Measuring efficiency of federally qualified health centers : a multi model approach using data envelopment analysis (DEA) & latent class analysis (LCA)

Abstract

There are 1300 federally qualified health centers (FQHCs) in the United States providing the health care to underserved and uninsured population. These FQHCs serve the patients irrespective of their ability to pay. Using the resources effectively, these FQHCs can provide better health care. In this study of prenatal care, we measure the efficiencies of the FQHCs using data envelopment analysis (DEA). As in service industry, where quality is of at most importance, we used two different DEA approaches considering quality called the Two model DEA approach by (Shimshak, D., and Lenard, M.L.,2007) and Quality adjusted DEA approach by (Sherman, H.D., and Zhu, J, 2006). Efficient frontiers are determined by using these DEA approaches. There are differences that exists across FQHCs due to various factors to include demographic characteristics of patients visited the FQHCs, operational characteristics of health centers. Latent class analysis is performed before performing the DEA to classify the FQHCs into different classes based on the regional and population measures. Four different models namely aggregated Shimshak and Lenard and aggregated Sherman and Zhu models (DEA model is run on the whole sample), partitioned S and L and partitioned S and Z models (DEA model is run individually by class) have been used to determine the efficiencies of the FQHCs. Using the S and L approach, it is found that the FQHCs that formed the efficient frontier is of smaller FQHCs whereas the S and Z approach has a mix of small and large FQHCs. Based on the results determined, more insights are provided on the FQHCs and the models used in the analysis.