Sex differences in interdisciplinary pain rehabilitation outcomes: a systematic review

Chronic pain

Chronic pain is defined as non-cancer pain lasting >3 months. This includes, but is not limited to, musculoskeletal pain (e.g. back, neck, joint, widespread pain), abdominal pain, neuropathic pain, headache, perioperative pain, and inflammatory pain conditions.

Interdisciplinary pain rehabilitation programs (IPRPs)

Interdisciplinary pain rehabilitation programs (IPRPs) are defined as comprehensive treatment programs that include a common philosophy of rehabilitation and active patient involvement [12]. Multiple healthcare providers participate in treatment and all care is provided at the same facility. For this review, rehabilitation programs must be delivered by an interdisciplinary team that includes at least three professions, such as physician, psychologist, physical therapist, and occupational therapist.

Depressed mood

Depressed mood refers to symptoms of low mood, absence of positive affect (loss of interest and enjoyment), and a range of associated emotional, cognitive, physical, and behavioral symptoms. Validated measures of depressed mood used in this review include the Beck Depression Inventory (BDI), the Center for Epidemiologic Studies Depression Scale (CES-D), and the Hospital Anxiety and Depression Scale (HADS).

Pain catastrophizing

Pain catastrophizing is broadly defined as an exaggerated negative cognitive and emotional response to pain that may include magnification of the threat value of pain, repetitive negative thinking (rumination) about pain-related information, and a sense of helplessness in the face of pain. Validated measures of pain catastrophizing often seen in IPRP literature are the Pain Catastrophizing Scale (PCS), the Coping Strategies Questionnaire-Catastrophizing subscale (CSQ-C), and the Pain-Related Self Statements Scale (PRSS).

Pain interference

Pain interference is defined as the average rating for responses regarding challenges in performing daily, social, or work-related tasks. Validated measures of pain interference used in this review include the Brief Pain Inventory Interference Scale (BPI), Pain Interference Index (PII), and the pain interference subscale of the Multidimensional Pain Inventory (MPI).

Pain intensity/severity

Pain intensity/severity is defined as the descriptive rating of the pain experience. Commonly used measures of pain intensity/severity include the Numerical Pain Rating Scale (NPRS), the Visual Analogue Scale (VAS), and the pain severity subscale of the Multidimensional Pain Inventory (MPI).

Sex

“Sex” refers to “biologically based differences” while the term “gender” refers to “socially based phenomena.” [31] However, much scientific literature conflates the terms “sex” with “gender;” it is the authors’ intension to appropriately use the word “sex” when describing differences in outcome data unless directly referencing the work of another author where the two words are used interchangeably.

Inclusion criteria

1. Treatment studies with at least 50 adult participants (to provide sufficient statistical power for analysis of pre/post outcome measures) experiencing chronic pain

2. Report data from an interdisciplinary pain rehab program (IPRP)

3. Peer reviewed and published in English in full

4. Mention sex- or gender-related outcomes in the manuscript title or abstract

5. Involve a measure of depression, pain catastrophizing, pain interference, and/or pain intensity/pain severity

6. Report pre- and post-treatment outcomes

Exclusion criteria

Qualitative or observational studies along with protocols, methodology studies, reviews (as the purpose of this study was to analyze primary literature that exists on this topic), conference abstracts, and dissertations were not included.

Eligibility for inclusion/exclusion was assessed via independent duplicate manual screening of article titles and abstracts by the first two authors. Disagreements were resolved by discussion. The same process was repeated when reviewing all relevant full-text articles. A supplementary search method, including a review of reference lists of included full-text articles (e.g. bread crumb search) was conducted by the first author with inclusion criteria verified and agreed upon by the second author.

Participants

The number of study participants ranged from 66 to 14,666 [28, 32, 33]. Studies were roughly equal in having more male or female participants (n = 10) with a greater number of female participants. Participants were most represented by studies taking place in the United States (n = 7); other countries of study origin included Sweden (n = 4), Germany (n = 2), Canada (n = 1), France (n = 1), Norway (n = 1), and the United Kingdom (n = 1). Most study participants had pain in multiple sites, except for two studies examining outcomes for patients with chronic low back pain (CLBP) [34], one study looking at chronically disabled work related spinal disorder (CDWRSD) [35], and another study specifically looking at patients with fibromyalgia [32]. Patients were generally referred from the community for treatment except for Murphy et al., which looked a population of veterans in the United States [27]. Not all studies included a mean length of time that patients had experienced chronic pain; Racine et al. [29] reported a mean length of chronic pain for study participants of 10 years and Silvemark et al. [25] reported a mean length of chronic pain for study participants of 6 years.

Interventions

A brief summary of the included study demonstrated that the most frequent professions included in IPRPs were physicians, psychologists, and physical therapists (PTs). The shortest program duration was 3 weeks [27, 32, 34, 36], [37], [38] and the longest program duration was 57 weeks [39]. Programs were a mixture of inpatient and outpatient treatment modalities, as well as a mixture of group and individual formats. Medical treatment involved assessment of fitness for program inclusion, education and guidance, opioid tapers, ongoing monitoring of symptoms throughout program, and medication management. Behaviorally focused psychological interventions, including cognitive-behavioral therapy (CBT) and acceptance and commitment therapy (ACT), were the only specific psychological interventions named. Other, non-specific psychological components of treatment included empowerment, biofeedback, stress management, and vocational reintegration. Elements of physical training included education, cardiovascular exercises, active stretching, proprioceptive exercises, posture, balance and coordination, and activity moderation. In addition, two programs also used social work services and included elements of diet and nutrition [40] while Murphy et al. utilized recreational therapy and added aqua therapy to treatment options [27].

Outcomes

Out of the 17 articles, all 17 used a measure of pain intensity/severity, 13 used a measure of depression, 5 used a measure of pain catastrophizing, and 6 used a measure of pain interference.

Pain intensity/severity

All 17 studies included in this review used an outcome measure of pain intensity/severity. Overall, most studies found no differences in pain intensity outcomes – both men and women reported significant improvement in pain intensity after completing an IPRP with no differences between the sexes (Table 2). This is consistent with the Racine et al. [29] summary of literature trends discussed previously. Only five out of 17 studies showed any significant sex differences for pain intensity [27, 28, 38, 40, 41]. Racine et al. [29] demonstrated significant sex by time interactions emerged for pain intensity (p = 0.019) outcomes with small to moderate magnitude effect sizes (η² = 0.03). Specifically, female participants, but not their male counterparts, showed pre- to post-treatment improvements in pain intensity.

The remaining studies with significant findings indicated significant sex differences only at follow-up; however, findings were mixed. Three studies indicated that improvements in pain intensity on the Numeric Pain Rating Scale (NPRS) were only maintained by male patients when measured again at follow-up [26, 27, 40]. In contrast, Pieh et al. [39] found that women’s pain levels on the NPRS improved to a greater level of magnitude at follow-up compared to men. Spinord et al. found that women maintained significant improvement in pain intensity using the multidimensional pain inventory (MPI) at one-year follow up; however, both men and women maintained significant improvement on the NPRS [40]. The remaining nine studies found no sex differences in pain intensity treatment outcomes.

Depression

Out of the 13 studies that included an outcome measure of depression, 10 of these studies did not find sex differences in depression outcomes after completing an IPRP (Table 3). Out of the three studies that did find sex differences [34, 37], both men and women showed improvement in their depressive symptoms, but women reported a greater magnitude of change in depressive symptoms compared to men at time of follow-up. Booster sessions for all participants, but especially men, were suggested [34, 37]. Gerdle et al. [28] reported that women demonstrated larger changes in depressive symptoms at 12 months follow up; however, these differences were not clinically significant. The remaining 10 studies found no sex differences.

Pain catastrophizing

Five of the included studies included an outcome measure of pain catastrophizing (Table 4). Overall, there was a significant reduction in pain catastrophizing for both sexes from pre- to post-treatment, and from pre- to follow up after completion of treatment with no significant differences between men and women. All included studies found no sex differences in pain catastrophizing outcomes at IPRP discharge. However, one study [38] included a 3 month follow up and found that the reduction in the helplessness scale of the PCS was only maintained by men at the follow up time point. Women returned to levels found at pre-intervention which were substantially higher than men. The authors also indicated that catastrophizing mediated the effect of gender differences on increasing pain after treatment. As this was the only study out of the five to include follow up data, further research into the long-term efficacy of IPRPs on pain catastrophizing by sex and PCS subscales appears warranted. This was the only outcome measure included in this study that did not have any sex differences.

Pain interference

Six of the included studies used an outcome measure of pain interference and all studies reported a significant treatment effect, with male and female participants of IPRPs showing reduced pain related interference between initial testing and discharge (Table 5). Findings were mixed in terms of sex differences and appear to be influenced by the measures used. In the three studies that utilized the MPI, women and men did not differ significantly regarding reductions in pain related interference. When looking at articles using the PDI, Pieh et al. [39] found that while there was no statistical difference between sex at baseline, women showed a slightly larger reduction in pain related disability via PDI scores. This study indicated that women benefited slightly more from treatment. Huffman et al. [36] also found that women were slightly more likely to experience a greater magnitude of reduced interference compared to men. While all individuals in the study showed an initial positive treatment effect, being male was associated with worsening PDI scores over time. It is difficult to say the reason that these outcomes related to use of the MPI or the PDI are inconsistent. While the MPI and PDI are both measuring the construct of interference, it is unclear if one measure may be more sensitive to the impact of sex than the other. Future research utilizing comparing both measures in men and women may help elucidate these findings.