Objective: To investigate the association of clinical workload and the deci
sion to perform procedures on infants in the neonatal intensive care unit (
Design: Prospective cohort study over one academic year, observing infants
exposed to housestaff working under various levels of clinical workload.
Subjects: All 31 housestaff rotating on the NICU service during the academi
c year 1993 to 1994 were observed. A total of 785 infants were admitted to
Setting: One academic Level III intensive care nursery.
Measurements and Main Results: Clinical workload was operationalized as num
ber of NICU infants cared for by the individual houseofficer on call each n
ight. The procedures of interest were number of umbilical artery catheters
(UACs), intubations, lumbar punctures (LPs), and peripheral phlebotomy perf
ormed by the houseofficer on-call. Using multiple linear regression approac
hes, controlling for the average severity of-illness of each of the NICU in
fants, the experience and residency program of the houseofficer on call, an
d the individual attending, we found that increased clinical workload (numb
er of NICU infants) resulted in a significantly greater probability that an
admitted infant received an umbilical artery catheter (p =.02), but result
ed in less probability that any NICU infant received a lumbar puncture (p =
.0001) or peripheral phlebotomy (p =.0002). The decision to intubate an inf
ant was not affected by the workload in the NICU.
Conclusions: The clinical workload of housestaff in the NICU can affect dec
isions to perform procedures on infants in the NICU. For equivalently sever
ely ill infants, there is a greater chance of receiving a UAC and less chan
ce of being phlebotomized or receiving an LP when workload is high. Attendi
ng neonatologists need to be sensitive to possible effects of workload on p
atient care in the NICU.