BOSTON – Many infants with a fever of unknown origin can be safely discharged from the hospital sooner than the routinely recommended 48 hours, a study has shown.
A review of 11 years’ worth of data found that just 0.5% of these infants’ blood and cerebrospinal fluid cultures took more than 24 hours to come back positive. These patients all had concerning signs on their clinical exam or initial labs, suggesting that infants with a low-risk profile can probably be safely discharged by 24 hours.
"Institutions that are hospitalizing all of these infants for 48 hours can think about changing that to 36 or even 24 hours for infants who have normal labs and look clinically well," Vikram Fielding-Singh said at the annual meeting of the Pediatric Academic Societies.
Mr. Fielding-Singh, a medical student at Stanford (Calif.) University, reviewed the university’s inpatient records from 1999 to 2010. All patients in his analysis were 30 days old or younger, and underwent a diagnostic work-up for fever of unknown origin; tests included blood and CSF cultures.
The study comprised 1,880 cultures taken from 1,145 infants. Of these patients, 483 were hospitalized for fever without a known source.
Of the total group, 31 infants (3%) had positive blood or CSF cultures. Only six (0.5%) of these took more than 24 hours to return a positive result. In the subset of infants hospitalized for a fever work-up, 24 (5%) had positive cultures. Again, just six (1%) returned a positive result after 24 hours.
All of these infants had warning signs of a serious illness, either on clinical exam or in other lab work, Mr. Fielding-Singh said. No infant with normal labs at presentation had a positive culture that took more than 24 hours to return a positive result.
Five infants had bacteremia – four with a concurrent urinary tract infection – and one had bacterial meningitis. All of these patients had concerning lab findings – a positive urine culture result, elevated white blood cell levels, or an increased absolute band count. Five also had other abnormal findings, including pleocytosis, a chest radiograph showing infiltrate, or an elevated C-reactive protein level.
Mr. Fielding-Singh also performed a sequential review of one of every four hospitalizations (121) for the fever of unknown origin work-up. More than half (56%) were considered low risk according to their labs and clinical evaluation, he said.
Dr. Alan R. Schroeder, who supervised the project, said it showed that physicians can "safely do less" with many patients.
"It’s our moral obligation to look at practices that put children at risk and that aren’t necessarily evidence based," said Dr. Schroeder, a pediatric critical care specialist at the Santa Clara Valley Medical Center, San Jose, Calif. "This is a practice that is almost rote for many institutions. These babies get all kinds of tests and are hospitalized for 48 hours almost without much thought."
A routine 48-hour admission for every infant with a presumed fever without source "puts a pretty big burden" on hospital resources, Dr. Schroeder said in an interview, not to mention putting small patients at great risk for nosocomial complications.
"If you could lessen the hospital stay [for] more than half of these kids, that should have a positive impact on resource utilization and certainly on the iatrogenic harms that can occur to these little babies," he said.
Neither Mr. Fielding-Singh nor Dr. Schroeder reported having any relevant financial disclosures.