R Kaestner1, A Racine, T Joyce. 1. Baruch College, New York, New York, USA. robert_kaestner@baruch.cuny.edu
Abstract
OBJECTIVE: To test whether socioeconomic differences in the ratio of infant hospitalizations to births, a proxy for infant hospitalization rates, and hospital lengths of stay for infants narrowed between 1988 and 1992: a period of large increases in the numbers of low-income infants enrolled in Medicaid. RESEARCH DESIGN: Before and after comparison of socioeconomic differences in the ratio of infant hospitalizations to births (ie, infant hospitalization rates) and lengths of stay between 1988 and 1992. By use of ICD-9 codes, hospitalizations were categorized as mandatory or discretionary. The difference between the 2 is that discretionary hospitalizations are potentially avoidable with appropriate primary care. Difference-in-differences techniques were used to assess the differential change in the rates of hospitalizations and lengths of stay for infants from low-income, compared with high-income, zip codes. SETTING AND PARTICIPANTS: Discharges of infants <2 years of age at 326 nonfederal, short-term, general, and other specialty hospitals in 8 states. OUTCOME MEASURES: Ratios of discretionary and mandatory hospitalizations to births (ie, hospitalization rates) and hospital lengths of stay of infants <2 years of age. RESULTS: Infants from the poorest zip codes had ratios of discretionary hospitalizations to births (discretionary hospitalization rate) that were 3.1% points higher than infants from the wealthiest zip codes and ratios of mandatory hospitalizations to births (mandatory hospitalization rates) that were 0.2% points higher. Poor versus nonpoor differences in lengths of stay were 0.3 and 1.9 days for discretionary and mandatory hospitalizations, respectively. No narrowing in the socioeconomic gradients about ratios of hospitalizations to births (ie, rates of hospitalization) or lengths of stay was observed. CONCLUSIONS: Expansions in the Medicaid program from 1988 to 1992 did not result in a decrease in ratios of discretionary hospitalizations to births (ie, discretionary hospitalization rate) or hospital length of stay for infants from low-income areas.
OBJECTIVE: To test whether socioeconomic differences in the ratio of infant hospitalizations to births, a proxy for infant hospitalization rates, and hospital lengths of stay for infants narrowed between 1988 and 1992: a period of large increases in the numbers of low-income infants enrolled in Medicaid. RESEARCH DESIGN: Before and after comparison of socioeconomic differences in the ratio of infant hospitalizations to births (ie, infant hospitalization rates) and lengths of stay between 1988 and 1992. By use of ICD-9 codes, hospitalizations were categorized as mandatory or discretionary. The difference between the 2 is that discretionary hospitalizations are potentially avoidable with appropriate primary care. Difference-in-differences techniques were used to assess the differential change in the rates of hospitalizations and lengths of stay for infants from low-income, compared with high-income, zip codes. SETTING AND PARTICIPANTS: Discharges of infants <2 years of age at 326 nonfederal, short-term, general, and other specialty hospitals in 8 states. OUTCOME MEASURES: Ratios of discretionary and mandatory hospitalizations to births (ie, hospitalization rates) and hospital lengths of stay of infants <2 years of age. RESULTS:Infants from the poorest zip codes had ratios of discretionary hospitalizations to births (discretionary hospitalization rate) that were 3.1% points higher than infants from the wealthiest zip codes and ratios of mandatory hospitalizations to births (mandatory hospitalization rates) that were 0.2% points higher. Poor versus nonpoor differences in lengths of stay were 0.3 and 1.9 days for discretionary and mandatory hospitalizations, respectively. No narrowing in the socioeconomic gradients about ratios of hospitalizations to births (ie, rates of hospitalization) or lengths of stay was observed. CONCLUSIONS: Expansions in the Medicaid program from 1988 to 1992 did not result in a decrease in ratios of discretionary hospitalizations to births (ie, discretionary hospitalization rate) or hospital length of stay for infants from low-income areas.