Thousands of sickly newborns could be saved each year if officials closed some of the smaller neonatal intensive care units in the U.S., according to a new study that suggests larger hospitals are better able to treat the infants.

Researchers found that extremely premature babies were up to twice as likely to survive when treated at a busy, advanced-care centre instead of one of the many community hospitals that have opened NICUs in recent years. 

'Size really matters.'— Study's lead author Ciaran Phibbs says hospitals that handle more babies had best survival rates

Even among the most advanced centres, those that handled the most babies had the best survival records, said Ciaran Phibbs, the lead author of the study appearing in Thursday's New England Journal of Medicine.

"Size really matters," said Phibbs, a Stanford University health economist.

Earlier studies found conflicting results when reviewing the relationship between neonatal deaths and number of infants treated by a hospital. But Phibbs' study is the largest and best of its kind, experts said.

"It's quite persuasive," said Dr. David Goodman, a neonatal care specialist at Dartmouth Medical School in Hanover, N.H., who was not involved in the study.

Higher death rate in smaller units

The study reviewed nearly 48,000 premature births and fetal deaths in California from 1991 through 2000, using birth and death certificates and hospital records. Researchers focused on babies with very low birth weights of one to three pounds.

California's top neonatal intensive care units, called Level 3 NICUs, offer the full range of neonatal care and surgery. They had the best survival rates.

Those that treated more than 100 premature babies each year had the lowest death rate, about 18 per cent. In similar facilities that treated 50 to 100 infants a year, about 20 per cent died.

The trend continued in less comprehensive neonatal units. The lower the level of care and the smaller the number of babies, the higher the death rate. In Level 2 NICUs that saw 10 or fewer tiny babies a year, more than 31 per cent died.

The effect was seen across race and size differences in the infants studied, and in both genders.

24-hour anesthesiology, experience help 

Based on the results, researchers estimated that consolidating intensive care units could prevent 21 per cent of deaths among especially small infants.

Why the difference? Hospitals with busy neonatal units also have advanced obstetrics care, including around-the-clock anesthesiology and other services to quickly handle emergencies, Phibbs said. They also have more practice.

"If your clinical group takes care of a lot of patients year after year, you start to identify trends you would miss with less volume. Experience lends knowledge and wisdom, to not just physicians but the entire team," said Dr. Aaron Caughey, a study co-author from the University of California at San Francisco.

Rural concerns

Historically, the expertise and technology needed to run an NICU have been scarce, and care was concentrated at academic medical centres. In some states, laws or regulations confined services to a few hospitals, but in the past 20 years, the number of NICUs boomed — from 578 in 1985 to 838 in 2004, one industry survey found.

One reason is more doctors specializing in neonatal care, Goodman said. Hospitals have also recognized that such care can be lucrative. Virtually all NICU patients have public or private insurance, and hospitals charge a lot for this kind of care. "NICUs make money for hospitals," Caughey said.

Debby Rogers, the vice-president of quality and emergency services for the California Hospital Association, said closing NICUs carries some risk.

Residents of some rural areas have limited access to advanced medical services, and it's better for them to have limited NICU care than no such care at all, she said.

Phibbs said his team considered that argument, and found 92 per cent of the births in 2000 occurred in urban areas with more than 100 such deliveries each year.