
encouraging at baseline (56.28%). Whereas at six months (14.57%) and at 12 months
(11.56%) the percentage of assessments returned was disappointing. It could be
argued that the disappointing response at these stages may be due to a number of
factors. Firstly, it is possible that the effect of face-to-face contact achieved during
the recruitment phase of the study contributed to the higher response rates in the
prospective patient participant group (01PG) at baseline as the assessments used at
this stage were handed directly to participants and, in a large proportion of the cases,
were handed directly back to the HCF once complete. Secondly, as the six-month and
the 12-month response rates relied upon the prospective patient participants receiving
assessments by post, to be completed on home sick leave and then returning them by
post, the response rates at six and at 12 months are more likely to reflect the well
established difficulties associated with postal surveys.
Finally, throughout the follow-up period there were consistent difficulties
associated with the ‘tracking’ of prospective patient participants and a significant
number of assessments were returned “undelivered” at six months and at 12 months.
The difficulties associated with achieving sufficient returns within a military study are
well recognised (Fear et al, 2010). While best efforts were made to maintain the
patient contact database using information obtained from the DMICP database and
directly from participants, the return of undelivered assessments represents a
significant difficulty.
In contrast to the prospective patient participant group (01PG) the response
rates achieved by the retrospective patient (01RG) and prospective partner group
(02PG) at six months and 12 months tended to be consistently higher and it seems
likely that this may be due to their initial committment to the study. While the
recruitment rates achieved in the retrospective patient (02RG) and prospective partner
(02PG) groups were significantly lower than those achieved in the prospective patient
group (01PG), it may be that, once committed to participating in the study, these
groups were more willing to complete the 12 month follow up.
74
Observations
1. Where possible, future research should employ direct assessment of
participants at the DMRC in order to minimise these difficulties and to
maximise participation.
2. Where the research requires the use of a study design incorporating less direct
methods of assessment, e.g. with geographically distant retrospective
participants or partners, the research team should consider seeking consent
from those participants to contact them by telephone during the study in order
to establish and maintain a research relationship likely to improve
participation.
Demographics
Ethnicity and gender
The majority of prospective patient participants (01PG) and prospective partner
participants (02PG) identified themselves as being ‘White: British’, ‘White: Irish’ or
‘White: Other’ and given the small numbers of those recruited to the study and
participating in the follow-up who identify themselves as belonging to the ‘Asian or
Asian British: Other’ or the ‘Black or Black British: African’ categories it is difficult
to draw any reliable conclusions based on the ethnicity of participants. One
observation that was made in respect of ethnicity was that approximately half of those
reporting themselves as abstainees from alcohol identified themselves as being ‘Asian
or Asian British: Other’ and the ‘Black or Black British: African’
As almost all of participants in the patient study were male (98.6%), and only
one of those female participants initially recruited onto the study (n=3) completed the
prospective participant assessments (baseline and six-months only), it is impossible to
draw any conclusions based on gender within the patient participant group. Likewise
in the prospective partner group (02PG) there were no male partner participants
recruited onto the study.
It is difficult to make any sound recommendations in respect of the
observations regarding the ethnic and gender bias of this study. As the study design
incorporated face-to-face recruitment of consecutive admissions in order to facilitate
maximum participation, maximum response, and to minimise problems associated
with selection bias the low numbers of female personnel and personnel from other
ethnicities remains a limitation of the study and could not be avoided.
75
Age, relationship status and financially dependent children
Within the prospective patient participant group (01PG), the mean age of participants
was 26.33 years and within the prospective partner group (02PG) it was 28.4 years.
Within each of these groups the youngest participant was 19 years of age and 10.2%
of the patient group was aged 20 years or less. The mean age of patient participants
(both prospective and retrospective groups) is representative of that of the general UK
Regular Forces14 and, as previously commented, this is most likely a reflection of the
‘junior’ nature (in both age and rank) of those deployed in forward areas on
operations.
Almost two-thirds of the patient participants (64.8%, n=125) regarded
themselves as being in a ‘committed relationship’ and in order to estimate how well
established these relationships were, the prospective partner participants were asked
how long they had been with their partner prior to his or her injury: 71.4% (n=15) of
the partners reported having been in their relationship for 48 months or less. While,
again, it should be noted that the number of partner participants responding to this
question makes it difficult to draw any reliable and generalizable conclusions
regarding the durability of the relationships of the population as a whole, given the
relative age of participants, it seems reasonable to suggest that the majority of
participants were in ‘new’ relationships.
Twenty-six percent (n=32) of prospective patient participants reported having
financially dependent children (mean=1.81 children, which is roughly in line with the
national UK average) and the mean age of children within the group was 6.5 years.
It seems likely therefore that given: the relatively low age of prospective
patient participants and their partners; the fact that the majority of participants in
relationships appear to be in relatively new relationships and that those with
financially dependent children have relatively young families, the additional
difficulties placed upon those relationships and families associated with coping with
combat-related injury only serve to compound the social and personal challenges
faced by both patient and partner participants.
Observations
1. The appointment of, or redeployment of an existing member of military staff
as a ‘Family Liaison Officer’ in order to support the current service offered by
14 At 1st April 2011, the mean age of those serving in the UK Regular Forces as ‘Other Ranks’ was twenty-nine years and as
‘Officers’, thirty-seven years of age.
76
the Social Work Department would provide a valuable link between patients,
their families and potential sources of support in the voluntary sector.
2. Further research establishing the impact of combat-related injury on the wider
family, i.e. the children and parents of those with combat-injuries who may be
regarded as ‘cross-generational’ carers would further enhance understanding
of the ‘ripple effect’ of trauma.
Highest level of education and employment
Within the prospective patient group (01PG), 70.2% (n=87) of participants report the
completion of secondary school as their highest level of education attainment.
Further examination of this sub-group indicates that 13.7% (n=17) of those
participants left school with no formal educational qualifications, while 39.5% (n=49)
achieved GSCE/CSE or O-Level equivalent qualifications and 19.9% (n=21) achieved
A-Level qualifications. A further 29.8% (n=37) of prospective patient participants
report post-secondary educational achievement within further and higher education
and are qualified to degree and to post-graduate level.
In contrast, only 23.8% (n=5) of the prospective partner group (02PG) report
the completion of secondary school as their highest level of educational attainment
and a significantly higher percentage of partners indicate that they have achieved the
completion of studies in further or higher education (51.8%, n=14) gaining diploma
(23.8%), degree (29.6%) or post-graduate level qualification (4.8%). While it may be
the case that the level of academic attainment reported here is not representative of the
partner population as a whole, it is consistent with the higher percentage of partner
participants within the study reporting the nature of their employment as ‘professional
or managerial’ (94%, n=9).
While the vast majority of patient participants are employed as regular
members of the armed forces (91.2%, n=118) only 67% (n=14) of the prospective
partner group report being employed in a full-time capacity.
It seems increasingly clear that the outcome of rehabilitation for those
individuals with disabilities incompatible with continued military service is medical
discharge. It also seems reasonable to suggest, therefore, that the future employability
of those patient participants deemed unfit for continued military service will be
dependent upon: the severity of their injuries; the nature of any resulting disability;
their existing skill set and their previous educational and professional qualifications as
77
well as, at a general level, their individual motivation and the level of support they are
offered prior to discharge. It may be argued that those with higher levels of disability
and lower levels of educational qualifications may be regarded as being less
‘employable’ in the future.
Primary outcome variables
PTSD and common mental health disorders
The civilian version of the Post-Traumatic Checklist (PCL-C) was used to estimate
the prevalence of PTSD within each of the study groups. Using a cut-off score of 50
as the normative threshold provided an estimate of the prevalence of PTSD within the
patient participant group (01PG) of 10.18% at baseline, 20.69% at six months and
18.18% at 12 months. Within the partner participant group (02PG) the estimate of the
prevalence of PTSD over the same period was 7.14%, 33.33% and 10.00%
respectively.
In addition to the estimation of PTSD within the study, the prevalence of
common mental health disorder was also established using the General Health
Questionnaire (GHQ-12). The GHQ-12 measure is used to assess psychiatric
‘caseness’ in participants and, while it is not sensitive enough to identify individual
pathologies, it does provide a useful means by which to estimate prevalence. Within
the patient participant group (01PG) the prevalence of common mental health disorder
was estimated as being, 66.36% at baseline, 57.14% at six months and 52.18% at 12
months. Likewise, within the prospective partner group (02PG) the prevalence of
common mental health disorder was estimated as being, 65.00% at baseline, 66.66%
at six months and 55.55% at 12 months.
In respect of the prospective patient participants, the prevalence of PTSD was
significantly higher than that reported in previous studies of UK troops returning from
operations (minus combat-related injury) i.e. Hotopf et al (2006) and this serves to
confirm the findings from earlier studies (e.g., Kulka et al, 1990 and Michaels et al,
1999) that indicate that bodily injury is a risk factor – rather than a protective one –
for PTSD.
In addition to these findings, the high levels of common mental health disorder
identified in both patient and partner participant groups further serve to confirm that
comorbidity in PTSD is the norm rather than the exception (O’Donnell et al, 2004)
and that a significant proportion of those individuals with a diagnosis of PTSD also
78
meet the criteria for at least one other psychiatric condition (Creamer et al, 2001).
Once again, while the use of GHQ-12 does not serve to identify what these individual
psychopathologies may be, the literature confirms that other symptoms from which
soldiers may be at risk include depression, anxiety, anger, sleep disturbance,
somatisation, substance misuse, dissociation and sexual problems (Goff et al, 2007).
In order to provide an effective estimate of the durability of any rehabilitative
gain in respect of PTSD and common mental health disorder a strict panel of only
those participants who responded to the questionnaires at all three stages (baseline, six
months and 12 months) was required. Due to poorer than expected response rates at
six and 12 months the panel of prospective patient participants consisted of only 12
individuals. The strict panel of prospective partner participants completing both PCL-
C and GHQ-12 at all three stages consisted of only four individuals and has been
excluded from this discussion.
Repeated Measures Analysis (RMA) was performed on mean PCL-C, and
GHQ-12 scores for the panel and the average levels of scores over the three time
periods fail to show any significant pattern. However, an analysis of those primary
outcome measures using the median scores rather than the mean scores (intended to
eliminate the bias introduced by the most extreme scores on either end of the scale)
indicates that it may be reasonable to conclude that over the initial six-month period
of the follow up there is a worsening of symptoms of PTSD, but that by 12 months
these symptoms return to baseline levels and that there has been a steady
improvement in general mental health of the prospective patient participants as
indicated by the GHQ-12. These findings are consistent with existing research
reporting delayed onset of symptoms (Hoge et al, 2004).
Direct comparison of prospective patient participant (01PG) and prospective
partner participants (02PG) mean PCL-C and GHQ-12 scores at baseline, six months
and 12 months appears to demonstrate that partner participants achieve higher mean
PCL-C scores at baseline and six months and consistently higher GHQ-12 scores than
patient participants.
Resilience (DRS-15 and CDRIS-10)
The Dispositional Resilience Scale (DRS-15) was utilised within the prospective
patient participant group (01PG) to establish a baseline measure of resilience in
participants. The administration of the DRS-15 was conducted through the
79
prospective patient baseline assessment (01PGBA) within one week of admission to
the DRMC. Given the timing of the administration of the DRS-15, however, a degree
of caution must be exercised in its interpretation, and it is reasonable only to suggest
that the individual DRS-15 scores achieved provide a baseline measure of resilience
‘at admission’ rather than as a baseline assessment of the resilience of the individual
participants as a whole. A simple analysis of the total time elapsed from ‘incident’ to
‘admission’ to the DMRC demonstrates that (for ninety-five percent of the group)
between one and 27 weeks may have elapsed (mean time between ‘injury’ and
‘admission’=12.74 weeks). It seems likely that during this time a combination of
personal and environmental factors will have had an effect upon overall resilience.
The comparison of the mean DRS-15 score achieved by the prospective patient group
(01PG) at admission to the DMRC (mean=26.24) and the norms cited by the scale
author (Bartone et al, 2007) for a generally comparable group of U.S. military
academy students (mean= 29.15) further supports this conclusion.
The Connor-Davidson Resilience Scale (CDRIS-10) was used with the patient
participant group (01PG) to estimate the relative levels of resilience for the population
at each of the three stages. A comparison of the mean CDRIS-10 scores at each stage
(when taken as independent samples) indicates that levels of resilience decline from
baseline (mean=31.33, n=109) to six months (mean=28.72, n=29) but stabilise
between six months and 12 months (mean=23, n=23). A further analysis of mean
levels of CDRIS-10 scores (paired samples tests: baseline to six months; six months
to 12 months and baseline to 12 months) confirms that there is a statistically
significant worsening of resilience at six months (these results were confirmed by
Wilcoxon Signed Ranks Test (asymptotic significance=0.036)).
Observations
1. A detailed evaluation of the ‘Post-Traumatic Rehabilitation Course (PTRC)
currently offered at the DMRC in order to establish its effectiveness in the
promotion of psychological resilience and management of those identified as
being ‘at risk’ in respect of Post-Traumatic Stress Disorder and Common
Mental Health Disorders and may provide a useful insight into promoting
resilience in the military cohort.
80
2. Further research is required within the general service population in order to
identify those factors influencing the development of resilience in UK military
personnel.
Health and lifestyle variables
Alcohol consumption
There is limited evidence to suggest that over the follow-up period there is a general
reduction in alcohol consumption as evidenced by a decrease in the maximum levels
of intake in the prospective patient participants. There is also evidence to support the
conclusion that rates of hazardous or harmful alcohol use decrease (as evidenced by a
fall in mean and median scores) over the same period. An initial analysis of the use of
alcohol appears to suggest that those individuals that reported themselves as
abstaining from the consumption of alcohol have higher mean PCL-C, and GHQ-12
scores and there is tentative evidence that suggests that the consumption of alcohol
(which includes the associated social aspects and benefits) may result in lower
reported levels of PTSD and raised levels of resilience at baseline. One explanation
of this phenomenon may be that levels of PTSD and common mental health disorder
are masked by the use of alcohol and that alcohol is employed as a coping
mechanism. In respect of the apparent positive effect of alcohol on resilience, it is
possible that the social aspect of alcohol and benefits that derive from the use of
alcohol within the military environment reinforce the positive effects of comradeship
and that the sense of ‘membership of the military family’ that this brings has a
positive effect of resilience.
Observation
1. The use of alcohol could potentially mask the symptoms of PTSD and
common mental health disorder in those with combat-injury. In addition to
the existing alcohol screening conducted at the DMRC by the Mental Health
team, alcohol screening conducted at ward level as part of the induction
process at each inpatient admission using the WHO’s AUDIT measure (Ware
et al, 1993) may provide an early indication of the adoption of alcohol as a
coping strategy in patients.
Organisations and social activities before injury
An initial analysis of the social activities engaged in by patient participants prior to
their injury suggested that ‘going to the gym with friends’ may have a positive effect
81
on their level of resilience at baseline but that it may have a negative effect on their
mental health in the longer term. In contrast, those patient participants who did not
report ‘going to the gym with friends’ had lower levels of resilience at baseline but
appeared to fare significantly better in terms of their mental health (GHQ-12).
Analysis of the primary outcome measures (PCL-C, GHQ-12, DRS-15 and CDRIS-
10) as independent samples at baseline, six months and 12 months indicated that the
effect might be statistically significant (see Table 38). While a statistically significant
difference was observed between those who reported ‘going to the gym with friends’
and those who did not report ‘going to the gym with friends’ prior to injury, these
results were NOT based on a consistent panel sample. The subsequent analysis of this
phenomenon using a consistent panel of participants failed to verify the finding.
While the subsequent analysis of these findings did not support the original
observation, one possible explanation offered of this initial observation was that
‘going to the gym with friends’ again promoted a sense of comradeship or
‘membership of the military family’ in participants as well as physical fitness and it
was these effects combined that promoted resilience at baseline. Participant
‘expectation’ was cited as an explanation for the observed difference in mean GHQ-
12 scores at baseline and it was suggested that those participants who reported ‘going
to the gym with friends’ prior to their injury had more difficulty in adjusting to their
injuries and to the alteration of their body-image and that perceived physical
limitations had a more profound effect on them than it had on those who had not
reported ‘going to the gym with friends’ prior to their injury.
Observation
2. Further research is required within the general service population in order to
confirm these findings and to further identify those factors influencing the
development of resilience in military personnel.
Relationships and social support
While there was some increase in relationship dissatisfaction reported by individuals
within the prospective patient group (01PG), the general level of satisfaction was
unchanged at 12 months. Use of the Social Support Inventory (SSI) to further assess
‘perceived and received criticism’ and patient participants’ perception of the ‘social
support’ offered by their partner indicated that the patient participants consistently
perceived their relationships as becoming less critical over the 12-month followed up.
82
In contrast, the analysis of the ‘social support’ sub-scale of the SSI indicated that
patient participants were generally less satisfied with the emotional and practical
support offered by their partners at six-months and at 12-months. While there was no
evidence to support the suggestion that there was a correlation between the reduction
in satisfaction measured and the primary outcome variables, the strict sample used to
analyse the differences in SSI over the 12 months was limited in size (n=11) and it is
recommended that these results be verified with a larger panel.
Qualitative findings (Partner study)
While it is difficult to draw any reliable conclusions from the qualitative data
provided by the partner participants, a number of themes emerge. While the partner
participants report a significant level of distress, as indicated by GHQ-12 and PCL-C
scores, their care is not the primary concern of the DMRC. The relationship between
the DMRC and the partners of those service-personnel treated there, may however be
contributing to those reported levels of distress. It must also be noted that the number
of partners responding to the open questions asked at baseline, six months and 12
months was very small and the effect of negativity bias must be considered.
Generally speaking, the partner participants reported a positive experience of
the care at the DMRC at baseline: 66.7% (n=10) reported that they found the staff to
be ‘very supporting’ or ‘supportive’: 80% (n=12) reported their perception that the
care provided to their partner was ‘good’ or ‘very good’ and 60% (n=9) reported that
they felt that the level of care provided made it ‘easier’ or ‘much easier’ to care for
their partner at home. These levels of satisfaction were not significantly different at
six months or 12 months.
In respect of the support offered to partner participants, a small number (n=4)
report having had no contact with staff at the DMRC and a similar number report not
having received any form of support. One consistent response indicates that the
provision of support does not extend to partners who were neither ‘wives’ nor ‘live-in
girlfriends’. Conversely, a number of partner participants report positive experiences
of support and identify individual members of staff whose willingness to “include and
explain” was significant in terms of providing positive emotional support.
Significantly, neither medical nor nursing staff were identified as providing positive
emotional support by partners and no distinction was made between civilian and
military staff in terms of this kind of support.
83
Partner participants consistently report a less positive experience of the
DMRC in relation to the provision of information regarding the care of their partner,
his or her recovery and the psychological impact of his or her injury (see Figure 32).
It should be recognised, however, that the provision of clinical information without
the consent of the patient would constitute a breach of patient confidentiality, and it
may be that patient participants were unwilling to give consent for the sharing of
clinical information with their partners.
Recommendations
We would like to make the following recommendations:
1. Extend the provision of care to the ‘wider’ family, e.g. to the partners of those
with combat-injuries, in order to further strengthen mutual social support and
to minimise the impact of those injuries on partner health.
2. Develop strategies designed to improve the longer-term employability of
patients following their discharge from the Defence Medical Rehabilitation
Centre (DMRC) and/or medical discharge from the Armed Forces.
3. Enhance the participation of partners of those with combat-injuries in the
delivery of their care, i.e. participation in consultation, direct care delivery and
through a more effective passage of information.
4. Develop and implement best-practice strategies designed to promote the
psychological resilience of those with combat-injuries in order to minimise the
risk of post-traumatic stress and common mental health disorders.
5. Identify and implement strategies designed to monitor alcohol consumption
and identify hazardous or harmful alcohol use by those with combat-injuries
during each period of inpatient admission at the DMRC.
84
References
Alexander DA (1996). Trauma research: a new era? Invited editorial. Journal of
Psychosomatic Research, 41(1), 1-5.
Alexander DA, Elliott KGB, Atcheson S & Hutchison J (1995). Accidental injury:
psychological outcomes and their prognostic indicators. Abstract, 4th European
Conference on Traumatic Stress, Paris.
American Congress of Rehabilitation Medicine. Definition of mild traumatic brain
injury. Journal of Head Trauma Rehabilitation, 8, 86-7.
American Psychiatric Association (1980). Diagnostic and Statistical Manual of
Mental Disorders (DSM III). Washington DC: American Psychiatric
Association.
Babor TF, Biddle-Higgins JC, Saunders JB & Monteiro MG (2001). AUDIT: The
Alcohol Use Disorders Identification Test: Guidelines for Use in Primary
Health Care. Geneva, Switzerland: World Health Organization.
Bartone PT, (2007). Test-retest reliability of the dispositional resilience scale-15, a
brief hardiness scale. Psychological Reports, 101, 943-944.
Batten SV & Pollack SJ (2008). Integrative outpatient treatment for returning service
members. Journal of Clinical Psychology, 64(8), 928-39.
Blanchard EB, Jones-Alexander J, Buckley TC & Forneris CA (1996). Psychometric
properties of the PTSD Checklist (PLC). Behaviour Research and Therapy,
34(8), 669-73.
Brewin CR, Andrews B & Valentine JD (2000). Meta-analysis of risk factors for
posttraumatic stress disorder in trauma-exposed adults. Journal of Consulting
and Clinical Psychology, 68(5), 748-66.
Brown ES, Fulton MK, Wilkeson A & Petty F (2000). The psychiatric sequelae of
civilian trauma. Comprehensive Psychiatry, 41(1), 19-23.
Browne T, Hull L, Horn O, Jones M, Murphy D, Fear NT, Greenberg N, French C,
Rona RJ, Wessely S & Hotopf M (2007). Explanations for the increase in
85
mental health problems in UK reserve forces who have served in Iraq. British
Journal of Psychiatry, 190, 484-9.
Chalder T, Berelowitz G, Pawlikowska T, Watts L, Wessely S, Wright D & Wallace
EP (1993). Development of a fatigue scale. Journal of Psychosomatic Research,
37(2), 147-53.
Collins RC & Kennedy MC (2008). Serving families who have served: Providing
family therapy and support in interdisciplinary polytrauma rehabilitation.
Journal of Clinical Psychology: In Session, 64(8), 993-1003.
Connor KM and Davidson JRT (2003). Development of a new resilience scale: The
Connor-Davidson Resilience Scale (CD-RISC), Depression and Anxiety, 18, 76-
82.
Cook JM, Riggs DS, Thompson R, Coyne JC & Sheikh JI (2004). Posttraumatic stress
disorder and current relationship functioning among World War II ex-prisoners
of war. Journal of Family Psychology, 18(1), 36-45.
Creamer M, Burgess P & McFarlane AC (2001). Post-traumatic stress disorder:
findings from the Australian National Survey of Mental Health and Well-being.
Psychological Medicine, 31(7), 1237-47.
Department of Health (1999). National service framework for mental health: modern
standards and service models. London: Department of Health.
Dirkzwager AJ, Bramsen I, Adèr H & van der Ploeg HM (2005). Secondary
traumatization in partners and parents of Dutch peacekeeping soldiers. Journal
of Family Psychology, 19(2), 217-26.
Engelhard IM, van den Hout MA, Weerts J, Arntz A, Hox JJ & McNally RJ (2007).
Deployment-related stress and trauma in Dutch soldiers returning from Iraq.
Prospective study. British Journal of Psychiatry, 191, 140-5.
Erbes CR, Polusny MA, Macdermid S & Compton JS (2008). Couple therapy with
combat veterans and their partners. Journal of Clinical Psychology, 64(8), 972-
83.
86
Fayers PM & Machin D (2001). Quality of Life. Assessment, Analysis and
Interpretation. Chichester: John Wiley & Sons.
Figley CR (1998). Burnout in Families: The Systemic Costs of Caring. CRC Press:
LLC.
Fletcher K (2007). Combat Stress (The Ex-Services Mental Welfare Society),
Veterans and Psychological Trauma. In: H Lee & E Jones (eds) War and
Health: Lessons from the Gulf War. John Wiley & Sons Ltd, Chapter 5.
French LM & Parkinson GW (2008). Assessing and treating veterans with traumatic
brain injury. Journal of Clinical Psychology, 64(8), 1004-13.
Goff BS, Crow JR, Reisbig AM & Hamilton S (2007). The impact of individual
trauma symptoms of deployed soldiers on relationship satisfaction. Journal of
Family Psychology, 21(3), 344-53.
Goldberg D & Williams P (1988). A User's Guide to the General Health
Questionnaire. Windsor, UK: NFER-Nelson.
Healthcare Commission (2009). Defence Medical Services: A review of the clinical
governance of the Defence Medical Services in the UK and overseas. London:
Commission for Healthcare Audit and Inspection.
Hill JJ, Mobo BH & Cullen MR (2009). Separating deployment-related traumatic
brain injury and posttraumatic stress disorder in veterans: Preliminary findings
from the Veterans Affairs traumatic brain injury screening program. American
Journal of Physical Medicine and Rehabilitation, 88(8), 605−614.
Hoge CW, Castro CA, Messer SC, McGurk D, Cotting DI & Koffman RL (2004).
Combat duty in Iraq and Afghanistan, mental health problems, and barriers to
care. New England Journal of Medicine, 351, 13-22.
Hoge CW, Toboni HE, Messer SC, Bell N, Amoroso P & Orman DT (2005). The
occupational burden of mental disorders in the U.S. military: psychiatric
hospitalizations, involuntary separations, and disability. American Journal of
Psychiatry, 162(3), 585-91.
87
Holeva V, Tarrier N & Wells A (2001). Prevalence and predictors of acute stress
disorder and PTSD following road traffic accidents: thought control strategies
and social support. Behavioural Therapy, 32, 65-81.
Hooley JM & Teasdale JD (1989). Predictors of relapse in unipolar depressives:
expressed emotion, marital distress and perceived criticism. Journal of
Abnormal Psychology, 98, 229-35.
Hotopf M & Wessely S (2005). Can epidemiology clear the fog of war? Lessons
from the 1990-91 Gulf War. International Journal of Epidemiology, 34(4), 791-
800.
Hotopf M, Hull L, Fear NT, Browne T, Horn O, Iversen A, Jones M, Murphy D,
Bland D, Earnshaw M, Greenberg N, Hughes JH, Tate AR, Dandeker C, Rona
R (2006). The health of UK military personnel who deployed to the 2003 Iraq
War: a cohort study. Lancet 367(9524), 1-12.
Iversen AC, Fear NT, Ehlers A, Hacker-Hughes J, Hull L, Earnshaw M, Greenberg N,
Rona R, Wessely S & Hotopf M. (2008). Risk factors for post-traumatic stress
disorder among UK Armed Forces personnel. Psychological Medicine, 36(4),
511-22.
Klein S & Alexander DA (2005). A prospective longitudinal study: the “Ripple
Effect” of trauma on victims’ partners. Final Report prepared for Tenovus-
Scotland.
Klein S & Alexander DA (2006). Epidemiology and presentation of post-traumatic
disorders. Psychiatry, 5, 225-7.
Klein S, Alexander DA, Hutchinson JD, Simpson JA, Simpson JM & Bell JS (2002).
The Aberdeen Trauma Screening Index: an instrument to predict post-accident
psychopathology. Psychological Medicine, 32(5), 863-71.
Koren D, Norman D, Cohen A, Berman J & Klein EM (2005). Increased PTSD risk
with combat-related injury: a matched comparison study of injured and
uninjured soldiers experiencing the same combat events. American Journal of
Psychiatry, 162(2), 276-82.
88
Kulka RA, Schlenger WE, Fairbank JA, Hough RL, Jordan BK, Marmar CR & Weiss
DS (1990). Trauma and the Vietnam War Generation: Report of Findings from
the National Vietnam Veterans Readjustment Study. New York, Brunner/Mazel.
Lapierre CB, Schwegler AF & Labauve BJ (2007). Posttraumatic stress and
depression symptoms in soldiers returning from combat operations in Iraq and
Afghanistan. Journal of Traumatic Stress, 20, 933-43.
Luthar SS, Ciccheti D & Becker B (2000). The construct of resilience: A critical
evaluation and guidelines for future work. Child Development, 71, 543-62.
Michaels AJ, Michaels CE, Moon CH, Smith JS, Zimmerman MA, Taheri PA &
Peterson C (1999). Posttraumatic stress disorder after injury: impact on general
health outcome and early risk assessment. Journal of Trauma, 47, 460–7.
National Audit Office (2010). Ministry of Defence: Treating injury and illness arising
on military operations. London: National Audit Office.
O’Donnell ML, Bryant RA, Creamer M & Carty J (2008). Mental health following
traumatic injury: toward a health system model of early psychological
intervention. Clinical Psychology Review, 28(3), 387-406.
O’Donnell ML, Creamer M, Bryant RA, Schnyder U & Shalev A (2003).
Posttraumatic disorders following injury: an empirical and methodological
review. Clinical Psychology Review, 23, 587-603.
O'Donnell ML, Creamer M & Pattison P (2004). Posttraumatic stress disorder and
depression following trauma: understanding comorbidity. American Journal of
Psychiatry, 161(8), 1390-6.
Ozer EJ, Best SR, Lipsey TL & Weiss DS (2003). Predictors of posttraumatic stress
disorder and symptoms in adults: a meta-analysis. Psychological Bulletin,
129(1), 52-73.
Robinson BC (1983). Validation of a Caregiver Strain Index. Journal of Gerontology,
38, 344-8.
89
Sammons MT & Batten SV (2008). Psychological service for returning veterans and
their families: Evolving conceptualization of the sequelae of war-zone
experiences. Journal of Clinical Psychology: 64(8), 921-927.
Sayer NA, Chiros CE, Sigford B, Scott S, Clothier B, Pickett T & Lew HL (2008).
Characteristics and rehabilitation outcomes among patients with blast and other
injuries sustained during the Global War on Terror. Archives of Physical
Medicine and Rehabilitation, 89(1), 163-70.
Unwin C, Blatchley N, Coker W, Ferry S, Hotopf M, Hull L, Ismail K, Palmer I,
David A, & Wessely S (1999). Health of UK servicemen who served in Persian
Gulf War. Lancet, 353, 169-78.
Vasterling JJ, Verfaellie M & Sullivan KD (2009). Mild traumatic brain injury and
posttraumatic stress disorder in returning veterans: perspectives from cognitive
neuroscience. Clinical Psychology Review, 29(8):674-84.
Ware JE, Kosinski M, Bjorner JB, Turner-Bowker DM, Gandek B & Maruish ME
(2008) SF-36v2® Health Survey: administration guide for clinical trial
investigators. Lincoln, RI: QualityMetric Incorporated.
Wessely S (2005). Risk, psychiatry and the military. British Journal of Psychiatry,
186, 459-466.
90
Aberdeen Centre for Trauma Research, Faculty of Health & Social Care, The Robert Gordon University, Garthdee Road, ABERDEEN AB10 7QG
Tel no: +44 (0)1224 263100 Fax no: +44 (0)1224 263109