Nonrespondents

The NESARC-III target population was the US noninstitutionalized civilian population 18 years or older, including residents of selected group quarters (eg, group homes, workers’ dormitories). As detailed elsewhere,¹⁹ we used probability sampling to select respondents randomly. Primary sampling units were individual counties or groups of contiguous counties; secondary sampling units, groups of US Census–defined blocks; and tertiary sampling units, households within the secondary sampling units. Finally, eligible adults within sampled households were randomly selected. We oversampled Hispanic, black, and Asian respondents, and in households with at least 4 eligible individuals who were ethnic or racial minorities, 2 respondents were selected (n = 1661). The total sample size consisted of 36 309 respondents. The screener- and person-level response rates were 72.0% and 84.0%, respectively, yielding a total NESARC-III response rate of 60.1%, comparable to those of most current US national surveys.^{16 ,20} Data were collected from April 2012 through June 2013 and analyzed in October 2014.
Data were adjusted for oversampling (including selection of 2 persons in selected households) and screener- and person-level nonresponse, then weighted through poststratification analyses to represent the US civilian population based on the American Community Survey 2012.²¹ Table 1 shows the weighted distribution of the NESARC-III population characteristics. These weighting adjustments were found to compensate adequately for nonresponse. When participants were compared with the total eligible sample, including nonrespondents, no significant differences were found in the percentages of Hispanic, black, or Asian respondents, population density, vacancy rate, proportion of the population in group quarters, or proportion of renters at the segment level. At the individual level, we found no differences in Hispanic ethnicity between respondents and the total eligible sample. Respondents included a slightly higher percentage of men (48.1% vs 46.2%), a greater percentage of those aged 60 to 69 years (13.7% vs 12.6%), and smaller percentages of those aged 40 to 49 (18.1% vs 18.3%) and 30 to 39 (16.7% vs 17.4%) years than in the eligible sample.
Interviewer field methods, detailed elsewhere,¹⁹ involved initial structured home study, in-person training, ongoing supervision, and random respondent callbacks to verify data. Oral informed consent was electronically recorded, and respondents received $90.00 for survey participation. Protocols were approved by the institutional review boards of the National Institutes of Health and Westat (the contractor for the NESARC-III).

Three authors (L.S.R., M.T., and J.B.S.) independently extracted the following data from each article using a standardized form: study design; geographic location; years of survey; year in school; sample size; average age of participants; number and percentage of male participants; diagnostic or screening method used; outcome definition (ie, specific diagnostic criteria or screening instrument cutoff); and reported prevalence estimates of depression, depressive symptoms, or suicidal ideation. Whether students who screened positive for depression sought psychiatric or other mental health treatment also was extracted. When there were studies involving the same population of students, only the most comprehensive or recent publication was included.
The same 3 authors independently assessed the risk of bias of these nonrandomized studies using a modified version of the Newcastle-Ottawa scale, which assesses sample representativeness and size, comparability between respondents and nonrespondents, ascertainment of depressive or suicidal symptoms, and thoroughness of descriptive statistics reporting (complete details regarding scoring appear in eMethods 2 in the Supplement).^{17 (link)} Studies were judged to be at low risk of bias (≥3 points) or high risk of bias (<3 points). A fourth author (D.A.M.) resolved discrepancies through discussion and adjudication.