Objective. We studied factors affecting length of hospital stay and re
sulting hospital charges among patients managed by two separate groups
of neonatologists in the same academic health science center. Design.
Retrospective analysis of clinical and financial data base informatio
n. Methods. Neonatal intensive care was provided in the same acute car
e nursery in a large university children's hospital by: (1) neonatolog
ists in a full-time academic division (group A) and (2) a group associ
ated with a private managed care organization (group B). Clinical and
financial parameters of all neonates admitted in fiscal year 1994 were
compared for the two provider groups. Stepwise regression analysis wa
s used to evaluate factors influencing hospital charges and length of
hospital stay (LOS) and to adjust for differences in clinical variable
s between the two groups. Results. Group A physicians provided care fo
r 340 infants, while 137 were treated by those of group B. Group A inc
luded older patients, more outborns, more level III patients, more sep
sis, more intermittent positive pressure ventilation, and more patent
ductus arteriosus. The incidence of transient tachypnea of the newborn
was higher among group B patients. Hospital charges were primarily de
termined by LOS, which was similar for the two groups. When the data w
ere corrected for differences in risk and patient acuity, however, a s
ignificant relationship between physician group and LOS was demonstrat
ed, with LOS being an average of 7.8 days shorter for group A. A net r
eduction of $3 114 969 in hospital charges might have been realized ha
d group A physicians provided care for all study patients. Conclusions
. Hospital charges were determined by LOS. In this setting, academic n
eonatologists produced shorter LOS and comparable clinical outcomes, d
espite caring for a population at greater risk. The reduction in LOS c
ould have resulted in more than $3.1 million in annualized savings had
the academic group provided care for all of the study patients.