Abstract
Objectives
To determine the added risk of opioid problem use (OPU) in youth with marijuana/alcohol problem use (MAPU).
Method
475 youth (ages 14–21 years) with OPU+MAPU were compared to a weighted sample of 475 youth with MAPU only (i.e., no OPU) before and after propensity score matching on gender, age, race, level of care, and weekly use of marijuana/alcohol. Youth were recruited from 88 drug treatment sites participating in eight Center for Substance Abuse Treatment funded grants. At treatment intake, participants were administered the Global Appraisal of Individual Need to elicit information on demographic, social, substance, mental health, HIV, physical and legal characteristics. Odds ratios with confidence intervals were calculated.
Results
The added risk of OPU among MAPU youth was associated with greater comorbidity: higher rates of psychiatric symptoms and trauma/victimization; greater needle-use and sex-related HIV-risk behaviors and greater physical distress. The OPU+MAPU group was less likely to be African American or other race and more likely to be age 15–17 years, Caucasian; report weekly drug use at home and among peers; engage in illegal behaviors and be confined longer; have greater substance abuse severity and poly drug use; and use mental health and substance abuse treatment services.
Conclusions
These findings expand on the existing literature and highlight the substantial incremental risk of OPU on multiple comorbid areas, among treatment-seeking youth. Further evaluation is needed to assess their outcomes following standard drug treatment and to evaluate specialized interventions for this subgroup of severely impaired youth.
Introduction
Over the past decade, past year prevalence of using most substances including marijuana and alcohol has either decreased or reached a plateau among U.S. teenagers in schools [1] and in the community [2]. In contrast, use of non-heroin opioids, the second most commonly used illicit drug among youth has almost doubled during this period (5 to 9%) [1]. It has been shown that teen opioid users develop abuse and/or dependence within months of weekly or greater (i.e. regular) use [3] and at a faster rate when compared to adult- onset opioid use [4]. Correspondingly, there has been a ten-fold increase in teenage admissions to publicly funded substance abuse treatment programs for non-heroin opioid use problems (0.2 to 2.2%) [2]. Therefore, problem use of opioids among youth has important implications for public health, with a heightened risk for overdose, mortality, HIV/Hepatitis-C infections from injection-use, and associations with a criminal lifestyle [5–7].
In contrast to a more extensive body of literature characterizing the typical youth with cannabis and/or alcohol use problems, often drawn from large samples, studies characterizing treatment-seeking youth with opioid use problems have examined smaller samples and are only just emerging [3, 8–12]. These studies have consistently described users/abusers as being predominantly Caucasian, male and older teens who present with poor academic involvement, legal problems and polysubstance use; and who are at high risk for HIV/Hepatitis – C infection secondary to injection drug use (IDU) and unprotected sexual behaviors. However, all (except for Woody et al. [8]), were single site studies and most recruited from residential settings [3, 9–11]. Two of these studies compared opioid using/dependent youth to non-opioid using youth entering treatment. Clemmey et al. [10] compared 44 heroin-users [with or without other substance use disorders (SUD)] to non-heroin substance abusing youth (n=109) using the Global Appraisal of Individual Needs (GAIN). They found that the heroin users consistently reported more baseline problems across multiple domains, including symptoms of psychiatric disorders, and that they had worse outcomes post-residential treatment. In a separate study, Subramaniam et al. [3], compared 94 adolescents with an opioid use disorder (OUD, 53–57 % of whom also had a current cannabis and/or alcohol use disorder) to 72 teens with non-opioid cannabis/alcohol use disorders. They showed that while adolescents with an OUD were more likely to have academic, multiple substance use, depressive and needle-use HIV-risk problems, they were no different from the cannabis/alcohol use disorder group with regard to high rates of psychiatric, legal and sexual risk behaviors. A limit of this study was that co-occurring psychiatric problems were often placement criteria for the residential treatment program and may explain why this study did not observe any difference on them. While these prior studies have made valuable contributions to our knowledge of youth with OUD, and shed some light on how they differ from non-opioid abusing youth, generalizeability was limited by the use of single site data and/or entire or predominantly residential samples and in most cases, small samples. What is needed is a description of a larger and more diverse treatment sample of youth in order to evaluate the added risks of problem use of opioids in youth.
This paper examines the incremental risk of opioid problem use (OPU) in a treatment-seeking sample of youth (e.g., adolescents and young adults) with OPU plus, marijuana and/or alcohol problem use (MAPU) when compared to those with those with MAPU only (the modal profile in treatment [2, 13]). We anticipate that having OPU in addition to MAPU (i.e. having OPU+MAPU) will be associated with being Caucasian, having poor school performance, engaging in polysubstance use, reporting higher rates of illegal behaviors and HIV/Hepatitis-C infection-risk behaviors; thereby confirming features that have been consistently reported in single site studies. In addition, we expect to report new findings: 1) higher rates of co-occurring psychiatric problems likely due to affective distress states associated with opioid withdrawal [14, 15] and demoralization from a lifestyle of high rates of crime and opioid dependence [3]; 2) determine if rates of victimization/trauma and physical distress are higher among opioid problem users compared to those without, since opioids are analgesics and this choice may represent self-medication of emotional and/or physically painful conditions; and 3) describe treatment utilization rates in this population, which may provide information on mechanisms for early identification and improving abstinence outcomes. Results of this study will extend the literature, by including a large sample of participants entering either outpatient or residential substance use disorder (SUD) treatment across the US. These results will better inform targeted treatment planning and the appropriate allocation of program resources for OPU youth.
Methods
Data Source
Intake assessments from participants ages 11–21 years entering treatment for SUDs were pooled together from 88 sites participating in 8 Center for Substance Abuse Treatment (CSAT) funded demonstration projects [Strengthening Communities-Youth (SCY), Adolescent Residential Treatment (ART), Effective Adolescent Treatment (EAT), Targeted Capacity Expansion (TCE, include TCE/HIV), Assertive Adolescent Family Treatment, (AAFT), Young Offender Reentry Program (YORP), Drug Court (DC), Adolescent Treatment Models (ATM)] that recruited youth participants from January 1999-August 2007 [13, 16, 17]. The data were pooled and de-identified for secondary analyses here under the terms of data sharing agreements and the supervision of Chestnut Health System’s Institutional Review Board. A common feature in all studies is that all participants were administered the Global Appraisal of Individual Needs – Intake assessment (GAIN-I) [18].
Inclusion and Exclusion Criteria
To be included in this study, participants had to (a) be between ages 11–21 years; (b) meet past year Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) symptom criteria for cannabis or alcohol use disorders (dependence or abuse) or self report weekly or more frequent use of cannabis or alcohol in either the 90 days prior to treatment enrollment or (if coming from a controlled environment like detention or residential treatment) in the last 90 days they were in the community prior to the confinement; and (c) have entered adolescent or young adult treatment settings. The use of both categorical criteria for DSM abuse/dependence and a dimensional severity criteria of weekly use or greater, as used in previous studies [19] ensures that the groups represent the typical youth entering community-based treatments for problems associated with substance use [2]. Of the 12,951 youth in the original data set, 11,201 (71%) met these inclusion criteria. To focus the contrast on those with and without problem opioid use (dependence, abuse or weekly use), we excluded another 3,771 (13% of 12,951) who had used opioids infrequently at one or more times in the past year, but did not meet the preceding threshold of problem use since they denied problems associated with opioid use. This left a final data set of 7,430 participants who could be classified into one of the two study groups, described below.
Definition of Study Groups
Opioid Problem Use (OPU+MAPU) group
Participants in this group either met DSM-IV criteria for opioid dependence (24%) or abuse (38%) and/or self-reported weekly or more frequent use of opioids in the past 90 days controlled for confinement (80%) plus marijuana and/or alcohol problem use (i.e. dependence, abuse or weekly use, MAPU). Per the inclusion and exclusion criteria described above, by definition, this group had a marijuana and/or alcohol problem as well. A total of 513 (7%) participants were identified with OPU+MAPU.
Marijuana and/or Alcohol Problem Use (MAPU) Only Group
The remaining 6,917 (93%) participants had a marijuana and/or alcohol problem [met DSM_IV criteria for dependence (29%) or abuse (47%) and/or weekly use of marijuana and/or alcohol (91%)], but did not meet the above criteria for opioid problem use. Per above, this group was also limited to those with “no” past year opioid use.
Matching via Propensity Score Weighting
To help control for spurious findings and create a more rigorous comparison, two samples were analyzed both without and with matching based on propensity score weighting [20, 21]. The propensity score was calculated using logistic regression to predict being in the OPU+MAPU group or not as a function of core demographics (gender, race, age); substance use severity (dichotomous measures or weekly cannabis and alcohol use vs. <weekly use); level of care (outpatient vs. residential); and whether the site had a high concentration of opioid users (15 or more in the site vs. <15 in the site). The weight for the MAPU group was then calculated as the propensity score times the sample size of the OPU group with valid responses on all propensity variables (93%) over the sum of the MAPU weights. Adding the latter (a constant) does not affect the match but creates equal sample sizes per group and creates a slightly more conservative test. OPU+MAPU youth were assigned a weight of 1 and thus do not change regardless of whether the analysis is unmatched or matched. The final weighted equal-sized groups contained 475 participants each.
Assessment
Data was obtained from the Global Appraisal of Individual Needs – Intake (GAIN-I) [18], which was administered by trained interviewers at treatment intake. The GAIN is a standardized, comprehensive structured interview widely used in adolescent and young adult substance abuse treatment research and it takes 1–2 hours to administer. It consists of eight main sections: demographics/background, substance use, physical health, risk behaviors, mental health, environment risk, legal involvement and academic/vocational issues. For this study, responses to individual questions and a few composite GAIN scales were selected to provide lifetime, past-year and/or past 90-day information on several clinical characteristics reported in the literature (see introduction). Additional details of the GAIN including its psychometric properties are available at www.chestnut.org/li/gain. The following is a description of clinical characteristics examined in this study that were derived from select GAIN scales:
SUD Diagnosis
Responses to questions on the GAIN’s Substance Problem Scale (which includes 7 dependence symptoms, 4 abuse symptoms, 2 symptoms of substance induced disorders and 3 screener items including weekly use) which can be used as a dimensional measure of severity or to categorize participants based on whether they met past year DSM-IV criteria for dependence or abuse or weekly use. This could be done overall or for each substance in DSM-IV. Those who either met criteria for dependence or abuse or weekly use were combined to create the OPU and MAPU groupings.
Internalizing Psychiatric Disorders
Responses to the Internal Mental Distress Scale in the mental health section of the GAIN were dichotomized to yield information on whether or not participants met past-year DSM-IV symptom criteria for any of the internalizing disorders. For this study they consist of major depressive disorder (MDD), generalized anxiety disorder (GAD) and traumatic stress disorder [23].
Externalizing Psychiatric Disorders
Responses to the Behavioral Complexity Scale in the mental health section of the GAIN were dichotomized to yield information on whether or not participants met past-year DSM-IV symptom criteria for any of the externalizing disorders, attention deficit/hyperactivity disorder (ADHD) and conduct disorder (CD) [23].
Victimization
Responses reporting physical, emotional or sexual abuse were counted for occurring ever (lifetime) or in the past year. Responses from the 15-item General Victimization Scale (GVS) in the environment section of the GAIN were categorized as having moderate to high levels of victimization for participant scores of 4 or more. Moderate/high levels on GVS correlates with high mental distress and reported traumatic distress [24]. Current worry about victimization includes any concern that physical, emotional or sexual abuse might occur.
HIV-risk Behaviors
Needle Risk: Those scoring 1 or more on the 9-item Needle Problem Scale are considered to have moderate/high needle risk in the past year [18, 25]. Since needle use is so infrequent, any problems related to use were considered moderate or high risk. Sexual Risk: Those scoring 2 or more on the 12-item Sexual Risk Scale were considered to have moderate/high sexual risk in the past year [25].
Physical Distress
Participants scoring 0.14 or higher on the proportional 14-item Health Problem Scale (HPS) were considered to have moderate/high levels of health problems (e.g., weekly or more frequent days with health problem, health problems keeping you from meeting responsibilities at home, school or work etc) in the past year [18]. The HPS was designed to be comparable to the Addiction Severity Index (ASI) medical scale. Those who also reported poor health were included under any physical distress or health problem.
Treatment Utilization
Rates of past SUD and mental health treatment utilization were calculated for reports of receiving any of the four following treatments: any emergency room visits, nights at a hospital, night at residential treatment, or outpatient office visits.
Statistical Analyses
Propensity score weighting was used for all analyses (see above) to control for demographics, level of care, frequency of marijuana and alcohol use and concentration of opioid users in the site. For each dependent variable, the odds ratio (OR) with 95% confidence intervals (CI) were calculated to represent the likelihood that the ‘variable of interest’ was associated with having OPU. Given the large sample sizes, even small ORs can be statistically significant, so we have focused the narrative on those that may also be clinically significant (i.e. only OR ≥ 1.20 or O.R. ≤ 0.8) [26]. All analyses were conducted in SPSS version 14.0.2 [27].
Results
Demographic, Environmental, Academic/Vocational and Illegal Activity Characteristics
Table-1 compares the OPU+MAPU group with the unweighted and weighted MAPU group. The unweighted group is presented primarily as a reference to the literature (which has largely focused on whether there were any differences), while the weighted comparison looks at the difference after matching on substance use severity, level of care and program and is the focus of this paper. Even after matching, the OPU+MAPU group was significantly less likely than the weighted MAPU group to be African American or other race and significantly more likely to be Caucasian, age 15 to 17, have weekly alcohol/drug use in their home or among their social peers, to be employed, to have been in a controlled environment, and to be higher on each of the measures of crime and violence.
Table 1.
Demographics and Social Characteristics of Opioid Problem Users (OPU, n=475) and Marijuana/Alcohol Problem Users (MAPU, =475)
| MAPU | OPU+MAPU vs.MAPU Unweighted | OPU+MAPU vs. MAPU Weighted | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Unweighted (n=6471) | Weighted (n=475) | OPU+MAPU (n=475) | ||||||||
| N (Mean) | % (SD) | N (Mean) | % (SD) | N (Mean) | % (SD) | Odds ratiof | 95%Confidence Interval | Odds ratiof | 95%Confidence Interval | |
| I. Demographics | ||||||||||
| % Female | 1588 | 25% | 163 | 34% | 173 | 36% | 1.76 | (1.6 – 2) | 1.09 | (0.8 – 1.4) |
| Race | ||||||||||
| % African American | 1295 | 20% | 101 | 21% | 20 | 4% | 0.18 | (0 – 0.6) | 0.16 | (0 – 0.7) |
| % Caucasian | 2557 | 40% | 176 | 37% | 308 | 65% | 2.82 | (2.6 – 3) | 3.13 | (2.9 – 3.4) |
| % Mixed | 913 | 14% | 74 | 16% | 60 | 13% | 0.88 | (0.6 – 1.2) | 0.78 | (0.4 – 1.2) |
| % Other | 1706 (15.8) | 26% – 144% | 124 (16.2) | 26% (1.5) | 87 (16.3) | 18% (1.4) | 0.63 | (0.4 – 0.9) | 0.63 | (0.3 – 0.9) |
| Age | ||||||||||
| % < 15 years | 1210 | 19% | (16.2) | 10% | 31 | 7% | 0.30 | (0 – 0.7) | 0.65 | (0.2 – 1.1) |
| % 15–17 years | 4789 | 74% | (16.2) | 76% | 381 | 80% | 1.42 | (1.2 – 1.7) | 1.29 | (1 – 1.6) |
| % 18–21 years | 472 | 7% | (16.2) | 15% | 63 | 13% | 1.94 | (1.7 – 2.2) | 0.90 | (0.5 – 1.3) |
| Site Experience with Opioid users | ||||||||||
| Clients from sites with <15 OP users | 5746 | 89% | 314 | 0.66 | 287 | 0.60 | 0.19 | (0 – 0.4) | 0.78 | (0.4 – 1.2) |
| Clients from sites with 15+ OP users | 725 | 11% | 161 | 0.34 | 188 | 0.40 | 5.19 | (5 – 5.4) | 1.28 | (0.9 – 1.6) |
| II. Environment: | ||||||||||
| Family/Home: | ||||||||||
| % Single Parent family | 3203 | 50% | 232 | 49% | 217 | 46% | 0.86 | (0.7 – 1.1) | 0.88 | (0.6 – 1.2) |
| % Weekly Alcohol Use among those at home | 1704 | 26% | 135 | 28% | 174 | 37% | 1.62 | (1.4 – 1.8) | 1.46 | (1.2 – 1.7) |
| % Weekly Drug Use among those at home | 843 | 13% | 84 | 18% | 121 | 25% | 2.27 | (2.1 – 2.5) | 1.59 | (1.3 – 1.9) |
| % Ever Homeless/Runaway | 2011 | 31% | 189 | 40% | 262 | 55% | 2.72 | (2.5 – 2.9) | 1.85 | (1.6 – 2.1) |
| Social Peers | ||||||||||
| % Regular Alcohol Use by Social Peers | 3649 | 56% | 308 | 65% | 338 | 71% | 1.91 | (1.7 – 2.1) | 1.34 | (1.1 – 1.6) |
| % Regular Drug Use by Social Peers | 4919 | 76% | 375 | 79% | 404 | 85% | 1.79 | (1.5 – 2.1) | 1.52 | (1.2 – 1.9) |
| III. Academic/Vocational | ||||||||||
| % Either in school, High School Grad or GEDa | 5846 | 90% | 415 | 87% | 400 | 84% | 0.57 | (0.3 – 0.8) | 0.77 | (0.4 – 1.1) |
| % Employeda | 1906 | 29% | 138 | 29% | 175 | 37% | 1.40 | (1.2 – 1.6) | 1.43 | (1.2 – 1.7) |
| IV. Illegal Activity | ||||||||||
| Juvenile Justice/Police Involvement | ||||||||||
| % Lifetime involved | 5489 | 85% | 410 | 86% | 404 | 85% | 1.01 | (0.7 – 1.3) | 0.90 | (0.5 – 1.3) |
| % Currently involveda | 4600 | 71% | 335 | 71% | 316 | 67% | 0.81 | (0.6 – 1) | 0.83 | (0.6 – 1.1) |
| % In detention/jaila | 1674 | 26% | 140 | 30% | 141 | 30% | 1.21 | (1 – 1.4) | 1.01 | (0.7 – 1.3) |
| % On probation/parolea | 3131 | 48% | 235 | 50% | 229 | 48% | 1.21 | (0.8 – 1.2) | 0.94 | (0.7 – 1.2) |
| % Arresteda | 1436 | 22% | 112 | 24% | 106 | 22% | 1.21 | (0.8 – 1.2) | 0.93 | (0.6 – 1.2) |
| % In a Controlled Environmenta | 2452 | 38% | 232 | 49% | 285 | 60% | 1.21 | (2.3 – 2.6) | 1.57 | (1.3 – 1.8) |
| % 13+ Days in Controlled Environmenta | 1586 | 25% | 171 | 36% | 224 | 47% | 1.21 | (2.6 – 2.9) | 1.59 | (1.3 – 1.9) |
| Criminal Behaviors | ||||||||||
| Past-year Any Violence or Illegal Activity | 5246 | 81% | 398 | 84% | 431 | 91% | 2.29 | (2 – 2.6) | 1.90 | (1.5 – 2.3) |
| Any Act of Physical Violenceb | 4432 | 68% | 343 | 72% | 392 | 83% | 2.19 | (1.9 – 2.4) | 1.83 | (1.5 – 2.1) |
| Any Property crimesc | 3072 | 47% | 247 | 52% | 341 | 72% | 2.82 | (2.6 – 3) | 2.35 | (2.1 – 2.6) |
| Any Interpersonal crimesd | 2828 | 44% | 224 | 47% | 302 | 64% | 2.25 | (2.1 – 2.4) | 1.96 | (1.7 – 2.2) |
| Any Drug related crimese | 2943 | 45% | 241 | 51% | 323 | 68% | 2.55 | (2.3 – 2.7) | 2.06 | (1.8 – 2.3) |
| Any Past-year Driving under the influence | 1206 | 19% | 134 | 28% | 225 | 47% | 3.93 | (3.7 – 4.1) | 2.28 | (2 – 2.5) |
| Any Past-year Drug Dealing | 1457 | 23% | 142 | 30% | 270 | 57% | 4.51 | (4.3 – 4.7) | 3.07 | (2.8 – 3.3) |
| Any Past-year Trading Sex for Drugs/money | 64 | 1% | 9 | 2% | 27 | 6% | 6.18 | (5.7 – 6.6) | 3.09 | (2.3 – 3.9) |
| Any Past-year Gambling | 1040 | 16% | 83 | 17% | 114 | 24% | 1.65 | (1.4 – 1.9) | 1.50 | (1.2 – 1.8) |
Notes: OPU defined as meeting criteria for past year opioid dependence/abuse or weekly or greater opioid use in the past 90-days and have mar/alc problem use; MAPU defined as meeting criteria for past year marijuana/alcohol dependence/abuse or weekly or greater marijuana/alcohol use in the past 90-days and no opioid problem use; Bold odds ratio are significantly different than 1 at p<.05.
During the past 90 days
Physical assault of another person within the past year.
Self report of or arrests related to vandalism, forgery, bad checks, shoplifting, theft, robbery, auto theft.
Self report of or arrests related to assault, aggravated assault with a weapon, rape, murder, and arson.
Substance Use Characteristics
Table-2 shows that the two groups were significantly different on the majority of substance use characteristics. The OPU+MAPU group compared to the weighted MAPU group initiated substance use at an earlier age (i.e. prior to age 10) and were less likely to initiate use between ages 15–17 years (average age: 11.7 vs. 12.6 years, p= 0.00, data not shown in table). They were also more likely to have used substances for 5 or more years while the MAPU group was more likely to report 1–2 years of substance use. The OPU+MAPU group was more likely the MAPU group to report weekly use of tobacco, crack/cocaine and other substances. (Note that weekly cannabis and alcohol use were not significantly different between the groups because they were intentionally matched on these). The OPU+MAPU group was also more likely to meet past-year any substance dependence diagnosis (excluding nicotine). They were also more likely to meet past-year SUD diagnosis for any substance, cocaine, amphetamines, and other drugs (hallucinogens, inhalants, PCP, sedatives, and other substances such as OTC medications, nitrous oxide, steroids, etc.); and more likely to meet criteria for 3 or more SUD diagnoses. They also reported higher rates with lifetime and any past-week withdrawal symptoms. Within the OPU group 6% reported heroin only, 67% used prescription opioids only, 16% used both types of opioids and 10% reported no opioid use in the past 90 days prior to intake/confinement (but met criteria for past-year opioid abuse or dependence). Sixteen percent of the OPU group reported using heroin at least weekly, while 64%, reported using prescription opioids at this frequency (by definition 0% used opioids in the MAPU group). Twenty four percent of the OPU group met criteria for past year opioid dependence diagnosis and 17% for past year opioid abuse diagnosis (data not shown in Table-2).
Table 2.
Substance Use and Treatment Characteristics of Opioid Problem Users (OPU, n=475) and Marijuana/Alcohol Problem Users (MAPU, n=475)
| MAPU | OPU+MAPU vs.MAPU Unweighted | OPU+MAPU vs. MAPU Weighted | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Unweighted (n=6471) | Weighted (n=475) | OPU+MAPU (n=475) | ||||||||
| N | % | N | % | N | % | Odds ratiof | 95%Confidence Interval | Odds ratiof | 95%Confidence Interval | |
| V. Substance Use Characteristics | ||||||||||
| Age of First use of any substance | ||||||||||
| % < 10 years | 511 | 8% | 44 | 9% | 84 | 18% | 2.50 | (2.2 – 2.7) | 2.08 | (1.7 – 2.5) |
| % 10–14 years | 4792 | 74% | 357 | 75% | 351 | 74% | 0.99 | (0.8 – 1.2) | 0.93 | (0.6 – 1.2) |
| % 15–17 years | 1155 | 18% | 71 | 15% | 39 | 8% | 0.42 | (0.1 – 0.7) | 0.51 | (0.1 – 0.9) |
| % 18 or > years | 14 | 0% | 2 | 0.4% | 1 | 0.2% | 0.97 | (0 – 3) | 0.50 | (0 – 2.9) |
| Years of Substance Usea | ||||||||||
| % < 1 year | 406 | 6% | 15 | 3% | 5 | 1% | 0.16 | (0 – 1) | 0.33 | (0 – 1.3) |
| % 1–2 years | 2675 | 41% | 149 | 31% | 96 | 20% | 0.36 | (0.1 – 0.6) | 0.55 | (0.3 – 0.9) |
| % 3–4 years | 2073 | 32% | 173 | 36% | 166 | 35% | 1.14 | (0.9 – 1.3) | 0.94 | (0.7 – 1.2) |
| % 5 or > years | 1317 | 20% | 138 | 29% | 208 | 44% | 3.04 | (2.9 – 3.2) | 1.90 | (1.6 – 2.2) |
| % Weekly Drug Use in the Past 90 daysb | ||||||||||
| Any substance Use\3 | 6393 | 99% | 472 | 99% | 473 | 100% | 2.88 | (1.5 – 4.3) | 1.47 | (0 – 3.3) |
| Tobacco Use | 3360 | 52% | 274 | 58% | 335 | 70% | 2.21 | (2 – 2.4) | 1.75 | (1.5 – 2) |
| Marijuana Use | 4282 | 66% | 389 | 82% | 396 | 83% | 2.56 | (2.3 – 2.8) | 1.10 | (0.8 – 1.4) |
| Alcohol Use | 1451 | 22% | 240 | 51% | 259 | 55% | 4.15 | (4 – 4.3) | 1.17 | (0.9 – 1.4) |
| Heroin Use | 0 | 0 | 78 | 16% | NA | NA | NA | NA | ||
| Prescription Opioid Use | 0 | 0 | 302 | 64% | NA | NA | NA | NA | ||
| Crack/Cocaine Use | 185 | 3% | 29 | 6% | 149 | 31% | 15.58 | (15.3 – 15.8) | 7.00 | (6.6 – 7.4) |
| Other Drug Use | 346 | 5% | 48 | 10% | 260 | 55% | 21.33 | (21.1 – 21.5) | 10.74 | (10.4 – 11.1) |
| % Past Year Substance Diagnosis | ||||||||||
| Any Substance Dependence Diagnosis | 3237 | 50% | 294 | 62% | 419 | 88% | 7.47 | (7.2 – 7.8) | 4.62 | (4.3 – 5) |
| Any Substance Abuse Diagnosis | 2679 | 41% | 132 | 28% | 39 | 8% | 0.13 | (0 – 0.5) | 0.23 | (0 – 0.6) |
| Any Substance Use Disorder Diagnosisc | 5916 | 91% | 426 | 90% | 458 | 96% | 2.48 | (2 – 3) | 3.06 | (2.5 – 3.6) |
| Opioid use disorder | 0 | 0% | 0 | 0% | 192 | 40% | NA | NA | NA | NA |
| * Cannabis use disorder | 3117 | 48% | 202 | 43% | 203 | 43% | 0.81 | (0.6 – 1) | 1.01 | (0.8 – 1.3) |
| * Alcohol use disorder | 1746 | 27% | 139 | 29% | 163 | 34% | 1.41 | (1.2 – 1.6) | 1.26 | (1 – 1.5) |
| * Cocaine use disorder | 128 | 2% | 16 | 3% | 86 | 18% | 11.01 | (10.7 – 11.3) | 6.52 | (6 – 7.1) |
| * Amphetamine use disorder | 192 | 3% | 21 | 4% | 57 | 12% | 4.45 | (4.1 – 4.8) | 2.95 | (2.4 – 3.5) |
| * Other substance use disorderd | 318 | 5% | 33 | 7% | 203 | 43% | 14.48 | (14.3 – 14.7) | 9.98 | (9.6 – 10.4) |
| 3 or more SUDs | 267 | 4% | 31 | 7% | 140 | 29% | 9.71 | (9.5 – 9.9) | 5.93 | (5.5 – 6.4) |
| Withdrawal Severity | ||||||||||
| % reporting any lifetime withdrawal | 2673 | 41% | 234 | 49% | 385 | 81% | 6.07 | (5.8 – 6.3) | 4.38 | (4.1 – 4.7) |
| % Any withdrawal symptoms past week | 1864 | 29% | 144 | 30% | 185 | 39% | 1.58 | (1.4 – 1.8) | 1.47 | (1.2 – 1.7) |
| SUD Treatment Characteristics | ||||||||||
| Current Level Of Treatment | ||||||||||
| % in Any Outpatient | 5177 | 80% | 281 | 59% | 267 | 56% | 0.32 | (0.1 – 0.5) | 0.89 | (0.6 – 1.1) |
| % in Any Residential setting | 1134 | 18% | 151 | 32% | 150 | 32% | 2.17 | (2 – 2.4) | 0.99 | (0.7 – 1.3) |
| % in Any Post-Residential Continuing Care | 193 | 3% | 43 | 9% | 58 | 12% | 4.52 | (4.2 – 4.8) | 1.40 | (1 – 1.8) |
| % reporting any days on Medications for Substance problems at intake | 0 | 0 | 4 | 1% | NA | NA | NA | NA | ||
| Need Perception and Readiness to Change | ||||||||||
| Perceives alcohol/substance as a probleme | 1639 | 25% | 176 | 37% | 315 | 66% | 5.82 | (5.6 – 6) | 3.35 | (3.1 – 3.6) |
| Perceives need for ANY treatment | 4777 | 74% | 382 | 80% | 426 | 90% | 3.05 | (2.7 – 3.4) | 2.08 | (1.7 – 2.5) |
| % Prior SA Treatment Episodes | ||||||||||
| Any | 1906 | 29% | 201 | 42% | 279 | 59% | 3.42 | (3.2 – 3.6) | 1.95 | (1.7 – 2.2) |
| One | 1168 | 18.1% | 103 | 22% | 108 | 23% | 1.33 | (1.1 – 1.6) | 1.07 | (0.8 – 1.4) |
| Two or more | 737 | 11.4% | 98 | 21% | 171 | 36% | 4.39 | (4.2 – 4.6) | 2.18 | (1.9 – 2.5) |
Notes: OPU defined as meeting criteria for past year opioid dependence/abuse or weekly or greater opioid use in the past 90-days and have mar/alc problem use; MAPU defined as meeting criteria for past year marijuana/alcohol dependence/abuse or weekly or greater marijuana/alcohol use in the past 90-days and no opioid problem use; Bold odds ratio are significantly different than 1 at p<.05.
Included in Clinical Comorbidities count
Current age minus age of first use.
Controlled for Confinement During the past 90 days.
Excluding tobacco.
includes hallucinogens, inhalants, PCP, sedatives and other miscellaneous substances such as OTC medications, nitrous oxide, steroids, etc.
Do you currently feel that you have any problems related to alcohol or drug use?
Substance Treatment Characteristics
Even after propensity score matching, the OPU+MAPU group was more likely than the weighted MAPU group to be treated in a post-residential continuing care setting. The OPU+MAPU group was also more likely to perceive both having an alcohol/substance problem and the need for any treatment; to have utilized any SUD treatment in the past; and to report multiple (i.e. 2 or more) prior treatment episodes.
Other Comorbid Conditions: Mental Health, HIV-Risk and Physical Health Problems
Table-3 shows that the OPU+MAPU group was more likely to endorse multiple problems on mental health, HIV-risk and physical health domains. The OPU+MAPU group was more likely than the weighted MAPU group to meet symptom criteria for any past-year any psychiatric disorder, internalizing disorder, or externalizing disorder, or both disorders; each of the individual internalizing and externalizing disorders; and high levels of victimization. Correspondingly, the OPU+MAPU group was also more likely to endorse higher rates of past mental health treatment utilization. Needle-use and sexual HIV-risk behaviors were more prevalent in the OPU+MAPU group, as were moderate to severe physical health problems. While the OPU+MAPU group was more likely than the unweighted MAPU group to have been pregnant, the weighted comparison lost significance.
Table 3.
Mental Health, HIV-Risk and Physical Health Characteristics of Opioid Problem Users (OP users, n=475) and Marijuana/Alcohol Problem Users (MAP users, n=475)
| MAPU | OPU+MAPU vs.MAPU Unweighted | OPU+MAPU vs.MAPU Weighted | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Unweighted (n=6471) | Weighted (n=475) | OPU+MAPU (n=475) | 95%Confidence Interval | |||||||
| N | % | N | % | N | % | Odds ratiof | 95%Confidence Interval | Odds ratiof | ||
| VI. Mental Health Characteristics: | ||||||||||
| A. % Past Year Symptoms of Internalizing Disorders | ||||||||||
| * Any Internalizing Disordera | 2535 | 39% | 238 | 50% | 363 | 76% | 5.01 | (4.8 – 5.2) | 3.21 | (4.7 – 5.3) |
| Major Depression Disorder | 2081 | 32% | 200 | 42% | 326 | 69% | 4.63 | (4.4 – 4.8) | 3.03 | (2.8 – 3.3) |
| Generalized Anxiety Disorder | 711 | 11% | 78 | 16% | 171 | 36% | 4.55 | (4.3 – 4.8) | 2.86 | (2.6 – 3.2) |
| Traumatic Stress Disorderb | 1351 | 21% | 141 | 30% | 245 | 52% | 4.04 | (3.8 – 4.2) | 2.51 | (2.2 – 2.8) |
| Suicidal Thoughts or Actions | 1318 | 20% | 124 | 26% | 211 | 44% | 3.12 | (2.9 – 3.3) | 2.25 | (2 – 2.5) |
| Any self injurious behaviors | 481 | 7% | 45 | 10% | 96 | 20% | 3.13 | (2.8 – 3.4) | 2.39 | (1.9 – 2.9) |
| B. % Past Year Symptoms of Externalizing Disorders/Problems | ||||||||||
| * Any Externalizing Disorder | 3805 | 59% | 311 | 66% | 409 | 86% | 4.34 | (4.1 – 4.6) | 3.25 | (2.9 – 3.6) |
| Conduct Disorder | 3178 | 49% | 271 | 57% | 378 | 79% | 4.02 | (3.8 – 4.2) | 2.92 | (2.6 – 3.2) |
| Attention Deficit-Hyperactivity Disorder | 2636 | 41% | 226 | 48% | 345 | 73% | 3.87 | (3.7 – 4.1) | 2.94 | (2.7 – 3.2) |
| C. % Past Year Internalizing/Externalizing Problems | ||||||||||
| Either | 4271 | 66% | 351 | 74% | 440 | 93% | 6.45 | (6.1 – 6.8) | 4.41 | (4 – 4.8) |
| Internal only | 451 | 7% | 38 | 8% | 29 | 6% | 0.87 | (0.5 – 1.3) | 0.75 | (0.2 – 1.2) |
| External only | 1734 | 27% | 113 | 24% | 77 | 16% | 0.53 | (0.3 – 0.8) | 0.63 | (0.3 – 0.9) |
| Both | 2084 | 32% | 200 | 42% | 333 | 70% | 4.96 | (4.8 – 5.2) | 3.23 | (3 – 3.5) |
| D. % Physical, Sexual, or Emotional Victimization | ||||||||||
| Any Lifetime History of Victimization | 3991 | 62% | 331 | 70% | 390 | 82% | 2.85 | (2.6 – 3.1) | 2.00 | (1.7 – 2.3) |
| Lifetime High Levels of Victimizationc | 2796 | 43% | 246 | 52% | 327 | 69% | 2.92 | (2.7 – 3.1) | 2.07 | (1.8 – 2.3) |
| * Past-Year Victimization | 2325 | 36% | 191 | 40% | 266 | 56% | 2.28 | (2.1 – 2.5) | 1.91 | (1.6 – 2.2) |
| Current worry about victimization | 1091 | 17% | 93 | 20% | 139 | 29% | 2.05 | (1.8 – 2.3) | 1.71 | (1.4 – 2) |
| E. % Any Prior Mental Health Treatment | 2426 | 37% | 203 | 43% | 312 | 66% | 3.18 | (3 – 3.4) | 2.56 | (2.3 – 2.8) |
| VII. % HIV-Risk Characteristics | ||||||||||
| Any Past Year Needle Use | 73 | 1% | 7 | 1% | 103 | 22% | 24.28 | (24 – 24.6) | 18.71 | (17.9 – 19.5) |
| * Past Year Moderate/High Needle Risk | 31 | 0% | 3 | 1% | 101 | 21% | 55.99 | (55.6 – 56.4) | 38.80 | (37.6 – 40) |
| Any Past-Year Sexual activity | 4966 | 77% | 394 | 83% | 424 | 89% | 2.52 | (2.2 – 2.8) | 1.71 | (1.3 – 2.1) |
| * Past Year Moderate/High Sexual Risk | 4018 | 62% | 346 | 73% | 402 | 85% | 3.37 | (3.1 – 3.6) | 2.07 | (1.7 – 2.4) |
| VIII. % Physical Health Problems | ||||||||||
| Any Physical Distress/health Problemd | 1824 | 28% | 144 | 30% | 203 | 43% | 1.90 | (1.7 – 2.1) | 1.72 | (1.4 – 2) |
| * Past Year Mod/High health problem | 1711 | 26% | 133 | 28% | 177 | 37% | 1.66 | (1.5 – 1.9) | 1.53 | (1.3 – 1.8) |
| Past Year Been Pregnant/Responsible for Pregnancy | 448 | 7% | 41 | 9% | 53 | 11% | 1.70 | (1.4 – 2) | 1.36 | (0.9 – 1.8) |
Notes: OPU defined as meeting criteria for past year opioid dependence/abuse or weekly or greater opioid use in the past 90-days and have mar/alc problem use; MAPU defined as meeting criteria for past year marijuana/alcohol dependence/abuse or weekly or greater marijuana/alcohol use in the past 90-days and no opioid problem use; Bold odds ratio are significantly different than 1 at p<.05.
Included in Clinical Comorbidities count
General anxiety disorder, major depression disorder, suicide thoughts or actions, traumatic stress disorder.
Post traumatic distress, acute traumatic distress or disorders of extreme stress not otherwise specified.
Reporting 4 or more of the following: types of victimization, traumagenic factors (e.g., multiple people, someone they trusted, fearing for life, sexual penetration, people didn’t believe them) or continuing fear it will reoccur.
Count of various types of physical problems endorsed by respondent.
Figure 1 illustrates by group the count of 11 past year problems endorsed in tables 1 to 3 including: cannabis use disorder, alcohol use disorder, cocaine use disorder, amphetamine use disorder, other substance use disorder, any internalizing disorder, any externalizing disorder, physical sexual or emotional victimization, needle use risk, moderate/high sexual risk and moderate/high health problem. The OPU+MAPU group endorsed significantly more comorbid conditions [5.1 compared to 3.4, F=9.3(948), p=.002]. As seen in Figure 1, 54% of the MAPU group reported 3 or fewer problems, 26% reported 5 or more problems and none reported 9 or more problems. By comparison, 24% of the OPU+MAPU group reported 3 or fewer problems, and more than 55% reported 5 or more problems. Thus, while comorbidity is ubiquitous and distributions overlap, the OPU group has more problems.
Figure 1.
Number of Major Clinical Problems* in the Past Year by Group
* Count of 11 past year problems endorsed in tables 1 to 3: cannabis use disorder, alcohol use disorder, cocaine use disorder, amphetamine use disorder, other substance use disorder, any internalizing disorder, any externalizing disorder, physical sexual or emotional victimization, needle use risk, moderate/high sexual risk, moderate/high health problem.
Discussion
This is the first study we know of that has evaluated the added risk of opioid problem use with marijuana/alcohol problem use among youth entering SUD treatment across multiple outpatient and residential sites in the U.S. The majority of the results confirmed findings from existing single-site studies of teen users of heroin [9, 10] or those with opioid dependence/abuse [3, 8, 11, 12]. Overall, this study found that having OPU was associated with higher rates of problems on virtually all of the psychosocial and clinical domains examined, with very few risk associations losing significance when comparisons were made to the weight adjusted MAPU group. The OPU+MAPU group was more likely than the weighted MAPU group to have multiple (on average five) clinically comorbid conditions. The OPU+MAPU group was less likely to be African American or other race and more likely than the weighted MAPU group to be older teens, i.e. 15–17 years old, Caucasian, report longer duration of substance use, have higher rates of SUD diagnoses (except for cannabis use disorder by design), have greater psychiatric symptom severity, report higher levels of victimization and HIV-risk behaviors (sexual risk and needle-use), and engage in higher rates of illegal behaviors.
This study showed that the added risk of OPU was associated with higher rates of each of the internalizing and/or externalizing psychiatric disorders examined, which is in contrast to findings of no such group differences in psychiatric disorders in a recent study [3] comparing OUD with non-OUD youth with cannabis/alcohol use disorders. This finding is most likely due to this study’s low proportion of residential patients (32%) where psychiatric comorbidity is an admission criterion. The novel findings of higher rates of lifetime and recent victimization may be conceived of either as a consequence of dependence on opioids exposing these youth to situations associated with greater risk for trauma/victimization (e.g. intoxication, selling drugs, trading drugs for sex or money); or as an attempt to self-medicate via psychological disassociation from psychological pain or seeking relief from physical pain (with opioids with established analgesic properties) or both. This assumption is supported by study results showing that having OPU was more likely to be associated with physical distress.
Additional new findings from the study include that 64% of the OPU+MAPU youth were using prescription opioids (in comparison to16% reporting heroin use) and that the added risk of OPU was associated with higher levels of homelessness and regular substance use by family members and peers. These findings suggest exposure to and/or modeling of deviant behaviors similar to findings reported previously by Nurco et al. [28]. The OPU+MAPU group was also more likely to be confined and spend long periods in confinement prior to the current episode of treatment, indicating higher severity of substance and legal problems. Homelessness may be conceived either as an antecedent or a consequence of their protracted substance use careers. However, a causal analysis of early-onset drug use associated problems and opioid use problems could not be examined in this cross-sectional study and warrants future research. The higher rates of confinement may be a consequence of higher rates of criminal activity (e.g. driving under the influence, trading sex for drugs/money, drug-dealing); or alternatively, suggest a more effective referral process for residential treatment initiated while they were in the juvenile justice system.
Another new area of exploration showed that the OPU youth were more likely to have used both mental health and SUD treatment services and to verbalize perceived need for current SUD treatment. It is likely that their prior treatments and current placement at higher intensity treatment settings may have influenced/primed their motivation to change. The implications of high treatment utilization on cost to society need to be explored in future studies.
Limitations
While corroborating information from collateral informants and the use of urine drug testing may have improved the validity of data, GAIN self-reports have been found to correlate highly with biometric measures of substance use [29]. The GAIN only provides diagnostic impressions and not “true” DSM-IV psychiatric diagnoses which may have resulted in either an over-inclusion of participants in each of the psychiatric disorders or under inclusion due do to denial/misreporting. Although we used propensity score to minimize the impact of significant demographic/social/treatment factors on the results, this method is still quasi-experimental and does not guarantee “true” group differences or account for the influence of high rates of polysubstance use. By not including 13% of the sample who reported infrequent/non-“problem use” of opioids in our comparisons, we were unable to comment on their characteristics, which may have important treatment implications. In this cross-sectional study, we were unable to examine causal relationships or outcomes for OP vs. MAP users. Although the data was pooled from several treatment sites from across the country, further study using a representative sample of the US population, may improve generalizeability.
Clinical Implications
The substantial added risk and comorbidities associated OPU may be a risk for or a consequence of opioid dependence and associated withdrawal symptoms; high crime; trauma and emotional distress; physical health problems; and needle-use and sexual risk behaviors. These findings highlight the importance of developing specialized and comprehensive treatments plans that integrate interventions that target the associated comorbidities. They also support the need for a) prevention studies to address environmental and mental health risk factors among these youth, b) further exploration of the complex relationships of the various risk factors with structural equation modeling and c) outcome studies to determine the influence of baseline comorbid risk factors such as environmental factors, polysubstance use, multiple psychiatric disorders, HIV/HCV infection-risk and criminal behaviors on opioid dependence treatment outcomes. Also worthy of investigation is whether more widespread use of effective medications for opioid dependent youth [8, 12] may improve other outcomes in addition to opioid use in this complex and multiply disabled subgroup of youth with problem use of opioids.
Acknowledgments
This development of this paper was supported by research grant K12DA000357 (Subramaniam, P.I.) from the National Institute on Drug Abuse (NIDA) and American Academy of Child and Adolescent Psychiatry, prior to Dr. Subramaniam’s employment at NIDA. This project was also supported by the Center for Substance Abuse Treatment (CSAT), Substance Abuse and Mental Health Services Administration (SAMHSA) contract 270-07-0191 using data provided by the following grants and contracts from CSAT (TI11424, TI11871, TI11874, TI11888, TI11894, TI13190, TI13305, TI13308, TI13309, TI13313, TI13322, TI13323, TI13340, TI13344, TI3345, TI13354, TI13356, TI13601 TI14090, TI14188, TI14189, TI14196, TI14214, TI14252, TI14254, TI14261, TI14267, TI14271, TI14272, TI14283, TI14311, TI14315, TI14355, TI14376, TI15348, TI15413, TI15415, TI15421, TI15433, TI15446, TI15447, TI15458, TI15461, TI15466, TI15467, TI15469, TI15475, TI15478, TI15479, TI15481, TI15483, TI15485, TI15486, TI15489, TI15511, TI15514, TI15524, TI15527, TI15545, TI15562, TI15577, TI15586, TI15670, TI15671, TI15672, TI15674, TI15677, TI15678, TI15682, TI15686, TI16386, TI16400, TI16414, TI16904, TI16928, TI16939, TI16992, TI17046, TI17055, TI17070, TI17071, TI17433, TI17434, TI17446, TI17476, TI17484, TI17490, TI17523, TI17604, TI17605; TI17638; TI17728; TI17761; TI17763; T17765, TI17769, TI17779, TI17786, TI17788, TI17812, TI17825, TI17830 TI18406; Contract 207-98-7047, Contract 277-00-6500, Contract 270-2003-00006). The authors would like to thank Melinda Tracy for preparation of the manuscript. The content and opinions are those of the authors and do not reflect official positions of the contributing project directors, the National Institute on Drug Abuse, the American Academy of Child and Adolescent Psychiatry, or Center for Substance Abuse Treatment/Substance Abuse and Mental Health Services Administration.
Footnotes
Conflict of interest statement: Dr. Subramaniam received salary support from Mountain Manor Treatment Center, Baltimore, MD. Dr. Stitzer receives funding from Pfizer Pharmaceuticals Inc. to conduct an investigator-generated research project that involves Chantix. Dr. Dennis is the director of the Global Appraisal of Individual Needs (GAIN) Coordinating Center (GCC) and Ms. Ives works at the GCC.
References
- 1.Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. Monitoring the Future national survey results on adolescent drug use: 1975–2006: Volume I, Secondary school students. Bethesda, MD: National Institute on Drug Abuse; 2007. (NIH Publication No. 07-6205) [Google Scholar]
- 2.Substance Abuse and Mental Health Services Administration Office of Applied Studies. Treatment Episode Dataset (TEDS), 1992–2005 [Computerfile] 2007 [cited 2007 April 28, 2008]; Available from: http://www.icpsr.umich.edu/cocoon/SAMHDA/SERIES/00056.xml?token=1.
- 3.Subramaniam GA, Stitzer MA, Woody GE, Fishman MJ, Kolodner K. Clinical Characteristics of Treatment Seeking Adolescents with Opioid Versus Cannabis/Alcohol Use Disorders. Drug AlcoholDepend. 2009;99(1):141–9. doi: 10.1016/j.drugalcdep.2008.07.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mills KL, Teesson M, Darke S, Ross J, Lynskey M. Young people with heroin dependence: findings from the Australian Treatment Outcome Study (ATOS) J Subst Ab Treatment. 2004;27(1):67–73. doi: 10.1016/j.jsat.2004.05.001. [DOI] [PubMed] [Google Scholar]
- 5.Substance Abuse and Mental Health Services Administration Office of Applied Studies. Drug Abuse Warning Network, 2005: National estimates of drug-related emergency department visits. Rockville, MD: Author; 2007. [Google Scholar]
- 6.Center for Disease Control and Prevention. Viral Hepatitis and injection drug users. Fact Sheet Series. 2002 [cited September 15, 2007]; Available from: http://www.cdc.gov/IDU/hepatitis/viral_hep_drug_use.htm.
- 7.Nurco DN, Hanlon TE, Kinlock TW. Recent Research on the Relationship between Illicit Drug Use and Crime. Behav Science and Law. 1991;9:221–42. [Google Scholar]
- 8.Woody G, Poole S, Subramaniam GA, Dugosh K, Bogenschutz M, Abbott P, et al. Extended Medication-Assisted Therapy Produces Better Outcomes Than Detoxification Among Opioid-Addicted Youth. JAMA. 2008;300(17):2003–11. doi: 10.1001/jama.2008.574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hopfer CJ, Mikulich SK, Crowley TJ. Heroin use among adolescents in treatment for substance use disorders. J Am Acad Child Adolesc Psychiatry. 2000 Oct;39(10):1316–23. doi: 10.1097/00004583-200010000-00021. [DOI] [PubMed] [Google Scholar]
- 10.Clemmey P, Payne L, Fishman M. Clinical characteristics and treatment outcomes of adolescent heroin users. J PsychoactDrugs. 2004;36(1):85–94. doi: 10.1080/02791072.2004.10399726. [DOI] [PubMed] [Google Scholar]
- 11.Gordon SM, Mulvaney F, Rowan A. Characteristics of adolescents in residential treatment for heroin dependence. Am J Drug Alcohol Abuse. 2004 Aug;30(3):593–603. doi: 10.1081/ada-200032300. [DOI] [PubMed] [Google Scholar]
- 12.Marsch LA, Bickel WK, Badger GJ, Stothart ME, Quesnel KJ, Stanger C, et al. Comparison of pharmacological treatments for opioid-dependent adolescents: a randomized controlled trial. Arch Gen Psychiatry. 2005 Oct;62(10):1157–64. doi: 10.1001/archpsyc.62.10.1157. [DOI] [PubMed] [Google Scholar]
- 13.Dennis ML, White M, Ives ML. Individual Characteristics and Needs Associated with Substance Misuse of Adolescents and Young Adults in Addiction Treatment. In: Leukefeld C, Gullotta T, Tindall MS, editors. Handbook on Adolescent Substance Abuse Prevention and Treatment: Evidence-Based Practice. In press. [Google Scholar]
- 14.Janiri L, Martinotti G, Dario T, Reina D, Paparello F, Pozzi G, et al. Anhedonia and substance-related symptoms in detoxified substance-dependent subjects: a correlation study. Neuropsychobiology. 2005;52(1):37–44. doi: 10.1159/000086176. [DOI] [PubMed] [Google Scholar]
- 15.Handelsman L, Aronson MJ, Ness R, Cochrane KJ, Kanof PD. The dysphoria of heroin addiction. Am J Drug Alcohol Abuse. 1992;18(3):275–87. doi: 10.3109/00952999209026067. [DOI] [PubMed] [Google Scholar]
- 16.Stewart-Sabin C, Chaffin M. Culturally competent substance abuse treatment for American Indian and Alaska Native youths. In: Stevens SJ, Morral AR, editors. Adolesent Substance Abuse Treatment in the United States: Exemplary Models from a National Evaluation Study. Binghampton, NY: Haworth Press; 2003. pp. 155–82. [Google Scholar]
- 17.Dennis ML, Ives ML, White MK, Muck RD. The strengthening communities for youth (SCY) initiative: A cluster analysis of the services received, their correlates and how they are associated with outcomes. J PsychoactDrugs. 2008;40:3–16. doi: 10.1080/02791072.2008.10399757. [DOI] [PubMed] [Google Scholar]
- 18.Dennis ML, Titus JC, White M, Unsicker J, Hodgkins D. Bloomington, IL: Chestnut Health Systems; 2003. Global Appraisal of Individual Needs (GAIN): Administration guide for the GAIN and related measures. (Version 5) Retrieved from http://www.chestnut.org/LI/gain/index.html on May 7, 2008. [Google Scholar]
- 19.Colder CR, Chassin L. The psychosocial characteristics ofalcohol users versus problem users: data from a study of adolescents at risk. Dev Psychopathol. 1999 Spring;11(2):321–48. doi: 10.1017/s0954579499002084. [DOI] [PubMed] [Google Scholar]
- 20.Dehejia RH, Wahba S. Propensity Score-Matching Methods for Nonexperimental Causal Studies. Review Econ Stats. 2002;84(1):151–61. [Google Scholar]
- 21.Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70(1):41–55. [Google Scholar]
- 22.Dennis ML, Chan YF, Funk R. Development and validation of the GAIN Short Screener (GSS) for internalizing, externalizing, and substance use disorders and crime/violence problems among adolescents and adults. Am J Addictions. 2008;15:80–91. doi: 10.1080/10550490601006055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chan Y-F, Dennis ML, Funk R. Prevalence and Comorbidity of Major Internalizing and Externalizing Problems amongAdolescents and Adults Presenting to Substance Abuse Treatment. J Subst Ab Treatment. 2008;34:14–24. doi: 10.1016/j.jsat.2006.12.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Titus JC, Dennis ML, White WL, Scott CK, Funk RR. Gender differences in victimization severity and outcomes among adolescents treated for substance abuse. Child Maltreat. 2003 Feb;8(1):19–35. doi: 10.1177/1077559502239612. [DOI] [PubMed] [Google Scholar]
- 25.Dennis ML, White M, Ives MI. Individual characteristics and needs associated with substance misuse of adolescents and young adults in addiction treatment. In: Leukefeld Carl, Gullotta Tom, StatonTindall Michele., editors. Handbook on Adolescent Substance Abuse Prevention and Treatment: Evidence-Based Practice. Chapter 3. New London, CT: Child Family Agency Press; 2009. pp. 45–72. [Google Scholar]
- 26.Kraemer HC, Morgan GA, Leech NL, Gliner JA, Vaske JJ, Harmon RJ. Measuresof Clinical Significance. J Am Acad Child Adolesc Psychiatry. 2003;42(12):1524–9. doi: 10.1097/00004583-200312000-00022. [DOI] [PubMed] [Google Scholar]
- 27.SPSS. Statistical Program for the Social Sciences. Chicago: SPSS; 2006. Version 14.0.2. [Google Scholar]
- 28.Nurco DN, Kinlock TW, O’Grady KE, Hanlon TE. Differential contributions of family and peer factors to the etiology of narcotic addiction. Drug Alcohol Depend. 1998 Aug 1;51(3):229–37. doi: 10.1016/s0376-8716(98)00041-6. [DOI] [PubMed] [Google Scholar]
- 29.Lennox RD, Dennis ML, Scott CK, Funk RR. Combining psychometric and biometric measures of substance use. Drug Alcohol Depend. 2006;83:95–103. doi: 10.1016/j.drugalcdep.2005.10.016. [DOI] [PMC free article] [PubMed] [Google Scholar]