Perinatal mortality: the role of hospital of birth.

OBJECTIVE

Patterns of perinatal mortality in a state with no organized system or guidelines for regionalized perinatal health care were examined.

STUDY DESIGN

Vital statistics on live births, fetal deaths, and neonatal deaths in Arkansas for 1985 through 1989 were analyzed by birth hospital level of obstetric care, birth weight, and county of residence. Birth weight-specific fetal and neonatal mortality rates (NNMR) were compared by level of obstetric care at the birth facility; multivariate logistic regression models estimated the odds of NNMR by level of obstetric hospital, adjusted for other risk factors.

RESULTS

Infants of < 2000 gm birth weight tend to have lower NNMR when born at level III hospitals than when born at level II facilities. The proportion of very low birth weight (< 1500 gm) and moderately low birth weight (1500 to 2499 gm) infants delivered at level III hospitals in Arkansas is lower than that in other states with formal systems of perinatal regionalization.

CONCLUSIONS

Although Arkansas's NNMR compares favorably with the national average, further reductions would be likely were a formal system of perinatal regionalization implemented in this state.