The effect of distance on the use of emergency hospital services in a Spanish region with high population dispersion: a multilevel analysis.

BACKGROUND: There is geographic variability in the use of hospital emergency services. The effect of distance to the hospital on this variability is unknown.
OBJECTIVES: (a) To analyze the independent effect of distance on use of hospital emergency services; (b) to describe the variability among municipalities in use of hospital emergency services and to analyze how much of this variability is explained by distance. RESEARCH
DESIGN: Weighted cross-sectional data from the 2003 Regional Health Survey of Castile and Leon were linked with municipal-level data from the 2001 Census, municipal health resources, and distance from municipality to hospital.
SUBJECTS: : Sample of 4281 adults residing in 179 municipalities of the region of Castile and Leon. MEASURES: Using multilevel logistic regression models with random intercept, we analyzed the association between distance to hospital and use of hospital emergency services.
RESULTS: The proportion of the sample using hospital emergency services in the last year was 14.4%. The multivariate analysis showed a significant inverse association between distance to hospital and use of emergency services (P=0.001). Use of hospital emergency services varied widely across municipalities (variance 0.484; standard error 0.132). Some 12.8% of the variability is attributable to differences among municipalities. The model explained 31.6% of the variability.
CONCLUSION: : Distance is a barrier to accessing hospital emergency services. There is large variability among municipalities in the use of emergency services not explained by the model. Variables related with the capacity of primary care facilities to resolve emergencies may reduce part of the observed variability.