The NSDUH uses cross-sectional designs to produce national drug use estimates. Our results reflect associations, and they cannot be applied to about 2% of institutionalized or homeless individuals that are not covered by the survey’s sampling plan. All drug use estimates are based on self-reports, which may be influenced by recall or reporting errors. The NSDUH also has strengths. One unique strength is about its use of the same standardized assessments of CUD during the studied years to examine CUD over time. The NSDUH also implements consistency checks in data management to ensure data quality, and develops statistical computation and analysis weights to minimize response inconsistency and nonresponse bias (Gfroerer et al., 2002; Harrison et al., 2007).
