Analyses of Substance Abuse and Treatment Need Issues

Changes Affecting NHSDA Estimates of Treatment Need for 1994-1996

• 1991-93 Treatment Need Estimates
• Questionnaire Changes Between 1991 and 1996 Affecting Treatment Need Components
• Developing a New Treatment Need Algorithm for 1994-1996
• Description of the New Treatment Need Algorithm for 1994-B - 1996
• Estimates of Treatment Need
• Table 1:Estimates of Number of Persons Needing and Receiving Treatment for Drug Abuse Problems: NHSDA 1991-96
• Appendix 1 DSM-III-R criteria for substance dependence and corresponding questions in the 1991-1993 NHSDA that represent them.
• Appendix 2 DSM-III-R criteria for substance dependence and corresponding questions from the 1994 NHSDA to represent them.
• Appendix 3 DSM-IV criteria for substance dependence and corresponding questionnaire items used in the 1995 and 1996 NHSDA to represent them.
• Appendix 4 DSM-IV Criteria for Substance Dependence and Corresponding Questions in the 1994-B NHSDA used to cover the DSM-IV criteria.
• Appendix 5 Prevalence Estimates of Individual DSM Criteria for 1994 and 1995 by Drug
• Appendix 5 cont’d
• Appendix 5 cont’d
• Table 1:Total Population Age 12 and Older with Drug Abuse Treatment Need, by Year and Demographic Characteristics, in Thousands
• Table 2:Percent of Population Age 12 and Older with Drug Abuse Treatment Need, by Year and Demographic Characteristics
• Table 3:Total Population Age 12 and Older with Level 2 Drug Abuse Treatment Need, by Year and Demographic Characteristics, in Thousands
• Table 4:Percent of Population Age 12 and Older with Level 2 Drug Abuse Treatment Need, by Year and Demographic Characteristics
• Table 5:Total Population Age 12 and Older Dependent on Any Illicit Drug by Year and Demographic Characteristics, in Thousands
• Table 6:Percent of Population Age 12 and Older Dependent on any Illicit Drug by Year and Demographic Characteristics
• Table 7: Estimates of Number of Persons Needing Treatment in Thousands: 1991-1996
• Appendix 7 Estimates and Standard Errors of the Number of Persons Needing Treatment for Drug Abuse Problems: 1994B-96

8. Changes Affecting NHSDA Estimates of Treatment Need for 1994-1996

By Joan Epstein and Joseph Gfroerer

The Office of Applied Studies (OAS) has used the National Household Survey on Drug Abuse (NHSDA) to produce estimates of the severity of substance abuse in the population. Between 1991 and 1993 estimates of dependence were based on questions designed to approximate the DSM-III-R definitions of drug dependence, while estimates of the need for treatment were based on these dependence questions in conjunction with questions that explore how drugs are used and recent treatment.

Beginning in 1994 modifications were made in the NHSDA that involved both changes in questions and editing procedures. Because of these changes, it was necessary for OAS to adjust the procedures for estimating dependence and treatment need. This paper describes the changes to the questionnaire and editing that affect the estimates of treatment need and the new algorithms for estimating treatment need that are being used for the 1994-1996 NHSDAs. A brief discussion of the logic of using these algorithms is also provided.

1991-93 Treatment Need Estimates

In the OAS algorithm for estimating treatment need from the surveys conducted in 1991, 1992, and 1993, persons with at least one of the following four conditions were defined as needing treatment at some time in the past year:

1. Drug Dependence

Dependent on any illicit drug in the past year based on an algorithm that follows the DSM-III-R definition. (See Appendix 1) The DSM-III-R defines a person as dependent for a substance if they meet 3 out of 9 criteria for that substance. The algorithm for estimating dependence combines items on symptoms and problems included in the NHSDA questionnaire to approximate 5 of the 9 DSM-III-R criteria for substance dependence. These five criteria are 2, 5, 6, 7, and 8. A person is defined as dependent if they meet 2 of these 5 criteria.

2. Heavy Drug Use

Any of following in the past year:

a. Used heroin at least once.

b. Used marijuana daily.

c. Frequent use (52+days/weekly) of some other illicit drug

3. Injection Drug Use

Administered heroin, cocaine or stimulants with a needle in the past year.

4. Treated for Drug Abuse

Received treatment for any illicit drug in the past year.

The OAS algorithm for estimating the number of persons needing treatment because of an especially severe drug problem (also referred to as level 2 drug abuse treatment need) was based on an individual having one of the following four conditions in the past year:

1. Drug Dependence

Dependent on any illicit drug except marijuana in the past year (defined the same as for total drug abuse treatment need).

2. Heavy Drug Use

Any of following in the past year:

a. Used heroin at least once.

b. Used marijuana daily AND dependent on marijuana (defined the same as for

total drug abuse treatment need).

c. Frequent cocaine use (52+days/weekly)

d. Daily use of any illicit drug except marijuana

3. Injection Drug Use

Administered heroin, cocaine or stimulants with a needle in the past year.

4. Treated for Drug Abuse at a Specialty Facility

Received treatment for any illicit drug at a specialty facility in the past year.

(A specialty facility includes treatment at a hospital (as an inpatient), a mental health center or drug treatment facility.)

Questionnaire Changes Between 1991 and 1996 Affecting Treatment Need Components

The new questionnaire adopted in 1994 contains both a core series of questions and a supplement. The core components remain constant from year to year while the supplemental part of the questionnaire can change from year to year. Several of the questions used to develop the estimates of treatment need are not from the core. Consequently, if there are changes to these questions, some inconsistency can occur in the treatment need estimates over time. Nevertheless, between 1994 and 1996 there were only minor differences in the questions which affect the treatment need estimates.

Of the four basic components in the treatment need definition, only heavy use of drugs other than heroin is based entirely on questions in the core part of the survey. In this case, the survey questions used to estimate frequency and recency of use with the new questionnaire were revised slightly in 1994. Between 1991 and 1993, the questions dealing with frequency of use focused on how often the drug was used in the past year. In the version of the survey fielded in 1994, the questions were changed to ask how many days in past 12 months the drug was used. Answer categories were also changed for some drugs in 1994. In the 1991 to 1993 version the questions on recency of use asked about the most recent use of drugs. The 1994 to 1996 version asked how long since last use. Fewer answer categories were used in the 1994 to 1996 version than in 1991 to 1993 version. Editing procedures for these core items also changed in 1994. Prior to 1994 all drug use data on the questionnaire was used in editing core drug use variables. In 1994 to 1996 data from core items were not edited to be consistent with data from non-core items since some non-core items changed from year to year.

In order to estimate the effects of changes in the questionnaire and variable construction, both the new and old versions of the survey were fielded in 1994. The "old" version of the survey, which is actually the same as the 1993 questionnaire is referred to as 1994-A, while the "new" version of the survey is referred to as 1994-B.

Comparisons between the new questionnaire (1994-B) and the old questionnaire (1994-A) have shown that estimates of the rate of lifetime illicit drug use from the new questionnaire are approximately 8% less than estimates from the old questionnaire. Estimates of the rate of past year illicit drug use are approximately 12% less with the new questionnaire. On the other hand, estimates of the rate of past month illicit drug use are approximately 5% higher with the new questionnaire. The new questionnaire produced approximately 29% higher estimates of the rate of weekly cocaine use and approximately 6% higher estimates of the rate of daily marijuana use.

The questions used to estimate dependence were changed significantly for the 1994-B survey. The questions used between 1991-1993 and the 1994-A surveys covered five of nine DSM-III-R criteria (See Appendix 1). The 1994-B dependence questions were originally designed to approximate the six of nine DSM-III-R criteria (See Appendix 2). However, for several reasons described below, they were actually used to approximate five of seven DSM-IV criteria. Although some of the criteria covered by the 1994-B dependence questions are the same as those in 1993 (DSM-III-R criteria 2, 6, and 7), the survey questions were different in both the way the questions are worded and the format in which they are asked. For example, the actual questions used to measure DSM-III-R criterion 6 (DSM-IV criterion 7) are different in the new than in the old questionnaire. (See Appendix 1 and 4). In addition, in the new questionnaire, for each drug, the respondent marks a box with a yes or no to indicate whether the drug caused the problem or symptom described by the question, while, in the old questionnaire, the respondent must list or circle the drugs that apply to an item.

In 1995, dependence questions were revised to approximate 6 of 7 DSM-IV criteria. (See Appendix 3)

Questions used to determine needle use and treatment also changed between the 1993 and 1994-B survey.

Developing a New Treatment Need Algorithm for 1994-1996

Several guiding principles were used in developing the new algorithm for estimating treatment need. The basic structure of the treatment need estimate has been retained, using the same four components in 1994-B - 1996 as used in 1991 - 1994-A. Specific decisions on which non-core items to separately use in the algorithm have been decided each year since 1994-B, with consideration given to consistency over time. Because the DSM-IV definition was the most recent definition of dependence and because the 1995-96 questions coincide more closely with the DSM-IV than with the DSM-III-R, the DSM-IV definition of dependence was adopted for the 1994-B - 1996 surveys. Analysis of 1994-B and 1995 data also suggested that comparable DSM-IV based estimates could be derived for 1994-B data (See Appendix 5). Although 1994-B to 1996 estimates will not be comparable to 1991-1993 estimates, it is possible to develop adjustment factors to make the 1991 to 1993 estimates comparable to 1994-B - 1996 estimates in order to describe trends.

Description of the New Treatment Need Algorithm for 1994-B - 1996

Under the new algorithm, the components for total treatment need are addressed as follows:

1. Dependence is defined as follows:

A person is identified as dependent on a specific drug in the 1994-B survey if he or she indicates having used that drug based on responses to questions in the core and the supplemental sections of the questionnaire and meets three of the five DSM-IV criteria covered by the 1994-B dependence questions; Appendix 4 shows the 1994-B dependence questions representing the five DSM-IV criteria.

A person is identified as dependent on a specific drug in the 1995-1996 surveys if he or she indicates having used that drug based on responses to questions in the core and the supplemental sections of the questionnaire and meet three of the six DSM-IV criteria covered by the dependence questions; Appendix 3 shows the dependence questions representing the six DSM-IV criteria used in the 1995-1996 NHSDAs. Note that in order to meet criterion 7, a person must report either emotional problems or physical problems.

2. Heavy Drug Use:

To be consistent with the way heavy drug use was defined previously, for the 1994-B - 1996 surveys, heavy drug use is defined using the same frequency of use cutoffs that were used for the 1991-1993 total treatment need estimates:

a. Used heroin at least once

b. Used marijuana daily

c. Frequent use (52+days/weekly) of some other illicit drug

Except for heroin, only questions from the core section of the questionnaire are used to determine heavy drug use. For heroin, a person is defined as using heroin in the past year if they report using heroin in the past year on any questions from the core part of the interview or if he or she reports using heroin in the past year from any of the following questions which are in the supplement portion of the questionnaire:

How long has it been since you last used a needle to inject heroin?

How long has it been since you last smoked heroin?

How long has it been since you last sniffed ("snorted") heroin powder through your nose?

3. Injection Drug Use:

Administered heroin, cocaine, or stimulants with a needle in the past year.

4. Treatment:

Received drug abuse treatment at any location in the past year.

Under the new algorithm the components for level 2 treatment need are addressed as follows:

1. Dependence is defined as follows:

Dependent on any illicit drug except marijuana during the past year, with dependence defined the same as for total treatment need in the new algorithm for 1994-B - 1996.

2. Heavy Drug Use:

To be consistent with the way level 2 treatment need was defined previously, for the 1994-B - 1996 surveys, heavy drug use is defined using the same frequency of use cutoffs that were used for the 1991-1993 level 2 treatment need estimates:

a. Used heroin at least once.

b. Used marijuana daily AND dependent on marijuana.

c. Frequent cocaine use (52+days/weekly)

d. Daily use of any illicit drug except marijuana

3. Injection Drug Use:

Administered heroin, cocaine, or stimulants with a needle in the past year. (Same as for total treatment need.)

4. Treatment:

Received treatment for any illicit drug at a specialty facility in the past year.

(A specialty facility includes treatment at a hospital (as an inpatient), a mental health center or drug treatment facility) (Same as for Level 2 treatment need defined for the 1991 to 1993 survey).

Estimates of Treatment Need

In Appendix 6, rates and totals are given by age, by race and by sex for 1991 to 1996 for total drug abuse treatment need, and level 2 drug abuse treatment need. Estimates for any illicit drug dependence are also given. Estimates of treatment need for 1991 to 1994-A are created using the old definition of total substance abuse treatment need. Estimates for 1994-B to 1996 are created using the new definition of treatment need. Estimates of totals that are adjusted to account for underestimation by the NHSDA due to underreporting and/or undercoverage of hard-core drug use are also included in Appendix 6. These 1991-1996 estimates include a ratio adjustment, which links NHSDA data on arrests and treatment with outside sources of data (primarily the Uniform Crime Reports and the National Drug and Alcohol Treatment Unit Survey (NDATUS)). Because data on arrests were not collected using the 1994-B questionnaire, 1994-B ratio adjustments could not be calculated using 1994-B data. To adjust the 1994-B estimates, overall ratio adjustments from 1995 were applied to the 1994-B estimates. In Appendix 7, standard errors are given for the ratio adjusted numbers of total drug abuse treatment need for 1994-B to 1996.

In addition to the ratio adjustment to account for underreporting and undercoverage, factors to adjust for the questionnaire change in 1994-B have been developed to enable measurement of trends from 1991 to 1996. The denominator of these factors is the estimate of treatment need using the old methodology and is an average of the treatment need estimates from the 1994-A NHSDA and the 1993 NHSDA. The numerator is the estimate of treatment need using the new methodology and is calculated using data from the 1994-B NHSDA. The adjustment factors for trend consistency are given in the note in the table below. These adjustment factors were applied to the 1991 to 1993 estimates, to make these estimates comparable to the 1994-96 estimates.

The ratio adjusted estimates of persons treated from the NHSDA, given in the table below, equal the independent estimates of treated persons based on the NDATUS (now known as the Uniform Facility Data Set (UFDS)). This is because the ratio adjustment procedure essentially reweights the NHSDA data on treated persons so that it is forced to equal the independent estimate from NDATUS/UFDS. It should also be mentioned that the independent estimates are calculated using a combination of NDATUS/UFDS data and data from the Drug Services Research Survey (DSRS). NDATUS/UFDS collects data on clients in treatment on a single day. These counts are inflated into annual estimates of persons treated based on DSRS estimates of length of stay in treatment and number of treatment episodes per year per client.

Table 1:Estimates of Number of Persons Needing and Receiving Treatment for Drug Abuse Problems: NHSDA 1991-96

Note:  Estimates for 1991-96 are ratio-adjusted to partially account for underestimation due to underreporting and undercoverage in the NHSDA. Estimates for 1991-93 are also adjusted for trend consistency, to account for the change in the NHSDA questionnaire in 1994. Adjustment factors for trend consistency were 1.19020 for total treatment need and 1.21125 for Level 2 treatment need.

Source:  Office of Applied Studies, SAMHSA. Unpublished data from the National Household Survey on Drug Abuse and Uniform Facility Data Set.

Appendix 1 DSM-III-R criteria for substance dependence and corresponding questions in the 1991-1993 NHSDA that represent them.

Appendix 2 DSM-III-R criteria for substance dependence and corresponding questions from the 1994 NHSDA to represent them.

Appendix 3 DSM-IV criteria for substance dependence and corresponding questionnaire items used in the 1995 and 1996 NHSDA to represent them.

Appendix 4 DSM-IV Criteria for Substance Dependence and Corresponding Questions in the 1994-B NHSDA used to cover the DSM-IV criteria.

Appendix 5 Prevalence Estimates of Individual DSM Criteria for 1994 and 1995 by Drug

Criteria 1-7 correspond to the criterion number in the DSM-IV. Criterion 8 is only in the 1993 DSM-III-R.

Appendix 5 cont’d

Criteria 1-7 correspond to the criterion number in the DSM-IV. Criterion 8 is only in the 1993 DSM-III-R.

Appendix 5 cont’d

Note:

1994-B 3 of 5 uses 1994 data to define a person as dependent for a specific drug in 1994 if they meet 3 of the 5 DSM-IV criteria covered by 6 of the 1994 dependence questions.

1995 3 of 5 uses 1995 data to defines a persons as dependent for a specific drug if they meet 3 of 5 DSM-IV criteria covered by the 1994 dependence questions

1995 3 of 6 uses 1995 data to define a person as dependent for a specific drug if they meet 3 of 6 DSM-IV criteria covered by the 7 1995 dependence questions.

1996 3 of 5 uses 1996 data to define a persons as dependent for a specific drug if they meet 3 of 5 DSM-IV criteria covered by the 1994 dependence questions

1996 3 of 6 uses 1996 data to define a person as dependent for a specific drug if they meet 3 of 6 DSM-IV criteria covered by the 7 1996 dependence questions.

In order to be dependent for alcohol a person must have taken 6 or more drinks in the past year.

Appendix 6

Table 1:Total Population Age 12 and Older with Drug Abuse Treatment Need, by Year and Demographic Characteristics, in Thousands

Table 2:Percent of Population Age 12 and Older with Drug Abuse Treatment Need, by Year and Demographic Characteristics

Note:  Estimates are not ratio-adjusted to partially account for underestimation due to underreporting and undercoverage in the NHSDA.

Source:  Office of Applied Studies, SAMHSA. National Household Survey on Drug Abuse, 1991-1996.

Appendix 6 cont’d

Table 3:Total Population Age 12 and Older with Level 2 Drug Abuse Treatment Need, by Year and Demographic Characteristics, in Thousands

Table 4:Percent of Population Age 12 and Older with Level 2 Drug Abuse Treatment Need, by Year and Demographic Characteristics

Note:  Estimates are not ratio-adjusted to partially account for underestimation due to underreporting and undercoverage in the NHSDA.

Source:  Office of Applied Studies, SAMHSA. National Household Survey on Drug Abuse, 1991-1996.

Appendix 6 cont’d

Table 5:Total Population Age 12 and Older Dependent on Any Illicit Drug by Year and Demographic Characteristics, in Thousands

Table 6:Percent of Population Age 12 and Older Dependent on any Illicit Drug by Year and Demographic Characteristics

Note:Estimates are not ratio-adjusted to partially account for underestimation due to under reporting and undercoverage in the NHSDA.

Source:Office of Applied Studies, SAMHSA. National Household Survey on Drug Abuse, 1991-1996.

Appendix 6 cont’d

Table 7: Estimates of Number of Persons Needing Treatment in Thousands: 1991-1996

Note:Estimates are ratio-adjusted to partially account for underestimation due to under reporting and undercoverage in the NHSDA.

Source:Office of Applied Studies, SAMHSA. National Household Survey on Drug Abuse, 1991-1996.

Appendix 7 Estimates and Standard Errors of the Number of Persons Needing Treatment for Drug Abuse Problems: 1994B-96

Note:Estimates are ratio-adjusted to partially account for underestimation due to underreporting and undercoverage in the NHSDA.

Source:Office of Applied Studies, SAMHSA. Unpublished data from the National Household Survey on Drug Abuse and Uniform Facility Data Set.

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.