2006 National Survey on Drug Use & Health:  National Results

Appendix B: Statistical Methods and Measurement

B.1 Target Population

An important limitation of estimates of drug use prevalence from the National Survey on Drug Use and Health (NSDUH) is that they are only designed to describe the target population of the survey—the civilian, noninstitutionalized population aged 12 or older. Although this population includes almost 98 percent of the total U.S. population aged 12 or older, it excludes some important and unique subpopulations who may have very different drug use patterns. For example, the survey excludes active military personnel, who have been shown to have significantly lower rates of illicit drug use. Also, persons living in institutional group quarters, such as prisons and residential drug use treatment centers, are not included in NSDUH, yet they have been shown in other surveys to have higher rates of illicit drug use. Also excluded are homeless persons not living in a shelter on the survey date; they are another population shown to have higher than average rates of illicit drug use. Appendix D describes other surveys that provide data for these populations.

B.2 Sampling Error and Statistical Significance

This report includes tables for national estimates (see Appendices F and G) that were drawn from a more comprehensive set of tables referred to as "detailed tables."⁸ The national estimates, along with the associated standard errors (SEs), were computed for all detailed tables, including those in this report, using a multiprocedure package, SUDAAN^® Software for Statistical Analysis of Correlated Data. SUDAAN was designed for the statistical analysis of data collected using stratified, multistage cluster sampling designs, as well as other observational and experimental studies involving repeated measures or studies subject to cluster correlation effects (RTI International, 2004). The final, nonresponse-adjusted, and poststratified analysis weights were used in SUDAAN to compute unbiased design-based drug use estimates.

The sampling error (i.e., the standard error or SE) of an estimate is the error caused by the selection of a sample instead of conducting a census of the population. The sampling error may be reduced by selecting a large sample and/or by using efficient sample design and estimation strategies, such as stratification, optimal allocation, and ratio estimation.

With the use of probability sampling methods in NSDUH, it is possible to develop estimates of sampling error from the survey data. These estimates have been calculated using SUDAAN for all estimates presented in this report using a Taylor series linearization approach that takes into account the effects of NSDUH's complex design features. The sampling errors are used to identify unreliable estimates and to test for the statistical significance of differences between estimates.

B.2.1 Variance Estimation for Totals

Although the SEs of estimates of means and proportions can be calculated appropriately in SUDAAN using a Taylor series linearization approach, SEs of estimates of totals may be underestimated in situations where the domain size is poststratified to data from the U.S. Census Bureau. Because of this underestimation, alternatives for estimating SEs of totals were implemented.

Estimates of means or proportions, _d, such as drug use prevalence estimates for a domain d, can be expressed as a ratio estimate:

,     D

where _d is a linear statistic estimating the number of substance users in the domain d and _d is a linear statistic estimating the total number of persons in domain d (both users and nonusers). The SUDAAN software package is used to calculate direct estimates of _d and _d and also can be used to estimate their respective SEs. A Taylor series approximation method implemented in SUDAAN provides estimates for _d and its SE.

When the domain size, _d, is free of sampling error, an appropriate estimate of the SE for the total number of substance users is

.     D

This approach is theoretically correct when the domain size estimates, _d, are among those forced to match their respective U.S. Census Bureau population estimates through the weight calibration process (Chen et al., 2007). In these cases, _d is not subject to a sampling error induced by the NSDUH design. For a more detailed explanation of the weight calibration process, see Section A.3.2 in Appendix A.

For estimated domain totals, _d, where _d is not fixed (i.e., where domain size estimates are not forced to match the U.S. Census Bureau population estimates), this formulation still may provide a good approximation if it can be assumed that the sampling variation in _d is negligible relative to the sampling variation in _d. This is a reasonable assumption for most cases in this study.

For various subsets of estimates, the above approach yielded an underestimate of the variance of a total because _d was subject to considerable variation. In 2000, an approach was implemented to reflect more accurately the effects of the weighting process on the variance of total estimates. This approach consisted of calculating SEs of totals for all estimates in a particular detailed table using the formula above when a majority of estimates in a table were among domains in which _d was fixed during weighting or if it could be assumed that the sampling variation in _d was negligible. SEs of totals in detailed tables, where the majority of estimates were among domains in which _d was subject to considerable variability, were calculated directly in SUDAAN. Starting with the 2005 NSDUH and continuing in the 2006 NSDUH, a "mixed" method approach was implemented for all detailed tables to improve on the accuracy of SEs. This method had been applied to only a select number of tables in the 2004 NSDUH. This approach assigns the method of SE calculation to domains (subgroups for which the estimates were calculated) within tables so that all estimates among a select set of domains with fixed _d were calculated using the formula above, and all other estimates were calculated directly in SUDAAN, regardless of other estimates within the same table. The set of domains considered controlled (i.e., those with a fixed _d) was restricted to main effects and two-way interactions in order to maintain continuity between years. Domains consisting of three-way interactions may be controlled in 1 year but not necessarily in preceding or subsequent years. The use of such SEs did not affect the SE estimates for the corresponding proportions presented in the same sets of tables because all SEs for means and proportions are calculated directly in SUDAAN. As a result of the use of this mixed-method approach, the SEs for the total estimates within many detailed tables were calculated differently from those in prior NSDUH reports.

Table B.1 at the end of this appendix contains a list of domains with a fixed _d. This table includes both the main effects and two-way interactions and may be used to identify the method of SE calculation employed for estimates of totals in the various tables of this report. For example, Table G.13 in Appendix G of this report presents estimates of illicit drug use among persons aged 18 or older within the domains of gender, Hispanic origin and race, education, and current employment. Estimates among the total population (age main effect), males and females (age by gender interaction), and Hispanics and non-Hispanics (age by Hispanic origin interaction) were treated as controlled in this table, and the formula above was used to calculate the SEs. The SEs for all other estimates, including white and black or African American (age by Hispanic origin by race interaction) were calculated directly from SUDAAN. It is important to note that estimates presented in this report for racial groups are among non-Hispanics. For instance, the domain for whites is actually non-Hispanic whites and is therefore a two-way interaction.

B.2.2 Suppression Criteria for Unreliable Estimates

As has been done in past NSDUH reports, direct survey estimates produced for this study that are considered to be unreliable due to unacceptably large sampling errors are not shown in this report and are noted by asterisks (*) in the tables containing such estimates. The criteria used for suppressing all direct survey estimates were based on the relative standard error (RSE) (defined as the ratio of the SE over the estimate), nominal (actual) sample size, and effective sample size for each estimate.

Proportion estimates () within the range [0 < < 1], rates, and the corresponding estimated number of users were suppressed if

RSE[-ln ()] > .175 when ≤ .5

or

RSE[-ln(1 - )] > .175 when > .5.

Using a first-order Taylor series approximation to estimate RSE[-ln()] and RSE[-ln(1 - )] the following equation was derived and used for computational purposes:

> .175 when ≤ .5     D

or

> .175 when > .5.     D

The separate formulas for ≤ .5 and > .5 produce a symmetric suppression rule; that is, if is suppressed, 1 – will be suppressed as well. See Figure B.1 for a graphical representation of the required minimum effective sample sizes as a function of the proportion estimated. When .05 < < .95, the symmetric properties of the rule produce local minimum effective sample sizes at = .2 and again at = .8, such that an effective sample size of greater than 50 is required; this means that estimates would be suppressed for these values of unless the effective sample sizes were greater than 50. Within this same interval of .05 < < .95, a local maximum effective sample size of 68 is required at = .5. So, to simplify requirements and maintain a conservative suppression rule, estimates of between .05 and .95, which had effective sample sizes below 68, were suppressed.

Below is a graph. Click here for the text describing this graph.

Figure B.1 Required Effective Sample as a Function of the Proportion Estimated

In addition, a minimum nominal sample size suppression criterion (n = 100) that protects against unreliable estimates caused by small design effects and small nominal sample sizes was employed. Prevalence estimates also were suppressed if they were close to 0 or 100 percent (i.e., if < .00005 or if ≥ .99995).

Estimates of other totals (e.g., number of initiates) along with means and rates that are not bounded between 0 and 1 (e.g., mean age at first use and incidence rates) were suppressed if the RSEs of the estimates were larger than .5. Additionally, estimates of the mean age at first use were suppressed if the sample size was smaller than 10 respondents. Also, the estimated incidence rate and number of initiates were suppressed if they rounded to 0.

The suppression criteria for various NSDUH estimates are summarized in Table B.2 at the end of this appendix.

B.2.3 Statistical Significance of Differences

This section describes the methods used to compare prevalence estimates in this report. Customarily, the observed difference between estimates is evaluated in terms of its statistical significance. Statistical significance is based on the p value of the test statistic and refers to the probability that a difference as large as that observed would occur due to random variability in the estimates if there were no difference in the prevalence estimates for the population groups being compared. The significance of observed differences in this report is reported at the .05 level. When comparing prevalence estimates, the null hypothesis (no difference between prevalence estimates) was tested against the alternative hypothesis (there is a difference in prevalence estimates) using the standard difference in proportions test expressed as

,     D

where ₁ = first prevalence estimate, ₂ = second prevalence estimate, var (₁) = variance of first prevalence estimate, var (₂) = variance of second prevalence estimate, and cov (₁, ₂) = covariance between ₁ and ₂. In cases where significance tests between years were performed, the prevalence estimate from the earlier year (e.g., 2002, 2003, 2004, or 2005) becomes the first prevalence estimate, and the prevalence estimate from the later year (e.g., 2003, 2004, 2005, or 2006) becomes the second prevalence estimate.

Under the null hypothesis, Z is asymptotically distributed as a normal random variable. Therefore, calculated values of Z can be referred to the unit normal distribution to determine the corresponding probability level (i.e., p value). Because the covariance term between the two estimates is not necessarily zero, SUDAAN was used to compute estimates of Z along with the associated p values using the analysis weights and accounting for the sample design as described in Appendix A. A similar procedure and formula for Z were used for estimated totals; however, it should be noted that because it was necessary to calculate the SE outside of SUDAAN for domains forced by the weighting process to match their respective U.S. Census Bureau population estimates, the corresponding test statistics also were computed outside of SUDAAN.

When comparing population subgroups across three or more levels of a categorical variable, log-linear chi-square tests of independence of the subgroups and the prevalence variables were conducted first to control the error level for multiple comparisons. If the chi-square test indicated overall significant differences, the significance of each particular pairwise comparison of interest was tested using SUDAAN analytic procedures to properly account for the sample design. Using the published estimates and SEs to perform independent t tests for the difference of proportions usually will provide the same results as tests performed in SUDAAN. However, where the significance level is borderline, results may differ for two reasons: (1) the covariance term is included in SUDAAN tests, whereas it is not included in independent t tests; and (2) the reduced number of significant digits shown in the published estimates may cause rounding errors in the independent t tests.

As part of a comparative analysis discussed in Chapter 9, prevalence estimates from the Monitoring the Future (MTF) study, sponsored by the National Institute on Drug Abuse (NIDA), were presented for recency measures of selected substances (see Tables 9.1 and 9.2). The analyses focused on prevalence estimates for 8^th and 10^th graders and prevalence estimates for young adults aged 19 to 24 for 2002 through 2006. Estimates for the 8^th and 10^th grade students were calculated using MTF data as the simple average of the 8^th and 10^th grade estimates. Estimates for young adults aged 19 to 24 were calculated using MTF data as the simple average of three modal age groups: 19 and 20 years, 21 and 22 years, and 23 and 24 years. Published results were not available from NIDA for significant differences in prevalence estimates between years for these subgroups, so testing was performed using information that was available.

For the 8^th and 10^th grade average estimates, tests of differences were performed between 2006 and the 4 prior years. Estimates for persons in grade 8 and grade 10 were considered independent, simplifying the calculation of variances for the combined grades. Across years, the estimates for 2006 involved samples independent of those in 2002, 2003, and 2004, but from 2005 to 2006 the sample of schools overlapped 50 percent, creating a covariance in the estimates. Design effects published in Johnston et al. (2007b) for adjacent and nonadjacent year testing were used. For the 19- to 24-year-old age group, tests of differences were done assuming independent samples across years, which is appropriate for comparisons of 2003 and 2005 with 2006. However, this results in conservative tests for comparisons of 2002 and 2004 data with 2006 data because it does not take into account covariances associated with repeated observations from the longitudinal samples. Estimates of covariances were not available.

As an example, the difference between the 2005 and 2006 averages of prevalence estimates for persons in grades 8 and 10 can be expressed as

₂ – ₁,

where ₁ = (₁₁ + ₁₂)/2, ₁₁ and ₁₂ are the prevalence estimates for the 8^th and 10^th grades, respectively, for 2005; and ₂ is defined similarly for 2006. The variance of a prevalence estimate can be written as

     D

where n is the sample size and D is the appropriate design effect obtained from the sampling design. In the MTF study, design effects were available for comparisons between adjacent year (i.e., 2005 vs. 2006) estimates and nonadjacent year (i.e., 2002 vs. 2006, 2003 vs. 2006, and 2004 vs. 2006) estimates; therefore, the variance of the difference between 2 years of estimates for a particular grade can be expressed as

     D

where i = 1 indexes the 8^th grade, i = 2 indexes the 10^th grade, D_i is the design effect appropriate for comparisons between estimates of the 2 years (with separate design effect parameters for adjacent and nonadjacent years), and the n_ji are the sample sizes corresponding to the indexed year and grade prevalence estimates, i, j = 1,2. Because the 8^th and 10^th grade samples were independently drawn, the variance of the difference between the 8^th and 10^th grade averages can be expressed as

     D

The test statistic can therefore be written as

     D

where Z is asymptotically distributed as a standard normal random variable.

B.3 Other Information on Data Accuracy

The accuracy of survey estimates can be affected by nonresponse, coding errors, computer processing errors, errors in the sampling frame, reporting errors, and other errors not due to sampling. They are sometimes referred to as "nonsampling errors." These types of errors and their impact are reduced through data editing, statistical adjustments for nonresponse, close monitoring and periodic retraining of interviewers, and improvement in various quality control procedures.

Although these types of errors often can be much larger than sampling errors, measurement of most of these errors is difficult. However, some indication of the effects of some types of these errors can be obtained through proxy measures, such as response rates and from other research studies.

B.3.1 Screening and Interview Response Rate Patterns

In 2006, respondents continued to receive a $30 incentive in an effort to maximize response rates. Of the 151,288 eligible households sampled for the 2006 NSDUH, 137,057 were screened successfully, for a weighted screening response rate of 90.6 percent (Table B.3). In these screened households, a total of 85,034 sample persons were selected, and completed interviews were obtained from 67,802 of these sample persons, for a weighted interview response rate of 74.2 percent (Table B.4). A total of 11,750 (17.7 percent) sample persons were classified as refusals or parental refusals, 3,144 (3.7 percent) were not available or never at home, and 2,338 (4.3 percent) did not participate for various other reasons, such as physical or mental incompetence or language barrier (see Table B.4, which also shows the distribution of the selected sample by interview code and age group). Among demographic subgroups, the weighted interview response rate was highest among 12 to 17 year olds (85.5 percent), females (75.9 percent), blacks (77.9 percent), among persons in the Midwest (75.4 percent), and among residents of nonmetropolitan areas (76.8 percent) (Table B.5).

The overall weighted response rate, defined as the product of the weighted screening response rate and weighted interview response rate, was 67.2 percent in 2006. Nonresponse bias can be expressed as the product of the nonresponse rate (1 - R) and the difference between the characteristic of interest between respondents and nonrespondents in the population (P_r - P_nr). By maximizing NSDUH response rates, it is hoped that the bias due to the difference between the estimates from respondents and nonrespondents is minimized. Drug use surveys are particularly vulnerable to nonresponse due to the difficult nature of accessing heavy drug users. In a study that matched 1990 census data to 1990 NHSDA nonrespondents,⁹ it was found that populations with low response rates did not always have high drug use rates. For example, although some populations were found to have low response rates and high drug use rates (e.g., residents of large metropolitan areas and males), other populations had low response rates and low drug use rates (e.g., older adults and high-income populations). Therefore many of the potential sources of bias tend to cancel each other in estimates of overall prevalence (Gfroerer, Lessler, & Parsley, 1997).

B.3.2 Inconsistent Responses and Item Nonresponse

Among survey participants, item response rates were above 99 percent for most drug use items. However, respondents could give inconclusive or inconsistent information about whether they ever used a given drug (i.e., "yes" or "no") and, if they had used a drug, when they last used it; the latter information is needed to identify those lifetime users of a drug who used it in the past year or past month. In addition, respondents could give inconsistent responses to items such as when they first used a drug compared with their most recent use of a drug. These missing or inconsistent responses first are resolved where possible through a logical editing process. Additionally, missing or inconsistent responses are imputed using statistical methodology (Aldworth et al., 2007a). These imputation procedures in NSDUH are based on responses to multiple questions, so that the maximum amount of information is used in determining whether a respondent is classified as a user or nonuser, and if the respondent is classified as a user, whether the respondent is classified as having used in the past year or the past month. For example, ambiguous data on the most recent use of cocaine are statistically imputed based on a respondent's data for use (or most recent use) of tobacco products, alcohol, inhalants, marijuana, hallucinogens, and nonmedical use of prescription psychotherapeutic drugs. Nevertheless, editing and imputation of missing responses are potential sources of measurement error. For more information on editing and statistical imputation, see Sections A.3 and A.3.1 of Appendix A. Additional information on editing and statistical imputation procedures can be found online at http://www.drugabusestatistics.samhsa.gov/nsduh/methods.cfm#top.

B.3.3 Data Reliability

NSDUH research staff are conducting a study to assess the reliability of respondents' responses to the survey. An interview/reinterview method was employed in which 3,136 individuals were interviewed on two occasions during 2006 generally 5 to 15 days apart. The reliability of the responses will be assessed by comparing the responses of the first interview (time 1) to the responses from the reinterview (time 2). Preliminary analyses of data from approximately two thirds of the study's respondents show that, overall, there is a good level of consistency between measures of substance use and mental health between the two data collection time points. Results of the study will be published later.

B.3.4 Validity of Self-Reported Substance Use

Most substance use prevalence estimates, including those produced for NSDUH, are based on self-reports of use. Although studies have generally supported the validity of self-report data, it is well documented that these data often are biased (underreported or overreported). The bias varies by several factors, including the mode of administration, the setting, the population under investigation, and the type of drug (Aquilino, 1994; Brener et al., 2006; Harrison & Hughes, 1997; Tourangeau & Smith, 1996; Turner, Lessler, & Gfroerer, 1992). NSDUH utilizes widely accepted methodological practices for increasing the accuracy of self-reports, such as encouraging privacy through audio computer-assisted self-interviewing (ACASI) and providing assurances that individual responses will remain confidential. Comparisons using these methods within NSDUH have shown that they reduce reporting bias (Gfroerer, Eyerman, & Chromy, 2002). Various procedures, such as biological specimens (e.g., urine, hair, saliva), proxy reports (e.g., family member, peer), and repeated measures (e.g., recanting), have been used to validate self-report data (Fendrich, Johnson, Sudman, Wislar, & Spiehler, 1999). However, these procedures often are impractical or too costly for general population epidemiological studies (SRNT Subcommittee on Biochemical Verification, 2002).

A recent study cosponsored by the Substance Abuse and Mental Health Services Administration (SAMHSA) and NIDA examined the validity of NSDUH self-report data on drug use among persons aged 12 to 25. The study found that it is possible to collect urine and hair specimens with a high response rate in a general population survey, and that most youths and young adults reported their recent drug use accurately (Harrison, Martin, Enev, & Harrington, 2007). However, there was some reporting of differences in either direction, with some respondents not reporting use but testing positive, and some reporting use but testing negative. Technical and statistical problems related to the hair tests precluded presenting comparisons of self-reports and hair test results, while small sample sizes for self-reports and positive urine test results for opiates and stimulants precluded drawing conclusions about the validity of self-reports of these drugs. Further, inexactness in the window of detection for drugs in biological specimens and biological factors affecting the window of detection could account for some inconsistency between self-reports and urine test results.

B.4 Measurement Issues

Several measurement issues associated with the 2006 NSDUH may be of interest and are discussed in this section. Specifically, these issues include the methods for measuring incidence, nicotine (cigarette) dependence, substance dependence and abuse, serious psychological distress (SPD), depression, methamphetamine use, and income.

B.4.1 Incidence

In epidemiological studies, incidence is defined as the number of new cases of a disease occurring within a specific period of time. Similarly, in substance use studies, incidence refers to the first use of a particular substance.

In the 2004 NSDUH national results report (Office of Applied Studies [OAS], 2005b), a new measure related to incidence was introduced, and since then it has become the primary focus of Chapter 5 in this national results report. The incidence measure is termed "past year initiation" and refers to respondents whose date of first use of a substance was within the 12 months prior to their interview date. This measure is determined by self-reported past year use, age at first use, year and month of recent new use, and the interview date.

Since 1999, the survey questionnaire has allowed for collection of year and month of first use for recent initiates (i.e., persons who used a particular substance for the first time in a given survey year). Month, day, and year of birth also are obtained directly or are imputed for item nonrespondents as part of the data postprocessing. Additionally, the questionnaire call record provides the date of the interview. By imputing a day of first use within the year and month of first use, a specific date of first use, t_fu,d,i, can be used for estimation purposes.

Past year initiation among persons using a substance in the past year can be viewed as an indicator variable defined as follows:

,     D

where DOI_i, MOI_i, and YOI_i denote the day, month, and year of the interview, respectively, and t_fu,d,i denotes the date of first use.

The calculation of this estimate does not take into account whether a respondent initiated substance use while a resident of the United States. This method of calculation has little effect on past year estimates and allows for direct comparability with other standard measures of substance use because the populations of interest for the measures will be the same (i.e., both measures examine all possible respondents and are not restricted to those initiating substance use only in the United States).

One important note for incidence estimates is the relationship between main categories and subcategories of substances (e.g., illicit drugs would be a main category, and inhalants and marijuana would be subcategories in relation to illicit drugs). For most measures of substance use, any member of a subcategory is by necessity a member of the main category (e.g., if a respondent is a past month user of a particular drug, then he or she is also a past month user of illicit drugs in general). However, this is not the case with regard to incidence statistics. Because an individual can only be an initiate of a particular substance category (main or sub) a single time, a respondent with lifetime use of multiple substances may not, by necessity, be included as a past year initiate of a main category, even if he or she were a past year initiate for a particular subcategory because his or her first initiation of other substances could have occurred earlier.

In addition to estimates of the number of persons initiating use of a substance in the past year, estimates of the mean age of past year first-time users of these substances are computed. Unless specified otherwise, estimates of the mean age at initiation in the past 12 months have been restricted to persons aged 12 to 49 so that the mean age estimates reported are not influenced by those few respondents who were past year initiates at age 50 or older. As a measure of central tendency, means are influenced heavily by the presence of extreme values in the data, and this constraint should increase the utility of these results to health researchers and analysts by providing a better picture of the substance use initiation behaviors among the civilian, noninstitutionalized population in the United States. This constraint was applied only to estimates of mean age at first use and does not affect estimates of incidence.

Because NSDUH is a survey of persons aged 12 years old or older at the time of the interview, younger individuals in the sample dwelling units are not eligible for selection into the NSDUH sample. Some of these younger persons may have initiated substance use during the past year. As a result, past year initiate estimates suffer from undercoverage if a user assumes that these estimates reflect all initial users instead of only for those above the age of 11. For earlier years, data can be obtained retrospectively based on the age at and date of first use. As an example, persons who were 12 years old on the date of their interview in the 2006 survey may report having initiated use of cigarettes between 1 and 2 years ago; these persons would have been past year initiates reported in the 2005 survey had persons who were 11 years old on the date of the 2005 interview been allowed to participate in the survey. Similarly, estimates of past year use by younger persons (age 10 or younger) can be derived from the current survey, but they apply to initiation in prior years and not the survey year.

To get an impression of the potential undercoverage in the current year, reports of substance use initiation reported in 2006 by persons aged 12 or older were estimated for the years in which these persons would have been 1 to 11 years younger. These estimates do not necessarily reflect behavior by persons 1 to 11 years younger in 2006. Instead, the data for the 11 year olds reflect initiation in the year prior to the 2006 survey, the data for the 10 year olds reflect behavior between the 12^th and 23^rd month prior to the 2006 survey, and so on. A very rough way to adjust for the difference in the years that the estimate pertains to without considering changes in the population is to apply an adjustment factor to each age-based estimate of past year initiates. This adjustment factor can be based on a ratio of lifetime users aged 12 to 17 in 2006 to the same estimate for the prior applicable survey year. To illustrate the calculation, consider past year use of alcohol. In the 2006 survey, 105,862 persons 12 years old in 2006 were estimated to have initiated use of alcohol between 1 and 2 years earlier. These persons would have been past year initiates in the 2005 survey conducted on the same dates had the 2005 survey covered younger persons. The estimated number of lifetime users currently aged 12 to 17 was 10,255,011 for 2006 and 10,305,889 for 2005, indicating fewer overall initiates of alcohol use among persons aged 17 or younger in 2006. Thus, an adjusted estimate of initiation of alcohol use by persons who were 11 years old in 2006 is given by

.     D

This yielded an adjusted estimate of 105,339 persons 11 years old on a 2006 survey date and initiating use of alcohol in the past year:

     D

A similar procedure was used to adjust the estimated number of past year initiates among persons who would have been 10 years old on the date of the interview in 2004 and for younger persons in earlier years. The overall adjusted estimate for past year initiates of alcohol use by persons 11 years of age or younger on the date of the interview was 268,883, or about 6.1 percent of the estimate based on past year initiation by persons 12 or older only (268,883 ÷ 4,381,000 = 0.0614).

Based on similar analyses, the estimated undercoverage of past year initiates was 5.4 percent for cigarettes, 1.7 percent for marijuana, and 27.7 percent for inhalants.

The undercoverage of past year initiates aged 11 or younger also affects the mean age at first use estimate. An adjusted estimate of the mean age at first use was calculated using a weighted estimate of the mean age at first use based on the current survey and the numbers of persons aged 11 or younger in the past year obtained in the aforementioned analysis for estimating undercoverage of past year initiates. Analysis results showed that the mean age at first use was changed from 16.6 to 16.1 (or a decrease of 2.7 percent) for alcohol, from 17.1 to 16.6 (or a decrease of 3.0 percent) for cigarettes, from 17.4 to 17.3 (or a decrease of 0.7 percent) for marijuana, and from 15.7 to 14.3 (or a decrease of 8.8 percent) for inhalants.

B.4.2 Nicotine (Cigarette) Dependence

The 2006 NSDUH computer-assisted interviewing (CAI) instrumentation included questions designed to measure nicotine dependence among current cigarette smokers. Nicotine dependence is based on criteria derived from the Nicotine Dependence Syndrome Scale (NDSS) (Shiffman, Hickcox, Gnys, Paty, & Kassel, 1995; Shiffman, Waters, & Hickcox, 2004) and the Fagerstrom Test of Nicotine Dependence (FTND) (Fagerstrom, 1978; Heatherton, Kozlowski, Frecker, & Fagerstrom, 1991). The above-mentioned criteria were first used to measure nicotine dependence in NSDUH in 2003.

The conceptual roots of the NDSS (Edwards & Gross, 1976) are similar to those behind the American Psychiatric Association (APA) Diagnostic and Statistical Manual of Mental Disorders, 4^th edition (DSM-IV), concept of dependence (APA, 1994). The 2006 NSDUH contained 19 NDSS questions that addressed five aspects of dependence:

1. Smoking drive (compulsion to smoke driven by nicotine craving and withdrawal)
   

1. After not smoking for a while, you need to smoke in order to feel less restless and irritable.
   
2. When you don't smoke for a few hours, you start to crave cigarettes.
3. You sometimes have strong cravings for a cigarette where it feels like you're in the grip of a force you can't control.
   
4. You feel a sense of control over your smoking - that is, you can "take it or leave it" at any time.
   
5. You sometimes worry that you will run out of cigarettes.
   

2. Nicotine tolerance
   

1. Since you started smoking, the amount you smoke has increased.
   
2. Compared to when you first started smoking, you need to smoke a lot more now in order to be satisfied.
   
3. Compared to when you first started smoking, you can smoke much, much more now before you start to feel anything.
   

3. Continuous smoking
   

1. You smoke cigarettes fairly regularly throughout the day.
   
2. You smoke about the same amount on weekends as on weekdays.
   
3. You smoke just about the same number of cigarettes from day to day.
   
4. It's hard to say how many cigarettes you smoke per day because the number often changes.
   
5. It's normal for you to smoke several cigarettes in an hour, then not have another one until hours later.
   

4. Behavioral priority (preferring smoking over other reinforcing activities)
   

1. You tend to avoid places that don't allow smoking, even if you would otherwise enjoy them.
   
2. There are times when you choose not to be around your friends who don't smoke because they won't like it if you smoke.
3. Even if you're traveling a long distance, you'd rather not travel by airplane because you wouldn't be allowed to smoke.

5. Stereotypy (fixed patterns of smoking)
   

1. Do you have any friends who do not smoke cigarettes?
   
2. The number of cigarettes you smoke per day is often influenced by other things - how you're feeling, or what you're doing, for example.
   
3. Your smoking is not affected much by other things. For example, you smoke about the same amount whether you're relaxing or working, happy or sad, alone or with others.
   

Each of the five domains listed above can be assessed by a separate measure, but an average score across all domains also can be obtained for overall nicotine dependence (Shiffman et al., 2004). The NDSS algorithm for calculating this average score was based on the respondent's answers to 17 of the 19 questions listed above. The two items regarding nonsmoking friends (4b and 5a) were excluded due to frequently missing data.

To optimize the number of respondents who could be classified for nicotine dependence, imputation was utilized for all respondents who answered all but 1 of the 17 nicotine dependence questions that were used in the NDSS algorithm. The imputation was based on weighted least square regressions using the other 16 NDSS items as covariates in the model (Aldworth et al., 2007a).

Responses to items 1a-c, 1e, 2a-c, 3a-c, 4a, 4c, and 5c were coded from 1 to 5 where

1 = Not at all true of me
2 = Sometimes true of me
3 = Moderately true of me
4 = Very true of me
5 = Extremely true of me

Responses to items 1d, 3d, 3e, and 5b were reverse coded from 5 to 1 where

5 = Not at all true of me
4 = Sometimes true of me
3 = Moderately true of me
2 = Very true of me
1 = Extremely true of me

The NDSS score was calculated as the sum of the responses to the previous questions divided by 17. The NDSS score was only calculated for current cigarette smokers who had complete data (based on actual reporting and imputation) for all 17 questions.

A current cigarette smoker was defined as nicotine dependent if his or her NDSS score was greater than or equal to 2.75. If the NDSS score for a current cigarette smoker was less than 2.75 or the NDSS score was not defined, then the respondent was determined to be nondependent based on the NDSS. The threshold of 2.75 was derived by examining the distribution of scores in other samples of smokers administered the NDSS, including a contrast of scores obtained for nondependent smokers (chippers) versus heavy smokers (Shiffman, Paty, Kassel, Gnys, & Zettler-Segal, 1994).

The FTND is a multi-item measure of dependence, but much of its ability to discriminate dependent smokers derives from a single item that assesses how soon after waking that smokers have their first cigarette (Heatherton, Kozlowski, Frecker, Rickert, & Robinson, 1989). Because most nicotine is cleared from the bloodstream overnight, smokers typically wake in nicotine deprivation, and rapid movement to smoke is considered a sign of dependence. A current cigarette smoker was defined as nicotine dependent based on the FTND if the first cigarette smoked was within 30 minutes of waking up on the days that he or she smoked.

Using both the NDSS and the FTND measures described above, a current cigarette smoker was defined as having nicotine dependence in the past month if he or she met either the NDSS or FTND criteria for dependence.

B.4.3 Illicit Drug and Alcohol Dependence and Abuse

The 2006 NSDUH CAI instrumentation included questions that were designed to measure dependence on and abuse of illicit drugs and alcohol. For these substances,¹⁰ dependence and abuse questions were based on the criteria in the DSM-IV (APA, 1994).

Specifically, for marijuana, hallucinogens, inhalants, and tranquilizers, a respondent was defined as having dependence if he or she met three or more of the following six dependence criteria:

1. Spent a great deal of time over a period of a month getting, using, or getting over the effects of the substance.
   
2. Used the substance more often than intended or was unable to keep set limits on the substance use.
   
3. Needed to use the substance more than before to get desired effects or noticed that the same amount of substance use had less effect than before.
   
4. Inability to cut down or stop using the substance every time tried or wanted to.
   
5. Continued to use the substance even though it was causing problems with emotions, nerves, mental health, or physical problems.
   
6. The substance use reduced or eliminated involvement or participation in important activities.
   

For alcohol, cocaine, heroin, pain relievers, sedatives, and stimulants, a seventh withdrawal criterion was added. A respondent was defined as having dependence if he or she met three or more of seven dependence criteria. The seventh withdrawal criterion is defined by a respondent reporting having experienced a certain number of withdrawal symptoms that vary by substance (e.g., having trouble sleeping, cramps, hands tremble).

For each illicit drug and alcohol, a respondent was defined as having abused that substance if he or she met one or more of the following four abuse criteria and was determined not to be dependent on the respective substance in the past year:

1. Serious problems at home, work, or school caused by the substance, such as neglecting your children, missing work or school, doing a poor job at work or school, or losing a job or dropping out of school.
   
2. Used the substance regularly and then did something that might have put you in physical danger.
   
3. Use of the substance caused you to do things that repeatedly got you in trouble with the law.
   
4. Had problems with family or friends that were probably caused by using the substance and continued to use the substance even though you thought the substance use caused these problems.
   

Criteria used to determine whether a respondent was asked the dependence and abuse questions included responses from the core substance use questions and the frequency of substance use questions, as well as the noncore substance use questions. Missing or incomplete responses in the core substance use and frequency of substance use questions were imputed. However, the imputation process did not take into account reported data in the noncore (i.e., substance dependence and abuse) CAI modules. Responses to the dependence and abuse questions that were inconsistent with the imputed substance use or frequency of substance use may have existed because different criteria and combinations of different criteria were used as skip logic for each substance.

For alcohol and marijuana, respondents were asked the dependence and abuse questions if they reported substance use on more than 5 days in the past year, or if they reported any substance use in the past year but did not report their frequency of past year use. Therefore, inconsistencies could have occurred where the imputed frequency of use response indicated less frequent use than required for respondents to be asked the dependence and abuse questions originally.

For cocaine, heroin, and stimulants, respondents were asked the dependence and abuse questions if they reported past year use in a core drug module or past year use in the noncore special drugs module. Thus, inconsistencies could have occurred when the response to a core substance use question indicated no use in the past year, but responses to dependence and abuse questions indicated substance dependence or abuse for the respective substance.

In 2005, there were two new questions added to the noncore special drugs module about past year methamphetamine use: "Have you ever, even once, used methamphetamine?" and "Have you ever, even once, used a needle to inject methamphetamine?" In 2006, there was an additional follow-up question added to the noncore special drugs module confirming prior responses about methamphetamine use: "Earlier, the computer recorded that you have never used methamphetamine. Which answer is correct?" The responses to these new questions were used in the skip logic for the stimulant dependence and abuse questions. Based on the decisions made during the methamphetamine analysis (see Section B.4.6), respondents who indicated past year methamphetamine use solely from these new special drug use questions (i.e., did not indicate methamphetamine use from the core drug module or other questions in the special drugs module) were categorized as NOT having past year stimulant dependence or abuse. Furthermore, if these same respondents were categorized as not having past year dependence on or abuse of any other substance (e.g., pain relievers, tranquilizers, or sedatives for the psychotherapeutic drug grouping), then they were categorized as NOT having past year dependence on or abuse of psychotherapeutics, illicit drugs, illicit drugs or alcohol, and illicit drugs and alcohol.

Respondents might have provided ambiguous information about past year use of any individual substance, in which case these respondents were not asked the dependence and abuse questions for that substance. Subsequently, these respondents could have been imputed to be past year users of the respective substance. In this situation, the dependence and abuse data were unknown; thus, these respondents were classified as not dependent on or abusing the respective substance. However, such a respondent never actually was asked the dependence and abuse questions.

B.4.4 Serious Psychological Distress

For this 2006 NSDUH report, serious psychological distress (SPD) was measured using the K6 screening instrument for nonspecific psychological distress (Kessler et al., 2003a). In NSDUH reports prior to 2004, the K6 scale was used to measure serious mental illness (SMI). For a discussion of the reasons that the K6 was used to measure SPD instead of SMI for the 2004 and later NSDUH reports, as well as details on a methodological study of the measurement of SMI, see Section B.4.4 of Appendix B in the 2004 NSDUH national results report (OAS, 2005b).

The K6 consists of six questions that ask respondents how frequently they experienced symptoms of psychological distress during the 1 month in the past year when they were at their worst emotionally. The use of this scale for SPD (or SMI prior to 2004) was based on a methodological study designed to evaluate several screening scales for measuring SMI in NSDUH. These scales evaluated in this methodological study consisted of a truncated version of the World Health Organization (WHO) Composite International Diagnostic Interview Short Form (CIDI-SF) scale (Kessler, Andrews, Mroczek, Üstün, & Wittchen, 1998), the K10/K6 scale of nonspecific psychological distress (Kessler et al., 2003a), and a truncated version of the WHO Disability Assessment Schedule (WHO-DAS) (Rehm et al., 1999). Overall, the K6 scale exhibited sound psychometric properties.

The six questions comprising the K6 scale are given as follows:

DSNERV1

Most people have periods when they are not at their best emotionally. Think of 1 month in the past 12 months when you were the most depressed, anxious, or emotionally stressed. If there was no month like this, think of a typical month.

During that month, how often did you feel nervous?

1     All of the time
2     Most of the time
3     Some of the time
4     A little of the time
5     None of the time
DK/REF
Response categories are the same for the following questions:

DSHOPE

During that same month when you were at your worst emotionally…how often did you feel hopeless?

DSFIDG

During that same month when you were at your worst emotionally…how often did you feel restless or fidgety?

DSNOCHR

During that same month when you were at your worst emotionally…how often did you feel so sad or depressed that nothing could cheer you up?

DSEFFORT

During that same month when you were at your worst emotionally…how often did you feel that everything was an effort?

DSDOWN

During that same month when you were at your worst emotionally…how often did you feel down on yourself, no good, or worthless?

To create a score, the six items (DSNERV1, DSHOPE, DSFIDG, DSNOCHR, DSEFFORT, and DSDOWN) on the K6 scale were coded from 0 to 4 so that "all of the time" was coded 4, "most of the time" 3, "some of the time" 2, "a little of the time" 1, and "none of the time" 0, with "don't know" and "refuse" also coded 0. Summing across the transformed responses resulted in a score with a range from 0 to 24. Respondents with a total score of 13 or greater were classified as having past year SPD (or SMI prior to 2004). This cut point was chosen to equalize false positives and false negatives.

In the 2003 NSDUH, the mental health module (i.e., the serious mental illness module) contained a truncated version of the CIDI-SF scale, the K10/K6 scale, and a truncated version of the WHO-DAS scale (in this order) to mirror the questions used by Kessler et al. (2003a). Thus, the module contained a broad array of questions from the CIDI-SF about mental health (i.e., panic attacks, depression, mania, phobias, generalized anxiety, posttraumatic stress disorder, and use of mental health services) that preceded the K6 items, and the four extra questions in the K10 scale were interspersed among the items in the K6 scale. In the 2004 NSDUH, the sample of respondents 18 or older was split evenly between the "long form" module, which included all items in the mental health module used in the 2003 NSDUH (sample A), and a "short form" module consisting only of the K6 items (sample B). The "short form" version was introduced to reduce interview time, removing questions that were not needed for estimation of SPD, and to provide space for a new module on depression. Inclusion of the "long form" version in half of the sample was to measure the impact on the K6 responses of changing the context of the K6.

Results from the 2004 NSDUH showed large differences between the two samples in both the K6 total score and the proportion of respondents with a K6 total score of 13 or greater. These differences were most pronounced in the 18 to 25 age group. These contextual differences suggest that the K6 scale is sensitive to item ordering in relation to other questions in the module; that is, respondents appear to respond to the K6 items differently depending on whether the scale is preceded by a broad array of other mental health questions.

Given the difference in K6 reporting between the A (long form) and B (short form) samples, the 2004 SPD estimates presented in the 2004 detailed tables and 2004 NSDUH national results report are based only on the A sample, which used a mental health module identical to that used in 2002 and 2003. In the 2005 and 2006 NSDUHs, only the "short form" SPD module was used; therefore, the 2004 SPD estimates presented in the 2005 and 2006 detailed tables and in the corresponding NSDUH national results reports are based on the B sample, so that the estimates are comparable. Note that the 2004 SPD estimates reported in the 2004 detailed tables (OAS, 2005a) are different from the 2004 SPD estimates reported in the 2005 and 2006 detailed tables (OAS, 2006a, 2007a), and SPD estimates reported in the 2005 and 2006 detailed tables are not comparable with estimates reported in previous years.

B.4.5 Major Depressive Episode

Beginning in 2004, modules related to major depressive episode (MDE) derived from DSM-IV (APA, 1994) criteria for major depression were included in the questionnaire. These questions permit estimates to be calculated of the lifetime and past year prevalence of MDE and treatment for MDE. Separate modules were administered to adults aged 18 or older and adolescents aged 12 to 17. The adult questions were adapted from the depression section of the National Comorbidity Survey–Replication (NCS-R; Harvard School of Medicine, 2005), and the adolescent questions were adapted from the depression section of the National Comorbidity Survey–Adolescent (NCS-A; Harvard School of Medicine, 2005). To make the modules developmentally appropriate for adolescents, there are minor wording differences in a few questions between the adult and adolescent modules. Revisions to the questions in both modules were made primarily to reduce its length and to modify the NCS questions, which are interviewer-administered, to the ACASI format used in NSDUH. In addition, some revisions, based on cognitive testing, were made to improve comprehension. Furthermore, even though titles similar to those used in the NCS were used for the NSDUH modules, the results of these items may not be directly comparable. This is mainly due to differing modes of administration in each survey (ACASI in NSDUH vs. computer-assisted personal interviewing [CAPI] in NCS), revisions to wording necessary to maintain the logical processes of the ACASI environment, and possible context effects resulting from deleting questions not explicitly pertinent to severe depression.

In 2004, a split-sample design was implemented where adults in sample B received the depression module while adult respondents in sample A did not. All adolescents were administered the adolescent depression module. In 2005 and 2006, all adult and adolescent respondents were administered their respective depression modules.

According to DSM-IV, a person is defined as having had MDE in his or her lifetime if he or she has had at least five or more of the following nine symptoms nearly every day in the same 2-week period, where at least one of the symptoms is a depressed mood or loss of interest or pleasure in daily activities (APA, 1994): (1) depressed mood most of the day; (2) markedly diminished interest or pleasure in all or almost all activities most of the day; (3) significant weight loss when not sick or dieting, or weight gain when not pregnant or growing, or decrease or increase in appetite; (4) insomnia or hypersomnia; (5) psychomotor agitation or retardation; (6) fatigue or loss of energy; (7) feelings of worthlessness; (8) diminished ability to think or concentrate or indecisiveness; and (9) recurrent thoughts of death or suicidal ideation. In addition to lifetime MDE, NSDUH measures past year MDE. Respondents who have had MDE in their lifetime are asked if, during the past 12 months, they had a period of depression lasting 2 weeks or longer while also having some of the other symptoms mentioned.

NSDUH measures the nine attributes associated with MDE as defined in DSM-IV with the following questions. Note that the questions shown are taken from the adult depression module. A few of the questions in the adolescent module were modified slightly to use wording more appropriate for youths. It should be noted that no exclusions were made for MDE caused by medical illness, bereavement, or substance use disorders.

1. Depressed mood most of the day

The following questions refer to the worst or most recent period of time when the respondent experienced any or all of the following: sadness, discouragement, or lack of interest in most things.

During that [worst/most recent] period of time…

1. … did you feel sad, empty, or depressed most of the day nearly every day?
   
2. … did you feel discouraged about how things were going in your life most of the day nearly every day?
   

2. Markedly diminished interest or pleasure in all or almost all activities most of the day

1. … did you lose interest in almost all things like work and hobbies and things you like to do for fun?
   
2. … did you lose the ability to take pleasure in having good things happen to you, like winning something or being praised or complimented?
   

3. Weight

In answering the next questions, think about the [worse/most recent] period of time.

1. Did you have a much smaller appetite than usual nearly every day during that time?
   
2. Did you have a much larger appetite than usual nearly every day?
   
3. Did you gain weight without trying to during that [worst/most recent] period of time?
        a. … because you were growing?
        b. … because you were pregnant?
        c. How many pounds did you gain?
   
4. Did you lose weight without trying to?
        a. … because you were sick or on a diet?
        b. How many pounds did you lose?
   

4. Insomnia or hypersomnia

1. Did you have a lot more trouble than usual falling asleep, staying asleep, or waking too early nearly every night during that [worst/most recent] period of time?
   
2. During that [worst/most recent] period of time, did you sleep a lot more than usual nearly every night?
   

5. Psychomotor agitation or retardation

1. Did you talk or move more slowly than is normal for you nearly every day?
   
2. Were you so restless or jittery nearly every day that you paced up and down or couldn't sit still?
   

6. Fatigue or loss of energy

1. During that [worst/most recent] period of time, did you feel tired or low in energy nearly every day even when you had not been working very hard?
   

7. Feelings of worthlessness

1. Did you feel that you were not as good as other people nearly every day?
   
2. Did you feel totally worthless nearly every day?
   

8. Diminished ability to think or concentrate or indecisiveness

1. During that [worst/most recent] time period, did your thoughts come much more slowly than usual or seem confused nearly every day?
   
2. Did you have a lot more trouble concentrating than usual nearly every day?
   
3. Were you unable to make decisions about things you ordinarily have no trouble deciding about?
   

9. Recurrent thoughts of death or recurrent suicidal ideation

1. Did you often think about death, either your own, someone else's, or death in general?
   
2. During that period, did you ever think it would be better if you were dead?
   
3. Did you think about committing suicide?
   

B.4.6 Revised Estimates of Methamphetamine Use

A challenge in measuring nonmedical use of prescription drugs comes when those drugs begin to be produced illegally. Drugs that have been manufactured by legitimate pharmaceutical companies under government regulation may become popular drugs of abuse, stimulating illegal production. In particular, most methamphetamine that currently is used nonmedically in the United States is produced by clandestine laboratories within the United States or abroad rather than by the legitimate pharmaceutical industry. Questions on methamphetamine use in NSDUH are first asked in the stimulants module in the core section of the questionnaire in the context of questions about nonmedical use of prescription stimulants. Therefore, one concern in measuring methamphetamine use in NSDUH is that some methamphetamine users may fail to report use if they do not recognize the drug when it is presented in the prescription drug context.

To address this concern, new questions were added to the special drugs module in the noncore section of the 2005 NSDUH to capture information from respondents who may have used methamphetamine but did not recognize it as a prescription drug and therefore did not report use in the core stimulants module. These new noncore questions differed from the methamphetamine use questions asked in the core stimulants module by asking about methamphetamine use outside of the context of prescription drug use. The new questions also included more descriptive information relevant to this drug. Respondents who did not indicate in the core stimulants module that they had used methamphetamine were asked to respond to 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 questions about the last time they used methamphetamine, whether they ever injected methamphetamine with a needle, and (if applicable) the last time they used a needle to inject methamphetamine. Answers to these questions were used to classify respondents as lifetime (i.e., ever used), past year, or past month users.

Findings from the methamphetamine analysis section (Ruppenkamp, Davis, Kroutil, & Aldworth, 2006) of the 2005 NSDUH Methodological Resource Book (RTI International, 2007a) suggested that estimates of methamphetamine use based only on core data could be lower than the true population prevalence. However, larger estimates of methamphetamine use based on both core and noncore answers could be a partial artifact of asking a second set of questions only from persons who did not report use the first time. Repeating questions for any drug only to those who did not report use the first time could artificially increase the positive responses. Doing so only for methamphetamine could result in a disproportionate reporting of that drug relative to the others in the survey. In addition, because the respondents reporting methamphetamine use in the new questions essentially had contradicted their prior responses, some may have made mistakes in answering the new questions.

For these reasons, additional follow-up items were included beginning with the 2006 NSDUH. In particular, these items sought to identify respondents who had failed to report methamphetamine use in response to the earlier question in the core stimulants module because they may not have considered methamphetamine to be a prescription drug. The new items added in 2006 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 ABOVE ITEM ANSWERED AS 2] 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
   

Respondents who reported "some other reason" for not having reported methamphetamine use in the core stimulants module but indicated use in the noncore questions were asked to specify this other reason.

Findings showed that it would be important to use data from these new consistency check questions in further investigating how best to estimate the prevalence of methamphetamine use in NSDUH (Ruppenkamp et al., 2006). In particular, respondents who confirmed in the first new 2006 follow-up question that they never used methamphetamine should not be counted as "additional" methamphetamine users based on their report of methamphetamine use in the noncore special drugs module. In addition, respondents who reported that they "made a mistake" in answering the earlier question about methamphetamine use in the core stimulants module would not be counted in prevalence estimates. As noted above, allowing respondents a second chance to report methamphetamine use could inflate the estimates for this drug relative to estimates for other drugs for which respondents were not asked a second set of questions.

The majority of respondents who should be included in estimates of the prevalence of methamphetamine based on the noncore special drugs questions consisted of those who both (a) confirmed in the first question that they used methamphetamine and (b) indicated in the second follow-up question that they had not reported methamphetamine use in the core stimulants module because they did not think of methamphetamine as a prescription drug. A smaller group of respondents who confirmed methamphetamine use in the noncore special drugs module also should be retained as methamphetamine users for prevalence estimation because they specified other similar reasons why they may not have recognized methamphetamine in the context of the earlier questions in the core stimulants module. More detailed documentation of how these methamphetamine data were edited will be provided in a forthcoming section of the 2006 NSDUH Methodological Resource Book (RTI International, 2007b).

To assess the impact of the new methamphetamine use questions, weighted estimates from 2006 were generated and compared for two different scenarios: (1) only methamphetamine data from the core stimulant module from 2006, and (2) core methamphetamine data and new methamphetamine use variables that were added to the special drugs module in 2005 and 2006 (taking into account the additional follow-up questions in 2006). Comparisons were made for the lifetime, past year, and past month measures of methamphetamine use. Prevalence estimates for scenario 2 were greater than those using only the core methamphetamine data. For example, the lifetime prevalence estimates of methamphetamine use among persons aged 12 or older increased from 4.62 percent based only on core data to 5.77 percent for core plus noncore data. See the column labeled "2006" in Table B.6 for a comparison of estimates for 2006 based on these two scenarios.

The methamphetamine use estimates for 2006 that are presented in this report and in the detailed tables are based both on the original methamphetamine items in the core stimulant module and the methamphetamine items in the special drugs module. For the purpose of examining trends in nonmedical methamphetamine use, a Bernoulli stochastic imputation procedure was used in conjunction with the predictive mean neighborhoods (PMN) method (described in Section A.3.1 of Appendix A in this report) to generate comparable estimates for prior years (i.e., 2002 through 2005).¹¹ An explanation of this imputation procedure is presented later in this section. See Table B.6 for the resulting "adjusted" estimates of lifetime, past year, and past month methamphetamine use for 2002 through 2005.

The 2005 and 2006 surveys also contained questions on how past year methamphetamine users obtained the methamphetamine that they last used. Respondents who reported past year methamphetamine use in the core stimulant or the noncore special drugs modules were asked these questions about obtaining the methamphetamine they last used. To assess the impact of respondents being routed to these source questions from both locations, weighted estimates for 2006 were generated and compared for the following two scenarios: (1) respondents routed to the source of methamphetamine questions from the core stimulants module only, and (2) respondents routed to the source of methamphetamine questions from either the core stimulants module or the noncore special drugs module (principally because they did not consider methamphetamine to be a prescription drug). This assessment revealed that an adjustment would be needed in order to compare 2006 estimates with 2005 estimates.

The 2006 estimates presented in this report and in the detailed tables for how past year methamphetamine users obtained the methamphetamine they used the last time were based on answers from respondents who reported methamphetamine use in the original core stimulant items and those who reported use in the special drugs module (principally because they did not consider methamphetamine to be a prescription drug). To generate comparable estimates for 2005, the past year source of methamphetamine estimates were adjusted by using the Bernoulli stochastic-adjusted past year methamphetamine variable. See Table B.7 for 2005 and 2006 estimates based on the different estimation methods.

In this report, estimates of the prevalence of methamphetamine use are based on data from the core and noncore methamphetamine items in 2006 and the adjusted estimates for 2002 through 2005 using the methods outlined below. These estimates are not comparable with those presented in previous NSDUH reports. However, the estimates of the numbers of past year initiates of methamphetamine use shown in this report are based only on responses to the age and date at first use questions from respondents who reported methamphetamine use in the original core stimulants items and are comparable with those in prior NSDUH reports. This procedure was necessary because data on age at first use, which are necessary to identify initiates, were not collected for noncore methamphetamine users in 2006. Starting with the 2007 NSDUH, age at first use of methamphetamine and frequency of use of the drug are being collected for persons reporting methamphetamine use in the noncore special drugs module and the core stimulant module.

Changes in estimates of methamphetamine use have the potential to affect estimates of nonmedical use of stimulants, nonmedical use of psychotherapeutic drugs, use of illicit drugs, and use of illicit drugs other than marijuana. The methamphetamine analysis reported in the forthcoming 2006 NSDUH Methodological Resource Book (RTI International, 2007b) revealed only negligible differences between core-only and core-plus-noncore estimates of use of illicit drugs or illicit drugs other than marijuana. Somewhat larger differences were found for estimates of nonmedical use of stimulants and psychotherapeutic drugs. No adjustment was made to these indicators in the present report pending the availability of further information on noncore methamphetamine users from 2007 and subsequent years (e.g., age at first use). Because full information on the new methamphetamine estimates is not yet available, methamphetamine estimates are not shown in the main tables in Appendix G.

The imputation-revised versions of the "core and noncore" methamphetamine recency variables were created by a complex combination of two imputation methods: predictive mean neighborhoods (PMN) and Bernoulli stochastic imputation (BSI). For a particular survey year, if the questionnaire covered the variable in question, then PMN was used to provide an imputation-revised version of that variable; otherwise, BSI was used. Core recency and lifetime variables were already imputed by methodologies discussed in Section A.3.1 of Appendix A in this report. Exhibit B.1 serves as a road map to the imputation methods used for the different variables in different survey years. Following standard NSDUH imputation procedures, lifetime use was imputed first, followed by recency.

The PMN and BSI methods are described briefly here. For step-by-step details on how the methods were applied, see the forthcoming methamphetamine analysis section in the 2006 NSDUH Methodological Resource Book (RTI International, 2007b).

The PMN method, which is used for most variables in NSDUH that undergo imputation, consists of a modeling step and a hot-deck step. During modeling, a neighborhood of potential donors is chosen for each item nonrespondent, and a final donor is randomly selected from that neighborhood. The neighborhood is formed by applying constraints to the set of item respondents; some of the constraints are based on predicted means from regression models. In the hot-deck step, the final donor is chosen so that its predicted mean(s) is (are) close to the predicted mean(s) of the item nonrespondent. For more information, see Section A.3.1 of Appendix A in this report.

BSI is a simpler version of PMN and can be used when the variable of interest is (1) dichotomous and (2) imputed on its own, not as part of a multivariate framework in which multiple variables need to be imputed simultaneously for consistency. As in PMN, logistic regression models are fit and predicted means are calculated. However, no neighborhoods are formed with BSI, and there is no hot-deck step. Once the predicted mean for the item nonrespondent is calculated, the imputation-revised value for the item nonrespondent is stochastically computed as follows: It is given the value of 1 with probability , and the value of 0 with probability 1 – .

As applied to these measures of methamphetamine prevalence, the data used to build the BSI regression models for the years when the relevant noncore variables were not collected came from the survey years when these items were collected. The PMN imputation was done for the survey years when the relevant variables were available. For example, 2006 data were used to build the model estimating the probability of noncore past year use given noncore lifetime use. Then, the parameter estimates from this model were used to calculate predicted means for each noncore lifetime user in the 2002-2005 survey years. Finally, these predicted means were used in the stochastic imputation of the noncore past year use variable for each noncore lifetime user in the 2002-2005 survey years.

Note that the BSI method is identical to the mean-centered univariate PMN imputation method for dichotomous variables.

B.4.7 Revised Income Questions

In the 2006 NSDUH, 3,847 (5.7 percent) of the sample of 67,802 respondents received a new reduced set of income questions designed to decrease the burden on respondents. Analyses were conducted to assess if the new questions had an effect on response variables representing personal income, family¹² income, and government assistance, relative to the old questions.

In the original income module, 10 source-of-income variables were included: Social Security, Supplemental Security Income, welfare cash assistance, welfare noncash assistance, wages, food stamps, child support, interest/investment income, other income, and the number of months receiving welfare. If a household contained other family members, then separate questions were asked to ascertain personal-level responses and other-family-level responses. These responses then were combined to create family-level responses.

The new set of income questions included only 6 of the 10 source-of-income variables; questions covering Social Security, child support, interest/investment income, and other income were omitted. In addition, separate questions to ascertain personal-level and other-family-level responses were no longer asked; all questions were asked at the family level only.

In both sets of income questions, personal and family-level questions were asked about actual annual income received at two levels of refinement.¹³

The respondents receiving the new income questions in 2006 consisted of two groups: (1) 2,050 were drawn from the 16,602 respondents in the first quarter, and (2) 1,797 were drawn from the 3,634 respondents who were assigned to a reliability study conducted within the main survey in the second, third, and fourth quarters. One difference between these two groups was the within-household sampling algorithm used to select respondents. In the main survey, respondents were selected according to an algorithm that allowed selection of 0, 1, or 2 persons in all households, but in the reliability study, respondents were restricted to those households in which only 1 person was selected.

An initial analysis was done to see whether the two groups needed to be analyzed in combination or separately. Using data from the 2004 NSDUH, it was shown that the two groups differed not only in the number of persons selected, but also in the number of persons eligible within a household. In the 2004 NSDUH, households with only one person eligible made up 8.7 percent of all households, but that percentage increased to 23.5 percent among households in which only one person was selected. Analyses of the 2004 survey on income and poverty variables, government assistance variables, and health insurance variables suggested that with some exceptions, the number selected within a household did not have much impact on the variables in question. However, these variables were greatly affected by whether one or more than one person in the household was eligible. Because the selection algorithm in the 2004 and 2006 NSDUHs is identical, these general conclusions are unlikely to differ in the 2006 NSDUH. Therefore, subsequent analyses dealing with the new income questions in the 2006 NSDUH needed to take into account that (1) household composition (in terms of number eligible) was likely to differ between the two groups of respondents, and (2) household composition was likely to have an effect on the income and related response variables of interest.

Analyses were conducted on the 2006 data to measure whether the new questions, relative to the old questions, had an effect on response variables representing personal income, family income, and government assistance. Results of the analyses suggested that the new income questions did not affect the reporting of personal income, family income, or government assistance response variables (except Supplemental Security Income). Based on subsequent analyses of the Supplemental Security Income variable, a decision was made to only reintroduce in a 2007 split sample questions about Social Security to the 2006 subset of six source-of-income variables because its omission in the 2006 survey appears to have caused some respondents to confuse Supplemental Security Income with Social Security. This revised module is expected to be fully implemented in 2008.

Simulation analyses were conducted on the 2005 data to measure the potential impact on imputation modeling procedures and imputation-revised estimates due to the new income questions. The simulation analyses indicated that the impact on imputation modeling procedures would be small and the impact on imputation-revised estimates would be negligible.

Finally, an analysis of the audit trail timing data from 2006 indicated that the mean time for all respondents to complete the income questions was reduced from 4.7 minutes for the old module to 3.7 minutes for the new module, and the median time was reduced from 4.2 to 3.2 minutes. Thus, the new income questions save about 1 minute of interview time in the 2006 and future NSDUHs. For further details, refer to the forthcoming 2006 NSDUH's new income questions analysis section included in the 2006 NSDUH Methodological Resource Book (Aldworth, Copello, Heller, Liu, & Robbins, 2007b).

B.5 Impact of Hurricanes Katrina and Rita on the NSDUH Sample

Hurricanes Katrina and Rita hit the Gulf Coast in the fall of 2005. At the end of August 2005, Hurricane Katrina caused large-scale damage and destruction in the coastal regions of Louisiana, Mississippi, and Alabama. In September 2005, Hurricane Rita devastated portions of Texas and Louisiana. The impact of the hurricanes on the NSDUH sample was evaluated, and a plan of action was developed and implemented for the 2006 survey.

The 2006 NSDUH quarter 1 (January to March) sample was supplemented with an additional segment in the seven areas determined to be hardest hit by the hurricanes. As a result, a total of 7,207 segments were fielded in the 2006 survey. In addition to supplementing the quarter 1 sample, field staff were reminded to apply standard procedures to handle unusual situations. Specifically, field staff were instructed to apply the residency rule for eligibility¹⁴ and to include displaced persons wherever they currently were residing. Finally, temporary housing units were included in the survey by applying the half-open interval rule.¹⁵ For more details on the 2006 sample supplement, see Morton et al. (2007).

Table B.2 Summary of 2006 NSDUH Suppression Rule

Estimate	Suppress if:
SE = standard error; RSE = relative standard error; deff = design effect. Source: SAMHSA, Office of Applied Studies, National Survey on Drug Use and Health, 2006.
Prevalence Rate, , with Nominal Sample Size, n, and Design Effect, deff	(1) The estimated prevalence rate, , is < 0.00005 or ≥ 0.99995, or (2)  > 0.175 when ≤ 0.5, or     D        > 0.175 when > 0.5, or     D (3) Effective n < 68, where Effective or     D (4) n < 100. Note: The rounding portion of this suppression rule for prevalence rates will produce some estimates that round at one decimal place to 0.0 or 100.0 percent but are not suppressed from the tables.
Estimated Number (Numerator of )	The estimated prevalence rate, , is suppressed. Note: In some instances when is not suppressed, the estimated number may appear as a 0 in the tables. This means that the estimate is greater than 0 but less than 500 (estimated numbers are shown in thousands).
Mean Age at First Use, , with Nominal Sample Size, n	(1) RSE > 0.5, or (2) n < 10.

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.