Objective: To compare depression prevalence estimates measured by the 8-item Patient Health Questionnaire (PHQ-8) with two administration modes in two national surveys.
Methods: Data on adults aged 18 years and older who participated in the 2006 Behavioral Risk Factor Surveillance System (BRFSS) (n = 198,678) and those who participated in the 2005-2006 National Health and Nutrition Examination Survey (NHANES) (n = 4,800) were analyzed.
Results: The crude PHQ-8 depression prevalence estimate using the diagnostic algorithm was higher in BRFSS with computer-assisted telephone interviewing (CATI) (9.16%, SE 0.15) than in NHANES with computer-assisted personal interviewing (CAPI) (6.28%, SE 0.59) (P < 0.001). After adjustment for demographic characteristics, the difference in the prevalence estimates remained (9.68% in BRFSS vs. 6.13% in NHANES, P < 0.001). Similar differences in the depression prevalence estimates using the PHQ-8 cutoff score ≥10 were detected (9.22% in BRFSS vs. 5.15% in NHANES, P < 0.001). Significant differences in the depression prevalence persisted in subgroups stratified by demographic characteristics and major health risk factors and outcomes.
Conclusions: The PHQ-8 administered by CATI yielded about 3.5% higher depression prevalence estimate than that by CAPI.