Health-care expenditure regressions are used in a wide variety of economic analyses including risk adjustment and program and treatment evaluations. Recent articles demonstrated that generalized gamma models (GGMs) and extended estimating equations (EEE) models provide flexible approaches to deal with a variety of data problems encountered in expenditure estimation. To date there have been few empirical applications of these models to expenditures. We use data from the US Medical Expenditure Panel Survey to compare the bias, predictive accuracy, and marginal effects of GGM and EEE models with other commonly used regression models in a cross-validation study design. Health-care expenditure distributions vary in the degree of heteroskedasticity, skewness, and kurtosis by type of service and population. To examine the ability of estimators to address a range of data problems, we estimate models of total health expenditures and prescription drug expenditures for two populations, the elderly and privately insured adults. Our findings illustrate the need for researchers to examine their assumptions about link functions: the appropriate link function varies across our four distributions. The EEE model, which has a flexible link function, is a robust estimator that performs as well, or better, than the other models in each distribution.
"Firstly, there is no one model that dominates in all respects and there seems to be a tradeoff between bias and precision (Veazie et al., 2003). Secondly, that the preferred model is likely to vary with the sample size of data on which the model is estimated (Deb and Burgess, 2003) and will also vary across datasets (Hill and Miller, 2010). It has also been noted that a more flexible model is not necessarily an adequate replacement for the correct model (Manning et al., 2005). "
[Show abstract][Hide abstract] ABSTRACT: This paper extends the literature on modelling healthcare cost data by applying the Generalised Beta of the Second Kind (GB2) distribution to UK data. A quasi-experimental design, estimating models on a subset of the data and evaluating performance on another subset, is used to compare this distribution with its nested and limiting cases. We nd that the GB2 may be a useful tool for choosing an appropriate distribution to apply, with the Beta-2 (B2) distribution and Generalised Gamma (GG) distribution performing the best with this dataset.
"The distribution of strictly positive expenditures is typically skewed, kurtotic, and heteroskedastic . A variety of models have been used to analyze expenditure/cost data, with many analysts now presenting more than one model in their reports  . We report the most popular MEPS regression model, GLM. "
[Show abstract][Hide abstract] ABSTRACT: To inform policymakers of the importance of evaluating various methods for estimating the direct medical expenditures for a low-incidence condition, head and neck cancer (HNC).
Four methods of estimation have been identified: 1) summing all health care expenditures, 2) estimating disease-specific expenditures consistent with an attribution approach, 3) estimating disease-specific expenditures by matching, and 4) estimating disease-specific expenditures by using a regression-based approach. A literature review of studies (2005-2012) that used the Medical Expenditure Panel Survey (MEPS) was undertaken to establish the most popular expenditure estimation methods. These methods were then applied to a sample of 120 respondents with HNC, derived from pooled data (2003-2008).
The literature review shows that varying expenditure estimation methods have been used with MEPS but no study compared and contrasted all four methods. Our estimates are reflective of the national treated prevalence of HNC. The upper-bound estimate of annual direct medical expenditures of adult respondents with HNC between 2003 and 2008 was $3.18 billion (in 2008 dollars). Comparable estimates arising from methods focusing on disease-specific and incremental expenditures were all lower in magnitude. Attribution yielded annual expenditures of $1.41 billion, matching method of $1.56 billion, and regression method of $1.09 billion.
This research demonstrates that variation exists across and within expenditure estimation methods applied to MEPS data. Despite concerns regarding aspects of reliability and consistency, reporting a combination of the four methods offers a degree of transparency and validity to estimating the likely range of annual direct medical expenditures of a condition.
Value in Health 01/2014; 17(1):90-7. DOI:10.1016/j.jval.2013.10.004 · 3.28 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Despite spatial econometrics is now considered a consolidated discipline,
only in recent years we have experienced an increasing attention to the
possibility of applying it to the field of discrete choices (e.g. Smirnov, 2010
for a recent review) and limited dependent variable models. In particular, only
a small number of papers introduced the above-mentioned models in Health
Economics. The main purpose of the present paper is to review the different
methodological solutions in spatial discrete choice models as they appeared in
several applied fields by placing an emphasis on the health economics
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.