Misuse of Prescription Drugs

8. Discussion: Substantive Findings and Methodological Issues and Challenges

This chapter summarizes and discusses the substantive findings from the National Survey on Drug Use and Health (NSDUH) on the misuse of prescription psychotherapeutic drugs, as presented in Chapters 2 through 7. It provides additional information on methamphetamine use by comparing estimates from NSDUH with treatment admissions data from the Treatment Episode Data Set (TEDS) (Office of Applied Studies [OAS], 2006a). The chapter concludes with a consideration of some of the methodological issues and challenges involved in collecting NSDUH data on the misuse of prescription psychotherapeutic drugs.

8.1. Summary and Discussion of Findings

Perhaps the most important finding of this study is that the nonmedical use of prescription psychotherapeutic drugs—and of pain relievers in particular—is now second only to marijuana use among the Nation's most prevalent drugs. Based on combined data for 2002, 2003, and 2004, an annual average of 14.8 million persons aged 12 or older misused prescription psychotherapeutic drugs in the past year compared with 25.5 million who used marijuana in that period. An estimated 11.3 million misused prescription pain relievers, 5.8 million used cocaine, and 5.0 million misused prescription tranquilizers.

In addition, the annual average number of past year initiates of pain reliever misuse based on combined data for 2002, 2003, and 2004 exceeded the number for marijuana (2.4 million vs. 2.1 million). The average number of past year initiates of tranquilizers (1.1 million new users) ranked third, ahead of cocaine (1.0 million).

An estimated 2 million persons met the criteria for dependence on or abuse of prescription psychotherapeutics in the past year. Among those who met the criteria for prescription psychotherapeutic dependence or abuse, only an estimated 290,000 persons received treatment for illicit drug use in the past year.

The remainder of this section summarizes additional key findings for trends in the misuse of prescription psychotherapeutics (Section 8.1.1), demographic and behavioral correlates of past year misuse (Section 8.1.2), geographic and State-level differences in rates of past year misuse (Section 8.1.3), lifetime misuse of specific psychotherapeutics (Section 8.1.4), initiation of psychotherapeutic drug misuse in the past year (Section 8.1.5), misuse of psychotherapeutic drugs and other drug use (Section 8.1.6), and dependence, abuse, and treatment for psychotherapeutic drug misuse (Section 8.1.7).

8.1.1 Trends in Misuse of Prescription Psychotherapeutics

Although the prevalence of past year prescription psychotherapeutic drug misuse overall did not change significantly from 2002 to 2004, statistically significant changes were noted in specific demographic groups, for other periods of use, and for specific psychotherapeutic drugs. In particular, lifetime misuse of the pain reliever OxyContin^® increased overall and in each of the three age groups highlighted in NSDUH (12 to 17, 18 to 25, and 26 or older). Among young adults aged 18 to 25, misuse of any prescription psychotherapeutic in the lifetime and the past month increased mainly because of increases in the misuse of pain relievers and tranquilizers. Lifetime misuse of stimulants decreased among the population aged 12 or older and among adults aged 26 or older. Lifetime and past year misuse of any stimulant and methamphetamine also decreased among youths aged 12 to 17.

In addition, findings from Chapter 3 indicate that the prevalence of lifetime nonmedical use from 2002 to 2004 increased for pain relievers containing hydrocodone (including Vicodin^®, Lortab^®, or Lorcet^®; generic hydrocodone; and other products specified by respondents). Significant increases in lifetime prevalence also were seen for pain relievers containing oxycodone (including Percocet^®, Percodan^®, or Tylox^®; OxyContin^®; plus other products specified by respondents). These increases generally occurred in each age group. Similarly, the lifetime prevalence of misuse of the tranquilizers Xanax^®, alprazolam, Ativan^®, or lorazepam increased from 2002 to 2004 among persons aged 12 or older.

The number of new past year initiates of prescription psychotherapeutic drug misuse remained stable from 2002 to 2004 overall (Chapter 4). Long-term trends in the numbers of initiates by calendar year generally showed higher numbers of initiates in recent years for prescription psychotherapeutic drugs. Methamphetamine and sedatives showed exceptions to this pattern. For these drugs, there tended to be higher numbers of initiates in the 1970s and early 1980s than in the most recent years. These data must be viewed with caution, however, because problems with recall tend to lead to underestimation of initiates in earlier years (Gfroerer et al., 2004).

8.1.2 Demographic and Behavioral Correlates of Past Year Misuse of Prescription Psychotherapeutics

Demographic comparisons revealed that young adults aged 18 to 25 had the highest rates of misuse of each class of prescription psychotherapeutics except sedatives; for sedatives, youths aged 12 to 17 and young adults had similar rates (0.5 percent). However, both age groups had rates of sedative misuse that were significantly higher than those for adults aged 26 or older (0.3 percent). Detailed age breakdowns revealed that the highest rates of use were at age 19 for any prescription psychotherapeutic drug, pain relievers, and stimulants; at age 21 for methamphetamine; and at ages 19 to 21 for tranquilizers.

An important finding is that nonmedical use of some prescription psychotherapeutic drugs among youths aged 12 to 17 was higher among females than among males. In the overall population aged 12 or older, males generally had higher rates of misuse of pain relievers and stimulants than did females, but the rates among males and females were roughly equal for tranquilizers and sedatives. Similarly, methamphetamine use was more common among males than females. Among youths aged 12 to 17, however, the rates of nonmedical use were higher among females than males for any prescription psychotherapeutic drug, pain relievers, stimulants, and tranquilizers.

In general, whites had higher rates of misuse of prescription psychotherapeutic drugs than did blacks or Asians. The ranking of Hispanics in the rate of prescription psychotherapeutic drug misuse varied by drug type. In particular, the rate of past year methamphetamine use was more prevalent among whites than among blacks or Hispanics.

Employment and education appeared to be protective against the nonmedical use of prescription psychotherapeutic drugs as has been found for use of illicit drugs in general (OAS, 2005b). Among persons aged 18 or older, rates of nonmedical use of prescription psychotherapeutic drugs were generally lowest among persons who completed college and highest among persons with less than 12 years of education or those who had some college education but did not graduate. Unemployed adults had higher rates of nonmedical use than those who were employed full time or part time. Persons in other employment categories, including those not in the workforce, had particularly low rates of misuse. Among young adults aged 18 to 22, those not enrolled as a college or university undergraduate had higher rates of nonmedical use of any prescription psychotherapeutic drug, pain relievers, tranquilizers, and methamphetamine than those enrolled full time.

Among females aged 15 to 44, rates of misuse of prescription psychotherapeutic drugs were lower among those who were currently pregnant than among those who were not. Statistically significant differences in this direction were observed for any psychotherapeutic drug, pain relievers, tranquilizers, and stimulants. This was generally true for women aged 18 to 25 and those aged 26 to 44. Among females aged 15 to 17, however, rates of misuse of prescription psychotherapeutic drugs were higher for those who were pregnant than those who were not, although the differences were not statistically significant. This pattern may reflect the characteristics of teenage females who become pregnant and may be associated with a general pattern of deviance.

Early onset of nonmedical use also was associated with higher likelihood of past year use for some types of prescription psychotherapeutic drugs. Among persons aged 26 or older, rates of past year misuse of any prescription psychotherapeutic drug, pain relievers, and stimulants were higher for persons who first used the respective drugs when they were under age 16 than among those who first used them at age16 to 20.

8.1.3 Geographic and State-Level Differences in Past Year Misuse of Prescription Psychotherapeutics

The rate of past year nonmedical use of any prescription psychotherapeutic was higher in the West than in the South, the Northeast, or the Midwest; the Midwest had a higher rate than the Northeast or South; and the rate in the South was higher than that in the Northeast. Consistent with this, the West had the highest rates of misuse of stimulants, methamphetamine, and pain relievers, whereas the Northeast had the lowest rates of use of these drugs. For tranquilizers, the South had a higher rate of misuse than the Northeast, Midwest, and West.

Rates of nonmedical use of prescription psychotherapeutic drugs also varied from State to State (Chapter 7). When States were classified into quintiles based on their levels of prescription psychotherapeutic drug misuse, several States ranked in the top quintile for multiple drugs. These are presented in Table  8.1. States in Table  8.1 are first ranked according to the number of prescription psychotherapeutic drugs for which the States were in the top quintile. Otherwise, States are listed alphabetically if they were in the top quintile for the same overall number of prescription psychotherapeutic drugs.

8.1.4 Lifetime Misuse of Specific Prescription Psychotherapeutics

Among all lifetime nonmedical users of pain relievers, the specific pain relievers that were misused most commonly were Darvocet^®, Darvon^®, or Tylenol^® with codeine; Vicodin^®, Lortab^®, or Lorcet^®; and Percocet^®, Percodan^®, or Tylox^®. Among persons who first used pain relievers nonmedically in the 12 months prior to the survey, however, the ranking of these drugs was somewhat different. Specifically, Vicodin^®, Lortab^®, or Lorcet^® was the pain reliever with the highest prevalence among new initiates of pain reliever misuse (50.3 percent). Although 62.7 percent of lifetime misusers of pain relievers had used a drug in the Darvocet^®, Darvon^®, or Tylenol^® with codeine category, only 34.3 percent of past year initiates used a drug in this category.

Similarly, the most commonly misused tranquilizers among lifetime nonmedical users were Valium^® or diazepam; Xanax^®, alprazolam, Ativan^®, or lorazepam; and Klonopin^® or clonazepam, in that order. Among past year initiates, the most commonly misused tranquilizer was in the category of Xanax^®, generic alprazolam, Ativan^®, or generic lorazepam.

Differences in the rates of misuse among lifetime nonmedical users and new initiates also can be due to changes in drug availability. For example, the sedative with the highest lifetime prevalence of misuse was methaqualone, Sopor^®, or Quaalude^®, which is no longer legally available in the United States. Among new initiates, the sedative with the highest prevalence of misuse was Restoril^® or temazepam.

8.1.5 Initiation of Prescription Psychotherapeutic Misuse in the Past Year

Youths aged 12 to 17 and young adults aged 18 to 25 were more likely than older adults aged 26 or older to initiate nonmedical use of prescription psychotherapeutic drugs in the past year. The average age at first use among persons who initiated nonmedical use in the past year was 23.4 years for any prescription psychotherapeutic drug, 20.5 years for methamphetamine, 21.8 years for any stimulant, 22.8 years for pain relievers, 24.5 years each for OxyContin^® and tranquilizers, and 29.5 years for sedatives.

8.1.6 Misuse of Prescription Psychotherapeutics and Other Drug Use

Data on the use of other drugs in addition to prescription psychotherapeutics indicate that, in the majority of cases, misuse of prescription drugs is part of an overall pattern of drug use and does not occur in isolation. A large majority of nonmedical users of prescription pain relievers, tranquilizers, stimulants, methamphetamine, and sedatives have used other drugs as well, both in their lifetime and in the past year. Use of other drugs was about 2.5 times more likely among persons who have misused prescription drugs than among those who have not. Lifetime nonmedical users of pain relievers, when compared with lifetime nonmedical nonusers, were 2.6 times more likely to have used marijuana in their lifetime, almost 7 times more likely to have used cocaine, and about 20 times more likely to have used heroin.

8.1.7 Dependence, Abuse, and Treatment for Prescription Psychotherapeutic Drug Misuse

As noted previously, an annual average of 2 million persons met the criteria for dependence on or abuse of prescription psychotherapeutic drugs that were misused in the past year. This estimate was based on combined data for 2002, 2003, and 2004. The following estimated numbers of persons were dependent on or abusing specific prescription psychotherapeutics: 1.4 million persons for pain relievers, 505,000 for tranquilizers, 424,000 for stimulants, and 147,000 for sedatives. Methamphetamine dependence and abuse are not specifically measured in NSDUH or characterized in the Diagnostic and Statistical Manual of Mental Disorders, 4^th edition (DSM-IV) (American Psychiatric Association [APA], 1994). Among past year users, the rate of dependence or abuse was 12.7 percent for pain relievers, 10.1 percent for tranquilizers, 14.4 percent for stimulants, and 17.3 percent for sedatives.

Similar to the pattern of higher prevalences of nonmedical use among females aged 12 to 17 compared with their male counterparts, females in this age group had a higher rate of dependence or abuse than did males for any prescription psychotherapeutic drug, pain relievers, tranquilizers, and stimulants. Patterns of racial or ethnic differences in the prevalence of past year dependence or abuse involving prescription psychotherapeutic drugs were similar to those observed for the overall prevalence in the past year (Section 8.1.2).

Several studies have reported that the likelihood of drug dependence and abuse is higher among persons who start using drugs early in life than among those who begin use at a later age (e.g., Anthony & Petronis, 1995; Gfroerer, Wu, & Penne, 2002; Grant & Dawson, 1998; OAS, 2002a, 2004a, 2005a, 2005b). As indicated in Chapter 6, the findings of this study indicate that the early onset of nonmedical use is associated with higher rates of past year dependence on or abuse of prescription psychotherapeutic drugs. Among adults aged 26 or older, for example, 6.6 percent of lifetime nonmedical pain reliever users who initiated such use under the age of 16 met the criteria for dependence or abuse, whereas only 3.3 percent of those who initiated use at or after the age of 16 met these criteria.

As noted previously, treatment for illicit drug use problems in the past year was reported by an estimated 290,000 persons who met the dependence or abuse criteria in the past year for any prescription psychotherapeutic drugs. Of these, 249,000 received specialty treatment (treatment for a drug problem delivered in a hospital inpatient, rehabilitation, or mental health center setting). An estimated 209,000 persons dependent on or abusing prescription psychotherapeutic drugs in the past year reported that their current or most recent treatment episode in the past year had included treatment for problems with psychotherapeutic drug misuse. By drug type, the approximate numbers receiving such treatment were 135,000 for pain relievers, 50,000 for tranquilizers, 64,000 for stimulants, and 34,000 for sedatives.

8.2. Methamphetamine as a Special Issue

The finding from NSDUH that nonmedical use of methamphetamine remained stable from 2002 to 2004 (or showed declines among some age groups) runs counter to the current widespread concern about increases in problems associated with the drug. Notably, the number of primary methamphetamine or amphetamine treatment admissions reported to TEDS increased from approximately 44,000 in 1994 to approximately 150,000 in 2004 (OAS, 2006a).¹ Over that period, there is little evidence of an increase in the rate of methamphetamine misuse in the population in NSDUH despite the potential effects of methodological changes between 1998 and 1999 and between 2001 and 2002 (see Appendix C).

The findings of three additional analyses of NSDUH data for 2002 through 2004, however, indicate points of comparability between NSDUH estimates and the TEDS treatment data. First, although the prevalence of past month methamphetamine use in NSDUH did not change significantly from 2002 to 2004, the proportion of the past month users who met the criteria for stimulant dependence or abuse in the past year² increased from 10.6 to 22.3 percent, and the proportion of past month methamphetamine users meeting the dependence or abuse criteria for any illicit drug increased from 27.5 to 59.3 percent. Figure 8.1 presents the estimated numbers of current (i.e., past month) methamphetamine users according to their patterns of past year dependence or abuse.

Second, NSDUH's estimates of the numbers of persons receiving specialty treatment for a stimulant use problem during their last or current treatment episode in the past year are similar to the TEDS counts of admissions to treatment for primary stimulant problems (Figure 8.2). The trend from 2002 to 2004 is upward for both, and the estimates are in the same general range when it is recognized that (a) NSDUH is counting persons, whereas TEDS is counting admissions; and (b) NSDUH allows designation of multiple drug problems in this indicator, whereas only one drug is named as the primary problem in TEDS.

Figure 8.1 Past Month Methamphetamine Use, by Past Year Dependence and Abuse, among Persons Aged 12 or Older: 2002-2004

     D

Source: SAMHSA, Office of Applied Studies, National Survey on Drug Use and Health, 2002, 2003, and 2004.

Figure 8.2 Primary Stimulant Admissions to Treatment (TEDS) and Specialty Treatment for a Stimulant Use Problem during Last Treatment Episode in the Past Year (NSDUH), 2002-2004

     D

Source: SAMHSA, Office of Applied Studies, National Survey on Drug Use and Health and Treatment Episode Data Set, 2002, 2003, and 2004.

Third, State-by-State differences in NSDUH estimates of rates of past year nonmedical methamphetamine use generally correspond to the State-by-State differences in the rates of methamphetamine or amphetamine treatment admissions reported by TEDS (Figure 8.3). Both tend to be high in the West and low in the Northeast. When the 50 States and the District of Columbia were assigned to high, medium, or low categories based on their rankings on prevalence rates³ and treatment admission rates, the category assignments agreed for 41 States, and no States were assigned to the high category on one measure and the low category on the other.

Thus, despite the apparent differences between in the trends for methamphetamine treatment and prevalence indicators, there are points of convergence. In particular, findings from the first analysis suggest that increases in treatment admissions reported in TEDS for methamphetamine could reflect progressive increases in problems among methamphetamine users, despite the overall stable trend in NSDUH.

Figure 8.3 Rate of Past Year Nonmedical Methamphetamine Use and Rate of Treatment Admissions for Primary Problems with Methamphetamine or Amphetamine Use

Percentage Using Methamphetamine Nonmedically in Past Year (NSDUH): Average of 2002-2004	Primary Methamphetamine/Amphetamine Treatment Admissions per 100,000 Population (TEDS): Average of 2001-2003

D

Source: SAMHSA, Office of Applied Studies, National Survey on Drug Use and Health, 2002-2004, and Treatment Episode Data Set, 2001-2003.

It also should be recognized that only a fraction of users of any drug meet the criteria for dependence or abuse and receive treatment. Thus, it is possible for treatment admissions to increase without an increase in the number of users. Such a pattern might be seen, for example, if the potency of a drug increases over time, or if changes in the route of administration increase the effects of a drug.

These types of changes have occurred in methamphetamine or amphetamine treatment admissions. From 1993 to 2003, the percentage of these admissions in which smoking was the primary route of administration increased from 15.0 to 56.2 percent (OAS, 2006b). At the same time, admissions in which the drug was inhaled ("snorted") showed an almost corresponding decrease from 42.7 to 15.1 percent. Admissions involving injection of the drug also declined slightly, from 29.0 to 21.6 percent.

Moreover, high-purity methamphetamine from Mexico, much of which is in the form of "ice" (crystal methamphetamine), has begun to replace methamphetamine produced in domestic clandestine laboratories in recent years. This latter change has occurred as law enforcement action has closed many such laboratories and legal and regulatory actions have become more common to reduce the availability of precursor chemicals, such as pseudoephedrine (National Drug Intelligence Center [NDIC], 2006). The increase in high-purity methamphetamine is consistent with the increase in smoking as a route of administration among treatment admissions, particularly with the increase in the availability of "ice."

8.3. Methodological Issues and Challenges

This section discusses the NSDUH definition of nonmedical use of prescription psychotherapeutic drugs and issues associated with that definition. It also outlines practical and conceptual challenges to the collection of accurate and useful data on prescription drug misuse and the strategies that the Substance Abuse and Mental Health Services Administration (SAMHSA) has followed in confronting these challenges. In addition, the section includes a discussion of changes to the survey questions on methamphetamine introduced in 2005 and 2006 to address methodological concerns.

8.3.1 Definition of Nonmedical Use and Impact on Prevalence Estimates

As discussed in Chapter 1, NSDUH defines nonmedical use of prescription pain relievers, tranquilizers, stimulants, or sedatives as use of these types of drugs that was not prescribed for the respondent or that the respondent took only for the experience or feeling that the drug caused. Patterns of prescription drug misuse may vary widely and may affect the likelihood of respondents' reporting the behavior in the survey. The latter part of the definition, use "for the experience or feeling the drug caused," covers "recreational use" regardless of how the drug is obtained, the frequency of use, or the presence or absence of symptoms of dependence or abuse.

On the other hand, the first part of the definition, "use without a prescription of the respondent's own," triggers a positive response regardless of the user's intent. Thus, some respondents who report nonmedical use may have been given the drugs by friends or family when they had an actual medical need for such medications. Although it might be questioned whether this latter use constitutes "illicit drug use," individuals in these situations are not likely to have been evaluated by a physician to determine the legitimacy of their need or the appropriateness of the drug and dosage. Consequently, psychotherapeutic medications taken without a doctor's prescription might be unsafe or ineffective for the individual's condition; the dosage might not be appropriate; potential interactions with other medications or medical conditions would not have been evaluated professionally; the user would not have been counseled by a physician or pharmacist about the appropriate use of the medication; and monitoring of adverse consequences generally would be lacking.

With medications that pose a risk of dependence, such issues are important, both for persons taking psychotherapeutic medications without a prescription and for persons who have a legitimately prescribed medication and knowingly provide it to others or unknowingly become a source of diverted medications. In particular, several of the psychotherapeutic medications included in NSDUH are classified by the Drug Enforcement Administration (DEA) as Schedule II drugs under the Controlled Substances Act (CSA).⁴ Examples of Schedule II drugs in NSDUH include (but are not limited to) pain relievers, such as oxycodone (including OxyContin^® and Percocet^®), morphine, and methadone; stimulants, such as methamphetamine, amphetamines (e.g., Biphetamine^®, Dexedrine^®), and Ritalin^® or methylphenidate; and sedatives, such as pentobarbital (Nembutal^®) and secobarbital (Seconal^®, Tuinal^®). Because Schedule II drugs have the highest potential for dependence or abuse among those that are approved for medical use in the United States, use of these drugs for a medical need but without a doctor's prescription could constitute risky substance use.

Table  8.2 summarizes the two criteria for nonmedical use in NSDUH and the types of situations associated with them. However, it is not possible to determine from the NSDUH data what proportion of cases fit the combinations shown in the table.

Another scenario that challenges the definition of nonmedical use is use of a legitimately prescribed drug for an actual medical need but more frequently or in a greater quantity than that specified on the label or by the physician, or use for a longer time period than the physician specified. Theoretically, survey respondents would report such instances as nonmedical use if their motives fit the rubric of use "for the experience or feeling the drug caused." However, respondents may not be reliable in making such determinations of intent and may tend to underreport nonmedical use in such circumstances. Situations of this type are important, however, because they may represent the beginnings of misuse, particularly in cases where the individual initially uses the drug under medical supervision but then begins to develop addictive patterns of use. In these cases, interpretations of intent may be subject to distortion not only in survey reporting but also in individuals' understanding of their own behavior.

Other scenarios of nonmedical use to consider include use of a prescription psychotherapeutic drug for a reason other than the purpose for which it was prescribed, such as using a stimulant to facilitate studying for a college examination. Another related scenario is use of a prescription psychotherapeutic drug not for the direct experience or feeling caused by the drug itself but in order to enhance, reduce, or otherwise modify the direct effects, side effects, or aftereffects of another drug. For example, diazepam (Valium^®) might be used to reduce the anxiety associated with the aftereffects of cocaine use.

Scenarios that may represent problematic use but may not necessarily constitute nonmedical use include use of a leftover prescription for a similar type of medical problem but not the one for which the medication was originally prescribed and accidental misuse. For example, a person may have been prescribed a pain reliever when recovering from surgery but occasionally takes leftover medication for other pain relief; once all of the medication has been taken, he or she does not ask the doctor for another prescription unless a new medical need should arise. This type of use probably would not constitute misuse if done with the knowledge or general guidance of a patient's doctor; due to issues mentioned previously, however, such as potential interactions with other drugs, this could be problematic if done on a patient's own. This type of use also could make the transition to misuse if a patient overstates his or her need for more medication.

In addition, accidental ingestion of a drug would not be considered misuse. In the 1990s, for example, flunitrazepam (Rohypnol^®, or "Roche" or "roofies"), became known as a "date rape" drug because of situations in which victims of sexual assaults unknowingly drank beverages in which the odorless and tasteless drug had been introduced. Similarly, a person may have accidentally taken an extra dose because he or she forgot having already taken the medication. In both of these situations, the type of use lacks either deliberate awareness of taking the drug or deliberate noncompliance with medical instructions that characterize other examples of misuse that were discussed previously.

The current NSDUH definition of misuse of prescription-type psychotherapeutic drugs has been in use since 1994. Although it is conceptually complex, the current definition is a simplification of the earlier definition and was developed to make the questions more understandable to respondents. Prior to 1994, questions about misuse of prescription-type drugs were asked using the term "nonmedical use," a phrase still used in NSDUH reports but no longer used in the NSDUH questionnaire. As illustrated in the question format for each class of psychotherapeutic drugs, respondents to NSDUH questionnaires prior to 1994 were presented with four criteria for nonmedical use:

We are interested in the nonmedical use of these prescription-type drugs. Nonmedical use of these drugs is any use on your own, that is, either:

1. without your own prescription from a doctor, or
   
   
2. in greater amounts than prescribed, or
   
   
3. more often than prescribed, or
   
   
4. for kicks, to get high, to feel good, or curiosity about the pill's effect, or for any reasons other than a doctor said you should take them.
   
   

Because this question is very complex and includes many concepts, several versions that decomposed the concepts into separate questions were tested as part of the 1994 instrument redesign. In the item testing, separate "yes/no" questions were asked about each of the ways one could use a drug "nonmedically" (e.g., use without a prescription), then a global "yes/no" question about "nonmedical use" was asked that included all of the ways a drug could be used nonmedically. The study found that (1) estimates of prescription-type drug misuse were much higher when the original question was decomposed into several questions, and (2) most respondents who admitted use without a prescription, or use in greater amounts or more often than prescribed, did not admit to "nonmedical use" in the global question. Most respondents who reported use, "for kicks or to get high," however, did admit to "nonmedical use." It was concluded that if a single question was adopted, the concepts of use without a prescription and use for recreational purposes should be included, but the words "nonmedical use" should not be included (Hubbard, Pantula, & Lessler, 1992).

Further testing was done in a cognitive laboratory (Caspar & Biemer, 1999) to refine the final wording for the 1994 instrument. Cognitive laboratory respondents reported that the wording "for kicks or to get high" was specific to particular subcultures and to particular drugs. After additional testing, the wording "…not prescribed for you or only for the experience or feeling it caused" was adopted (Caspar, Hubbard, Kennedy, Wayne, & Biemer, 1993).

8.3.2 Changes in Prescription Drug Status: The Case of Methamphetamine

One potential problem in measuring prescription psychotherapeutic drug misuse is that specific pharmaceuticals may change status. Not only may new drugs be introduced and others discontinued, but drugs could move from nonprescription to prescription status if abuse potential were recognized after initial marketing. In addition, it is possible for drugs generally manufactured by legitimate pharmaceutical companies with government approval to become popular as abuse substances and to be produced illegally outside that setting. This has happened with methamphetamine. As indicated in Chapter 1, most methamphetamine that is currently being used nonmedically is produced by clandestine laboratories within the United States or abroad rather than by the legitimate pharmaceutical industry. Consequently, the drug is distributed primarily through illicit trafficking rather than through pharmacies. It could be argued that methamphetamine has moved from being a prescription-type drug that can be used legitimately but also can be misused to being principally an illicit street drug, for which any use likely constitutes misuse.

This evolution in the production and availability of methamphetamine presents challenges for measuring the prevalence of misuse in NSDUH. In responding to the questions in NSDUH, some users who are not aware of methamphetamine's history as a prescription medication may fail to recognize it—and not report misusing it—when methamphetamine is presented in the context of prescription-type drugs. This risk of underreporting may be increased by the depictions of tablets and capsules on the "pill cards." If this were true, NSDUH would be underestimating the prevalence of methamphetamine misuse.

In 2005, new items were added to the NSDUH instrument to address these concerns. These new items were added in a segment of the instrument separate from the prescription drug modules and were administered only to respondents who did not indicate in the stimulants module that they had used methamphetamine. This question also included additional slang terms for methamphetamine and information on how it is commonly used. These respondents received the following item:

Methamphetamine, also known as crank, ice, crystal meth, speed, glass, and many other names, is a stimulant that usually comes in crystal or powder forms. It can be smoked, "snorted," swallowed or injected. Have you ever, even once, used Methamphetamine?

Respondents who answered "Yes" to this question then were asked a question to classify them as past month, past year, or lifetime users.

Review of the early data for 2005 revealed additional cases of methamphetamine use based on responses to the new items. That might be expected, however, given that respondents who did not previously report that they misused methamphetamine were given an extra opportunity to report use. Therefore, follow-up items were added in 2006 to verify that respondents who previously reported never using methamphetamine had not made a mistake and to clarify why they failed to report methamphetamine use in the prescription stimulants module. These items are as follows:

Earlier, the computer recorded that you have never used Methamphetamine, Desoxyn or Methedrine. Which answer is correct?

1	I have never, even once, used Methamphetamine, Desoxyn or Methedrine
2	I last used methamphetamine [time period]

[IF 'YES' TO ABOVE ITEM] Why did you report earlier that you had never used methamphetamine?

1	The earlier question asked about prescription drugs, and I didn't think of methamphetamine as a prescription drug
2	I made a mistake when I answered the earlier question about ever using methamphetamine
3	Some other reason

These follow-up items are necessary, in part, to reduce the risk of increasing the reports of methamphetamine use simply by repeating questions. Repeating questions for any drug would likely result in increased numbers of positive responses, and doing so only for methamphetamine could result in a disproportionate reporting of that drug relative to the others in the survey. Thus, the follow-up item sought to identify respondents who had failed to report methamphetamine use in response to the earlier question because they did not consider methamphetamine to be a prescription drug. The 2005 NSDUH report of national findings (OAS, in press) will present information on responses to the new items added in 2005, data from the early review of the new 2006 items, and an assessment of the impact of these responses on the survey estimates of the prevalence of methamphetamine use and other indicators that may be affected.

8.3.3 Challenges in Modifying the NSDUH Instrument

NSDUH researchers seek to collect the most accurate data possible on illicit drug use, including use of illicit substances (e.g., marijuana, cocaine) and nonmedical use of prescription psychotherapeutic drugs. Since the survey's inception, several methodological changes have been made to improve the estimates (Appendix C). SAMHSA recognizes, however, that the current instrument still has a number of limitations. It would be desirable, for example, to collect more information on specific psychotherapeutic drugs to determine not only lifetime use, as is currently done, but also age at first use and the period of most recent use.

Similarly, it might be desirable to ask about legitimate medical use with a prescription in order to provide contextual information for the statistics on misuse. It also might be desirable to break the definition of nonmedical use into its component parts and ask separately about use without a prescription and use for the feeling or experience the drug caused.

In addition, a more comprehensive and current list of specific prescription drugs could be displayed, based upon information from the "other-specify" responses, and to ask questions about these drugs individually. Specifically, write-in responses to the "other-specify" questions in the four therapeutic drug class modules of the prescription drug section of the instrument provide a means of tracking the emergence of new nonmedical usage patterns, which may reflect introduction of new pharmaceuticals with abuse potential. Ultimately, the new substances that are most frequently reported could be considered for addition to the "pill cards." Conversely, specific psychotherapeutic drugs that may have been chosen for inclusion in a module based on their prevalence in 1998 (i.e., immediately prior to the last major survey redesign in 1999) but are no longer commonly reported could be candidates for exclusion from a "pill card" or psychotherapeutic drug module.

However, there are obstacles to modifying the NSDUH instrument to address such limitations. These obstacles, which reflect competing objectives, are chiefly of two types: (1) the need to keep questions consistent—if not identical—in order to allow comparison across survey years, and (2) time limits in administering the survey.

8.3.3.1 Trend Comparisons

Maintaining consistent methods to ensure comparable measurement in support of trend analysis has been a significant issue with epidemiological surveys that are used as surveillance systems. In SAMHSA's experience, even slight changes in NSDUH questions can result in differences in survey estimates, and decisions to revise items in the survey instrument are never taken lightly.

When methodological changes have been made, SAMHSA has always attempted to measure their effects independent of the effects of any change in actual prevalence. When the survey questionnaire was redesigned in 1994, for example, SAMHSA used a split-sample technique that allowed the effects of the methods change to be measured. The results of this procedure provided a means of adjusting estimates from prior years to make them comparable with those generated after the change (OAS, 2000a). In addition, a key feature of the redesigned questionnaire in 1994 involved dividing the questionnaire into "core" and "noncore" sections. This core/noncore structure was maintained when the questionnaire was again redesigned as a computer-assisted interview in 1999. Core sections, such as key demographic characteristics and drug use prevalence questions, were designed to stay relatively constant from 1 year to the next in order to permit measurement of trends in drug use. In contrast, the content of noncore sections could change considerably across years to measure new topics of interest or to rotate topics in or out of the interview. In noncore sections, therefore, questions or entire sections could be added or deleted, or the wording of existing questions could change from 1 year to the next.

More recent experiences with major methodological changes in NSDUH have not had such successful outcomes. As discussed in Appendix C, trends were disrupted due to methodological improvements between 1998 and 1999 and between 2001 and 2002.

Although the survey ideally should provide the best possible estimate of drug use at any single point in time, proposed improvements must be balanced against the need to maintain consistency for trend analysis. When a new drug is added to a category, for example, it may be expected to increase the estimated prevalence for that category and each of the aggregate drug groupings to which the drug contributes (e.g., pain relievers, any psychotherapeutic drug, and any illicit drug). What would not be known, however, is whether any observed increases in prevalence due to the introduction of questions about new drugs might reflect longer-term trends that began to occur more than 1 year prior to the introduction of the new questions.

In addition, such changes can have other, more subtle effects. For example, adding items about new pharmaceuticals in the stimulant category might influence responses to existing questions on drugs in that category by changing the context in which these existing questions appear.

In seeking to maintain consistency of survey items for measurement of trends, consideration also must be given to the actual meaning of the questions and the behaviors to be measured. With frequent changes in the prescription drugs that are available on the market, long-term tracking of the same set of specific medications in a given psychotherapeutic class would not necessarily ensure consistency in measuring the overall problem of use of drugs in that class because it would fail to take into account the introduction of new pharmaceuticals and disappearance of obsolete ones. Failure to adapt the list of specific drugs would likely result in gradual erosion of the accuracy of measurement of misuse of that class of drugs, as respondents would be presented with a list of increasingly obsolete drugs and are not asked about new drugs that are increasingly being misused. Thus, it is essential that NSDUH and other epidemiological surveys periodically update the substances listed to ensure continuous tracking of the general phenomenon.

In addition, changes may be needed to accommodate shifts in the street names given to drugs as new generations of users emerge (Morral, McCaffrey, & Chien, 2003). Understanding of drug categories, such as "tranquilizers" and "sedatives," also may not be the same across generations of users. Further, categories such as "tranquilizers" and "sedatives" may not be understood by respondents as conceptually distinct categories of drugs.

8.3.3.2 Survey Length

The other main problem in modifying the NSDUH instrument is the time required for survey questions. With survey administration currently averaging 62 minutes, adding new items frequently requires deletion of existing items. Thus, the need for new questions must be balanced against the time required to administer them.

The approach used to collect information on lifetime use of specific pharmaceuticals within psychotherapeutic drugs represents a compromise between asking only a general question about use of any drug in the class and asking questions specific to each major drug in the class. Detailed information about initiation of use and most recent use of specific psychotherapeutics is collected only for OxyContin^® (since 2004) and methamphetamine (since 1999). Asking respondents when they last used each of the specific pharmaceuticals included in NSDUH would be desirable, but the amount of time required could be prohibitive, particularly for lifetime misusers of many psychotherapeutics.

8.3.3.3 NSDUH Approach for Collecting New Drug Information

When it is deemed necessary to add questions about new drugs, the core/noncore organization in NSDUH allows for the introduction of new items to noncore sections. Because the noncore sections are administered after respondents have completed the section of the core interview that asks about their misuse of prescription psychotherapeutic drugs, placement of questions about new drugs in noncore sections would avoid possible distortion of estimates for existing drugs. This approach has been applied in previous or current surveys to measure the use of "specialty" cigarettes (such as bidis and clove cigarettes) and use of cigars that contain marijuana ("blunts").

Since the questions about specific psychotherapeutic drugs were last revised for NSDUH, new psychotherapeutics with abuse potential have appeared on the market. Further, some psychotherapeutic drugs not listed on the pill cards have begun to be reported in substantial numbers in the "other-specify" items for the four therapeutic drug classes covered in the survey. In recognition of the possible emergence of new patterns of psychotherapeutic drug misuse, the 2006 NSDUH instrument was modified to add questions regarding misuse of Adderall^® (a prescription stimulant) and Ambien^® (a prescription sedative). To avoid influencing responses to existing items, these items, and the new methamphetamine items described previously, were added in noncore sections of the survey instrument. Appendix B provides further information about the specific psychotherapeutic drugs covered in NSDUH.

In methodological discussions, SAMHSA has devoted considerable attention to balancing the need to limit the time required to administer the survey, to keep items consistent to ensure that data can be used for trend analysis, and to accommodate needed changes, such as adding questions about new drugs as they emerge. One way to balance these objectives would be to postpone making optional changes of methodology until a planned periodic redesign, perhaps every 10 to 12 years. Of course, major methodological problems would need to be addressed more promptly. Both for year-to-year tweaking and the less frequent redesign efforts, split-sample and other techniques could be used to measure the impact of methodological changes, providing the capability to preserve long-term trend assessment.

In summary, the NSDUH estimates of misuse of prescription-type psychotherapeutic drugs are generally strong, and NSDUH is the most comprehensive source of epidemiological information on such misuse in the general U.S. population aged 12 or older. Although there are recognized limitations to the data that are collected, SAMHSA statisticians are continually seeking to overcome those limitations to the extent feasible, while taking into consideration the need to maintain consistent methods to allow trend analysis and the need to avoid lengthening the interview to an extent that would become burdensome for the respondents and interviewers.

¹ TEDS data on methamphetamine and amphetamine admissions are aggregated because some States do not distinguish between the two. Based on data from States that make the distinction, approximately 86 percent of these admissions involve methamphetamine as the primary drug of abuse.

² As noted previously, specific criteria related to dependence or abuse for methamphetamine are not established in DSM-IV (APA, 1994) and are not included in NSDUH.

³ Because the prevalence map in Figure 8.3 groups States into three groups of 17 States each (including the District of Columbia) based on their prevalences, the appearance of this map differs from the prevalence map for methamphetamine in Figure 7.5 in Chapter 7. In Figure 7.5, States were grouped according to the 10 States with the highest prevalences, the 10 States with the lowest prevalences, and remaining States with intermediate prevalences.

⁴ Title II of the Comprehensive Drug Abuse Prevention and Control Act of 1970, 21 U.S.C. Section 802 (1996). For details, see http://www.dea.gov/pubs/csa/802.htm.

This page was last updated on June 03, 2008.

SAMHSA, an agency in the Department of Health and Human Services, is the Federal Government's lead agency for improving the quality and availability of substance abuse prevention, addiction treatment, and mental health services in the United States.