Significance of integration and use of multiple data sources for understanding substance use and mental health disorders

Reuter et al.'s discussion raises the potential for misinterpreting quality and representativeness of National Survey on Drug Use Health data more generally, nature utility numerous general population surveys conducted worldwide relied upon public health policy decisions al. [1] evaluate prevalence frequent heroin use from (NSDUH), a survey (GPS) at Center Behavioral Statistics Quality (CBHSQ). CBHSQ Substance Abuse Mental Services Administration (SAMHSA) is lead Federal government agency behavioral dissemination in United States [2, 3]. The study emphasizes that underreporting selective non-response NSDUH can to underestimation. Claiming these rare stigmatized behaviors are prone imprecise estimates erratic year-to-year fluctuations because small sample sizes, authors recommend considering auxiliary sources help improve [1]. It widely recognized GPSs face challenges, especially when measuring problematic drug such as heroin, numbers, poor access people with substance disorders, self-under-reporting legal concerns and/or stigma. Nonetheless, authors’ decisions. tracks population-level indicators through various different levels, personal interviews emergency room visits, facilities- administrative-based records [2]. These provide scientific factual information academic other researchers, state local officials, policymakers, who further understanding pressing emerging issues (e.g. [4, 5]). are, therefore, essential evidence-based action by helping guide formulation intervention design. Begun 1971, largest every year; it GPS civilian, noninstitutionalized, 12 years older living (US) [6]. Although primary source conditions among US households, restriction noninstitutionalized persons results exclusion approximately 3% [7]. several chronic diseases hospitalizations were similar GPSs, Interview Survey, Nutrition Examination Risk Factor Surveillance System, Medical Expenditure Panel [8]. For some conditions, however, there was considerable variation underestimating instability mental disorder low extensive illicit drugs cocaine also concern [9]. Despite differences inaccuracies estimates, population-based considered be most useful surveillance tools providing periodic representative data, if repeated regular basis [10]. One fundamental principle base programmatic evidence derived highest-quality provides trends reliable within statistical standards established its methodology. example, cumulative incidence period rates disorders national, state, sub-state together their temporal variation. A unique feature area which along insights possible merging [11]. Moreover, investigating how complement collections, both inform informed NSDUH, Warning Network (DAWN), [12] Information System (BHSIS) [13], Disorders Prevalence Study (MDPS) [14]. DAWN monitors department visits secondary warning system emergence new substances combinations [12]. BHSIS levels facilities services available treatment, characteristics clients served [13]. MDPS program, will include non-household people—the incarcerated, institutionalized, homeless Therefore, although continuous monitoring program management, relying multiple decision making could an important step forward robust long-term strategies than do currently alternatives, interventions. None. author thanks staff, particularly Division Data Collection (DSDC), thoughtful reviews helpful inputs. opinions expressed this article author's own not reflect view Administration, Department Human Services, or government. Krishnan Radhakrishnan: Conceptualization.