Abstract
Aims
To estimate the influence of non-medical use of prescription opioids (NMUPO) on heroin initiation among U.S. veterans receiving medical care.
Design
Using a multivariable Cox regression model, we analyzed data from a prospective, multi-site, observational study of HIV-positive and an age/race/site matched control group of HIV-negative veterans in care in the United States. Approximately annual behavioral assessments were conducted and contained self-reported measures of NMUPO and heroin use.
Setting
Veterans Health Administration (VHA) infectious disease and primary care clinics in Atlanta, Baltimore, New York, Houston, Los Angeles, Pittsburgh, and Washington, DC.
Participants
A total of 3,396 HIV-infected and uninfected patients enrolled in the Veterans Aging Cohort Study who reported no lifetime NMUPO or heroin use, had no opioid use disorder diagnoses at baseline, and who were followed between 2002 and 2012.
Measurements
The primary outcome measure was self-reported incident heroin use and the primary exposure of interest was new onset NMUPO. Our final model was adjusted for sociodemographics, pain interference, prior diagnoses of post-traumatic stress disorder and/or depression, and self-reported other substance use.
Findings
Using a multivariable Cox regression model, we found that NMUPO was positively and independently associated with heroin initiation (adjusted hazard ratio [AHR] = 5.43, 95%CL: 4.01, 7.35).
Conclusions
New onset non-medical use of prescription opioids (NMUPO) is a strong risk factor for heroin initiation among HIV-infected and uninfected veterans in the United States who reported no previous history of NMUPO or illicit opioid use.
Keywords: veterans, nonmedical prescription drug use, heroin, opioid-related disorders, longitudinal study
INTRODUCTION
Heroin use and opioid use disorders are serious public health problems (1). The U.S. Centers for Disease Control and Prevention (CDC) reported that the total number of heroin-attributable overdose deaths increased five-fold from 2001 to 2013, with 8,257 heroin-related deaths in 2013 alone (2). The substantial increase in heroin use and heroin-attributable overdose rates may be linked to the growth in sales, use, and misuse of prescription opioids (3). Over the last decade, the number of opioid prescriptions dispensed in the U.S. increased by 48% (4, 5). Concomitantly, the non-medical use of prescription opioids (NMUPO)—operationally defined in the National Survey on Drug Use and Health as the use of prescription opioids “without a prescription of the individual's own or simply for the experience or feeling the drugs cause”—is increasingly common (6). The last 15 years has seen sharp rises in NMUPO (6–8), with 4.5 million individuals reporting NMUPO in the past month in 2013 (9). Canadian data estimates that approximately 4.8% of the general population used prescription opioids non-medically in 2009 and as the demand for prescription opioids has increased, so too has the availability of diverted prescription opioids.(10) In the past two decades, the use of opioids has also increased significantly in Europe.(11) In particular, there is a high prevalence of misuse of diverted opioids among drug using populations in England and Scotland.(11)
Persons who engage in NMUPO may be at a higher risk for transitioning to heroin use, in part because heroin has become more accessible and less expensive than prescription opioids in many US settings (12–14). A nationally representative survey found that four out of five recent heroin initiates reported prior NMUPO, and the rate of heroin initiation among prior nonmedical prescription opioid users was approximately 19 times greater than those who did not report nonmedical use (15). Heroin initiation is particularly troubling because of the additional risks associated with heroin use, including unknown purity and contaminants, overdose, and injection-associated infections and vascular disease (12, 16). Transition from NMUPO to heroin also represents an increasingly prominent pathway leading to opioid-related mortality (17).
U.S. military veterans represent a particularly high-risk population for illicit substance use and abuse (18–20). Chronic pain is a significant problem among veterans (21), and is commonly treated with opioid analgesics (22). Veterans also have high rates of mental health conditions that further increase the risk for NMUPO (23). However, it is poorly understood whether NMUPO plays a role in veterans' risk for initiating heroin use, particularly those veterans from newer eras, including Operation Enduring Freedom and Operation Iraqi Freedom (24).
To address this critical public health problem, we examined the relationship between new onset NMUPO and heroin initiation among veterans. We aimed to: 1) identify risk factors for heroin initiation among participants enrolled in the Veterans Aging Cohort Study; 2) calculate the crude incidence rate of heroin initiation in this high risk population; and 3) compare the hazard of heroin initiation between participants with self-reported NMUPO and no prior NMUPO.
METHODS
Data sources
This research utilizes data from the Veterans Aging Cohort Study (VACS); detailed data collection and survey methodology for VACS are described elsewhere (25–31). The VACS is an ongoing, prospective cohort study of HIV-infected and uninfected veterans receiving medical care at 8 Veterans Health Administration (VHA) sites located throughout the United States. Since June 2002, VACS has enrolled over 7,000 patients from the infectious disease or general medical clinics in Atlanta, Baltimore, Houston, Los Angeles, Pittsburgh, Washington D.C., and multiple sites in New York City. Participants in VACS are similar to other veterans receiving care within the VA, with the exception of participants being older and more predominantly black (30). This comprehensive, longitudinal database contains variables from surveys completed approximately every 18 months, available from 2002-2012. The survey data is also linked to robust VA electronic medical records (EMR) containing data on prescribed medications, medical and substance use diagnoses, and laboratory results for each patient. The VACS was approved by the institutional review boards at each participating VHA Medical Center and affiliated academic institutions.
Participant eligibility
The flow chart illustrating the selection of eligible participants is illustrated in Figure 1. Of 7,324 potentially eligible VACS participants, 2,792 (38.1%) were excluded since they reported “yes” to ever using a prescription opioid non-medically or heroin use at baseline. Of these participants, we excluded those who reported any injection drug use at baseline, or who were previously diagnosed with opioid dependence based on linked EMR records (ICD-9 code 304.0) (n= 399). Finally, we excluded participants who were missing or had invalid NMUPO and heroin use responses in all five follow-up surveys (n = 186) or only completed the baseline survey (n=551). Using data from the VHA Patient Treatment File, the Beneficiary Identification Records Locating System (which tracks VHA death benefits, the Medicare Vital Status file, and the Social Security National Death index, we identified 203 (36.9%) deaths among those participants who only completed the baseline survey. We compared the characteristics of those who did not complete at least one follow-up visit due to death or other reasons with those who were eligible (see Supplemental File). We accounted for biases arising from potential differential loss to follow-up in weighted statistical analyses (see below). The final analytic sample consisted of 3,396 veterans.
Figure 1.
Flow chart of eligible VACS participants, 2002-2012
Measures
We operationalized the main exposure of NMUPO using responses to two different survey questions. In the first two of five survey waves, participants were shown a list of the following substances: “marijuana, cocaine/crack, stimulants, heroin, and prescription opioids (morphine, codeine, Vicodin, Percocet, OxyContin)” and asked “For each of the following drugs, please fill in the oval that best indicates how often in the past 12 months you have used each drug.” Those participants who reported any use of prescription opioids were categorized as reporting NMUPO. In the final three follow-up survey waves participants were asked, “Now think only about the past 12 months. On average, how many days each week in the past 12 months did you use any prescription pain reliever that was not prescribed for you or that you took only for the experience or feeling that it caused (32)?” Again, participants reporting any frequency of use were categorized as reporting NMUPO. Participants who experienced the outcome of interest (i.e., heroin initiation) in the same year that they initiated NMUPO were included in the analysis; thus, new onset NMUPO was considered to occur prior to or concurrently with heroin initiation.
Self-reported heroin initiation was ascertained based on respondent endorsement of following time-updated survey item: “How often in the past year have you used each drug – Heroin?” A heroin initiation event was defined as change in respondent's answer to the previous survey question from “Never”, to a response indicating some frequency of heroin use in the previous year.
Selection of other independent variables and potential confounders of the relationship between NMUPO and heroin initiation were chosen for analysis based on the extant literature (22, 28, 33–36). Demographic and clinical characteristics were ascertained from the baseline survey data and included age, sex, race, marital status, and gross annual income. HIV status was identified using VA Immunology Case Registry and hepatitis C virus status was determined using ICD-9 codes and laboratory data. Previous posttraumatic stress disorder (PTSD) and depression diagnoses were ascertained from the following questions: “has the doctor ever told you that you have PTSD?” and “has the doctor ever told you that you have depression?” Pain interference in daily life was ascertained from responses to the following time-updated survey item: “During the past 4 weeks, how much did pain interfere with your normal work (including both work outside the home and housework) (37)?” The response options for this question included a scale from “not at all” to “extremely”, which we dichotomized into “no interference” (“not at all”) and “any interference” (any other response). Past-year use of marijuana, cocaine and methamphetamines was dichotomized as yes/no. Alcohol use was characterized using the Alcohol Use Disorders Identification Test (AUDIT); a score of 4 and above indicated unhealthy alcohol use for men, and a threshold of 3 for women (38). Responses to questions regarding substance use (including marijuana, cocaine, and methamphetamines), unhealthy alcohol use, prior PTSD and depression diagnoses, and pain interference in daily life were updated at each survey and were considered time-dependent covariates. Consistent with prior methods, receipt of outpatient prescribed opioids from the VHA was ascertained from linked pharmacy records, and long-term prescription opioid use was defined as ≥90 days of continuous use allowing for a 30 day gap between fill and refill (27, 29, 39). We created a time-updated variable for past year receipt of prescription opioids for each wave of data, and categorized it into “none”, “short-term”, and “long-term”.
Statistical analyses
First, we used chi-square tests to examine the baseline correlates of heroin initiation. Additionally, in a post hoc analysis, we examined interaction terms between NMUPO and race. Next, unadjusted Kaplan–Meier analysis and the log-rank test were used to calculate the incidence of heroin initiation and compare time to heroin use, stratifying by prior or concurrent NMUPO.
To analyze the factors associated with heroin initiation, we used Cox proportional hazards regression to estimate crude hazard ratios (CHRs) and 95% confidence intervals (CI) for each variable. To determine the independent relationship between new onset NMUPO and heroin initiation, we then constructed a multivariable Cox model, including all variables assessed in bivariable analyses. All variables were found to meet the proportional hazards assumption for the Cox regression models (i.e., none exhibited significant deviance from this assumption at p < 0.05) (40).
We created inverse probability of censoring weights (IPCWs) to account for potential biases arising from differential dropout (41). The use of IPCW re-weights the sample such that the contribution of participants who remain in the study, but who share characteristics of those who dropout, are inflated (42). Weights were obtained through fitting a weighted pooled logistic regression model for dropout, including baseline and time varying predictors of dropping out. We included a robust sandwich estimator in all Cox regression models to account for potential clustering heterogeneity by study site (43). Analyses were conducted using SAS version 9.4.
Sensitivity analysis
In order to assess the validity of our exclusion criteria and ensure to our sample was restricted to heroin naïve participants, we re-ran the adjusted Cox regression model excluding those participants who were HCV positive at baseline (n = 701), under the assumption that the majority could have contracted HCV via injecting.
RESULTS
Sample Composition
The baseline demographic, clinical and substance use characteristics of the 3,396 eligible VACS participants are described in Table 1. The mean age was 49.7 (SD = 10.6), 2,106 (62.5%) were Black, and 327 (9.7%) were Hispanic. Approximately 1500 participants (45%) were HIV-infected. At baseline, 102 (3.8%) participants had a past year opioid prescription from the VHA. Table 1 describes the characteristics of participants who reported new onset NMUPO during the study. Past year stimulant and past year cocaine use were significantly associated with NMUPO in our study sample (p < 0.05). However, receipt of a prescription opioid from the VHA was not significantly associated with NMUPO. There were no substantial differences between the distribution of baseline characteristics among those participants who did not complete at least one follow-up visit or were missing exposure data and those participants who were eligible for our study (see Table S1, Supplementary File).
Table 1.
Baseline characteristics associated with new onset non-medical use of prescription opioids among veterans participating in VACS, 2002-2012.
| Characteristic | Total (%) (N = 3396) | Reported NMUPO |
P – value* | |
|---|---|---|---|---|
| Yes (%) (N = 1416) | No (%) (N = 1980) | |||
| Sex | 0.468 | |||
| Male | 3171 (93.4) | 1317 (41.5) | 1854 (58.5) | |
| Female | 225 (6.6) | 99 (44.0) | 126 (56.0) | |
| Age | 0.001 | |||
| ≤42 years | 846 (24.9) | 356 (42.1) | 490 (57.9) | |
| 43–49 years | 880 (25.9) | 387 (44.0) | 493 (56.0) | |
| 50–56 years | 884 (26.0) | 392 (44.3) | 492 (55.7) | |
| ≥57 years | 786 (23.1) | 281 (35.8) | 505 (64.2) | |
| HIV | <0.001 | |||
| Yes | 1539 (45.3) | 694 (45.1) | 845 (54.9) | |
| No | 1857 (54.7) | 722 (38.9) | 1135 (61.1) | |
| HCV | 0.034 | |||
| Yes | 701 (20.6) | 317 (45.2) | 384 (54.8) | |
| No | 2695 (79.4) | 1099 (40.8) | 1596 (59.2) | |
| Race | <0.001 | |||
| White | 808 (24.0) | 430 (53.2) | 378 (46.8) | |
| Black | 2106 (62.5) | 800 (38.0) | 1306 (62.0) | |
| Hispanic | 327 (9.7) | 106 (32.4) | 221 (67.6) | |
| Other | 128 (3.8) | 57 (44.5) | 71 (55.5) | |
| Education | 0.001 | |||
| High School or Less | 1301 (38.3) | 499 (38.4) | 802 (61.6) | |
| Some College or Greater | 2095 (61.7) | 917 (43.8) | 1178 (56.2) | |
| Gross Annual Income | 0.7112 | |||
| <$6,000 | 532 (16.2) | 220 (41.4) | 312 (58.7) | |
| $6,000 – $11,999 | 695 (21.2) | 295 (42.5) | 400 (57.6) | |
| $12,000 – $24,999 | 856 (26.1) | 344 (40.2) | 512 (59.8) | |
| $25,000 – $49,999 | 815 (24.8) | 474 (58.2) | 341 (41.8) | |
| ≥$50,000 | 383 (11.7) | 170 (44.4) | 213 (55.6) | |
| Marital Status | <0.425 | |||
| Married/Living with Partner | 1128 (33.5) | 481 (42.6) | 647 (57.4) | |
| Divorced/Separated/Widowed | 1257 (37.3) | 508 (40.4) | 749 (59.6) | |
| Never Married | 984 (29.2) | 421 (42.8) | 563 (57.2) | |
| Pain Interference in Daily Life | <0.001 | |||
| Yes | 1966 (58.2) | 955 (48.6) | 1011 (51.4) | |
| No | 1415 (41.9) | 457 (32.3) | 958 (67.7) | |
| Opioid Rx, Past Year | 0.330 | |||
| None | 3265 (96.1) | 1357 (41.6) | 1908 (58.4) | |
| Short-term | 102 (3.0) | 43 (42.2) | 59 (57.8) | |
| Long-term | 29 (0.9) | 16 (55.2) | 13 (44.8) | |
| Marijuana Use, Past Year | 0.165 | |||
| Yes | 654 (19.3) | 289 (44.2) | 365 (55.8) | |
| No | 2730 (80.7) | 1125 (41.2) | 1605 (58.8) | |
| Stimulant Use, Past Year | 0.012 | |||
| Yes | 53 (1.6) | 31 (58.5) | 22 (41.5) | |
| No | 3341 (98.4) | 1384 (41.4) | 1957 (58.6) | |
| Cocaine Use, Past Year | <0.024 | |||
| Yes | 421 (12.5) | 197 (46.8) | 224 (53.2) | |
| No | 2958 (87.5) | 1213 (41.0) | 1745 (59.0) | |
| Unhealthy alcohol use, Past Year (AUDIT-C score ≥ 3 or 4) | 0.239 | |||
| Yes | 864 (25.4) | 375 (43.4) | 489 (56.6) | |
| No | 2532 (74.6) | 1041 (41.1) | 1491 (58.9) | |
Note:
From Chi-square tests; HIV=Human Immunodeficiency Virus; HCV=Hepatitis C Virus
The mean proportion of participants who initiated or continued non-medical use of prescription opioids across all follow-up waves after baseline was 6.9%. There was no clear secular trend in the proportion of participants who reported NMUPO over time (Mantel-Haenszel test for trend p = 0.208).
Risk Factors for Heroin Initiation
Of the total sample, 500 (14.7%) participants initiated heroin use over the 10-year study period. Being black, male, 43–49 years of age, less educated, and having a lower income were significantly associated with heroin initiation (all p<0.001). Marijuana, cocaine, stimulant, and unhealthy alcohol use in the past year were significantly associated with heroin initiation. (all p<0.05, see Table 2).
Table 2.
Baseline characteristics associated with heroin initiation among veterans participating in VACS, 2002-2012.
| Characteristic | Total (%) (N = 3396) | Initiated Heroin |
P – value* | |
|---|---|---|---|---|
| Yes (%) (N = 500) | No (%) (N = 2896) | |||
| Sex | <0.001 | |||
| Male | 3171 (93.4) | 487 (15.4) | 2684 (84.6) | |
| Female | 225 (6.6) | 13 (5.8) | 212 (94.2) | |
| Age | <0.001 | |||
| ≤42 years | 846 (24.9) | 105 (12.4) | 741 (87.6) | |
| 43–49 years | 880 (24.9) | 186 (21.1) | 694 (78.9) | |
| 50–56 years | 884 (26.0) | 138 (15.6) | 746 (84.4) | |
| ≥57 years | 786 (23.1) | 71 (9.0) | 715 (91.0) | |
| HIV | <0.001 | |||
| Yes | 1539 (45.3) | 279 (18.1) | 1260 (81.9) | |
| No | 1857 (54.7) | 221 (11.9) | 1636 (88.1) | |
| HCV | <0.001 | |||
| Yes | 701 (20.6) | 170 (24.3) | 531 (75.8) | |
| No | 2695 (79.4) | 330 (12.2) | 2365 (87.8) | |
| Race | <0.001 | |||
| White | 808 (24.0) | 76 (9.4) | 732 (90.6) | |
| Black | 2106 (62.5) | 363 (17.2) | 1743 (82.8) | |
| Hispanic | 327 (9.7) | 38 (11.6) | 289 (88.4) | |
| Other | 128 (3.8) | 22 (17.2) | 106 (82.8) | |
| Education | <0.001 | |||
| High School or Less | 1301 (38.3) | 235 (18.1) | 1066 (81.9) | |
| Some College or Greater | 2095 (61.7) | 265 (12.7) | 1830 (87.3) | |
| Gross Annual Income | <0.001 | |||
| <$6,000 | 532 (16.2) | 116 (21.8) | 416 (78.2) | |
| $6,000 – $11,999 | 695 (21.2) | 128 (18.4) | 567 (81.6) | |
| $12,000 – $24,999 | 856 (26.1) | 119 (13.9) | 737 (86.1) | |
| $25,000 – $49,999 | 815 (24.8) | 94 (11.5) | 721 (88.5) | |
| ≥$50,000 | 383 (11.7) | 28 (7.3) | 355 (92.7) | |
| Marital Status | <0.001 | |||
| Married/Living with Partner | 1128 (33.5) | 126 (11.2) | 1002 (88.8) | |
| Divorced/Separated/Widowed | 1257 (37.3) | 200 (15.9) | 1057 (84.1) | |
| Never Married | 984 (29.2) | 171 (17.4) | 813 (82.6) | |
| Pain Interference in Daily Life | 0.115 | |||
| Yes | 1966 (58.2) | 305 (15.5) | 1661 (84.5) | |
| No | 1415 (41.9) | 192 (13.6) | 1223 (86.4) | |
| Opioid Rx, Past Year | 0.565 | |||
| None | 3265 (96.1) | 481 (14.7) | 2784 (85.3) | |
| Short-term | 102 (3.0) | 13 (12.8) | 89 (87.2) | |
| Long-term | 29 (0.9) | 6 (20.7) | 23 (79.3) | |
| Marijuana Use, Past Year | <0.001 | |||
| Yes | 654 (19.3) | 126 (19.3) | 528 (80.7) | |
| No | 2730 (80.7) | 373 (13.7) | 2357 (86.3) | |
| Stimulant Use, Past Year | 0.005 | |||
| Yes | 53 (1.6) | 15 (28.3) | 38 (71.7) | |
| No | 3341 (98.4) | 485 (14.5) | 2856 (85.5) | |
| Cocaine Use, Past Year | <0.001 | |||
| Yes | 421 (12.5) | 144 (34.2) | 277 (65.8) | |
| No | 2958 (87.5) | 354 (12.0) | 2604 (88.0) | |
| Unhealthy alcohol use, past year (AUDIT-C score ≥ 3 or 4) | 0.006 | |||
| Yes | 864 (25.4) | 152 (17.6) | 712 (82.4) | |
| No | 2532 (74.6) | 348 (13.7) | 2184 (86.3) | |
Note:
From Chi-square tests; HIV=Human Immunodeficiency Virus; HCV=Hepatitis C Virus
Figure 2 illustrates the Kaplan Meier curves for time to heroin initiation, which differed significantly between participants who reported prior/concurrent NMUPO and non-users; the log-rank test was significant at p<0.001. Of the 500 heroin initiates, 77% reported previous or concurrent NMUPO (p<0.001). The crude incidence rate for heroin initiation in our entire study sample was 2.60 per 100 person years, while the crude incidence rate for heroin was 4.82 per 100 person years among those reporting NMUPO and 1.02 per 100 person years among non-users, respectively (p < 0.001). Of the participants who reported new onset NMUPO, 27.3% initiated heroin by the end of the 10-year study period. Figure S1 (see Supplemental File) depicts the proportion initiating heroin use among participants who engaged in NMUPO at each time point.
Figure 2.
Relationship between prior/concurrent NMUPO and first-time initiation of heroin among VACS participants (2002-2012)
The results from both the unadjusted and adjusted Cox regression models using IPCW are shown in Table 3. The crude hazard for heroin initiation was significantly higher for males than for females and the crude hazards for black and Hispanic participants were significantly higher than for white participants. Those participants with a previous PTSD or depression diagnoses had a significantly higher hazard of heroin initiation, when compared to those without these diagnoses. The crude hazard ratios of heroin initiation among past year marijuana, stimulant and cocaine users were also significantly higher, than those participants who reported no use. For participants reporting concurrent or prior NMUPO versus none, the crude hazard ratio was 5.15 (95%CI: 3.89–6.81).
Table 3.
Inverse probability weighted Cox proportional hazard model of factors associated with time to self-reported incident heroin use among NMUPO/heroin naïve veterans participating in VACS, 2002-2012**
| Characteristic | Unadjusted HR* (95% CI) | P - value | Adjusted HR* (95% CI) | P - value |
|---|---|---|---|---|
|
| ||||
| NMUPO (any NMUPO versus none)† | 5.15 (3.89–6.81) | <0.001 | 5.43 (4.01–7.35) | <0.001 |
| Sex (ref: Female)¶ | 2.91 (1.18–7.19) | 0.021 | 2.61 (1.08–6.29) | 0.033 |
| Age (ref: ≤42 years)¶ | ||||
| 43 – 49 years | 1.79 (1.56–2.07) | <0.001 | 1.67 (1.44–1.93) | <0.001 |
| 50 – 56 years | 1.36 (1.20–1.53) | <0.001 | 1.33 (1.16–1.52) | <0.001 |
| ≥57 years | 0.76 (0.54–1.05) | 0.093 | 0.95 (0.64–1.42) | 0.817 |
| HIV (infected versus uninfected)¶ | 1.65 (1.44–1.89) | <0.001 | 1.17 (1.04–1.32) | 0.012 |
| HCV (infected versus uninfected)¶ | 2.19 (1.63–2.95) | <0.001 | 1.46 (1.24–1.72) | <0.001 |
| Race (ref: White)¶ | ||||
| Black | 1.88 (1.42–2.49) | <0.001 | 1.95 (1.46–2.61) | <0.001 |
| Hispanic | 1.39 (1.08–1.79) | 0.010 | 1.49 (1.09–2.03) | 0.011 |
| Other | 2.18 (1.85–2.57) | <0.001 | 1.86 (1.49–2.30) | <0.001 |
| Education (ref: High School or Less)¶ | ||||
| Some College or Greater | 0.67 (0.61–0.75) | <0.001 | 0.72 (0.59–0.87) | <0.001 |
| Gross Annual Income (ref: <$6,000)¶ | ||||
| $6,000 – $11,999 | 0.79 (0.59–1.07) | 0.131 | 1.01 (0.74–1.37) | 0.956 |
| $12,000 – $24,999 | 0.58 (0.49–0.67) | <0.001 | 0.88 (0.71–1.10) | 0.268 |
| $25,000 – $49,999 | 0.50 (0.35–0.70) | <0.001 | 0.78 (0.63–0.97) | 0.028 |
| ≥$50,000 | 0.28 (0.16–0.49) | <0.001 | 0.52 (0.30–0.90) | 0.019 |
| Marital Status (ref: Married/Living w/Partner)¶ | ||||
| Divorced/Separated/Widowed | 1.50 (1.19–1.90) | <0.001 | 1.14 (0.86–1.51) | 0.363 |
| Never Married | 1.74 (1.32–2.31) | <0.001 | 1.36 (0.99–1.86) | 0.061 |
| PTSD (ever diagnosis vs. none)‡ | 1.70 (1.43–2.03) | <0.001 | 1.25 (1.02–1.54) | 0.031 |
| Depression (ever diagnosis vs. none)‡ | 1.42 (1.16–1.72) | <0.001 | 0.99 (0.80–1.24) | 0.951 |
| Pain Interference in Daily Life (any vs. none)‡ | 1.26 (0.98–1.63) | 0.075 | 0.85 (0.65–1.12) | 0.254 |
| Opioid Rx (ref: none)‡ | ||||
| Short term | 1.44 (1.08–1.93) | 0.012 | 1.65 (1.43–1.90) | <0.001 |
| Long term | 1.06 (0.85–1.34) | 0.601 | 1.00 (0.68–1.48) | 0.996 |
| Unhealthy alcohol use, Past Year (AUDIT-C score ≥ 3 or 4)‡ | 1.18 (0.97–1.44) | 0.093 | 1.04 (0.85–1.29) | 0.688 |
| Marijuana Use, Past Year (ref: none)‡ | 1.34 (1.14–1.56) | <0.001 | 0.85 (0.66–1.09) | 0.207 |
| Cocaine Use, Past Year (ref: none)‡ | 2.94 (2.23–3.87) | <0.001 | 1.74 (1.10–2.76) | 0.019 |
| Stimulant Use, Past Year (ref: none) | 5.00 (3.19–7.83) | <0.001 | 2.12 (1.05–4.27) | 0.036 |
Note:
refers to time-updated covariates;
at baseline;
refers to any report of new onset NMUPO prior or concurrent to heroin initiation or censoring; HIV=Human Immunodeficiency Virus; HCV=Hepatitis C Virus; PTSD=Posttraumatic Stress Disorder; NMUPO = Nonmedical Use of Prescription Opioids;
Using robust sandwich estimator
In the multivariable Cox regression analysis, NMUPO remained positively and significantly associated with heroin initiation (adjusted hazard ratio [AHR] = 5.43, 95%CI: 4.01–7.35). In the fully adjusted and weighted model, those who reported stimulant use and cocaine use in the last year had AHRs of 2.12 (95%CI: 1.05–4.27) and 1.74 (95%CI: 1.10–2.76), respectively, compared to those participants who reported no past-year use. Receipt of a short-term opioid prescription from the VHA increased the hazard of heroin initiation by 65% (AHR = 1.65, 95%CI: 1.43–1.94) in our fully adjusted model. Other factors independently associated with heroin initiation are shown in Table 3. In post hoc analyses, the effect of NMUPO on heroin initiation varied significantly by race: (AHR = 5.46, 95%CI 3.72–8.02) for black participants, (AHR=7.67, 95%CI 4.23–13.88) for Hispanic participants, and (AHR=4.89, 95%CI 2.52–9.37) for white participants.
Sensitivity Analysis
After removing the 701 (20.6%) HCV-infected participants, the association between NMUPO and heroin initiation remained similar to that in the original model. Specifically, NMUPO had an AHR of 6.21 (95%CI: 4.54–8.51).
DISCUSSION
In this study, we characterized the relationship between new onset NMUPO and heroin initiation in a population of US military veterans receiving medical care in the VHA. Our results indicate a strong association between prior or concurrent NMUPO and initiation of heroin use. The observed effect was robust to covariate adjustment and a sensitivity analysis. To our knowledge, this study is the first to demonstrate an effect of NMUPO on risk for heroin initiation prospectively among veterans receiving medical care.
Given these findings, it is important for clinicians to be cognizant of risk factors that are associated with transitioning to heroin initiation among veterans and other high-risk populations. Our results, corroborated by other literature (12, 44, 45), indicate that being a racial minority, having lower education and income levels, and reporting other substance use (particularly cocaine, stimulant and alcohol use) are all associated with heroin initiation. HIV and HCV infection were both predictors of heroin initiation in our fully adjusted model. This is cause for concern, as the transmission of infectious diseases may result as a result of initiation of injection of heroin or prescription opioids (46). Finally, the differences in the effect of NMUPO on heroin initiation between black, white, and Hispanic participants may be due to disparities in access to appropriate pain management, and/or reduced access to opioid prescriptions among minorities (47). Future research is needed to further explore the role of reduced access to prescription opioids as a risk factor for heroin initiation in minority populations.
Our study suggests that the identification and treatment of nonmedical prescription opioid use in a veteran population could be an important strategy for preventing heroin initiation. Guidelines, including those developed by the Department of Veterans Affairs and Department of Defense, recommend that prescribers of long-term opioids regularly reassess treatment effectiveness, adverse effects, and adherence to therapy; monitor for evidence of opioid misuse or substance abuse; and consider written treatment agreements and periodic urine drug testing (48). It is important that continued attention be given to the development and refinement of screening procedures to identify problematic prescription opioid use among veterans receiving care.
Our study eligibility criteria resulted in the exclusion of a number of participants with opioid prescriptions from the VHA. For example, participants who reported any injection drug use and participants who had a previous diagnosis of an opioid dependence at baseline were excluded, both of which are factors associated with prescription opioid receipt in the VACS sample (39). For this reason, the proportion of the study sample who received an opioid prescription may be lower than the prevalence of opioid prescriptions in other VA populations, which averaged about 7.7%, with a range of 0.26 to 21.8% in 2012 (49).
Nonetheless, the finding that receipt of a short-term opioid prescription was independently associated with an increased hazard of heroin initiation adds to the literature demonstrating a strong correlation between therapeutic exposure to opioid analgesics and their abuse.(50) Collectively, these results supported recently published CDC guidelines recommending that, when opioids are used to treat acute pain, physicians should prescribe no greater quantity that needed for the expected duration of the severe pain.(51) These strategies may also reduce the total volume of diverted opioids for non-medical use. Finally, our findings suggest that clinicians should evaluate risk factors for heroin initiation (e.g., history of a substance use disorder) prior to initiating opioid therapy. The potential to reduce NMUPO and subsequent opioid use disorders and heroin use by reducing the prescribing of opioids is an especially important consideration for those countries where opioid prescribing is on the rise. If the rates of opioid prescribing and NMUPO continue to increase in Europe, future regulatory responses may be needed to prevent rates of opioid misuse that are seen in North America (52).
Our study had a number of limitations. First, our results may not be generalizable to all veterans receiving care in the VHA. Due to its design, the study enrolled individuals who are likely at higher risk for heroin initiation than the general veteran population (e.g., those with HIV and/or HCV infection). Notably, the rate of heroin initiation observed in this study, even among those non-exposed to NMUPO (1.0 per 100 person-years) is higher than the rate observed among US adults (0.11 per 100 person-years in 2011).(15) Future work is needed to examine the rate of heroin use and initiation among lower-risk populations of veterans and veterans not receiving care in the VHA. Another major limitation of this study was the fact that there were different versions of the survey question assessing nonmedical use of prescription opioid use. To mitigate potential information bias arising from the fact that participants may have misunderstood the question in the first two versions of the survey (as referring to medical use of prescription opioids), we adjusted for receipt of an opioid prescription in all analyses. Third, as with any survey questions reporting on substance use behaviors, it is likely that participants underreported their previous heroin and NMUPO. We attempted to address this issue by excluding participants who reported previous injection drug use and those with an ICD9 code for opioid dependence, as well as running sensitivity analyses excluding participants with hepatitis C. Additionally, we were only able to ascertain new onset NMUPO concurrent or prior to heroin initiation. Finally, differential loss to follow up was a potential source of bias in this study. This was addressed by using inverse probability of censoring weights to account for possible biases arising from differential loss to follow-up.
Despite these limitations, our study fills an important gap in the literature by comparing the demographic, clinical and substance use characteristics associated with heroin initiation among US veterans receiving medical care in the VHA. Identifying why particular veterans engage in NMUPO, and why a portion of them then transition to heroin initiation, are possible next steps in developing effective NMUPO screening strategies. For example, previous studies involving college students have found that the single leading reason for nonmedical use was to relieve pain (53–56). However, in our final model, pain interference in daily life was not a significant predictor of heroin initiation. Given that the experiences and the demographics of a veteran population are substantially different from young college students, further research needs to be conducted in order to elucidate motivations for NMUPO among veterans who receive medical care in the VHA. In sum, recognizing that NMUPO is a strong risk factor for heroin initiation suggests the urgent need for improved screening and assessment.
Supplementary Material
ACKNOLWEDGEMENTS
This work was supported by grants from the National Institute on Alcohol Abuse and Alcoholism (NIAAA: U10-AA013566, U01-AA020795, U01-AA020790, U24-AA020794, U10-AA013566, and P01-AA019072), the National Institute of Allergy and Infectious Diseases (P30-AI042853), and in kind by the US Department of Veterans Affairs. Dr. Julie Gaither is supported by the National Institute on Drug Abuse (F31-DA035567). E. Jennifer Edelman is a Yale-Drug Abuse, Addiction, and HIV Research Scholar (K12-DA033312). Dr. Brandon Marshall is supported by the National Institute on Drug Abuse (R03-DA037770). Dr. Silvia Martins is supported by the National Institute on Drug Abuse (R01-DA037866). Dr. Stephen Crystal is supported by AHRQ awards 1U19HS021112 and R18-HS023258. Dr. Robert Kerns is supported by a Center of Innovation grant from the Health Services Research and Development Service of the Department of Veterans Affairs (CIN 13-047). The sponsors had no role in the study design; the collection, analysis and interpretation of data; the writing of the report; and in the decision to submit the article for publication. We would like to acknowledge the veterans who participate in the Veterans Aging Cohort Study (VACS) and the study coordinators and staff at each VACS site and at the West Haven Coordinating Center. We would also like to thank Melissa Skanderson for her assistance and support during data acquisition. The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs.
Footnotes
Declaration of interest: none
REFERENCES
- 1.Volkow ND. Testimony before House Committee on Energy and Commerce, Subcommittee on Oversight and Investigations: National Institute on Drug Abuse. 2015. What science tells us about opioid abuse and addiction. [Google Scholar]
- 2.Hedegaard H, Chen L-H, Warner M. Drug-poisoning deaths involving heroin: United States, 2000–2013. Hyattsville, MD: 2015. [PubMed] [Google Scholar]
- 3.Unick GJ, Rosenblum D, Mars S, Ciccarone D. Intertwined epidemics: national demographic trends in hospitalizations for heroin- and opioid-related overdoses, 1993–2009. PloS one. 2013;8(2):e54496. doi: 10.1371/journal.pone.0054496. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Manchikanti L, Fellow B, Ailinani H, Pampati V. Therapeutic use, abuse, and nonmedical use of opioids: a ten-year perspective. Pain Physician. 2010;13:401–35. [PubMed] [Google Scholar]
- 5.Fischer B, Keates A, Buhringer G, Reimer J, Rehm J. Non-medical use of prescription opioids and prescription opioid-related harms: why so markedly higher in North America compared to the rest of the world? Addiction. 2014;109(2):177–81. doi: 10.1111/add.12224. [DOI] [PubMed] [Google Scholar]
- 6.Results from the 2012 National Survey on Drug Use and Health: Summary of National Findings. Substance Abuse and Mental Health Services Administration (SAMHSA); Rockville, MD: 2014. [PubMed] [Google Scholar]
- 7.Joranson D, Ryan K, Gilson A, Dahl J. Trends in medical use and abuse of opioid analgesics. JAMA. 2000;283:1710–4. doi: 10.1001/jama.283.13.1710. [DOI] [PubMed] [Google Scholar]
- 8.Gilson AM, Ryan KM, Joranson DE, Dahl JL. A reassessment of trends in the medical use and abuse of opioid analgesics and implications for diversion control: 1997–2002. Journal of pain and symptom management. 2004;28(2):176–88. doi: 10.1016/j.jpainsymman.2004.01.003. [DOI] [PubMed] [Google Scholar]
- 9.Paulozzi L, Franklin G, Kerlikowske R, Jones C, Ghiya N, Popovic T. Prescription drug overdoses - a U.S. epidemic. Centers for Disease Control and Prevention; 2012. [Google Scholar]
- 10.Voon P, Callon C, Nguyen P, Dobrer S, Montaner JS, Wood E, et al. Denial of prescription analgesia among people who inject drugs in a Canadian setting. Drug and alcohol review. 2015;34(2):221–8. doi: 10.1111/dar.12226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.van Amsterdam J, Phillips L, Henderson G, Bell J, Bowden-Jones O, Hammersley R, et al. Ranking the harm of non-medically used prescription opioids in the UK. Regulatory toxicology and pharmacology : RTP. 2015;73(3):999–1004. doi: 10.1016/j.yrtph.2015.09.014. [DOI] [PubMed] [Google Scholar]
- 12.Peavy KM, Banta-Green CJ, Kingston S, Hanrahan M, Merrill JO, Coffin PO. “Hooked on” prescription-type opiates prior to using heroin: results from a survey of syringe exchange clients. J Psychoactive Drugs. 2012;44(3):259–65. doi: 10.1080/02791072.2012.704591. [DOI] [PubMed] [Google Scholar]
- 13.Cicero TJ, Ellis MS, Surratt HL, Kurtz SP. The Changing Face of Heroin Use in the United States: A Retrospective Analysis of the Past 50 Years. JAMA Psychiatry. 2014;71(7):821–6. doi: 10.1001/jamapsychiatry.2014.366. [DOI] [PubMed] [Google Scholar]
- 14.Jones CM. Heroin use and heroin use risk behaviors among nonmedical users of prescription opioid pain relievers - United States, 2002–2004 and 2008–2010. Drug Alcohol Depend. 2013;132(1–2):95–100. doi: 10.1016/j.drugalcdep.2013.01.007. [DOI] [PubMed] [Google Scholar]
- 15.Muhuri PK, Gfroerer JC, Davies MC. Associations of nonmedical pain reliever use and initiation of heroin use in the United States. Rockville, MD: 2013. [Google Scholar]
- 16.Irish C, Maxwell R, Dancox M, Brown P, Trotter C, Verne J, et al. Skin and soft tissue infections and vascular disease among drug users, England. Emerg Infect Dis. 2007;13(10):1510–1. doi: 10.3201/eid1310.061196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dart RC, Surratt HL, Cicero TJ, Parrino MW, Severtson SG, Bucher-Bartelson B, et al. Trends in opioid analgesic abuse and mortality in the United States. N Engl J Med. 2015;372(3):241–8. doi: 10.1056/NEJMsa1406143. [DOI] [PubMed] [Google Scholar]
- 18.Robins LN, Helzer JE, Hesselbrock M, Wish E. Vietnam veterans three years after Vietnam: how our study changed our view of heroin. The American journal on addictions / American Academy of Psychiatrists in Alcoholism and Addictions. 2010;19(3):203–11. doi: 10.1111/j.1521-0391.2010.00046.x. [DOI] [PubMed] [Google Scholar]
- 19.Bray RM, Hourani LL. Substance use trends among active duty military personnel: findings from the United States Department of Defense Health Related Behavior Surveys, 1980–2005. Addiction. 2007;102(7):1092–101. doi: 10.1111/j.1360-0443.2007.01841.x. [DOI] [PubMed] [Google Scholar]
- 20.McFall ME, Mackay PW, Donovan DM. Combat-related posttraumatic stress disorder and severity of substance abuse in Vietnam veterans. J Stud Alcohol. 1992;53(4):357–63. doi: 10.15288/jsa.1992.53.357. [DOI] [PubMed] [Google Scholar]
- 21.Lew HL, Otis JD, Tun C, Kerns RD, Clark ME, Cifu DX. Prevalence of chronic pain, posttraumatic stress disorder, and persistent postconcussive symptoms in OIF/OEF veterans: polytrauma clinical triad. J Rehabil Res Dev. 2009;46(6):697. doi: 10.1682/jrrd.2009.01.0006. [DOI] [PubMed] [Google Scholar]
- 22.Golub A, Bennett AS. Prescription opioid initiation, correlates, and consequences among a sample of OEF/OIF military personnel. Substance use & misuse. 2013;48(10):811–20. doi: 10.3109/10826084.2013.796988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Seal KH, Shi Y, Cohen G, Cohen BE, Maguen S, Krebs EE, et al. Association of mental health disorders with prescription opioids and high-risk opioid use in US veterans of Iraq and Afghanistan. JAMA. 2012;307(9):940–7. doi: 10.1001/jama.2012.234. [DOI] [PubMed] [Google Scholar]
- 24.Robins LN. Vietnam veterans' rapid recovery from heroin addiction: a fluke or normal expectation? Addiction. 1993;88:1041–54. doi: 10.1111/j.1360-0443.1993.tb02123.x. [DOI] [PubMed] [Google Scholar]
- 25.Barry DT, Goulet JL, Kerns RK, Becker WC, Gordon AJ, Justice AC, et al. Nonmedical use of prescription opioids and pain in veterans with and without HIV. Pain. 2011;152(5):1133–8. doi: 10.1016/j.pain.2011.01.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Green TC, Kershaw T, Lin H, Heimer R, Goulet JL, Kraemer KL, et al. Patterns of drug use and abuse among aging adults with and without HIV: A latent class analysis of a US Veteran cohort. Drug Alcohol Depend. 2010;110(3):208–20. doi: 10.1016/j.drugalcdep.2010.02.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gaither JR, Goulet JL, Becker WC, Crystal S, Edelman EJ, Gordon K, et al. Guideline-concordant management of opioid therapy among human immunodeficiency virus (HIV)-infected and uninfected veterans. J Pain. 2014;15(11):1130–40. doi: 10.1016/j.jpain.2014.08.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Edelman EJ, Gordon K, Becker WC, Goulet JL, Skanderson M, Gaither JR, et al. Receipt of opioid analgesics by HIV-infected and uninfected patients. J Gen Intern Med. 2013;28(1):82–90. doi: 10.1007/s11606-012-2189-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Weisberg DF, Gordon KS, Barry DT, Becker WC, Crystal S, Edelman EJ, et al. Long-term prescription of opioids and/or benzodiazepines and mortality among HIV-infected and uninfected patients. J Acquir Immune Defic Syndr. 2015;69(2):223–33. doi: 10.1097/QAI.0000000000000591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Justice AC, Dombrowski E, Conigliaro J, Fultz SL, Gibson D, Madenwald T, et al. Veterans Aging Cohort Study (VACS): Overview and description. Med Care. 2006;44(8 Suppl 2):S13–24. doi: 10.1097/01.mlr.0000223741.02074.66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Justice AC, Lasky E, McGinnis KA, Skanderson M, Conigliaro J, Fultz SL, et al. Medical disease and alcohol use among veterans with human immunodeficiency infection: a comparison of disease measurement strategies. Med Care. 2006;44(8):S52–S60. doi: 10.1097/01.mlr.0000228003.08925.8c. [DOI] [PubMed] [Google Scholar]
- 32.Swartz JA, Lurigio AJ. Screening for serious mental illness in populations with co-occurring substance use disorders: Performance of the K6 scale. Journal of substance abuse treatment. 2006;31(3):287–96. doi: 10.1016/j.jsat.2006.04.009. [DOI] [PubMed] [Google Scholar]
- 33.Boscarino JA, Rukstalis M, Hoffman SN, Han JJ, Erlich PM, Gerhard GS, et al. Risk factors for drug dependence among out-patients on opioid therapy in a large US health-care system. Addiction. 2010;105(10):1776–82. doi: 10.1111/j.1360-0443.2010.03052.x. [DOI] [PubMed] [Google Scholar]
- 34.Dobscha SK, Morasco BJ, Duckart JP, Macey T, Deyo RA. Correlates of prescription opioid initiation and long-term opioid use in veterans with persistent pain. Clin J Pain. 2013;29(2):102–8. doi: 10.1097/AJP.0b013e3182490bdb. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Khosla N, Juon HS, Kirk GD, Astemborski J, Mehta SH. Correlates of non-medical prescription drug use among a cohort of injection drug users in Baltimore City. Addict Behav. 2011;36(12):1282–7. doi: 10.1016/j.addbeh.2011.07.046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Kolodny A, Courtwright DT, Hwang CS, Kreiner P, Eadie JL, Clark TW, et al. The prescription opioid and heroin crisis: a public health approach to an epidemic of addiction. Annu Rev Public Health. 2015;36:559–74. doi: 10.1146/annurev-publhealth-031914-122957. [DOI] [PubMed] [Google Scholar]
- 37.Salyers MP, Bosworth HB, Swanson JW, Lamb-Pagone J, Osher FC. Reliability and validity of the SF-12 health survey among people with severe mental illness. Med Care. 2000;38(11):1141–50. doi: 10.1097/00005650-200011000-00008. [DOI] [PubMed] [Google Scholar]
- 38.Bradley KA, Bush KR, Epler AJ, Dobie DJ, Davis TM, Sporleder JL, et al. Two brief alcohol-screening tests from the alcohol use disorders identification test (AUDIT) Arch Intern Med. 2003;163:821–9. doi: 10.1001/archinte.163.7.821. [DOI] [PubMed] [Google Scholar]
- 39.Becker WC, Gordon K, Jennifer Edelman E, Kerns RD, Crystal S, Dziura JD, et al. Trends in any and high-dose opioid analgesic receipt among aging patients with and without HIV. AIDS and behavior. 2015 doi: 10.1007/s10461-015-1197-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Christensen E. Multivariate survival analysis using Cox's regression model. Hepatology. 1987;7(6):1346–58. doi: 10.1002/hep.1840070628. [DOI] [PubMed] [Google Scholar]
- 41.Robins JM, Finkelstein DM. Correcting for noncompliance and dependent censoring in an AIDS clinical trial with inverse probability of censoring weighted (IPCW) log-rank tests. Biometrics. 2000;56:779–88. doi: 10.1111/j.0006-341x.2000.00779.x. [DOI] [PubMed] [Google Scholar]
- 42.Cain LE, Cole SR. Inverse probability-of-censoring weights for the correction of time-varying noncompliance in the effect of randomized highly active antiretroviral therapy on incident AIDS or death. Stat Med. 2009;28(12):1725–38. doi: 10.1002/sim.3585. [DOI] [PubMed] [Google Scholar]
- 43.Graubard BI, Korn EL. Regression Analysis with Clustered Data. Statistics in medicine. 1994;13:509–22. doi: 10.1002/sim.4780130514. [DOI] [PubMed] [Google Scholar]
- 44.Lipari RN, Hughes A. The NSDUH report: trends in heroin use in the United States: 2002 to 2013. Substance Abuse and Mental Health Services Administration (SAMHSA); Rockville, MD: 2015. [PubMed] [Google Scholar]
- 45.Mars SG, Bourgois P, Karandinos G, Montero F, Ciccarone D. “Every `never' I ever said came true”: transitions from opioid pills to heroin injecting. Int J Drug Policy. 2014;25(2):257–66. doi: 10.1016/j.drugpo.2013.10.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Guarino H, Marsch LA, Deren S, Straussner SL, Teper A. Opioid use trajectories, injection drug use, and hepatitis C virus risk among young adult immigrants from the former Soviet Union living in New York City. Journal of addictive diseases. 2015;34(2–3):162–77. doi: 10.1080/10550887.2015.1059711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Ringwalt C, Roberts AW, Gugelmann H, Skinner aC. Racial disparities across provider specialties in opioid prescriptions dispensed to medicaide beneficiaries with chronic noncancer pain. Pain Med. 2015;16:633–40. doi: 10.1111/pme.12555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Krebs EE, Ramsey DC, Miloshoff JM, Blair MJ. Primary care monitoring of long-term opioid therapy among veterans with chronic pain. Pain Med. 2011;12:740–6. doi: 10.1111/j.1526-4637.2011.01099.x. [DOI] [PubMed] [Google Scholar]
- 49.Healthcare Inspection - VA Patterns of Dispensing Take-Home Opioids and Monitoring Patients on Opioid Therapy. 2014 [Available from: http://www.va.gov/oig/pubs/VAOIG-14-00895-163.pdf.
- 50.Cicero TJ, Surratt H, Inciardi JA, Munoz A. Relationship between therapeutic use and abuse of opioid analgesics in rural, suburban, and urban locations in the United States. Pharmacoepidemiology and drug safety. 2007;16(8):827–40. doi: 10.1002/pds.1452. [DOI] [PubMed] [Google Scholar]
- 51.Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain -- United States, 2016. MMWR Morb Mortal Wkly Rep. 2016;65(1) doi: 10.15585/mmwr.rr6501e1. [DOI] [PubMed] [Google Scholar]
- 52.Voon P, Kerr T. “Nonmedical” prescription opioid use in North America: a call for priority action. Substance Abuse Treatment, Prevention, and Policy. 2013;8(39) doi: 10.1186/1747-597X-8-39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.McCabe SE, Cranford JA, Boyd CJ, Teter CJ. Motives, diversion and routes of administration associated with nonmedical use of prescription opioids. Addict Behav. 2007;32(3):562–75. doi: 10.1016/j.addbeh.2006.05.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Sung HE, Richter L, Vaughan R, Johnson PB, Thom B. Nonmedical use of prescription opioids among teenagers in the United States: trends and correlates. J Adolesc Health. 2005;37(1):44–51. doi: 10.1016/j.jadohealth.2005.02.013. [DOI] [PubMed] [Google Scholar]
- 55.McCabe SE, West BT, Teter CJ, Boyd CJ. Medical and nonmedical use of prescription opioids among high school seniors in the United States. Arch Pediatr Adolesc Med. 2012;166(9):797–802. doi: 10.1001/archpediatrics.2012.85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Boyd CJ, Young A, Grey M, McCabe SE. Adolescents' nonmedical use of prescription medications and other problem behaviors. The Journal of adolescent health : official publication of the Society for Adolescent Medicine. 2009;45(6):543–50. doi: 10.1016/j.jadohealth.2009.03.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.