Table 1.

Main characteristics of the studies included in the systematic review

Ref.	Study aims	Sample size (n)	Participants’ origin	Mean age (SD)	Women (%)/Men (%)	Chronic pain + PTSD + Opioid intake participants (%)	Variables and measures	Study design	Statistical analyses	Summary of results related to current study aims	Conclusions related to current study aims
Bilevicius et al[40], 2018	To assess independent and combined contributions of PTSD and chronic pain conditions (digestive, nerve, and musculoskeletal pain) on opioid use disorder (OUD)	n = 36309	Civilian, non-institutionalised United States residents	Not given	44/56	3	Chronic pain: ICD-10; PTSD: AUDADIS-5; Opioids (abuse): AUDADIS-5	Retrospective cohort study	Multivariate logistic regression	PTSD was associated with OUD for musculoskeletal pain [adjusted odds ratio (AOR1): 4.2 95%CI: 2.54-7.12, P < 0.001) and nerve pain (AOR1: 3.1 95%CI: 1.93-5.10, P < 0.001), but not for digestive pain (AOR1: 1.8, 95%CI: 0.85-3.82, P = 0.124)	Comorbidity between PTSD and musculoskeletal/nerve chronic pain is a vulnerability risk factor for OUD. These patients should be carefully screened for opioid use, regardless of whether they are seeking a prescription
Han et al[41], 2017	To analyse the characteristics (including PTSD) associated with long-term opioid dosing trends among veterans with chronic musculoskeletal pain	n = 79015	Veterans’ health care	29.80 (9.10)	11/89	49	Chronic pain: ICD-9CM + NRS; PTSD: (ICD-9-CM); Opioids: Dispensing pattern (daily dose, total number of days, and number of prescription episodes)	Retrospective cohort study	General estimating equation (GEE), GEE-logistic models	PTSD, major depression disorder, and substance abuse disorder were associated with 30% increased odds of high-dose opioid prescribing. Veterans with these disorders had both higher log (ME/d) opioid dose (adjusted mean difference: 0.038, 0.057, 0.063; all P values < 0.0001) and greater odds of high-dose prescribing (adjusted OR, 1.31; 1.36; 1.32; P = 0.008, 0.001, 0.002) than those without each diagnosis, respectively. Excluding PTSD did not modify the predicted dosing trends of the average log (βt = 0.03, P < 0.001; βt2 = −0.003, P = 0.013) or the odds of high-dose prescribing (βt = 0.19, P = 0.035; βt2 = −0.03, P = 0.060)	Veterans who were dispensed opioids for an extended duration or who had a mental health diagnoses tend to receive high opioid dose therapy regardless of PTSD diagnosis. Future studies are needed to assess the potential impact of opioid dosing trends on clinical outcomes and effectiveness of emerging intervention programs targeting high-risk opioid prescribing
Hudson et al[39], 2017	To examine the pharmacoepidemiology of opioid use among veterans with chronic pain who are regular users of veterans' health care	n = 1397946	Inpatients/or outpatients veterans' health care	34.10 (9.70)	12/88	10	Chronic pain: ICD-9 + NRS; PTSD: NEPEC; Opioids: Dispensing pattern (PBM records)	Retrospective cohort study	Multivariate logistic regression	The percentage of chronic opioid users with PTSD (58%) was higher than nonchronic users (41%) or those with no opioid use (29%). Veterans with PTSD were more likely to receive opioids chronically (OR: 1.22; 95%CI: 1.20–1.25; P = 0.01) than those with a diagnosis of major depressive or tobacco use disorders	A diagnosis of PTSD was strongly associated with chronic prescription and use of opioids
Liebschutz et al[37], 2010	To analyse the characteristics associated with prescription drug use disorder (PDUD) in patients with chronic pain	n = 597	Primary care	45.51 (9.16)	Not given	14	Chronic pain: GCPS PTSD: CIDI, v.2.1 (PTSD module); Opioids (abuse): CIDI v.2.1 (Drug disorders module)	Cross-sectional study	Multivariate logistic regression	The percentage of PTSD (28%) in the PDUD group was higher than in those with no substance use disorder (17%) but not in those with a substance use disorder (33%). PDUD patients had the highest percentage of PTSD (52%). Along with other variables, PTSD was associated with greater odds of both PDUD and substance abuse disorder (OR: 1.93; 95%CI: 1.09–3.43; P = 0.01)	PTSD was associated with a high probability of PDUD. Physicians treating patients with pain should screen for PTSD to help identify those at the highest risk of PDUD
Macey et al[42], 2011	To examine variables (including PTSD) associated with opioid prescription in chronic pain patients	n = 762	Veterans' health care service.	34.00 (8.80)	15/85	40	Chronic pain: NRS; PTSD: ICD9-CM; Opioids: Dispensing pattern (days of prescription within 12 mo)	Retrospective cohort study	Multivariate logistic regression	Veterans prescribed opioids long-term had a greater prevalence of PTSD, major depressive disorder, and nicotine use disorder, but the proportion of users with both a short-term and long-term prescription was higher for those with PTSD (56% and 69%, respectively). A PTSD diagnosis was associated with increased opioid prescriptions (OR = 1.42, 95%CI: 1.04-1.96)	PTSD is associated with a greater likelihood of receiving a prescription for an opioid medication, which suggests the need for improvement in implementing guideline-level pain care for these veterans
Outcalt et al[43], 2013	To examine health care utilization and dispensed medication (including opioids) among veterans with both PTSD and chronic pain	n = 40716	Primary care, pain-related specialty or mental health pain veteran outpatients	Not given	5/95	Not given	Chronic pain: ICD-9; PTSD: ICD-9+PC-PTSD screen; Opioid: Pharmacy prescription dispensing data	Retrospective cohort study	Negative binomial regression and sensitivity analyses	Adjusted rates of opioid medication prescriptions were significantly higher for the chronic pain and PTSD group (mean = 1.47, SD = 0.86) than either of the comparison groups [pain only (mean = 0.92, SD = 0.51) and PTSD only (mean = 0.17, SD = 0.10) (P < 0.001)]	Opioids were more commonly prescribed for pain relief in veterans with both chronic pain and PTSD symptoms than in those with pain or PTSD alone. More integrated and streamlined treatments for this clinical population are needed
Seal et al[44], 2012	To analyse the effect of mental health disorders, particularly PTSD, on patterns of opioid prescription, associated risks, and adverse outcomes	n = 141029	Veterans’ health care	Not given	11/89	11	Chronic pain: ICD-9-CM; PTSD: ICD9-CM; Opioid: Dispensing pattern (≥ 20 consecutive days of prescription within 12 m	Retrospective cohort study	Poisson regression	Compared to veterans without a mental health diagnosis (6.5%) and veterans with mental health diagnoses other than PTSD (11.7%), 17.8% (adjusted RR, 2.58; 95%CI: 2.49-2.67) those with a PTSD diagnosis were significantly more likely to be prescribed opioids. Veterans with a drug use disorder and comorbid PTSD were more likely to be prescribed opioids than veterans with no mental health disorders (33.5% vs 6.5%; adjusted RR, 4.19; 95%CI: 3.84-4.57). Those with PTSD with prescription opioids were significantly more likely to be in the highest quintile by dose (22.7% vs 15.9%; adjusted RR, 1.42; 95%CI: 1.31-1.54), receive more than 1 type of opioid concurrently (19.8% vs 10.7%; adjusted RR, 1.87; 95%CI: 1.70-2.06), and obtain early opioid refills (33.8% vs 20.4%; adjusted RR, 1.64; 95%CI: 1.53-1.75)	Findings support further efforts to improve care of patients with comorbid pain and PTSD, given the heightened risk of self-medication with opioids and substance abuse in veterans with PTSD. Integrated treatments that simultaneously target mental health disorders and pain are effective for both problems and may decrease harm resulting from opioid therapy
Seal et al[45], 2018	To test the hypothesis that among veterans with chronic pain diagnoses, greater traumatic brain injury severity and mental health comorbidity independently predict subsequent initiation of short- and long-term opioid therapy	n = 53124	Veterans’ health care (veterans with traumatic brain injury)	32.50 (8.50)	7/93	17	Chronic pain: not given; PTSD: ICD9-CM Opioids: Dispensing pattern (days of prescription within 12 mo)	Retrospective cohort study	Multivariate logistic regression	A PTSD diagnosis (69.4%) was significantly associated with subsequent long-term opioid therapy (RR, 1.98; 95%CI: 1.67-2.34) and, to a lesser extent, short-term opioid therapy (RR, 1.23; 95%CI: 1.15-1.31) after controlling for sex, race/ethnicity, marital status, rank and education, military component, branch of service, number of deployments, antidepressant medication use, alcohol disorders, drug disorders, and self-rated pain disability	Comorbid mental health problems substantially increase risk when initiating short- and long-term opioid therapy in veterans with moderate to severe traumatic brain injury, PTSD, and depression. There is a need to provide enhanced education and interdisciplinary behavioural health support to primary care providers who care for veterans with complex chronic pain that includes TBI and comorbid mental health problems
Trevino et al[36], 2013	To examine opioid use and psychological characteristic (including PTSD) associated in traumatically injured individuals with chronic pain (secondary to physical injuries) 4 months post trauma	n = 101	Trauma surgery inpatients' service	Not given	Not given	Not given	Chronic pain: BPI-SF; PTSD: PCL-C; Opioid: Opiates use (yes/no)	Prospective cohort study	Multivariate analysis of covariance	There was a statistically significant difference between those using narcotics and those not using narcotics at 4 mo posttrauma on the combined dependent variables (pain, life interference, depression, anxiety, PTSD, and length of hospital stay [F (6, 68) = 2.7, P = 0.02; Wilks’ lambda = 0.81; partial eta squared = 0.2 Bonferroni adjusted level = 0.007). PTSD was higher in patients with chronic pain and opioid use (mean = 44, SD = 3.6) than in the non-opioid use groups (mean = 32, SD = 2.7)	Opioid prescription should be used carefully in traumatically injured patients, especially in individuals with comorbid psychological pathology such as PTSD. Viewing chronic pain as a disease that requires further diagnostic workup through assessments of psychological disorders can help to identify those who may not benefit from opioid use and identify those at higher risk for adverse outcomes
Wilsey et al[38], 2008	To analyse psychological factors (including PTSD) that are correlated with the propensity for prescription opioid abuse among patients with chronic pain	n = 113	Patients visiting emergency department or urgent care centre	Not given	42/58	34	Chronic pain: not given; PTSD: SCID; Opioid (propensity abuse): SOAPP	Cross-sectional study	Multiple regression analysis	A total of 81% of the patients were positive according to SOAPP for propensity for prescription opioid abuse. PTSD (34%) was significantly correlated with the SOAPP score (r = 0.26, P < 0.005) but did not significantly predict this score (β = 1.2, P = 0.39)	Improving outcomes for patients with marked psychopathology will require treatments that independently address chronic pain and psychopathology

OUD: Opioid use disorder; PTSD: Posttraumatic stress disorder; PDUD: Prescription drug use disorder; NRS: Numerical Rating Scale; GCPS: Graded Chronic Pain Scale; BPI-SF: Brief Pain Inventory Short Form; CIDI: Composite International Diagnostic Interview; SOAPP: Screener and Opioid Assessment for Patients with Pain.