Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jun 1.
Published in final edited form as: Addict Behav. 2024 Feb 24;153:107997. doi: 10.1016/j.addbeh.2024.107997

Long-term opioid therapy trajectories in veteran patients with and without substance use disorder

Sydney A Axson a,b,c, William C Becker d,e, Jessica Merlin f, Karl Lorenz g, Amanda M Midboe g,h, Anne C Black d,e
PMCID: PMC11080947  NIHMSID: NIHMS1974521  PMID: 38442438

1. Introduction

Millions of opioid prescriptions are dispensed in the United States annually, predominantly for patients with acute pain; however policies over the past decade have focused on addressing long-term prescribing for chronic pain (Dowell et al., 2016). Use of prescription opioids for the management of chronic pain is an evolving topic in healthcare with numerous federal agencies, including the Centers for Disease Control and Prevention (CDC) and the Veterans Health Administration of the Department of Veteran Affairs (VHA), supporting efforts to characterize opioid prescribing practices and implement risk mitigation strategies for the reduction of opioid-related harms (Dowell et al. (2016); (Guy et al., 2017; Sandbrink et al., 2020). Long-term opioid therapy (LTOT) and clinical management of chronic pain are often foci of such federal agencies’ policy efforts.

Management of LTOT can be complex based on patient presentations, such as when patients have comorbid substance use disorders (SUDs). It is well established that SUDs are associated with poor opioid-related outcomes. For example, patients with SUD are at increased risk for drug overdoses, particularly with opioids (Brady et al., 2017; Karmali et al., 2020; Park et al., 2016). It is also well established that outcomes can vary by receipt of opioid prescriptions and doses, including LTOT doses. Higher doses are associated with increased risk of opioid related mortality (Bohnert et al., 2011; Dunn et al., 2010; Frank et al., 2017; Gomes et al., 2011; Gordon et al., 2020). Similarly, high dose variability may also be associated with overdose risk (Glanz et al., 2019). However, research conducted in civilian and veteran patient populations suggests patients with SUDs, amongst other psychiatric conditions, may be more likely to be prescribed LTOT and higher doses compared to patients without psychiatric conditions (Braden et al., 2009; Dobscha et al., 2013; Edlund et al., 2010; Howe & Sullivan, 2014; Morasco et al., 2010; Quinn et al., 2017; Weisner et al., 2009). Therefore, the relationship between SUDs and LTOT warrants further investigation in an effort to better meet the needs of patients with SUDs on LTOT and mitigate the potential risks.

Previous research on veteran patient populations has identified distinct LTOT dose trajectories (Merlin, 2023; Rentsch et al., 2019). Recipients of these trajectories were clinically distinguishable from each other by types of pain, pain scores, prevalence of AUD and smoking, and incidence of OUD (Rentsch et al., 2019). Further, higher and escalating dose trajectories were found to be associated with higher odds of opioid overdose compared to a low, stable dose trajectory (Merlin, 2023). This work increases the understanding that dose trajectory, not just dose, is important for health outcomes and has clinical implications. However, it has yet to be determined if and how a veteran’s SUD diagnosis is associated with a particular LTOT dosage trajectory.

Thus, in this retrospective cohort study using data from health records of 285,772 patients seen at any VA facility between 2010 and 2017, we aimed to examine if membership in previously described LTOT trajectories differs with the presence of a SUD. We hypothesized that trajectory membership would differ by SUD status demonstrating that veterans with SUD were prescribed LTOT at higher doses compared to veterans without SUD.

2. Materials and Methods

2.1. Data and participants

We conducted a retrospective cohort study using data from the Corporate Data Warehouse (CDW), which houses clinical and pharmacy data for every patient seeking care at the VHA. Patients at least 18 years of age, actively engaged in care at the VHA (defined as ≥2 outpatient visits or ≥ 1 inpatient admission in the year prior to cohort entry) with incident LTOT between 2010 and 2017 were included. Incident LTOT was defined as 90 consecutive days of opioid receipt, allowing for a 30-day gap between fills. Patients with cancer (except non-melanoma skin cancer) were excluded, as were those who enrolled in Medicare or contract hospice within 90 days of cohort entry. This study received expedited approval from the Institutional Review Boards of VA Connecticut and VA Palo Alto/Stanford.

2.2. Study variables

Following methods used in Merlin et al., opioid receipt was extracted from the CDW and quantified by 30-day periods after LTOT initiation (Merlin, 2023). Opioid analgesics prescribed in this sample included levorphanol, meperidine, tapentadol, codeine, hydrocodone, oxycodone, oxycodone sustained action (SA), morphine, morphine SA, fentanyl, hydromorphone, oxymorphone, propoxyphene, pentazocine, tramadol and methadone (pills). Buprenorphine and liquid methadone were excluded from MEDD conversions as medications primarily used to treat OUD in the VA. Dosages were converted to morphine equivalent daily dose (MEDD) to allow for comparison of different opioid analgesics and the mean dose was calculated across days prescribed for each 30-day interval. Data were then aggregated into 90-day periods as the unweighted mean prescribed dose.

SUD status was determined by ICD-9 and ICD-10 codes and was treated as a dichotomous (yes/no) variable at time of LTOT initiation. SUDs associated with alcohol, amphetamines, cannabis, cocaine, hallucinogens, opioids, tobacco, sedatives, hypnotics, anxiolytics, tranquilizers, barbiturates were included. ICD codes associated with “other, specified drug dependence”, “combinations excluding opioids”, “unspecified drug dependence”, and “other, mixed or unspecified drug abuse”, were also included in the construction of this variable.

Mental health diagnoses relating to psychosis, depression, bipolar disorder, and post-traumatic stress disorder were also extracted at time of LTOT initiation and coded as a single dichotomous indicator of any diagnosis. Age was measured as a continuous variable and was recorded from the CDW at time of LTOT initiation. Sex and race were extracted at time of initiation and were measured as categorical variables.

2.3. Analysis

Descriptive statistics characterized the overall sample. Categorical variables were examined as frequencies and percentages. Continuous variables were examined using measures of central tendency (means, medians, range). Comparative statistics (Chi-square, t-test) were used to determine differences in cohort characteristics by SUD status.

Growth mixture modeling (GMM) was utilized to examine the relationship between SUD and LTOT trajectories. Merlin et al previously conducted GMM to determine LTOT receipt patterns using the PROC TRAJ procedure within SAS software (Enterprise Guide version 8.2 update 4) (Merlin, 2023). The final model specified 5 latent trajectories of LTOT receipt: (1) low-dose/stable, (2) low-dose/de-escalating, (3) moderate-dose, (4) moderate-dose/ escalating with quadratic downturn, and (5) high-dose/escalating with quadratic downturn. Cohort members in this study were assigned to the trajectory in which they had the highest posterior probability of membership. We replicated GMM methods for the current study and specified SUD and mental health diagnosis as risk factors for trajectory within SAS PROC TRAJ.

Each trajectory comprised patients who followed a similar LTOT dosing pattern over time. These trajectories are unique in their pattern over time, but MEDD observed in two trajectories can overlap. For example, MEDDs observed in the low-dose/de-escalating trajectory may, at some points in time resemble MEDDs observed in the low-dose/stable trajectory. The difference between the trajectories is that patients in the low-dose/de-escalating trajectory experienced a decrease in MEDD over time, while those in the low-dose/stable trajectory did not. The focus of this analysis is the entire trajectory, and not MEDDs observed at single points in time. We used chi-square to determine if the trajectories were significantly different from each other by cohort characteristics.

Lastly, a sub-analysis was conducted of patients with SUD (N=53,505) to examine their conditional odds of being in any trajectory, compared to the low dose trajectory, relative to the odds for patients with no SUD for each trajectory. For this sub-analysis the low-dose/stable trajectory was the referent group, as Merlin and colleagues previously identified the low-dose/stable trajectory as being the least risky regarding certain opioid related outcomes, like opioid overdose (Merlin, 2023). Additionally, we controlled for mental health status, as mental health diagnoses and SUD diagnoses are often collinear. Analyses were conducted in SAS Enterprise. Statistical significance was set at p < 0.01.

To assess the impact of emerging national opioid safety guidelines during the cohort entry period, in a sensitivity analysis, we divided the cohort by whether patients entered before 2014 or after, representing entry before or after full institution of the Opioid Safety Initiative opioid surveillance dashboard within VA (Lin et al., 2017), and compared trajectory membership probabilities by SUD status across the two time periods.

3. Results

The final analytic sample included 285,772 patients, which were majority male (93%) and white (77%), with a mean age of 64.04 (14.09) years (Table 1). At time of LTOT initiation, 19% (53,505) of the sample had a SUD diagnosis (Table 1). Patients with SUD were slightly younger in age, on average (p<0.001). While most patients in the sample were white, there were statistically significant differences by race and SUD status (p < 0.001). For example, 17% of patients with SUD were Black or African American compared to those without SUD, 14.4%. Over half of patients with SUD (52.9%) presented with a mental health diagnosis compared to 29.1% of those without SUD.

Table 1:

Characteristics of cohort by SUD status

Variable Total N= 285,772 SUD = Yes n = 53,505 SUD = No n = 232,267 p
Age, mean (SD) 64.04 (14.09) 60.0 (12.0) 65.0 (14.4) <0.001
Sex, n (%) 0.39
 Male 265,835 (93) 49,818 (93) 216,017 (93)
 Female 19,937 (7) 3,687 (7) 16,250 (7)
Race, n (%) < 0.001
 White 220,711(77.2) 40,619 (75.9) 180,092 (77.5)
 Black/AA 42,629 (15) 9,089 (17) 33,540 (14.4)
 Hispanic/Latino 11,498 (4) 1,845 (3.5) 9,653 (4.2)
 Asian 925 (0.3) 117 (0.2) 808 (0.4)
 Other/multi/unknown 10,009 (3.5) 1,835 (3.4) 8,174 (3.5)
Mental health diagnoses, n (%) <0.001
 Yes 95,833 (33.5) 28,278 (52.9) 67,555 (29.1)
 No 189,939 (66.5) 25,227 (47.1) 164,712 (70.9)
Open in a new tab

SUD: Substance use disorder; SD: Standard deviation; AA: African American; Mental health diagnoses include depression, psychosis, bipolar disorder, and post-traumatic stress disorder. Test difference between means using t-test. Test differences between categorical variables using Chi-square.

The five resulting LTOT trajectories (Figure 1) replicated patterns described by Merlin and colleagues (Merlin, 2023); 36% (102,941) of the cohort was assigned to the low-dose/stable trajectory, 28% (79,734) to the low-dose/de-escalating trajectory, and 23% (66,888) to the moderate-dose trajectory. Membership in the moderate-dose/escalating and the high-dose/escalating trajectories was less common, with 8% (22,645) and 5% (13,564) of the sample, respectively (Table 2). On average, those in low-dose/stable trajectory were older than those in the other trajectories (p<0.001). There were also differences in trajectory membership by sex and race (p<0.001). For example, of the low-dose/stable trajectory was 93% men compared to 95% of the high-dose/escalating trajectory. Within the low-dose/stable trajectory, 77% of patients were white, while 84% of patients in the high-dose/escalating trajectory were white. Finally, controlling for risk posed by mental health diagnoses, SUD was significantly associated with membership in higher-dose trajectories (p<.001 for each trajectory). Compared to those without SUD, those with SUD comprise higher proportions of the moderate-dose/escalating and high-dose/escalating than the low-dose/stable trajectory.

Figure 1. Opioid dosing trajectories.

Figure 1

Open in a new tab

Note. Trajectory intercepts represent mg MEDD start values and slope estimates represent mean linear and quadratic change in mg MEDD per 90-day interval. Intercepts (SE) and slopes (SE) for each trajectory: low dose/stable trend (intercept=15.97 (0.03); (2) moderate dose/stable trend (intercept=37.02 (0.03)); (3) low dose/de-escalating trend (intercept = 24.45 (0.05), linear slope=−4.45 (0.01)); (4) moderate dose/escalating trend (intercept=58.51 (0.11), linear slope = 3.81 (0.03), quadratic slope −0.26 (0.00)); high dose/escalating trend (92.46 (0.16), linear slope = 6.86 (0.05), quadratic slope −0.41 (0.00)).

Table 2:

Characteristics of cohort by LTOT trajectory

LTOT Trajectory
1 N = 102,941 2 N = 79,734 3 N = 66,888 4 N = 22,645 5 N = 13,564 p
Variable
Age, mean (SD) 65.3
(14.3)		 62.2
(14.9)		 64.4
(13.3)		 63.6
(13.0)		 63.8
(11.8)		 <0.001
SUD, n (%) <0.001
 Yes 16,529
(16.1)		 16,401
(20.6)		 12,884
(19.3)		 4,795
(21.2)		 2,896
(21.4)
 No 86,412
(83.9)		 63,333
(79.4)		 54,004
(80.7)		 17,850
(78.8)		 10,668
(78.6)
Sex, n (%) <0.001
 Male 95,522
(92.8)		 73,298
(92)		 62,863
(94)		 21,307
(94.1)		 12,845
(94.7)
 Female 7,419
(7.2)		 6,436
(8.0)		 4,025
(6.0)		 1,338
(5.9)		 719 (5.3)
Race, n (%) <0.001
 White 79,089
(76.8)		 59,257
(74.3)		 52,606
(78.7)		 18,334
(81.0)		 11,425
(84.2)
 Black/AA 15,555
(15.1)		 13,693
(17.2)		 9,443
(14.1)		 2,702
(11.9)		 1236
(9.1)
 Hispanic/Latino 4,232
(4.1)		 3,661
(4.6)		 2,355
(3.5)		 814 (3.6) 436 (3.2)
 Asian 354 (0.3) 325 (0.4) 162 (0.2) 50 (0.2) 34 (0.3)
 Other/multi/unknown 3,711
(3.6)		 2,798
(3.5)		 2,322
(3.5)		 745 (3.3) 433 (3.2)
Mental health diagnoses, n (%) <0.001
 Yes 31,002
(30.1)		 30,133
(37.8)		 21,906
(32.8)		 8,002
(35.3)		 4,790
(35.3)
 No 71,939
(69.9)		 49,601
(52.2)		 44,982
(67.2)		 14,643
(64.7)		 8,774
(64.7)
Open in a new tab

LTOT: Long term opioid therapy; SD: Standard deviation; SUD: Substance use disorder; AA: African American. Mental health diagnoses include depression, psychosis, bipolar disorder, and post-traumatic stress disorder.

1 = Low-dose/stable; 2 = Low-dose/de-escalating; 3 = moderate-dose/stable; 4 = moderate-dose/escalating; 5 = high-dose/escalating

Test differences between means using ANOVA. Test differences between categorical variables using Chi-square.

3.1. LTOT trajectories in patients with SUD

Of the 53,505 patients with a SUD, 16,529 (32%) were in the low-dose/stable trajectory, 16,401 (31%) in the low-dose/de-escalating, 12,884 (24%) in the moderate-dose, 4,795 (9%) in the moderate-dose/escalating and 2,896 (5%) in the high-dose/escalating trajectory (Table 2).

Patients with SUD had higher odds of being in the riskier trajectories, compared to the referent low-dose/stable trajectory (Table 3). Compared to low-dose/stable trajectory membership, having a SUD was associated with 20% greater odds of being in the low-dose/de-escalating or moderate-dose trajectory (aOR: 1.20, 99% CI: 1.16–1.25; aOR: 1.20, 99% CI: 1.15–1.24). Further, patients with a SUD had 32% greater odds of being in the moderate- dose/escalating trajectory and 32% greater odds of being the high-dose/escalating trajectory, compared to the low-dose/stable trajectory (aOR: 1.32, 99% CI: 1.25–1.39; aOR: 1.32, 99% CI: 1.24–1.41).

Table 3:

Adjusted odds ratios of LTOT trajectory membershipfor patients with SUD

Trajectory aOR 99% CI p
2 1.20 1.16–1.25 <.001
3 1.20 1.15–1.24 <.001
4 1.32 1.25–1.39 <.001
5 1.32 1.24–1.41 <.001
Open in a new tab

LTOT: Long-term opioid therapy; SUD: Substance use disorder; aOR: Adjusted odds ratio; CI: Confidence interval. Referent group = low-dose/stable trajectory. 2 = Low-dose/de-escalating; 3 = moderate-dose/stable; 4 = moderate-dose/escalating; 5 = high-dose/escalating

3.2. Sensitivity analysis

Inferences about the SUD-associated risk of membership in higher-dose trajectories did not differ by time of entry in the cohort; regardless of entry before or after 2014, patients with SUD had significantly greater odds of being in higher-dose trajectories than patients without SUD (Appendix Table 3).

4. Discussion

This retrospective cohort study examined LTOT trajectory membership in VHA patients with and without SUD. Patients with SUD had significantly higher odds of being in any other higher-dose trajectory compared to the low-dose/stable trajectory.

This study’s finding that patients with SUD are at greater odds of experiencing a higher-dose LTOT trajectory reflects findings from previous research examining longitudinal opioid prescribing to patients with SUD. Rentsch and colleagues similarly found significant associations between trajectory membership and SUD diagnosis, specifically opioid use disorder (OUD). They found veteran patients in a rapidly escalating trajectory were more likely to have an OUD diagnosis, compared to patients in a low dose trajectory (Rentsch et al., 2019). Further, their team identified four prescribed opioid trajectories (low, moderate, escalating, and rapidly escalating) in their development of prescription phenotypes among patients with and without HIV. While their sample included patients with opioid prescriptions of a least seven consecutive days, we instead focused on LTOT and SUD. Our work uniquely explores the relationship between LTOT and SUD in a large sample of veterans. Wei and colleagues also identified 5 opioid dose trajectories (low dose, consistent moderate dose, escalating dose, de-escalating dose, and consistent high dose), focusing on prescriptions to patients in the year before an incident OUD or overdose diagnosis (Wei et al., 2019). They also found significant differences in these trajectories by mental health, specifically depression and anxiety. However, analyses did not compare prescribing to those without OUD.

This study builds on previous research conducted by Merlin and colleagues, who found patients in three trajectories (moderate-dose, moderate-dose/escalating, and high-dose/escalating) had significantly higher risk of overdose, compared to the low-dose trajectory (Merlin, 2023).

We found patients with SUD have higher odds of being in those high-risk trajectories. Our results further highlight the importance of SUD treatment in reducing risk for mortality (Gaither et al., 2016). For clinicians, these findings underscore the importance of considering comorbid conditions, like SUD, in the context of LTOT over time, and suggest the value of accessible diagnostic records for all providers. Interdisciplinary and collaborative care may be necessary to effectively manage patient symptoms and improve opioid related outcomes; in recognition of this, VA is rolling out pain management teams nationally that include providers with addiction expertise Policies focused on curbing LTOT must consider the nuances of various dosing trajectories and the stratified risks linked to each trajectory for patients with and without SUD. Approaches like GMM, using longitudinal data, are useful in building this line of inquiry beyond common cross-sectional approaches.

There are limitations to this work. Directionality of the relationship between SUD status and LTOT could not be determined by analyses. That is, this analysis did not assess whether SUD status leads to higher dose trajectories or vice versa, or whether a third unmodeled variable, such as pain, accounts for the association; future study into the nature of the association is warranted. For this descriptive analysis both SUD and mental health were treated as dichotomous variables which masked the spectrum of severity that exists within conditions. Pain scores and other pain related outcomes were not consistently available in electronic health records and thus were not included in analyses. The presence of pain in the veteran population is well documented and use of prescription opioids continues to evolve (Edlund et al., 2014; Gordon et al., 2020; Haskell et al., 2012; Helmer et al., 2009; Hoerster et al., 2012; Kazis et al., 1998; Kerns et al., 2003; Nahin, 2017). Further, the sample was majority male – consistent with the veteran population -- and thus there are sex related differences relevant to this work that we were unable to explore. For example, women are more like to have some chronic pain conditions (e.g. fibromyalgia) and have a higher prevalence of psychological symptoms that influence their pain experience (Berkley, 1997; Fillingim, 2000; Fillingim et al., 1999; Fillingim et al., 1998; Unruh, 1996; Wolfe et al., 1995) Future work should address potential sex-related differences in the SUD-trajectory association to better tailor future interventions. Though beyond the scope of this work, it is also important to note there are interactions between these variables that this study did not address. For example, mental health diagnoses can be comorbid with chronic pain and associated with increased risk of developing chronic pain (Goesling et al., 2018; Gureje et al., 2001; Howe & Sullivan, 2014; Magni et al., 1994). While future work will benefit from moving beyond these limitations, this analysis was an important first step in examining the relationship between SUD and LTOT trajectories. Further, LTOT opioid trajectories were defined by VHA prescriptions. Therefore, we could not account for opioid prescriptions from non-VHA sources, nor could we know whether medications received were ingested, a common limitation in large pharmacoepidemiologic studies. Finally, the CDC issued notable guidelines regarding the treatment of chronic pain in 2016 (Dowell et al., 2016). Although sensitivity analyses suggested robust results, our study period and focus were not designed to fully explore the impact of those guidelines on opioid prescribing trends. Despite these limitations, this study’s large sample size, use of data from a 7-year period, and inclusion of SUD are valuable additions to the literature. The use of LTOT trajectories also allows for analyses to accommodate changing patient presentations and management patterns over time.

5. Conclusions

Overall, we found patients with SUD had higher odds of membership in riskier LTOT trajectories. These analyses, which used a large national dataset, build upon previous research related to LTOT trajectories by highlighting the unique risks patients with SUD may face in receiving higher-dose trajectories compared to those without SUD. Veterans on LTOT with SUD may require interdisciplinary care to manage symptoms and improve opioid-related outcomes; current VA initiatives incorporating addiction specialists in pain management teams facilitate an interdisciplinary approach. Characterizing opioid prescribing practices to patients with vs. without SUD may inform policies and care delivery for veteran patients. Future research is needed to further understand the nature of the SUD-associated risk for trajectory membership and potential moderators, including sex. Additional risk factors for membership in high-dose trajectories, including risks related to pain and specific SUD and mental health disorders should be explored, and may contribute to more personalized care for patients with co-morbid conditions and LTOT.

Supplementary Material

1

Highlights.

  • Growth mixture modeling estimated long-term opioid therapy trajectory membership

  • Veterans with SUD are disproportionally represented in higher dose opioid trajectories

Results highlight the importance of considering patient comorbidities that may compound risk for opioid-related adverse events

Appendix

Appendix Table 3:

Adjusted odds ratios of LTOT trajectory membership for patients with SUD entering the cohort pre- vs. post-2014

Trajectory aOR 99% CI p
 Pre-2014 entry
 N=213,971		 2 1.22 1.17–1.27 <.001
3 1.18 1.13–1.23 <.001
4 1.27 1.20–1.35 <.001
5 1.28 1.20–1.37 <.001
 Post-2014 entry
 N=71,801		 2 1.19 1.11–1.28 <.001
3 1.22 1.13–1.33 <.001
4 1.47 1.29–1.67 <.001
5 1.47 1.29–1.66 <.001
Open in a new tab

LTOT: Long-term opioid therapy; SUD: Substance use disorder; aOR: Adjusted odds ratio; CI: Confidence interval. Referent group = low-dose/stable trajectory. 2 = low-dose/de-escalating; 3 = moderate-dose; 4 = moderate-dose/escalating; 5 = high-dose/escalating

Footnotes

Declarations of interest: none

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Berkley KJ (1997). Sex differences in pain. Behav Brain Sci, 20(3), 371–380; discussion 435–513. 10.1017/s0140525×97221485 [DOI] [PubMed] [Google Scholar]
  2. Bohnert ASB, Valenstein M, Bair MJ, Ganoczy D, McCarthy JF, Ilgen MA, & Blow FC (2011). Association Between Opioid Prescribing Patterns and Opioid Overdose-Related Deaths. Jama-Journal of the American Medical Association, 305(13), 1315–1321. 10.1001/jama.2011.370 [DOI] [PubMed] [Google Scholar]
  3. Braden JB, Sullivan MD, Ray GT, Saunders K, Merrill J, Silverberg MJ, Rutter CM, Weisner C, Banta-Green C, Campbell C, & Von Korff M. (2009). Trends in long-term opioid therapy for noncancer pain among persons with a history of depression. General Hospital Psychiatry, 31(6), 564–570. 10.1016/j.genhosppsych.2009.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brady JE, Giglio R, Keyes KM, DiMaggio C, & Li G. (2017). Risk markers for fatal and non-fatal prescription drug overdose: a meta-analysis. Inj Epidemiol, 4(1), 24. 10.1186/s40621-017-0118-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dobscha SK, Morasco BJ, Duckart JP, Macey T, & Deyo RA (2013). Correlates of Prescription Opioid Initiation and Long-term Opioid Use in Veterans With Persistent Pain. Clinical Journal of Pain, 29(2), 102–108. 10.1097/AJP.0b013e3182490bdb [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Dowell D, Haegerich TM, & Chou R. (2016). CDC Guideline for Prescribing Opioids for Chronic Pain-United States, 2016. JAMA, 315(15), 1624–1645. 10.1001/jama.2016.1464 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dunn KM, Saunders KW, Rutter CM, Banta-Green CJ, Merrill JO, Sullivan MD, Weisner CM, Silverberg MJ, Campbell CI, Psaty BM, & Von Korff M. (2010). Opioid Prescriptions for Chronic Pain and Overdose A Cohort Study. Annals of Internal Medicine, 152(2), 85-+. 10.7326/0003-4819-152-2-201001190-00006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Edlund MJ, Austen MA, Sullivan MD, Martin BC, Williams JS, Fortney JC, & Hudson TJ (2014). Patterns of opioid use for chronic noncancer pain in the Veterans Health Administration from 2009 to 2011. Pain, 155(11), 2337–2343. 10.1016/j.pain.2014.08.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Edlund MJ, Martin BC, Devries A, Fan MY, Braden JB, & Sullivan MD (2010). Trends in Use of Opioids for Chronic Noncancer Pain Among Individuals With Mental Health and Substance Use Disorders: The TROUP Study. Clinical Journal of Pain, 26(1), 1–8. 10.1097/AJP.0b013e3181b99f35 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fillingim RB (2000). Sex, gender, and pain: women and men really are different. Curr Rev Pain, 4(1), 24–30. 10.1007/s11916-000-0006-6 [DOI] [PubMed] [Google Scholar]
  11. Fillingim RB, Maddux V, & Shackelford JA (1999). Sex differences in heat pain thresholds as a function of assessment method and rate of rise. Somatosens Mot Res, 16(1), 57–62. 10.1080/08990229970654 [DOI] [PubMed] [Google Scholar]
  12. Fillingim RB, Maixner W, Kincaid S, & Silva S. (1998). Sex differences in temporal summation but not sensory-discriminative processing of thermal pain. Pain, 75(1), 121–127. 10.1016/S0304-3959(97)00214-5 [DOI] [PubMed] [Google Scholar]
  13. Frank JW, Lovejoy TI, Becker WC, Morasco BJ, Koenig CJ, Hoffecker L, Dischinger HR, Dobscha SK, & Krebs EE (2017). Patient Outcomes in Dose Reduction or Discontinuation of Long-Term Opioid Therapy: A Systematic Review. Annals of Internal Medicine, 167(3), 181–191. 10.7326/M17-0598 [DOI] [PubMed] [Google Scholar]
  14. Gaither JR, Goulet JL, Becker WC, Crystal S, Edelman EJ, Gordon K, Kerns RD, Rimland D, Skanderson M, Justice AC, & Fiellin DA (2016). The Effect of Substance Use Disorders on the Association Between Guideline-concordant Long-term Opioid Therapy and All-cause Mortality. Journal of Addiction Medicine, 10(6), 418–428. 10.1097/Adm.0000000000000255 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Glanz JM, Binswanger IA, Shetterly SM, Narwaney KJ, & Xu S. (2019). Association Between Opioid Dose Variability and Opioid Overdose Among Adults Prescribed Long-term Opioid Therapy. Jama Network Open, 2(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Goesling J, Lin LA, & Clauw DJ (2018). Psychiatry and Pain Management: at the Intersection of Chronic Pain and Mental Health. Current Psychiatry Reports, 20(2). [DOI] [PubMed] [Google Scholar]
  17. Gomes T, Mamdani MM, Dhalla IA, Paterson JM, & Juurlink DN (2011). Opioid Dose and Drug-Related Mortality in Patients With Nonmalignant Pain. Archives of Internal Medicine, 171(7), 686–691. <Go to ISI>://WOS:000289867700017 [DOI] [PubMed] [Google Scholar]
  18. Gordon KS, Manhapra A, Crystal S, Dziura J, Edelman EJ, Skanderson M, Kerns RD, Justice AC, Tate J, & Becker WC (2020). All-cause mortality among males living with and without HIV initiating long-term opioid therapy, and its association with opioid dose, opioid interruption and other factors. Drug and Alcohol Dependence, 216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gureje O, Simon GE, & Von Korff M. (2001). A cross-national study of the course of persistent pain in primary care. Pain, 92(1–2), 195–200. 10.1016/S0304-3959(00)00483-8 [DOI] [PubMed] [Google Scholar]
  20. Guy GP, Zhang K, Bohm MK, Losby J, Lewis B, Young R, Murphy LB, & Dowell D. (2017). Vital Signs: Changes in opioid prescribing in the United States, 2006–2015. MMWR, 66(26), 697–704. 10.15585/mmwr.mm6626a4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Haskell SG, Ning Y, Krebs E, Goulet J, Mattocks K, Kerns R, & Brandt C. (2012). Prevalence of Painful Musculoskeletal Conditions in Female and Male Veterans in 7 Years After Return From Deployment in Operation Enduring Freedom/Operation Iraqi Freedom. Clinical Journal of Pain, 28(2), 163–167. 10.1097/AJP.0b013e318223d951 [DOI] [PubMed] [Google Scholar]
  22. Helmer DA, Chandler HK, Quigley KS, Blatt M, Teichman R, & Lange G. (2009). Chronic Widespread Pain, Mental Health, and Physical Role Function in OEF/OIF Veterans. Pain Medicine, 10(7), 1174–1182. 10.1111/j.1526-4637.2009.00723.x [DOI] [PubMed] [Google Scholar]
  23. Hoerster KD, Lehavot K, Simpson T, McFall M, Reiber G, & Nelson KM (2012). Health and Health Behavior Differences US Military, Veteran, and Civilian Men. American Journal of Preventive Medicine, 43(5), 483–489. 10.1016/j.amepre.2012.07.029 [DOI] [PubMed] [Google Scholar]
  24. Howe CQ, & Sullivan MD (2014). The missing ‘P’ in pain management: how the current opioid epidemic highlights the need for psychiatric services in chronic pain care. General Hospital Psychiatry, 36(1), 99–104. 10.1016/j.genhosppsych.2013.10.003 [DOI] [PubMed] [Google Scholar]
  25. Karmali RN, Ray GT, Rubinstein AL, Sterling SA, Weisner CM, & Campbell CI (2020). The role of substance use disorders in experiencing a repeat opioid overdose, and substance use treatment patterns among patients with a non-fatal opioid overdose. Drug Alcohol Depend, 209, 107923. 10.1016/j.drugalcdep.2020.107923 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kazis LE, Miller DR, Clark J, Skinner K, Lee A, Rogers W, Spiro A, Payne S, Fincke G, Selim A, & Linzer M. (1998). Health-related quality of life in patients served by the department of veterans affairs - Results from the Veterans Health Study. Archives of Internal Medicine, 158(6), 626–632. 10.1001/archinte.158.6.626 [DOI] [PubMed] [Google Scholar]
  27. Kerns RD, Otis J, Rosenberg R, & Reid MC (2003). Veterans’ reports of pain and associations with ratings of health, health-risk behaviors, affective distress, and use of the healthcare system. Journal of Rehabilitation Research and Development, 40(5), 371–379. 10.1682/Jrrd.2003.09.0371 [DOI] [PubMed] [Google Scholar]
  28. Lin LA, Bohnert AS, Kerns RD, Clay MA, Ganoczy D, & Ilgen MA (2017). Impact of the Opioid Safety Initiative on opioid-related prescribing in veterans. Pain, 158(5), 833–839. [DOI] [PubMed] [Google Scholar]
  29. Magni G, Moreschi C, Rigattiluchini S, & Merskey H. (1994). Prospective-Study on the Relationship between Depressive Symptoms and Chronic Musculoskeletal Pain. Pain, 56(3), 289–297. 10.1016/0304-3959(94)90167-8 [DOI] [PubMed] [Google Scholar]
  30. Merlin J, Black AC, Midboe AM, Troszak L, Asch S, Bohnert A, Fenton B, Giannitrapani K, Glassman P, Kerns R, Silveira M, Lorenz K, Abel E, Becker W. (2023). Long-term opioid therapy trajectories and overdose in patients with and without cancer. BMJ Oncology, 2(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Morasco BJ, Duckart JP, Carr TP, Deyo RA, & Dobscha SK (2010). Clinical characteristics of veterans prescribed high doses of opioid medications for chronic non-cancer pain. Pain, 151(3), 625–632. 10.1016/j.pain.2010.08.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Nahin RL (2017). Severe Pain in Veterans: The Effect of Age and Sex, and Comparisons With the General Population. Journal of Pain, 18(3), 247–254. 10.1016/j.jpain.2016.10.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Park TW, Lin LA, Hosanagar A, Kogowski A, Paige K, & Bohnert AS (2016). Understanding Risk Factors for Opioid Overdose in Clinical Populations to Inform Treatment and Policy. J Addict Med, 10(6), 369–381. 10.1097/ADM.0000000000000245 [DOI] [PubMed] [Google Scholar]
  34. Quinn PD, Hur K, Chang Z, Krebs EE, Bair MJ, Scott EL, Rickert ME, Gibbons RD, Kroenke K, & D’Onofrio BM (2017). Incident and long-term opioid therapy among patients with psychiatric conditions and medications: a national study of commercial health care claims. Pain, 158(1), 140–148. 10.1097/j.pain.0000000000000730 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rentsch CT, Edelman EJ, Justice AC, Marshall BDL, Xu K, Smith AH, Crystal S, Gaither JR, Gordon AJ, Smith RV, Kember RL, Polimanti R, Gelernter J, Fiellin DA, Tate JP, Kranzler HR, Becker WC, & Team VP (2019). Patterns and Correlates of Prescription Opioid Receipt Among US Veterans: A National, 18-Year Observational Cohort Study. Aids and Behavior, 23(12), 3340–3349. 10.1007/s10461-019-02608-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sandbrink F, Oliva EM, McMullen TL, Aylor AR, Harvey MA, Christopher ML, Cunningham F, Minegishi T, Emmendorfer T, & Perry JM (2020). Opioid Prescribing and Opioid Risk Mitigation Strategies in the Veterans Health Administration. J Gen Intern Med, 35(Suppl 3), 927–934. 10.1007/s11606-020-06258-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Unruh AM (1996). Gender variations in clinical pain experience. Pain, 65(2–3), 123–167. 10.1016/0304-3959(95)00214-6 [DOI] [PubMed] [Google Scholar]
  38. Wei YJJ, Chen C, Fillingim R, Schmidt SO, & Winterstein AG (2019). Trends in prescription opioid use and dose trajectories before opioid use disorder or overdose in US adults from 2006 to 2016: A cross-sectional study. Plos Medicine, 16(11). [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Weisner CM, Campbell CI, Ray GT, Saunders K, Merrill JO, Banta-Green C, Sullivan MD, Silverberg MJ, Mertens JR, Boudreau D, & Von Korff M. (2009). Trends in prescribed opioid therapy for non-cancer pain for individuals with prior substance use disorders. Pain, 145(3), 287–293. 10.1016/j.pain.2009.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wolfe F, Ross K, Anderson J, Russell IJ, & Hebert L. (1995). The prevalence and characteristics of fibromyalgia in the general population. Arthritis Rheum, 38(1), 19–28. 10.1002/art.1780380104 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1974521-supplement-1.docx (13.5KB, docx)

RESOURCES