Assessing Panamanian hospitals' performance with alternative frontier methods

Abstract

In the last decades, there has been a growing interest in measuring the efficiency of hospitals using different methodological approaches, mainly represented by data envelopment analysis (DEA) and stochastic frontier analysis (SFA). In this study, we estimate efficiency measures of performance for a sample of Panamanian public hospitals over an 11-year period (2005-2015) using both traditional methods (DEA and SFA) and compare them with efficiencies estimated with an alternative approach, the so-called StoNED (stochastic semi-nonparametric envelopment of data), which combines the virtues of those methods in a unified framework. One of the most interesting features of the public health system in Panama is that it is segmented, as hospitals are operating under two parallel management schemes (the Ministry of Health and the Social Security Fund), thus in our empirical analysis we will also focus on exploring the differences between hospitals operating under each regime. Our results show that there are certain divergences in the efficiency scores estimated with different methodologies, but for all of them it is possible to detect that Panamanian hospitals experienced a clear decrease in their efficiency levels throughout the period evaluated, being this much higher in the hospitals belonging to the Social Security Fund.

The Core of Healthcare Efficiency: A Comprehensive Bibliometric Review on Frontier Analysis of Hospitals

“…The outputs are more diversified than inputs in the applications. The most common healthcare products regarded in the hospital efficiency core applications are the number of discharges [ 7 , 70 , 71 , 75 , 83 ], the number of surgeries [ 67 , 77 , 78 , 79 , 83 , 86 ], and the total number of inpatient and outpatient treatments, diagnoses and healthcare income [ 68 , 77 , 81 ]. Childbirth-related outputs are antenatal checkups, deliveries (such as cesarean-section deliveries), abortions, post-natal care, and the number of male and female sterilizations [ 83 , 84 , 86 ].…”

“…In this regard, one can emphasize that there are several methodologies to evaluate the efficiency of hospitals. According to Cordero et al [ 7 ], efficiency scores of healthcare institutions can be estimated using either parametric or non-parametric methods. Both methodologies aim to assess production units and are based on productivity indicators, also known as technical efficiency (TE), that offer measures that characterize the operations of the units analyzed.…”

Efficiency evaluation of healthcare services in China based on stochastic multicriteria acceptability analysis and directional distance function

“…A commonly used technique for measuring healthcare service efficiency is the frontier efficiency analysis method (Du et al., 2014; Kohl et al., 2019; Mergoni and De Witte, 2022). Stochastic frontier analysis (SFA) and data envelopment analysis (DEA) are two mainstream frontier methods used for healthcare efficiency evaluation (O'Neill et al., 2008; Cordero et al., 2023). The former is a type of regression‐based approach that determines a specific form of the production function between input and output (Aigner et al., 1977).…”

Medida de la eficiencia en atención primaria: fronteras eficientes y modelos no paramétricos condicionados

“…Dentro de los modelos no paramétricos, uno de los más utilizados es el análisis envolvente de datos (Data Envelopment Analysis o DEA, por sus siglas inglés), ya que permite evaluar la eficiencia relativa de las unidades de decisión mediante la creación de una frontera de producción utilizando la mejor práctica dentro de los datos observados. Aunque también presenta algunos inconvenientes, como su sensibilidad ante la presencia de valores extremos o anómalos y la imposibilidad de realizar un análisis de sensibilidad probabilístico, esta debilidad se ha solventado en trabajos posteriores (30).…”

“…Ante los inconvenientes que presentan tanto los modelos paramétricos de frontera estocástica, como los modelos DEA no paramétricos, se puede plantear como alternativa la utilización de alguna de las combinaciones de ambos métodos que se han desarrollado en la literatura, como el análisis envolvente de datos estocástico. Sin embargo, el uso de estas metodologías en el sector sanitario es todavía muy incipiente y existen pocos trabajos al respecto (30).…”