Objective. To describe the structure and organization of pediatric int
ensive care units (ICUs) in the United States. Survey Methods. We dire
cted a mail survey to pediatric ICU medical and nursing directors and
hospital quality assurance officers. A total of 201 of 301 hospitals w
ith pediatric ICUs initially responded. Telephone confirmation of the
mail survey (n = 193) and telephone data collection for mail survey no
nresponders (n = 42) were also undertaken. Survey Results. The largest
proportion (40.0%) of pediatric ICUs had four to six beds per unit, w
hile only 6.0% had >18 beds per unit. The admissions per year averaged
528 +/- 24, and the mortality rates averaged 5.5 +/- 0.2%. Only 79.6%
of the pediatric ICUs had full-time medical directors. A pediatric in
tensivist was available to 73.2% of the units. Physician coverage for
24 hrs/day dedicated only to the pediatric ICU was present in 48.5% of
hospitals. As ICU size increased, the estimated mortality rates incre
ased, as did the percentages with full-time directors, pediatric inten
sivists, and 24 hrs/day dedicated coverage. Medical school affiliation
existed for 79.6% of pediatric ICU hospitals, and 81.1% of these hosp
itals were the primary teaching program sites for pediatrics. Other IC
Us caring for children were present in 30.2% of the hospitals. Survey
Application: The mail survey respondents were stratified using four fa
ctors: size, teaching status, intensivist status, and coordination of
care status. A total of 16 respondents were randomly selected for an o
ngoing outcomes study of the importance of these factors. Conclusions:
Substantial diversity exists in pediatric ICU structure and organizat
ion. Determining factors associated with quality of care is important
for improving outcomes.