1 / 28

Proxy Pattern-Mixture Analysis of Missing Health Expenditure Variables in the Medical Expenditure Panel Survey

Proxy Pattern-Mixture Analysis of Missing Health Expenditure Variables in the Medical Expenditure Panel Survey . Robert M. Baskin, Samuel H. Zuvekas and Trena M. Ezzati-Rice Division of Statistical Methods and Research Center for Financing, Access and Cost Trends. Purpose of Study.

loren
Télécharger la présentation

Proxy Pattern-Mixture Analysis of Missing Health Expenditure Variables in the Medical Expenditure Panel Survey

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Proxy Pattern-Mixture Analysis of Missing Health Expenditure Variables in the Medical Expenditure Panel Survey Robert M. Baskin, Samuel H. Zuvekas and Trena M. Ezzati-Rice Division of Statistical Methods and Research Center for Financing, Access and Cost Trends

  2. Purpose of Study • Use Fraction of Missing Information (FMI) to evaluate new item imputation methodology in Medical Expenditure Panel Survey (MEPS) • Expenditures for hospitals and office-based physicians from MEPS 2008 will be used.

  3. Medical Expenditure Panel Survey Components • HC -- Household Component • MPC -- Medical Provider Component • IC -- Insurance Component

  4. What is MEPS-HC Annual Survey of ~15,000 households: Provides national estimates of health care use, expenditures, insurance coverage, sources of payment, access to care and health care quality Permits studies of: • Distribution of expenditures and sources of payment • Role of demographics, family structure, insurance • Expenditures for specific conditions • Trends over time

  5. MEPS-HC Survey Design • Nationally representative sub-sample of responding households from previous year’s National Health Interview Survey (NHIS) • Covers civilian non-institutionalized population • Selected from ~ 200/400 NHIS PSUs • Five CAPI interviews cumulate data for 2 consecutive years • Overlapping panels for annual data • Two panels in field concurrently

  6. MEPS-HC Core Interview Content • Demographics • Health Status • Conditions • Employment • Health Insurance • Health Care Use & Expenditures

  7. Non-response in MEPS • Unit non-response • - weighting adjustment • Item non-response • - imputation • The following ignores unit non-response

  8. MEPS-MPC • Survey of medical providers that provided care to MEPS sample persons • Signed permission forms required to contact providers • Purpose is to collect data that can be difficult for HC respondents to report completely or accurately • Charges and payments • Dates of visit, diagnosis and procedure codes • Not designed as independent nationally representative sample of providers

  9. Primary Uses of MPC Data • Supplement or replace expenditure data reported in HC • Imputation source • Methodological studies

  10. MPC - Targeted Sample • All providers for households with Medicaid recipients • All hospitals and associated physicians • About ½ of office-based physicians • All home health agencies • All pharmacies

  11. Linking MPC to HC Data • Probabilistic record linkage approach • Primary variables used: • Date • Event Type • Medical condition(s) • Types of services

  12. Final MEPS Expenditure Data • General approach • MPC data used when available • HC data used when no MPC data available • Events with no expenditure data from MPC or HC are imputed • MPC data generally preferred donor

  13. Sources of Expenditure Data for Selected Event Types, 2008

  14. Method of Imputation • 1996-2007: Weighted Sequential Hotdeck within imputation cells • 2008: Office Based Visits used Predictive Mean Matching (PMM) • 2009: 4 Event Types will use PMM • -Office Based Visits • -Out Patient • -Emergency Room • -In Patient

  15. Predictive Mean Matching • For each event type recipients are classified into subgroups based on available predictors of total payments • For each subgroup four models are built based on donor data

  16. Four Models • Basic: all predictors in hotdeck • - no transformation • Expanded: add GPCI codes (Medicare geographic payment codes) and chronic conditions (e.g. diabetes) • - no transformation • - log of payments - square root of payments

  17. Model R-Squared2008 MEPS

  18. Proxy Pattern-Mixture Models • The stated purpose of the study is to use Proxy Pattern-Mixture models to evaluate the effect of missingness on the estimates of mean • - Little (1994) describes analyzing the data based on the pattern of missingness

  19. Proxy Pattern-Mixture Models • Likelihood based f(Y, X, M| θ,π)= f(Y, X | M, θ) f(M|π) - Y=dependent variable with missingness - X=covariates - M=missingness indicator

  20. Proxy Pattern-Mixture Assumptions • f(Y, X | M, θ) is estimable from respondents • f(M| Y, X, θ) is an increasing function of X + λY λ is assumed to be known – it is not estimable from the data

  21. Proxy Pattern-Mixture Assumptions • If f(M| Y, X, θ) is an increasing function of X + λY λ = 0 is equivalent to missing at random λ = 1 is equivalent to Heckman selection λ = ∞ is equivalent to Brown model

  22. Proxy Pattern-Mixture Estimate of Bias • If f(M| Y, X θ) is an increasing function of X + λY then the maximum likelihood estimate of the bias in estimating the mean using respondents is given by

  23. Percent Bias Estimate from Proxy Pattern-Mixture Analysis

  24. Proxy Pattern-Mixture Models and FMI “The FMI due to non-response is estimated by the ratio of between-imputation to total variance under multiple imputation. Traditionally one applies this under the assumption that data are MAR, but we propose its application under the pattern-mixture model where missingness is not necessarily at random.” (from Andridge and Little)

  25. FMI vs PPMA • The Pattern Mixture-Model estimates the bias in using the mean of respondents (complete case analysis) • FMI estimates the ‘uncertainty’ in using the mean including imputed values

  26. PMM Percent Bias Estimate and FMI

  27. Respondent Means vs Imputed Means

  28. Summary • Item imputation in MEPS is improved with use of available predictors • Under assumptions for Proxy Pattern-Mixture models MEPS item imputation evaluated well

More Related