National Institute of Mental Health (NIMH)

Statistical Methods and Measurement Caveats

National Survey on Drug Use and Health (NSDUH)

Population:

• NSDUH participants are representative of the civilian, non-institutionalized population aged 12 years old or older residing within the United States.
• The survey covers residents of households (persons living in houses/townhouses, apartments, condominiums; civilians living in housing on military bases, etc.) and persons in non-institutional group quarters (e.g., shelters, rooming/boarding houses, college dormitories, migratory workers’ camps, and halfway houses).
• The survey does not cover persons who, for the entire year, had no fixed address (e.g., persons experiencing homelessness and/or transient persons not in shelters); were on active military duty; or who resided in institutional group quarters (e.g., correctional facilities and hospitals ).
• Data regarding sex of the respondent was assessed using male and female categories only. Gender identity information was not collected in the survey.

Interview Response and Completion:

• In 2022, 52.0% of the selected NSDUH sample of people 18 or older did not complete the interview. This rate of non-response is higher than in years before 2020. Please see the Background on the 2022 NSDUH and the COVID-19 Pandemic section below for more information.
• Reasons for non-response to interviewing include the following: refusal to participate (29.3%); respondent unavailable, never at home, or did not respond to the web survey (18.2%); and other reasons such as partially completed but unusable interviews, physical/mental incompetence or language barriers (4.5%).
• People with suicidal behavior may disproportionately fall into these non-response categories. While NSDUH weighting includes non-response adjustments to reduce bias, these adjustments may not fully account for differential non-response by suicide behavior status. Prior to the 2020 NSDUH, this bias was deemed small and inconsequential due to low rates of item nonresponse and low prevalence estimates for outcomes imputed using the zero-fill method. With the increase in break-offs among adults in 2020 who completed the questionnaire via the web, the potential bias of this approach for handling missing data became of greater concern. Therefore, missing values in the variables associated with multiple outcomes, including for suicidal thoughts and behaviors, were statistically imputed beginning in 2021.

Data Suppression:

• For some groups, data are not reported due to low precision. Data may be suppressed in the above charts if the data do not meet acceptable ranges for prevalence estimates, standard error estimates, and sample size.

Background on the 2022 NSDUH and the COVID-19 Pandemic:

• Data collection methods for the 2022 NSDUH changed in several ways because of the COVID-19 pandemic: the 2022 NSDUH continued the use of multimode data collection procedures (both in-person and virtual data collection) that were first implemented in the fourth quarter of the 2020 NSDUH. Overall, 40.7% of interviews were completed via the web, and 59.3% were completed in person. For comparison, more than half of interviews in 2021 (54.6%) were completed via the web. In 2022, the weighted response rates for household screening and for interviewing were 25.5% and 47.4%, respectively, for an overall response rate of 12.1% for people aged 12 or older. Given the use of multimode data collection procedures through the entirety of the collection year and the rate of non-response, comparison of estimates from the 2022 NSDUH with those from prior years must be made with caution.

Please see the 2022 National Survey on Drug Use and Health Methodological Summary and Definitions report  for further information on how these data were collected and calculated.