For Some Premature Infants Smaller Hospitals Work Just Fine

Analysis of Patterns of obstetric referrals in eight states

In 1999, researchers at the Harvard University School of Public Health studied recent changes in obstetric referral patterns, their impact on infant mortality and how these patterns are associated with changes in the market environment.

The study obtained data for six states—Alabama, California, Illinois, Indiana, Pennsylvania and South Carolina—and merged it with data from other sources.

Due to delays in data provision by the states, researchers were unable to complete analysis during the grant period. By October 2002, they had completed a partial analysis of data from Pennsylvania.

Key Findings

In Pennsylvania, researchers found that:

  • Small neonatal intensive care units (fewer than 20 beds) perform no differently than larger centers (more than 20 beds) for infants who weigh 1,000 grams or more.
  • Larger units consistently perform better for infants who weigh less than 1,000 grams.
  • There was a 20 percent increase in the number of infants born weighing less than 1,000 grams at small neonatal intensive care units in the periods from 1985 to 1989 and from 1990 to 1994.
  • Newborn mortality rates decreased in small neonatal intensive care units, large units and hospitals without neonatal intensive care units between 1985 and 1989 and from 1990 to 1994. The reduction was largest for hospitals without neonatal intensive care units.

Key Conclusions

The researchers concluded that:

  • Regulatory action may be effective in ensuring perinatal regionalization.
  • Small neonatal intensive care units are effective for infants who weigh 1,000 grams or more.
  • Regional referral to large units is valid for infants who weigh less than 1,000 grams.
  • The rise in birth rates of infants who weigh less than 1,000 grams is a serious health policy issue.
  • The cost-access tradeoffs (i.e., having many small units provides quicker local access, but unnecessary duplication and cost) of multiple small urban neonatal intensive care units requires investigation.
  • The dramatic reduction in neonatal mortality rates in hospitals without neonatal intensive care units may be one benefit of an improved ability to stabilize and transfer these infants to hospitals with units.