Validation of a model for estimating state and local prevalence of serious mental illness

Predicting the Population Risk of Suicide Using Routinely Collected Health Administrative Data in Quebec, Canada: Model-Based Synthetic Estimation Study

BACKGROUND

Suicide is a significant public health issue. Many risk prediction tools have been developed to estimate an individual's risk of suicide. Risk prediction models can go beyond individual risk assessment; one important application of risk prediction models is population health planning. Suicide is a result of the interaction among the risk and protective factors at the individual, health care system, and community levels. Thus, policy and decision makers can play an important role in suicide prevention. However, few prediction models for the population risk of suicide have been developed.

OBJECTIVE

This study aims to develop and validate prediction models for the population risk of suicide using health administrative data, considering individual-, health system-, and community-level predictors.

METHODS

We used a case-control study design to develop sex-specific risk prediction models for suicide, using the health administrative data in Quebec, Canada. The training data included all suicide cases (n=8899) that occurred from January 1, 2002, to December 31, 2010. The control group was a 1% random sample of living individuals in each year between January 1, 2002, and December 31, 2010 (n=645,590). Logistic regression was used to develop the prediction models based on individual-, health care system-, and community-level predictors. The developed model was converted into synthetic estimation models, which concerted the individual-level predictors into community-level predictors. The synthetic estimation models were directly applied to the validation data from January 1, 2011, to December 31, 2019. We assessed the performance of the synthetic estimation models with four indicators: the agreement between predicted and observed proportions of suicide, mean average error, root mean square error, and the proportion of correctly identified high-risk regions.

RESULTS

The sex-specific models based on individual data had good discrimination (male model: C=0.79; female model: C=0.85) and calibration (Brier score for male model 0.01; Brier score for female model 0.005). With the regression-based synthetic models applied in the validation data, the absolute differences between the synthetic risk estimates and observed suicide risk ranged from 0% to 0.001%. The root mean square errors were under 0.2. The synthetic estimation model for males correctly predicted 4 of 5 high-risk regions in 8 years, and the model for females correctly predicted 4 of 5 high-risk regions in 5 years.

CONCLUSIONS

Using linked health administrative databases, this study demonstrated the feasibility and the validity of developing prediction models for the population risk of suicide, incorporating individual-, health system-, and community-level variables. Synthetic estimation models built on routinely collected health administrative data can accurately predict the population risk of suicide. This effort can be enhanced by timely access to other critical information at the population level.

Mapping Heat Vulnerability of a Community Mental Health Center Population

Individuals with serious mental illness are vulnerable to extreme heat due to biological, social, and place-based factors. We examine the spatial correlation of prevalence of individuals treated at a community mental health center to heat vulnerability. We applied a heat vulnerability index (HVI) to the catchment of the Connecticut Mental Health Center in New Haven, Connecticut. Geocoded addresses were mapped to correlate patient prevalence with heat vulnerability of census tracts. Census tracts closer to the city center had elevated vulnerability scores. Patient prevalence was positively correlated with HVI score (Pearson's r(44) = 0.67, p < 0.01). Statistical significance persists after correction for spatial autocorrelation (modified t-test p < 0.01). The study indicates that individuals treated at this community mental health center are more likely to live in census tracts with high heat vulnerability. Heat mapping strategies can help communicate risk and target resources at the local scale.

Validation of a small-area model for estimation of smoking prevalence at a subnational level

INTRODUCTION

Small-area estimation methods are an alternative to direct survey-based estimates in cases where a survey's sample size does not suffice to ensure representativeness. Nevertheless, the information yielded by small-area estimation methods must be validated. The objective of this study was thus to validate a small-area model.

METHODS

The prevalence of smokers, ex-smokers, and never smokers by sex and age group (15-34, 35-54, 55-64, 65-74, ≥75 years) was calculated in two Spanish Autonomous Regions (ARs) by applying a weighted ratio estimator (direct estimator) to data from representative surveys. These estimates were compared against those obtained with a small-area model applied to another survey, specifically the Spanish National Health Survey, which did not guarantee representativeness for these two ARs by sex and age. To evaluate the concordance of the estimates, we calculated the intraclass correlation coefficient (ICC) and the 95% confidence intervals of the differences between estimates. To assess the precision of the estimates, the coefficients of variation were obtained.

RESULTS

In all cases, the ICC was ≥0.87, indicating good concordance between the direct and small-area model estimates. Slightly more than eight in ten 95% confidence intervals for the differences between estimates included zero. In all cases, the coefficient of variation of the small-area model was <30%, indicating a good degree of precision in the estimates.

CONCLUSIONS

The small-area model applied to national survey data yields valid estimates of smoking prevalence by sex and age group at the AR level. These models could thus be applied to a single year's data from a national survey, which does not guarantee regional representativeness, to characterize various risk factors in a population at a subnational level.

Small-area models to assess the geographical distribution of tobacco consumption by sex and age in Spain

INTRODUCTION

Complete and accurate data on smoking prevalence at a local level would enable health authorities to plan context-dependent smoking interventions. However, national health surveys do not generally provide direct estimates of smoking prevalence by sex and age groups at the subnational level. This study uses a small-area model-based methodology to obtain precise estimations of smoking prevalence by sex, age group and region, from a population-based survey.

METHODS

The areas targeted for analysis consisted of 180 groups based on a combination of sex, age group (15-34, 35-54, 55-64, 65-74, and ≥75 years), and Autonomous Region. Data on tobacco use came from the 2017 Spanish National Health Survey (2017 SNHS). In each of the 180 groups, we estimated the prevalence of smokers (S), ex-smokers (ExS) and never smokers (NS), as well as their coefficients of variation (CV), using a weighted ratio estimator (direct estimator) and a multinomial logistic model with random area effects.

RESULTS

When smoking prevalence was estimated using the small-area model, the precision of direct estimates improved; the CV of S and ExS decreased on average by 26%, and those of NS by 25%. The range of S prevalence was 11-46% in men and 4-37% in women, excluding the group aged ≥75 years.

CONCLUSIONS

This study proposes a methodology for obtaining reliable estimates of smoking prevalence in groups or areas not covered in the survey design. The model applied is a good alternative for enhancing the precision of estimates at a detailed level, at a much lower cost than that involved in conducting large-scale surveys. This method could be easily integrated into routine data processing of population health surveys. Having such estimates directly after completing a health survey would help characterize the tobacco epidemic and/or any other risk factor more precisely.

Benchmarks for Needed Psychiatric Beds for the United States: A Test of a Predictive Analytics Model

The ideal balanced mental health service system presupposes that planners can determine the need for various required services. The history of deinstitutionalization has shown that one of the most difficult such determinations involves the number of needed psychiatric beds for various localities. Historically, such assessments have been made on the basis of waiting and vacancy lists, expert estimates, or social indicator approaches that do not take into account local conditions. Specifically, this study aims to generate benchmarks or estimated rates of needed psychiatric beds for the 50 U.S. states by employing a predictive analytics methodology that uses nonlinear regression. Data used were secured primarily from the U.S. Census’ American Community Survey and from the Substance Abuse and Mental Health Administration. Key predictors used were indicators of community mental health (CMH) service coverage, mental health disability in the adult population, longevity from birth, and the percentage of the 15+ who were married in 2018. The model was then used to calculate predicted bed rates based on the ‘what-if’ assumption of an optimal level of CMH service availability. The final model revealed an overall rate of needed beds of 34.9 per 100,000 population, or between 28.1 and 41.7. In total, 32% of the states provide inpatient psychiatric care at a level less than the estimated need; 28% at a level in excess of the need; with the remainder at a level within 95% confidence limits of the estimated need. These projections are in the low range of prior estimates, ranging from 33.8 to 64.1 since the 1980s. The study demonstrates the possibility of using predictive analytics to generate individualized estimates for a variety of service modalities for a range of localities.

Measuring disparities in police use of force and injury among persons with serious mental illness

OBJECTIVES

To measure disparities in experience of police use of force and injury among persons with serious mental illnesses.

METHODS

We gathered novel police use of force and suspect injury data from 2011 to 2017 from a nonrandom sample of nine police departments in the United States and used synthetic methods to estimate the share of the local population with serious mental illness. We estimate disparities using multi-level models estimated in a Bayesian framework.

RESULTS

Persons with serious mental illness constitute 17.0% of use of force cases (SD = 5.8) and 20.2% of suspects injured in police interaction (SD = 9.0) in sample cities. The risk that persons with serious mental illness will experience police use of force is 11.6 times higher (95% CI, 10.7-12.6) than persons without serious mental illness. Persons with serious mental illness are also at a higher risk of experiencing injury, 10.7 times (95% CI, 9.6-11.8), relative to persons without serious mental illness. These relative risk ratios are several times larger than racial and ethnic disparities estimated in the same cities.

CONCLUSION

Persons with serious mental are at a significantly elevated risk of experiencing police use of force and injury in police encounters than the general public. The disparities we estimate are several times higher than racial/ethnic disparities in force and injury. Efforts to reform police practices and reimagine public safety in the United States should address significant disparities in police use of force against those with serious mental illness.

Cognition and daily life functioning among persons with serious mental illness: A cluster analytic examination of heterogeneity on the Test of Grocery Shopping Skills

OBJECTIVE

To characterize variability in daily life functioning among individuals with serious mental illness based on a naturalistic performance measure of grocery shopping and standard neuropsychological tasks using cluster analytic methods.

METHODS

A naturalistic performance measure, the Test of Grocery Shopping Skills (TOGSS), and standard neuropsychological tasks, were completed by 191 participants with serious mental illness. Hierarchical cluster analytic techniques were used to explore functional subgroups based on naturalistic performance measure variables. Multivariate analyses of variance were utilized to compare subgroups on TOGSS variables and neuropsychological measures, respectively.

RESULTS

Two distinct functional subgroups emerged from the cluster analysis. On average, participants in cluster one were faster, more efficient, and more accurate compared to cluster two. Based on performance on neuropsychological tasks, cluster one had better verbal memory, visual attention, and processing speed, and executive functioning scores, compared to cluster two. The clusters did not differ on a measure of auditory working memory.

CONCLUSIONS AND IMPLICATIONS FOR PRACTICE

Naturalistic performance measures can assist with characterizing the heterogeneity in real life functioning among people with serious mental illness. Further work to illuminate the relationship between specific cognitive abilities and specific functional abilities is warranted and may assist with targeting effective treatment plans for functional recovery. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

The impact of geographic isolation on mental disability in the United States

Geographic isolation has long been hypothesized to have a role in the origins and development of mental disabilities. A considerable body of research has established such a correlation. However, study designs have limited researchers’ ability to establish a causal connection and rule out rival hypotheses.

This study, therefore, aims to assess the strength of the geographic isolation - mental disability relationship and to disentangle it from alternative possibilities, namely that it reflects socioeconomic status, social isolation, economic inequality, or reverse causation. The study employs an analysis of variations in the rates of mental disability throughout 2960 U.S. counties using both Census and CDC data. In addition to partial correlation and ordinary least square analyses, the study employs two-stage least squares regression with instrumental variables (2SLS-IV), a procedure that permits resolution of the problem of endogeneity involving the potential effects of unmeasured variables and reverse causation.

Results reveal that the initial bivariate effects of geographic isolation on rates of mental disability are robust after controls for socioeconomic status, income inequality, social isolation, and other predictors are introduced and when tested with the 2SLS-IV procedure. Most variation (54.4%) in county mental disability rates is accounted for by the independent effects of geographic isolation, socioeconomic status, income inequality, and other variables. The results presented, although not conclusive, supports more targeted service planning and more equitable resource investments in rural parts of the United States and other nations.

•

Isolation is a moderately strong predictor of mental disability.

•

Socioeconomic status has a moderate negative correlation with mental disability.

•

Income inequality compounds the effects of isolation and SES on mental disability.

•

Most of variation in MI disability rates explained by isolation, SES, and inequality.

Comparison of Methods for Estimating Prevalence of Chronic Diseases and Health Behaviors for Small Geographic Areas: Boston Validation Study, 2013

Introduction

Local health authorities need small-area estimates for prevalence of chronic diseases and health behaviors for multiple purposes. We generated city-level and census-tract–level prevalence estimates of 27 measures for the 500 largest US cities.

Methods

To validate the methodology, we constructed multilevel logistic regressions to predict 10 selected health indicators among adults aged 18 years or older by using 2013 Behavioral Risk Factor Surveillance System (BRFSS) data; we applied their predicted probabilities to census population data to generate city-level, neighborhood-level, and zip-code–level estimates for the city of Boston, Massachusetts.

Results

By comparing the predicted estimates with their corresponding direct estimates from a locally administered survey (Boston BRFSS 2010 and 2013), we found that our model-based estimates for most of the selected health indicators at the city level were close to the direct estimates from the local survey. We also found strong correlation between the model-based estimates and direct survey estimates at neighborhood and zip code levels for most indicators.

Conclusion

Findings suggest that our model-based estimates are reliable and valid at the city level for certain health outcomes. Local health authorities can use the neighborhood-level estimates if high quality local health survey data are not otherwise available.

Neighborhood sociodemographic predictors of Serious Emotional Disturbance (SED) in schools: demonstrating a small area estimation method in the National Comorbidity Survey (NCS-A) Adolescent Supplement

We evaluate the precision of a model estimating school prevalence of SED using a small area estimation method based on readily-available predictors from area-level census block data and school principal questionnaires. Adolescents at 314 schools participated in the National Comorbidity Supplement, a national survey of DSM-IV disorders among adolescents. A multilevel model indicated that predictors accounted for under half of the variance in school-level SED and even less when considering block-group predictors or principal report alone. While Census measures and principal questionnaires are significant predictors of individual-level SED, associations are too weak to generate precise school-level predictions of SED prevalence.

Validation of multilevel regression and poststratification methodology for small area estimation of health indicators from the behavioral risk factor surveillance system

Small area estimation is a statistical technique used to produce reliable estimates for smaller geographic areas than those for which the original surveys were designed. Such small area estimates (SAEs) often lack rigorous external validation. In this study, we validated our multilevel regression and poststratification SAEs from 2011 Behavioral Risk Factor Surveillance System data using direct estimates from 2011 Missouri County-Level Study and American Community Survey data at both the state and county levels. Coefficients for correlation between model-based SAEs and Missouri County-Level Study direct estimates for 115 counties in Missouri were all significantly positive (0.28 for obesity and no health-care coverage, 0.40 for current smoking, 0.51 for diabetes, and 0.69 for chronic obstructive pulmonary disease). Coefficients for correlation between model-based SAEs and American Community Survey direct estimates of no health-care coverage were 0.85 at the county level (811 counties) and 0.95 at the state level. Unweighted and weighted model-based SAEs were compared with direct estimates; unweighted models performed better. External validation results suggest that multilevel regression and poststratification model-based SAEs using single-year Behavioral Risk Factor Surveillance System data are valid and could be used to characterize geographic variations in health indictors at local levels (such as counties) when high-quality local survey data are not available.

'In general, how do you feel today?'--self-rated health in the context of aging in India

This thesis is centered on self-rated health (SRH) as an outcome measure, as a predictor, and as a marker. The thesis uses primary data from the WHO Study on global AGEing and adult health (SAGE) implemented in India in 2007. The structural equation modeling approach is employed to understand the pathways through which the social environment, disability, disease, and sociodemographic characteristics influence SRH among older adults aged 50 years and above. Cox proportional hazard model is used to explore the role of SRH as a predictor for mortality and the role of disability in modifying this effect. The hierarchical ordered probit modeling approach, which combines information from anchoring vignettes with SRH, was used to address the long overlooked methodological concern of interpersonal incomparability. Finally, multilevel model-based small area estimation techniques were used to demonstrate the use of large national surveys and census information to derive precise SRH prevalence estimates at the district and sub-district level. The thesis advocates the use of such a simple measure to identify vulnerable communities for targeted health interventions, to plan and prioritize resource allocation, and to evaluate health interventions in resource-scarce settings. The thesis provides the basis and impetus to generate and integrate similar and harmonized adult health and aging data platforms within demographic surveillance systems in different regions of India and elsewhere.

Self-rated health: small area large area comparisons amongst older adults at the state, district and sub-district level in India

We compared prevalence estimates of self-rated health (SRH) derived indirectly using four different small area estimation methods for the Vadu (small) area from the national Study on Global AGEing (SAGE) survey with estimates derived directly from the Vadu SAGE survey. The indirect synthetic estimate for Vadu was 24% whereas the model based estimates were 45.6% and 45.7% with smaller prediction errors and comparable to the direct survey estimate of 50%. The model based techniques were better suited to estimate the prevalence of SRH than the indirect synthetic method. We conclude that a simplified mixed effects regression model can produce valid small area estimates of SRH.

Peer Education: Participatory Qualitative Educational Needs Assessment

BACKGROUND

In the area of youth health, peers education is an approach to health promotion. Assess the training needs of peers educators clarifies the components, values, and quality of training protocols. Aim to that we conducted a participatory educational needs assessment of youth peer educators.

METHODS

Involving youth and key informants in direct collaboration with research team, a qualitative approach was planned based on grounded theory. For data collection a semi-structured guide questioning was designed. Sixteen focus group discussions and 8 in depth interview were held.

RESULTS

The majority of participants emphasized on the importance of mental health, life skills, AIDS prevention, contraception methods, and healthy nutrition as the main training topics. They were extremely interested into the comprehensive educational material among their participatory role in peer programs.

CONCLUSION

The training programs should be well defined based on the knowledge, skills and behavior of peers. During the implementation, training programs should be followed to meet the ongoing educational needs of service providers.

Modeling the geographic distribution of serious mental illness in New Zealand

PURPOSE

This study aims to estimate, apply, and validate a model of the risk of serious mental illness (SMI) in local service areas throughout New Zealand.

METHODS

The study employs a secondary analysis of data from the Te Rau Hinengaro Mental Health Survey of 12,992 adults aged 16 years and over from the household population. It uses small area estimation (SAE) methods involving: (1) estimation of a logistic model of risk of SMI; (2) use of the foregoing model for computing estimates, using census data, for District Board areas; (3) validation of estimates against an alternative indicator of SMI prevalence.

RESULTS

The model uses age, ethnicity, marital status, employment, and income to predict 92.2 % of respondents' SMI statuses, with a specificity of 95.9 %, sensitivity of 16.9 %, and an AUC of 0.73. The resulting estimates for the District Board areas ranged between 4.1 and 5.7 %, with confidence intervals from ±0.3 to ±1.1 %. The estimates demonstrated a correlation of 0.51 (p = 0.028) with rates of psychiatric hospitalization.

CONCLUSIONS

The use of SAE methods demonstrated the capacity for deriving local prevalence rates of SMI, which can be validated against an available indicator.

Declines in mental illness over the adult years: an enduring finding or methodological artifact?

OBJECTIVES

Epidemiological surveys have revealed dramatic declines in the prevalence of serious mental illness (SMI) as adults age. Analyses of these datasets have not determined whether this is attributable, on one hand, to declining incidence, shorter duration of condition, and lesser severity with age, or on the other hand, confounding variables and methodological biases. This study, thus, aims to test several such competing hypotheses.

METHODS

This study employs a secondary analysis of data from the 2001/2002 US National Comorbidity Replication Survey of 9282 adults, 18 and older, living in the household population of the 48 contiguous states, as well as supplemental datasets from Israel, New Zealand, and other sources.

RESULTS

One-year SMI prevalence rates in the US drop from 8.0% of adults 18-29 to 1.4% of the 65+, and to similar degrees in Israel and New Zealand. The drop in the US can be explained by the early onset of most mental illnesses, and declines in both incidence and duration of condition with age. Comorbidity also drops with age; however, the remaining diagnoses show a gradually increasing severity. Institutionalization explains a small portion of the declines, as does premature mortality of the seriously mentally ill that accounts for 16.3% of the decline.

CONCLUSION

The results reveal that a substantial portion of the declines are explainable in terms of declining incidence and improving recovery rates, with some reduction also attributable to institutionalization and premature mortality which removes some older and more seriously disabled adults from the epidemiological survey populations.

Disparities in the geography of mental health: implications for social work

This article reviews recent theory and research on geographic disparities in mental health and their implications for social work. It focuses on work emerging from the fields of mental health geography, psychiatric epidemiology, and social work, arguing that a wide range of spatial disparities in mental health are important to understand but that of greatest relevance are inequities, or disparities, that violate fundamental norms of fairness and social justice. Research is reviewed on geographic variations in subjective well-being and mental health, on personality (using the five-factor model), and on psychopathology as well as several studies on the disparate implementation of mental health policy and services. Critical is the need to simultaneously assess, on the one hand, differential patterns of mental health conditions and, on the other, the services and policies designed to address them--the fact that considering only one dimension often leads to unintended consequences. Many of the most outstanding disparities have been found to exist at the local level, between towns and neighborhoods, and are based on socioeconomic conditions. This review concludes by discussing the implications of geographic disparities in mental health for allocation decisions and for social work practice, including decisions about the most efficacious mix of services at both the community and clinical practice levels.

Disparities in the geography of serious mental illness in Israel

OBJECTIVE

The aim of this research is to test and apply a model of the disparities and variations in serious mental illness (SMI) to estimating prevalence in local areas throughout Israel.

METHODS

This study employs a secondary analysis of data from the 2003/2004 Israel National Health Survey of 4859 adults aged 21 and over from the household population of legal residents and citizens. It uses small area estimation methods (SAE), specifically to: (i) estimate and test a multivariate logistic model of disparities in the risk of serious mental illness; (ii) use the foregoing model for computing estimates, using census data, for local areas; (iii) validate these estimates against the rate of psychiatric hospitalizations.

RESULTS

The model uses standard demographic and socioeconomic variables to successfully predict 92.5% of respondents' statuses as SMI, with a sensitivity of 26.9%, specificity of 95.9%, and an AUC index of .797. The resulting estimates of the percentage of adults with an SMI in the 16 subdistricts ranged between 3.7% and 7.7%, with a national mean of 5.0%. The estimates have a partial correlation of .63 with rates of psychiatric hospitalization in Jewish localities, but elevated rates have not been validated in Arab localities.

CONCLUSION

The use of small area estimation methods demonstrated the capacity for deriving local prevalence rates of serious mental illness, ones that can be validated against psychiatric hospitalization for the majority population group in Israel.

Using a short screening scale for small-area estimation of mental illness prevalence for schools

We use data collected in the National Comorbidity Survey - Adolescent (NCS-A) to develop a methodology to estimate the small-area prevalence of serious emotional distress (SED) in schools in the United States, exploiting the clustering of the main NCS-A sample by school. The NCS-A instrument includes both a short screening scale, the K6, and extensive diagnostic assessments of the individual disorders and associated impairment that determine the diagnosis of SED. We fitted a Bayesian bivariate multilevel regression model with correlated effects for the probability of SED and a modified K6 score at the individual and school levels. Our results provide evidence for the existence of variation in the prevalence of SED across schools and geographical regions. Although the concordance between the modified K6 scale and SED is only modest for individuals, the school-level random effects for the two measures are strongly correlated. Under this model we obtain a prediction equation for the rate of SED based on the mean K6 score and covariates. This finding supports the feasibility of using short screening scales like the K6 as an alternative to more comprehensive lay assessments in estimating school-level rates of SED. These methods may be applicable to other studies aiming at small-area estimation for geographical units.

Promoting Self-Care Management among Persons with Serious Mental Illness and HIV

People with a serious mental illness (SMI) along with HIV have complex health conditions. This population also has high rates of poverty, difficulty in sustaining regular housing, and limited supportive networks. Typically, the combination of psychotropic and HIV medication regimens is complicated, changes frequently, and requires coordination among multiple providers. Furthermore, fragmented and divided primary health care and mental health care systems present substantial barriers for these individuals and for the public health nurses who care for them. In this paper, we present "real world" case studies of individuals with SMI and HIV and the self-care management strategies used by nurses to address medication and treatment management, build interpersonal skills, and develop sustainable health networks. The case studies can be used for quality improvement discussions among practicing public health nurses and for instructing nursing students in a self-care management approach.

Behavioral vaccines and evidence-based kernels: nonpharmaceutical approaches for the prevention of mental, emotional, and behavioral disorders

In the United States the rates for some mental, emotional, and behavioral problems (MEBs) have objectively increased over the past 20 to 50 years. The attributes of a public health approach to the treatment of MEBs are defined in this article. Multiple examples of how public health approaches might reduce or prevent MEBs using low-cost evidence-based kernels, which are fundamental units of behavior, are discussed. Such kernels can be used repeatedly, which then act as "behavioral vaccines" to reduce morbidity or mortality and/or improve human wellbeing. The author calls for 6 key policy actions to improve MEBs in young people.

Estimating prevalence of serious emotional disturbance in schools using a brief screening scale

Information about the prevalence of serious mental illness (SMI) among adults or serious emotional disturbance (SED) among youth in small domains such as counties, states, or schools is valuable for mental health policy planning purposes, but prohibitively expensive to collect with semi-structured surveys. Commonly used synthetic estimation methods yield imprecise estimates. An improved method is described here that combines information about socio-demographic covariates with screening scale scores obtained from a sample of individuals, using a prediction equation derived from a Bayesian multilevel regression model with bivariate outcomes fitted to a larger population survey. This method is illustrated using K6 screening scale scores to predict school-level prevalence of SED in the sample of 282 schools that participated in the National Comorbidity Survey Replication Adolescent Supplement. Respondents completed a diagnostic interview that was used to define DSM-IV SED. SED prevalence varied significantly across schools and was strongly correlated with aggregate K6 scores (rho = 0.70). Calculations suggest that near-maximum precision of school-level SED prevalence estimates could be attained with K6 samples of 200 students per school. This modeling approach holds great promise for generating accurate estimates of SMI/SED in small-area planning units based on K6 scores collected in ongoing health tracking surveys.