Advertisement

Leftover Prescription Opioids and Nonmedical Use Among High School Seniors: A Multi-Cohort National Study

  • Sean Esteban McCabe
    Correspondence
    Address correspondence to: Sean Esteban McCabe, Ph.D., Institute for Research on Women and Gender, University of Michigan, 204 S. State Street, Ann Arbor, MI 48109-1290.
    Affiliations
    Institute for Research on Women and Gender, University of Michigan, Ann Arbor, Michigan

    Substance Abuse Research Center, University of Michigan, Ann Arbor, Michigan
    Search for articles by this author
  • Brady T. West
    Affiliations
    Center for Statistical Consultation and Research, University of Michigan, Ann Arbor, Michigan

    Survey Research Center, Institute for Social Research, University of Michigan, Ann Arbor, Michigan
    Search for articles by this author
  • Carol J. Boyd
    Affiliations
    Institute for Research on Women and Gender, University of Michigan, Ann Arbor, Michigan

    Substance Abuse Research Center, University of Michigan, Ann Arbor, Michigan
    Search for articles by this author

      Abstract

      Purpose

      To (1) estimate the proportion of nonmedical users of prescription opioids (i.e., used prescription opioids in the past year without a doctor's orders) who used leftover medications from their own previous prescriptions; (2) assess substance use behaviors as a function of diversion source; and (3) identify the sources for these prescribed opioids.

      Methods

      We analyzed data collected via self-administered questionnaires from nationally representative samples of high school seniors (modal age, 18 years) as a part of the Monitoring the Future (MTF) study. The sample consisted of four cohorts (senior years of 2007–2010, n = 8,888), including 647 high school seniors who reported past-year nonmedical use of prescription opioids, of whom 53% were estimated to be women.

      Results

      An estimated 36.9% of past-year nonmedical users of prescription opioids obtained these opioid medications from their own previous prescriptions. Logistic regression analyses indicated that nonmedical users who used leftover medications from their previous prescriptions were primarily motivated to relieve physical pain, whereas nonmedical users who obtained medications from other sources had significantly higher odds of prescription opioid abuse and other substance use behaviors. Based on a subanalysis of nonmedical users who obtained prescription opioids from their previous prescriptions in 2010 (n = 51), approximately 27.1% obtained them from a dentist, 45.0% obtained them from an emergency room physician, and 38.3% obtained them from another physician.

      Conclusions

      Leftover prescription opioids from previous prescriptions represent a major source of nonmedical use of prescription opioids among high school seniors. These findings indicate that enhanced vigilance is needed when prescribing and monitoring prescription opioids among adolescents, to reduce leftover medications and nonmedical use.

      Keywords

      Implications and Contribution
      We examined the sources of prescription opioids among high school seniors who reported nonmedical use in the past year. The findings indicate that leftover prescription opioids from a previous prescription represent a major source of nonmedical use and enhanced vigilance is needed when prescribing prescription opioids and monitoring their use among adolescents.
      Approximately 13.0% of high school seniors in the U.S. reported nonmedical use of prescription opioids (NMUPO) (i.e., used prescription opioids without a doctor's orders) in their lifetime, and 8.7% of high school seniors reported NMUPO in the past 12 months, according to the 2010 Monitoring the Future (MTF) study [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ]. These lifetime and annual prevalence rates of NMUPO have increased significantly over the past 2 decades among high school seniors [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ]. The estimated number of emergency department visits involving NMUPO more than doubled between 2004 and 2008 for patients in the U.S. younger than 21 years of age [
      • Substance Abuse and Mental Health Services Administration
      The DAWN Report: trends in emergency department visits involving nonmedical use of narcotic pain relievers.
      ].
      Prescription opioids are highly efficacious when used appropriately for pain management; however, the prescribing rates of prescription opioids have nearly doubled among adolescents and young adults in the U.S. since 1994 [
      • Fortuna R.J.
      • Robbins B.W.
      • Caiola E.
      • et al.
      Prescribing of controlled medications to adolescents and young adults in the United States.
      ]. Prescription opioids were often prescribed for common conditions such as back pain to adolescents 15–19 years of age (21.5%) and young adults 20–29 years of age (33.4%). One potential consequence of an increase in prescribing rates of prescription opioids is an increase in the amount of leftover medication that could lead to NMUPO (owing to greater availability). A recent study conducted in Utah found that 72% of adult respondents aged 18 years and older who were prescribed an opioid had leftover medications, and 71% of those individuals kept the leftover medications [
      Centers for Disease Control and Prevention
      Adult use of prescription opioid pain medications: Utah, 2008.
      ]. Dentists, primary care physicians, and emergency medicine physicians serve as the leading prescribers of opioid analgesics among adolescents and young adults in the U.S. [
      • Moore P.A.
      • Nahouraii H.S.
      • Zovko J.G.
      • Wisniewski S.R.
      Dental therapeutic practice patterns in the U.S., I: anesthesia and sedation.
      ,
      • Moore P.A.
      • Nahouraii H.S.
      • Zovko J.G.
      • Wisniewski S.R.
      Dental therapeutic practice patterns in the U.S., II: analgesics, corticosteroids, and antibiotics.
      ,
      • Volkow N.D.
      • McLellan T.A.
      • Cotto J.H.
      • et al.
      Characteristics of opioid prescriptions in 2009.
      ].
      Among adolescent and young adult nonmedical users, the leading diversion sources of prescription opioids are their peers and leftover medications from previous prescriptions [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ,
      • McCabe S.E.
      • Boyd C.J.
      Sources of prescription drug for illicit use.
      ]. For the purposes of this investigation, “diversion” is defined as the exchange (i.e., selling, trading, loaning, giving away) of controlled medications that leads to their use by individuals other than those for whom they were prescribed. At least two studies have found that more than half of adolescent and young adult nonmedical users of prescription opioids were given these medications by a friend for free, and between 33% and 40% obtained leftover prescription opioids from their own previous prescriptions [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ,
      • McCabe S.E.
      • Boyd C.J.
      Sources of prescription drug for illicit use.
      ]. A better understanding of the diversion of prescription opioids would likely help guide prevention and intervention efforts to reduce NMUPO and related adverse consequences among adolescents and young adults.
      Past work shows that diversion sources, routes of administration, and motives of NMUPO are significantly associated with substance use and abuse [
      • McCabe S.E.
      • Boyd C.J.
      Sources of prescription drug for illicit use.
      ,
      • McCabe S.E.
      • Cranford J.A.
      • Boyd C.J.
      • Teter C.J.
      Motives, diversion and routes of administration associated with prescription opioids.
      ]. For instance, approximately seven of every 10 individuals who report lifetime NMUPO and administered the opioids intranasally screened positive for past-year drug abuse, whereas only 8% of those who report lifetime NMUPO for pain relief or who obtained opioid medication from family sources screened positive for past-year drug abuse [
      • McCabe S.E.
      • Cranford J.A.
      • Boyd C.J.
      • Teter C.J.
      Motives, diversion and routes of administration associated with prescription opioids.
      ]. To date, there is a lack of studies that examine the sources of leftover prescription opioids from previous prescriptions and whether nonmedical users who use leftover medications differ from other nonmedical users who obtain prescription opioids from other sources, in terms of substance use behaviors. As a result, it is important to distinguish between diversion sources in terms of associated substance use behaviors. Such information can be useful for developing interventions and identifying those who may need appropriate pain management and/or have an increased risk of substance abuse.
      The present study, which draws on national cross-sectional data from the MTF study, includes four cohorts of nationally representative samples of high school seniors in the U.S. and examines three research questions: What proportion of past-year nonmedical users of prescription opioids obtained leftover medications from their own previous prescription? Do substance use behaviors differ between nonmedical users who use leftover medications and those who obtain from others sources? What are the original sources of leftover prescription opioids that lead to subsequent NMUPO?

      Methods

      Participants and procedures

      The MTF study annually surveys a cross-sectional, nationally representative sample of high school seniors in approximately 135 public and private schools in the coterminous U.S., using self-administered paper–pencil questionnaires in classrooms. The MTF study uses a multistage sampling procedure. In Stage 1, geographic areas or primary sampling units are selected; in Stage 2, schools within primary sampling units are selected (with probability proportionate to school size); and in Stage 3, students within schools are selected. The student response rates for high school seniors ranged from 79%–85% between 2007 and 2010. Because so many questions are included in the MTF study, much of the questionnaire content is divided into six different questionnaire forms, which are randomly distributed. This approach results in six virtually identical subsamples. The measures most relevant for this study were introduced on Form 1 in 2007, so this study focuses on the subsamples receiving Form 1 within each year cohort between 2007 and 2010. Details about the MTF design and methods are available elsewhere [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ]. The University of Michigan Health Sciences Institutional Review Board granted approval for this study.
      The sample included 8,888 individuals in the MTF cohorts from 2007 through 2010 who completed Form 1 during the spring of their senior year, including 647 individuals who reported past-year NMUPO. The sample represents a population that is estimated to be 53% female, 70% white, 13% African-American, and 17% Hispanic students. The modal age of the individuals in the sample was 18 years.

      Measures

      The MTF study assesses demographic characteristics and standard measures of substance use behaviors such as nonmedical use of prescription medications, and use of marijuana and other drugs [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ].
      NMUPO was assessed by asking respondents on how many occasions (if any) in their lifetime (and past 12 months) they used prescription opioids on their own—that is, without a doctor telling them to take the medicine (e.g., Vicodin, OxyContin, Percodan, Percocet, Demerol, Dilaudid, Ultram, morphine, opium, codeine). The response scale ranged from: 1 = no occasions to 7 = 40 or more occasions. For purposes of this investigation, we dichotomized responses into “no use” versus “any use.”
      Diversion sources for NMUPO were assessed by asking respondents where they obtained the prescription opioids they used without a doctor's order during the past 12 months (and to mark all answers that applied). The list of diversion sources included: (1) bought on the Internet; (2) took from friend or relative without asking; (3) given for free by friend or relative; (4) bought from friend or relative; (5) from a prescription I had; (6) bought from drug dealer or stranger; and (7) other method. In 2010, respondents who used prescription opioids without a doctor's orders during the past 12 months and used medication from a previous prescription in their name were also asked from whom they received this prescription. The list of sources included: (1) a dentist; (2) an emergency room doctor; and (3) another doctor.
      Past-year drunkenness was measured using the following item: “On how many occasions (if any) have you been drunk or very high from drinking alcoholic beverages during the last 12 months?” The response scale and coding were the same as for NMUPO.
      Past-year marijuana use was measured using the following item: “On how many occasions (if any) have you used marijuana during the last 12 months?” The response scale and coding were the same as for NMUPO.
      Past-year other illicit drug use—including lysergic acid diethylamide, psychedelics other than lysergic acid diethylamide, heroin, cocaine, and crack—was measured with the following item for each drug: “On how many occasions (if any) have you used [SPECIFIED DRUG CLASS] …during the last 12 months?” The response scale and coding for each substance was the same as for NMUPO.
      Past-year nonmedical use of other prescription medications—including stimulants, tranquilizers and sedatives—was measured with the following item for each medication: “On how many occasions (if any) have you used [SPECIFIED DRUG CLASS] …during the last 12 months?” The response scale for each substance was the same as for NMUPO.
      Routes of administration for NMUPO were assessed with five items that asked which methods respondents used for taking prescription opioids not prescribed to them (and to mark all that applied). The binary items included: (1) intranasal (snorting or sniffing); (2) smoking; (3) injection; (4) orally (by mouth); and (5) other.
      Motives for NMUPO were assessed by asking respondents who reported NMUPO in the past 12 months to indicate the most important reasons for using prescription opioids on their own without a doctor's orders (and to mark all that applied). The list of 17 binary items included but was not limited to the following motives: “to relieve physical pain,” “to experiment,” and “to feel good or get high.”

      Data analysis

      The MTF study provides survey weights for responding cases in each of its public-use data files, and we used these weights in all analyses to ensure that estimates of population features were unbiased. We computed the estimated prevalence rates of diversion sources associated with NMUPO—across subgroups defined by demographic characteristics and substance use behaviors—using weighted cross-tabulations. We conducted Rao–Scott [
      • Rao J.N.K.
      • Scott A.J.
      On chi-squared tests for multi-way tables with cell proportions estimated from survey data.
      ] chi-square tests of homogeneity and design-based logistic regression analyses [
      • Heeringa S.G.
      • West B.T.
      • Berglund P.A.
      Applied survey data analysis.
      ] to determine whether diversion sources were significantly associated with substance use behaviors. We compared the following four mutually exclusive groups in the analyses: (1) NMUPO from a previous prescription only; (2) NMUPO from a previous prescription plus other source(s); (3) NMUPO from other source(s); and (4) nonusers. Logistic regression models included sex, cohort year, and school geographical region as covariates based on their significant associations with NMUPO identified in previous research [
      • McCabe S.E.
      • Boyd C.J.
      • Teter C.J.
      Illicit use of opioid analgesics by high school seniors.
      ]. The complex multistage sampling design used in the MTF study resulted in the need to account for effects of stratified cluster sampling on variance estimates. We adjusted estimated (linearized) variances of weighted estimates using average design effects provided by MTF study staff [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ] per the method outlined by West and McCabe [

      West BT, McCabe SE. Incorporating complex sample design effects when only final survey weights are available. Stata, in press.

      ], before the construction of confidence intervals. We divided weighted Pearson chi-square statistics by an average design effect factor of 2.0, per the recommendation of Johnston and colleagues [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ]. We performed all statistical analyses using commands for the analysis of complex sample survey data in Stata 11.2 software (StataCorp, College Station, TX).

      Results

      Prevalence of diversion sources associated with NMUPO

      An estimated 36.9% of past-year nonmedical users of prescription opioids obtained these opioid medications from their own leftover medication. The estimated prevalence of other diversion sources included: bought on the Internet (1.4%), took from friend or relative without asking (22.2%), given for free from friend or relative (55.0%), bought from a friend or relative (37.9%), bought from a drug dealer (19.4%), and other method (9.5%). Among past-year nonmedical users of prescription opioids, approximately 14.4% indicated they used opioids from their previous prescription only, 22.5% reported they used opioids from their previous prescription along with other source(s), and 63.2% obtained opioids only from other source(s). As illustrated in Table 1, the prevalence of diversion sources of NMUPO varied significantly by sex. The proportion of past-year nonmedical users of prescription opioids who obtained these opioid medications from their leftover medications was higher among females, whereas obtaining prescription opioids from other sources was more prevalent among males.
      Table 1Diversion sources of nonmedical prescription opioids among high school seniors nationally, overall and by sex, 2007–2010
      Diversion sourceOverall nonmedical users (n [%])Male nonmedical users (n [%])Female nonmedical users (n [%])Sex differences (χ2 [df], p Value)
      Previous prescription only104 (14.4%)40 (11.5%)59 (17.2%)29.04 (1), <.001
      Previous prescription plus other sources147 (22.5%)68 (20.8%)74 (25.1%)11.54 (1), <.001
      Other sources396 (63.2%)202 (67.7%)170 (57.7%)47.07 (1), <.001
      Sample sizes for sex vary from overall sample sizes because of missing data on the sex variable in the MTF data set. Source: 2007–2010 Monitoring the Future study; weighted estimates with average design effect correction of 2.0 applied to chi-square statistics.
      We conducted a subanalysis of the 2010 data to determine the original source of leftover prescription opioids that was associated with subsequent NMUPO. Among past-year nonmedical users who obtained prescription opioids from their own previous prescription (n = 51), an estimated 27.1% received the opioids from a dentist, 45.0% received them from an emergency room doctor, and 38.3% received them from another doctor. A more detailed breakdown indicates that among nonmedical users who obtained prescription opioids from their previous prescription, the original source of these medications used nonmedically was 11.4% from a dentist only, 7.3% from an emergency room doctor only, and 23.1% from another doctor only; 24.9% reported multiple doctor or dentist sources.

      Sources of NMUPO and other substance use behaviors

      As illustrated in Table 2, we used bivariate analyses to examine the associations among diversion sources of past-year NMUPO and substance use behaviors. Substance use behaviors included past-year drunkenness, marijuana use, illicit drug use other than marijuana use, and nonmedical use of other prescription medications. Rao–Scott [
      • Rao J.N.K.
      • Scott A.J.
      On chi-squared tests for multi-way tables with cell proportions estimated from survey data.
      ] chi-square analyses revealed significant associations between diversion sources of past-year NMUPO and each substance use measure (p < .001). Multiple logistic regression results reinforced the bivariate findings. The odds of substance use behaviors were considerably higher among past-year nonmedical users who obtained prescription opioids from sources besides their own previous prescriptions, compared with nonmedical users who only obtained leftover prescription opioids from previous prescriptions after adjusting for sex, cohort year, and school geographical region (p < .01). In contrast, the odds of substance use behaviors were generally lower among past-year nonusers relative to nonmedical users who only used leftover prescription opioids from previous prescriptions.
      Table 2Substance use behaviors as a function of diversion source of nonmedical prescription opioids, 2007–2010
      Diversion sourcePast-year drunkennessPast-year marijuana usePast-year illicit drug use other than marijuanaPast-year nonmedical use of other prescription medications
      %Adjusted OR (95% CI)%Adjusted OR (95% CI)%Adjusted OR (95% CI)%Adjusted OR (95% CI)
      Previous prescription only68.17.710.723.8
      Previous prescription plus other sources92.14.9 (1.6–15.0)a35.86.1 (1.8–20.7)a44.16.1 (2.1–18.0)b75.110.5 (4.2–26.6)b
      Other sources88.83.2 (1.4–7.1)a33.55.6 (1.8–17.4)a52.48.3 (3.0–22.6)b66.66.4 (2.9–14.0)b
      Nonuser40.3.3 (.1–.6)b4.0.4 (.1–1.3)3.9.3 (.1–.8)c4.4.1 (<.1–.3)b
      χ2 (df), p value237.8 (3), <.001423.2 (3), <.001855.7 (3), <.0011,280.4 (3), <.001
      The reference group is “Previous prescription only.” Sample sizes for the regression models ranged from 7,857 (Past-Year Drunkenness) to 8,442 (Past-Year Marijuana Use) because of missing data. The results represent estimated odds ratios in multivariate logistic regression models, adjusted for sex, cohort year, and school geographical region. Any past-year illicit drug use other than marijuana included lysergic acid diethylamide, psychedelics other than lysergic acid diethylamide, heroin, crack, and cocaine. Any past-year nonmedical use of other prescription medications included stimulants, tranquilizers, and sedatives. Source: 2007–2010 Monitoring the Future study; weighted estimates with an average design effect correction of 2.0 applied to chi-square statistics.
      CI = confidence interval; OR = odds ratio.
      a p < .01, b p < .001, and c p < .05, based on logistic regression results.
      As illustrated in Table 3, we also examined the associations among diversion sources of past-year NMUPO and specific behaviors related to NMUPO, such as route of administration, subjective high, and motives, for nonmedical users of prescription opioids. Design-adjusted chi-square analyses revealed several significant associations (p < .001). Multiple logistic regression results supported the bivariate findings; the odds of intranasal administration of prescription opioids and getting moderately or very high when using prescription opioids were significantly greater among nonmedical users who obtained prescription opioids from sources besides their own previous prescriptions, compared with nonmedical users who only obtained leftover prescription opioids from previous prescriptions, after adjusting for sex, cohort year, and school geographical region (p < .001). For example, the odds of using prescription opioids intranasally were more than six times greater among nonmedical users who obtained prescription opioids from sources besides their own previous prescriptions, compared with nonmedical users who only obtained leftover prescription opioids from their own previous prescriptions. In contrast, the odds of using prescription opioids to relieve physical pain were significantly lower among nonmedical users who only obtained prescription opioids from sources besides their own previous prescriptions, compared with nonmedical users who only obtained leftover prescription opioids from their own previous prescriptions (adjusted odds ratio = .2; 95% confidence interval = .1–.5; p < .001). The prevalence of using prescription opioids nonmedically to relieve physical pain among past-year nonmedical users who only obtained prescription opioids from sources besides their own previous prescriptions was 39.6%, compared with more than 71.3% of nonmedical users who only obtained leftover prescription opioids from their own previous prescriptions.
      Table 3Nonmedical prescription opioid use behaviors as a function of diversion source, 2007–2010
      Diversion sourceSnorted when using prescription opioidsModerately or very high when using prescription opioidsUsed prescription opioids to relieve physical pain
      %Adjusted OR (95% CI)%Adjusted OR (95% CI)%Adjusted OR (95% CI)
      Previous prescription only9.426.171.3
      Previous prescription plus other sources35.46.7 (2.0–22.2)a51.83.3 (1.4–7.5)a72.81.0 (.4–2.4)
      Other sources44.99.0 (2.9–28.0)b60.04.3 (2.0–9.1)b39.6.2 (.1–.5)b
      χ2 (df), p value278.4 (2), <.001243.0 (2), <.001439.2 (2), <.001
      The reference group is “from a prescription I had only.” Sample sizes for the regression models ranged from 605 (Relieve Pain) to 609 (Moderate or Very High) because of missing data. Multivariate logistic regression models include sex, cohort year, and school geographical region as covariates. Source: 2007–2010 Monitoring the Future study; weighted estimates with average design effect correction of 2.0 applied to chi-square statistics.
      Abbreviations are as in Table 2.
      a p < .01, b p < .001 based on logistic regression results.

      Discussion

      The present study found that leftover prescription opioids from previous prescriptions account for a substantial source of NMUPO among high school seniors in the U.S. We found that more than a third of past-year nonmedical users of prescription opioids obtained these medications from their own previous prescription(s). The findings of the present study are consistent with previous work among adolescents, which has shown that using one's previous prescription serves as a major source of NMUPO [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ,
      • McCabe S.E.
      • Boyd C.J.
      Sources of prescription drug for illicit use.
      ]. It has been estimated that 5%–23% of all prescription opioid doses dispensed are used nonmedically, and the findings from the present study and previous work indicate that leftover medication, peers, and family members have a role in the diversion and NMUPO among adolescents [
      • Johnston L.D.
      • O'Malley P.M.
      • Bachman J.G.
      • Schulenberg J.E.
      Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
      ,
      • McCabe S.E.
      • Boyd C.J.
      Sources of prescription drug for illicit use.
      ,
      • Katz N.P.
      • Birnbaum H.G.
      • Castor A.
      Volume of prescription opioids used nonmedically in the United States.
      ].
      There were notable sex differences in the prevalence of nonmedical use with leftover prescription opioids among high school seniors in the U.S. The higher rates of nonmedical use with leftover prescription opioids among females could be partially explained by higher rates of prescribing and medical use of opioid analgesics among female high school students [
      • Boyd C.J.
      • McCabe S.E.
      • Teter C.J.
      Medical and nonmedical use of prescription pain medication by youth in a Detroit-area public school district.
      ,
      • McCabe S.E.
      • Boyd C.J.
      • Young A.
      Medical and nonmedical use of prescription drugs among secondary school students.
      ]. For example, one study of secondary school students found that 53.1% of females and 35.0% of males reported lifetime medical use of prescription opioids [
      • McCabe S.E.
      • Boyd C.J.
      • Young A.
      Medical and nonmedical use of prescription drugs among secondary school students.
      ]. These sex differences in nonmedical use with leftover prescription opioids deserve more attention in future research and may provide useful information for developing prevention efforts.
      We found that nonmedical users of prescription opioids who obtained these medications from diversion sources other than their previous prescriptions had significantly higher odds of intranasal administration of prescription opioids and other substance use behaviors. In contrast, nonmedical users of prescription opioids who obtained these medications from diversion sources other than their previous prescriptions had significantly lower odds of using prescription opioids nonmedically to relieve physical pain. These findings suggest that individuals who use leftover prescription opioid medication nonmedically are most likely attempting to self-treat physical pain and could benefit from appropriate pain management. Prescribers such as dentists may have a key role in minimizing leftover medication and help reduce nonmedical use and abuse of prescription opioids through careful assessment, appropriate prescribing with concomitant education, vigilant monitoring, and referral for substance abuse treatment when indicated [
      • Denisco R.C.
      • Kenna G.A.
      • O'Neil M.G.
      • et al.
      Prevention of prescription opioid abuse: the role of the dentist.
      ]. Indeed, an evidence-based protocol for assessment and education among prescribers is needed, one aimed at preventing nonmedical use, ensuring safe and secure storage, and dictating safe disposal [
      • Boyd C.J.
      • Anderson K.G.
      • Rieckman T.
      ABC's of controlled medications: what patients need to know about their prescription pain medicine.
      ]. Patient and prescriber interventions must take place across a range of health care settings, and materials should consider the possibility of nonmedical use and diversion.
      Among nonmedical users who obtained prescription opioids from their previous prescriptions based on a subsample of the 2010 data, the leading sources included emergency room physicians, dentists, and other physicians. Based on prescription data from a national database in the U.S., dentists were the leading prescriber of prescription opioids among patients 10–19 years of age [
      • Volkow N.D.
      • McLellan T.A.
      • Cotto J.H.
      • et al.
      Characteristics of opioid prescriptions in 2009.
      ]. Dentists' prescribing practices after performing third-molar extractions were examined based on a survey of 563 practicing oral and maxillofacial surgeons from a sample from the American Dental Association Survey Center [
      • Moore P.A.
      • Nahouraii H.S.
      • Zovko J.G.
      • Wisniewski S.R.
      Dental therapeutic practice patterns in the U.S., I: anesthesia and sedation.
      ,
      • Moore P.A.
      • Nahouraii H.S.
      • Zovko J.G.
      • Wisniewski S.R.
      Dental therapeutic practice patterns in the U.S., II: analgesics, corticosteroids, and antibiotics.
      ]. Approximately 85.1% of U.S. surgeons usually prescribed opioid analgesics after third-molar extractions; the medication of choice was a hydrocodone and acetaminophen formulation [
      • Moore P.A.
      • Nahouraii H.S.
      • Zovko J.G.
      • Wisniewski S.R.
      Dental therapeutic practice patterns in the U.S., II: analgesics, corticosteroids, and antibiotics.
      ]. The median number of pills prescribed by dentists for all opioid formulations was 20 (range, 8–40) and labeling instructions for opioid analgesics were “as needed” or “as needed for pain” in 96% of the cases [
      • Moore P.A.
      • Nahouraii H.S.
      • Zovko J.G.
      • Wisniewski S.R.
      Dental therapeutic practice patterns in the U.S., II: analgesics, corticosteroids, and antibiotics.
      ]. There are potential problems associated with the “PRN” or “as needed” designation that licensed clinicians use when prescribing opioid analgesics; this designation may allow for greater liberties than the prescriber intends. Based on the findings of the present study, there is a need for more research to examine the sources of leftover medications and how prescribing practices may contribute to leftover controlled medications.
      This study has noteworthy limitations that should be considered when weighing its practical implications. First, because the present study represented secondary analyses, the survey items in the MTF limited the variables that could be examined. Survey items did not assess the diagnosis, dosage or type of medication prescribed, specify on how many occasions respondents had used each diversion source, or indicate the reasons why respondents had leftover opioid medication. Second, the results cannot be generalized to all adolescents because this sample only included high school seniors (modal age, 18 years) and did not include individuals who had dropped out of school or were not present in school on the day of survey administration. Third, the data are subject to potential bias introduced when assessing sensitive behaviors via self-report surveys administered in a school setting. The present study attempted to minimize these biases by informing potential respondents that participation was voluntary and assuring potential respondents that data would remain confidential [
      • Harrison L.
      • Hughes A.
      The validity of self-reported drug use: improving the accuracy of survey estimates. NIH Publication 97–4147, NIDA Research Monograph No. 167.
      ,
      • Johnston L.D.
      • O'Malley P.M.
      Issues of validity and population coverage in student surveys of drug use.
      ]. Fourth, the cross-sectional nature of the study presented some limitations in terms of making causal inferences. Finally, the sample size for the subanalysis of the sources of previous prescriptions in 2010 was small (n = 51), and future research in this area is needed with larger samples.
      We found that more than a third of high school seniors in the U.S. who reported past-year NMUPO had obtained leftover prescription opioids for nonmedical use from their previous prescriptions. Nonmedical users of prescription opioids who used leftover medications from their previous prescription were primarily motivated to relieve physical pain, whereas nonmedical users who obtained from other sources had significantly higher odds of prescription opioid abuse and other substance use behaviors. The leading sources of these leftover opioid medications included emergency room doctors, dentists, and other doctors. Taken together, these findings indicate that enhanced vigilance is needed when prescribing prescription opioids and monitoring their use among adolescents to reduce leftover medications and nonmedical use.

      Acknowledgments

      The development of this article was supported by Research Grants R01DA024678 and R01DA031160 from the National Institute on Drug Abuse , National Institutes of Health . The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Drug Abuse or the National Institutes of Health. The Monitoring the Future data were collected by Research Grant R01DA01411 from the National Institute on Drug Abuse, National Institutes of Health. The authors thank the Substance Abuse and Mental Health Data Archive for providing access to these data. Drs. Sean Esteban McCabe and Brady T. West had full access to all the data reported in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

      References

        • Johnston L.D.
        • O'Malley P.M.
        • Bachman J.G.
        • Schulenberg J.E.
        Monitoring the future national survey results on drug use, 1975–2010. Volume I: secondary school students.
        University of Michigan Institute for Social Research, Ann Arbor, MI2011
        • Substance Abuse and Mental Health Services Administration
        The DAWN Report: trends in emergency department visits involving nonmedical use of narcotic pain relievers.
        Office of Applied Studies, Rockville, MD2010 (Available at:) (Accessed 9 October 2011)
        • Fortuna R.J.
        • Robbins B.W.
        • Caiola E.
        • et al.
        Prescribing of controlled medications to adolescents and young adults in the United States.
        Pediatrics. 2010; 126: 1108-1116
        • Centers for Disease Control and Prevention
        Adult use of prescription opioid pain medications: Utah, 2008.
        MMWR Morb Mortal Wkly Rep. 2010; 59: 153-157
        • Moore P.A.
        • Nahouraii H.S.
        • Zovko J.G.
        • Wisniewski S.R.
        Dental therapeutic practice patterns in the U.S., I: anesthesia and sedation.
        Gen Dent. 2006; 54: 92-98
        • Moore P.A.
        • Nahouraii H.S.
        • Zovko J.G.
        • Wisniewski S.R.
        Dental therapeutic practice patterns in the U.S., II: analgesics, corticosteroids, and antibiotics.
        Gen Dent. 2006; 54: 201-207
        • Volkow N.D.
        • McLellan T.A.
        • Cotto J.H.
        • et al.
        Characteristics of opioid prescriptions in 2009.
        JAMA. 2011; 305: 1299-1301
        • McCabe S.E.
        • Boyd C.J.
        Sources of prescription drug for illicit use.
        Addict Behav. 2005; 30: 1342-1350
        • McCabe S.E.
        • Cranford J.A.
        • Boyd C.J.
        • Teter C.J.
        Motives, diversion and routes of administration associated with prescription opioids.
        Addict Behav. 2007; 32: 562-565
        • Rao J.N.K.
        • Scott A.J.
        On chi-squared tests for multi-way tables with cell proportions estimated from survey data.
        Ann Stat. 1984; 12: 46-60
        • Heeringa S.G.
        • West B.T.
        • Berglund P.A.
        Applied survey data analysis.
        Chapman and Hall, London2010
        • McCabe S.E.
        • Boyd C.J.
        • Teter C.J.
        Illicit use of opioid analgesics by high school seniors.
        J Subst Abuse Treat. 2005; 28: 224-229
      1. West BT, McCabe SE. Incorporating complex sample design effects when only final survey weights are available. Stata, in press.

        • Katz N.P.
        • Birnbaum H.G.
        • Castor A.
        Volume of prescription opioids used nonmedically in the United States.
        J Pain Palliat Care Pharmacother. 2010; 24: 141-144
        • Boyd C.J.
        • McCabe S.E.
        • Teter C.J.
        Medical and nonmedical use of prescription pain medication by youth in a Detroit-area public school district.
        Drug Alcohol Depend. 2006; 81: 37-45
        • McCabe S.E.
        • Boyd C.J.
        • Young A.
        Medical and nonmedical use of prescription drugs among secondary school students.
        J Adolesc Health. 2007; 40: 76-83
        • Denisco R.C.
        • Kenna G.A.
        • O'Neil M.G.
        • et al.
        Prevention of prescription opioid abuse: the role of the dentist.
        J Am Dent Assoc. 2011; 142: 800-810
        • Boyd C.J.
        • Anderson K.G.
        • Rieckman T.
        ABC's of controlled medications: what patients need to know about their prescription pain medicine.
        J Addict Nurs. 2011; 22: 1-3
        • Harrison L.
        • Hughes A.
        The validity of self-reported drug use: improving the accuracy of survey estimates. NIH Publication 97–4147, NIDA Research Monograph No. 167.
        Government Printing Office, Washington, DC1997
        • Johnston L.D.
        • O'Malley P.M.
        Issues of validity and population coverage in student surveys of drug use.
        NIDA Res Monogr. 1985; 57: 31-54