Utilization of public health centres in Portugal: effect of time costs and other determinants. Finite mixture models applied to truncated samples.

The impact of time costs on the utilization of medical care has been a subject of theoretical and empirical research since the early 1970s. The main goal of this paper is to show the effect of time costs on the number of visits to general practitioners (GP) in Portuguese public health centres. We measured the elasticity of primary health care utilization relative to the total time spent in the health centre and relative to travel time. We also provided evidence regarding the impact of an appointment delay on the utilization of public GP services. Our data resulted from the application of an endogenous sampling scheme, resulting in a truncated-at-zero data set. To model our dependent variable, number of visits, and accounting for the truncated nature of the data we used a finite mixture model specification. The data were obtained from the most recent implementation in Portugal of the 2003/2004 Europep Survey. The two-component negative binomial II finite mixture model led to the identification of two different latent classes of health centre users: a low-users class that comprises 88% of patients with an estimated utilization mean of 4.3 GP visits per year and a frequent-users class with an estimated utilization mean of 11.1 visits for the remaining 12% of the population. We failed to find any statistically significant elasticity of time cost utilization, when this variable is measured as the total time spent in the health centre. Regarding the effect of an appointment delay on health centre utilization we concluded that individuals respond to this variable by lowering the number of GP visits. This last finding may have policy implications, which will be discussed at the end of the paper. (c) 2005 John Wiley & Sons, Ltd.