Abstract
Objective The objective of this study was to examine the time trend in length of stay (LOS)
and explore potential differences in neonatal LOS by insurance type for preterm infants
in Arkansas between 2004 and 2010.
Study Design There were 18,712 preterm infants included in our analyses. Accelerated failure time
models were used to model neonatal LOS as a function of insurance type and discharge
year while adjusting for key maternal and infant characteristics, and complication/anomaly
indicators.
Results Before adjusting for the complication/anomaly indicators, the LOS for preterm infants
delivered to mothers in the Medicaid group was 3.2% shorter than those in the private
payer group. Furthermore, each subsequent year was associated with a 1.6% increase
in the expected LOS. However, after accounting for complications and anomalies, insurance
coverage differences in neonatal LOS were not statistically significant while the
trend in LOS persisted at a 0.59% increase for each succeeding year.
Conclusion All of the apparent differences in LOS by insurance type and more than half of the
apparent increase in LOS over time are accounted for by higher rates of complications
among privately insured preterm infants and increasing rates of complications for
all surviving preterm infants between 2004 and 2010.
Keywords
length of stay - preterm infants - accelerated failure time model - time trend
