Military veterans with and without post-traumatic stress disorder: results from a chronic pain management programme

1 Introduction

There is a high prevalence of co-morbid chronic pain and post-traumatic stress disorder (PTSD) in military veterans seeking pain management: 49% of patients at a US Veterans’ Administration (VA) health care facility [1]; 21.7% of Croatian homeland war veterans [2]; and in US Armed Forces deployed in Iraq and Afghanistan, 81.5% reported chronic pain and 68.2% PTSD [3]. Williamson et al. [4] quotes a Ministry of Defence statistic that up to 60% of medical discharges within the United Kingdom Armed Forces (UK AF) between 2001 and 2014 were for musculoskeletal injuries with 4% of regular personnel suffering from PTSD and 9% of deployed ex-regular veterans were found to experience PTSD. Comorbidity between medical conditions and mental health is significant, with medical conditions correlating with anxiety and mood disorders, including work loss. A multidisciplinary team approach as the treatment of choice is being suggested [5], [6].

There is growing interest in the relationship between chronic pain and post-traumatic stress disorder (PTSD) and identifying effective treatments when these conditions present together. Psychological models proposed to explain the association of chronic pain and post-traumatic stress remain largely hypothetical [7], [8], [9]. The recognition of interaction between chronic pain and post-traumatic stress has generated recommendations for integrated treatment [10], [11], [12], for example, behavioural activation using a collaborative approach involving primary care, mental health, and other clinicians [12], Cognitive-Behavioural Therapy (CBT), interdisciplinary pain programmes, and exposure-based interventions [11]. There are some important differences from standard chronic pain programmes, particularly the need to address the common myth among veterans of “no pain, no gain”, and their strategic use of pain to distract from post-traumatic stress symptoms [13].

Guidelines exist for treatment of chronic pain [14], [15] and for post-traumatic symptomatology [16], [17], but few studies report on combined and simultaneous treatment of chronic pain and PTSD. For example, Otis et al. [18] describes an integrated treatment model, combining cognitive processing therapy (designed for post-traumatic stress disorder (PTSD) and includes elements of CBT) and CBT for pain management in a trial for integrated treatment of veterans with comorbid chronic pain and PTSD. Despite a high drop-out rate, the integrated treatment model appears to be beneficial and further clinical trials and evaluation are in progress.

Health services often lack the expertise to assess and manage pain conditions in the veteran population [19] or to fully understand the complexities of comorbid chronic pain and PTSD [1], [10]. Veterans with both chronic pain and post-traumatic stress symptoms may be offered the treatments in sequence, but there are differences in clinical opinion on the best order of treatment. This can result in veterans being excluded from each service until their other problem is resolved, meaning they receive no treatment and the experience itself is likely to increase negative mood. Few data exist on UK veterans with comorbid chronic pain and post-traumatic stress symptoms [7], with studies mainly focused on mental health or musculoskeletal problems [4], [20], [21], but not the combination, which is common [22], [23], [24]. Data are not available on support and treatment provided by local services for UK veterans with physical health problems, such as musculoskeletal pain, and some individuals may require long-term support [4].

The first interdisciplinary veteran-specific pain management programme (PMP) in the UK was established with charitable funding in a private hospital in London in 2015. This study reports on the first 158 patients, investigating the differences in clinical outcomes between those with and without PTSD.

2 Methods

3 Results

Of the 164 veterans who started treatment, six dropped out; two because of employment demands, three for medical or mental health reasons, and one unknown. Information concerning the 158 veterans who completed treatment is detailed in Table 2. Overall attendance of sessions was 88%, with sessions missed for similar reasons as for drop out, flare-up of post-traumatic stress symptoms or because of difficulties with travel from distant parts of the country for single days.

Of the 148 veterans that were deployed, most served in Afghanistan and Iraq, The Falkland Islands and Northern Ireland. Sixteen veterans were not deployed during their service. All veterans reported pain in more than one site.

Baseline scores for the 158 who completed treatment are divided according to diagnosis of PTSD (Table 3). Clinical diagnosis of post-traumatic symptoms were borne out by scores on the IES-6, with the PTSD group (n=95) scoring well above the threshold of clinical concern, and those without scoring well below that threshold (n=63) (see Table 3). The PTSD group also reported poorer mental wellbeing (CORE-10) at the start of treatment, but there were no other differences.

Repeated measures ANOVA from day 1 to day 10 showed statistically significant improvements at the group level for all variables except the IES-6 (Table 4), but no interaction with PTSD group for any outcome; that is, both those with and without PTSD improved with treatment.

Additionally, at end of treatment the PTSD group when compared to the no PTSD group had made statistically significant reductions on the CORE-10 and on the IES-6 (see Table 5). However, there was substantial missing data at day 10 in both PTSD and No PTSD groups, 31 and 23, respectively, mainly due to administrative error [12], work [11], pain flare-up [8], and post-traumatic stress symptoms [19].

Figure 1 shows reliable changes [41] in confidence in activity despite pain (PSEQ) and pain catastrophising (PCS); those making reliable improvement far outnumbered those who reliably worsened, only a few of the latter for each outcome. The diagonal lines indicate the line of no change from pre- to post-treatment, with measurement error marked either side. Very few veterans scored worse after treatment. Also shown on both figures is a reference level from large-sample norms [41]: for the PCS it shows the proportion of the treated population who achieved or exceeded the threshold score of clinical concern (dark blue horizontal line); for the PSEQ it shows the mean score above which an population with chronic pain was more likely to be at work than not.

Fig. 1:

Clinically significant change in catastrophizing (PCS) and confidence in activity despite pain (PSEQ).

4 Discussion

Veterans with and without PTSD, treated in the PMP adapted for post-traumatic stress symptoms, made statistically and clinically significant improvements in pain, pain interference with life, and psychological health, with clear clinical improvement in post-traumatic stress symptoms and overall mental health for those scoring high at the start of the programme, in the non-clinical range at discharge. Results suggest that it is clinically appropriate to treat veterans with both conditions within the same treatment model. Comparison of the outcomes at day 10 of our veteran specific PMP with a 16-day residential UK PMP [42] with few military veteran participants shows similar improvements were achieved in both PMPs where the same scales were used. It is harder to compare findings from the UK, where armed forces are medically discharged back into the National Health System and from the USA, where military veterans are treated under multiple health insurance schemes, predominantly the Veterans’ Administration.

This study is a worthwhile addition to the scarce data on treatment of veterans presenting with high rates of chronic pain and PTSD. Although it cannot be extrapolated from the data presented, clinical experience of the treatment team puts emphasis on adapting usual PMP content and process to address veterans’ tendency to over-compensate physically to manage emotional distress, requiring flexibility in and adaptation of goal setting and paced progression towards goals. Staff members’ years of experience in pain management and team cohesion made such adaptations feasible. Additionally, veterans have a strong sense of shared identity and core values, so that engagement and group dynamics and cohesion benefited substantially from providing a veterans’ only PMP. Easier access to similar programmes would benefit the large numbers of UK military veterans with chronic pain.

There are several limitations to this study. There was no follow-up beyond the last day of the programme (albeit at 4–6 months from the start), so no evidence can be provided for maintenance of treatment gains in the longer term. Missing values were a problem for some scales, particularly the IES-6. Veterans with severe PTSD were excluded from the PMP, more because of difficulty providing sufficient psychological and psychiatric cover during treatment than because treatment was thought to be unsuitable, but it means that results cannot be extrapolated to veterans with chronic pain and any level of PTSD.

This paper supports the adaptation of pain management programmes for chronic pain to treat veterans with comorbid PTSD, as well as those without. Further research is required on whether there are measurable benefits to treating military veterans together rather than among civilian patients.