Infants' infectious disease rate still significant

March 1, 2008

An AHRQ investigation found too many US newborns are getting seriously sick.

Information from a relatively new data system on children's hospital use shows the nation's infectious-disease hospitalization rate for infants is still significant.

There was one such discharge for every 14 infants in 2003. Infectious disease was the primary diagnosis for 43% of all hospitalizations for infants under a year, according to the new report from the Kids' Inpatient Database (KID). Hospitalizations for births were not included in the study.

Overall, there were an estimated 7,011 infant infectious disease (ID) hospitalizations per 100,000 live births in 2003, said the report. And among those, by far the largest number of first-listed diagnoses was lower respiratory tract infections (LRTIs), with a rate of 4,135 hospitalizations.

The KID, according to its co-sponsor the federal Agency for Healthcare Research and Quality (AHRQ), is the only dataset on hospital use, outcomes, and charges targeted at children's hospital use. And the new analysis, published in the February issue of Pediatrics, is based on the first KID release that has samplings from more than three-quarters of the country-36 states in total. A KID release for the year 1997 covered 22 states, and a release for the year 2000 covered 27 states.

Claudia Steiner, MD, MPH, of AHRQ notes that the database does not capture what happens before a hospitalization and can't pinpoint whether some of the episodes were preventable. However, she notes that infectious disease "explains a pretty substantial proportion of infants that are admitted to the hospital." The findings justify more attention to the topic, she said.

The analysis, led by Centers for Disease Control and Prevention researchers, notes that a higher level of infant ID hospitalizations appears to be associated with a lower family income. About 36% of those infants came from zip code areas where the average annual income was under $36,000. For 58% of them, Medicaid was the primary payer.

The numbers also show that ID infant hospitalizations peaked during the second month of life, with almost 19% occurring then. Indeed, close to half of all ID hospitalizations for the first year happen in the first three months. After that, the rate drops off.

Steiner says the numbers point to areas for investigation: "Where are the places that you could make a difference?" Socio-economic factors could influence these rates. So could patient education on issues such as behavioral factors and increased vaccinations.

"Can we encourage more breastfeeding?" Steiner added. "Because, we know that makes a difference."

In other findings, the Pediatrics report says the median length of stay for infant ID hospitalizations was three days, and the median estimated hospital cost was $2,235. Total charges for these episodes, including hospital and other costs, were an estimated $3 billion.

Lower respiratory tract diseases account for 59% of these hospitalizations, the authors note. This is consistent with respiratory syncytial virus's (RSV) reputation as the most common viral cause of LRTIs in infants. RSV bronchiolitis is also the leading primary diagnosis for infants' hospital discharges.

"Studies showed that the rate of bronchiolitis-associated hospitalization increased among US infants during 1980 to 1996 and during 1999 to 2001, indicating a need for continued surveillance and research into potential vaccines and treatments," says the report.

The third leading cause of ID infant hospitalizations was septicemia, with a rate of 458 hospitalizations per 100,000 live births. The great majority of those, 309 per 100,000, occurred in infants less than a month old.

The authors note that septicemia occurs more frequently among low-birth-weight and preterm infants, and, "Nearly three fourths of in-hospital deaths among neonates were attributable to sepsis in the newborn."

The database, according to AHRQ, is developed through state data partners, such as hospital associations, private data organizations, and publicly funded organizations. Steiner says the KIDS dataset is scheduled to be released every three years, with the 2006 release due in a few months. The data sets, which are part of the Healthcare Costs and Utilization Project, are publicly available, albeit with some fees involved, through the AHRQ Web site,