Identifying Determinants of Preventable Hospital Admissions

Objectives: To discover the determinants of preventable hospital admissions. This research project will operate under the hypothesis that regional and local healthcare resources influence preventable hospital admissions. More specifically, that a higher density of primary care physicians and services in an area is associated with a decreased likelihood of preventable hospital admissions for that area. Materials and Methods: The 2006 National Inpatient Sample (NIS) was analyzed with the PQI tool provided by the Agency for Healthcare Research and Quality (AHRQ) to identify preventable hospitalizations. The PQI analysis of the NIS was then merged with the 2006 Area Resource File (ARF). A logistic regression analysis on the NIS + PQI + ARF file was then performed with Stata statistical software to find the association between population/physician ratios and preventable hospitalizations. Results: 901,403 of the 8,051,045 identifiable discharges in the 2006 NIS were preventable (11.20%). These preventable hospitalizations amounted to 4,951,335.27 total days spent in the hospital and $21.2 billion in charges. Logistic regression analysis of the population/physician ratio returned an odds ratio (OR) of 2.024 (p<0.001, 95% CI 1.963 to 2.087) for all preventable admissions. Of the 14 individual PQI's, 11 returned significant results with OR's>1.000 indicating that preventable admissions have a positive association to the population/physician ratio in an area. Conclusions: Preventable hospitalizations pose a significant problem in the United States and cost billions of dollars each year. The likelihood of a patient having a preventable hospitalization is positively associated with the population/physician ratio, as the number of people per physician in an area increases, so too does the likelihood of preventable hospitalization.