Objective: To evaluate an inpatient physician system initiated in lune 1996
for all patients of a health maintenance organization admitted to the gene
ral medicine service of an urban teaching hospital. In the new program, att
ending physician duties were transferred from the patient's own general int
ernist to another internist sewing on a hospital-based rotation.
Study Design: Cohort with historical controls.
Participants and Methods:. We compared the following measures before and af
ter the new inpatient physician program began: (1) hospital length of stay
and total charges, (2) outcomes related to quality of care, (3) primary car
e physician satisfaction, and (4) housestaff satisfaction. Differences befo
re and after initiation of the inpatient physician program were evaluated u
sing multivariate analyses to adjust for patient differences and secular tr
ends.
Results: There were 2265 patients discharged from the general medical servi
ce in the year following implementation of the inpatient physician program.
Postintervention average length of stay decreased from 3.5 to 3.0 days (P
<.001). In multivariate analyses, average length of stay was reduced by 0.3
days (P =.008), and total hospital charges were reduced an average of $426
per admission (P=.001). In-hospital mortality rates, percentage of patient
s discharged home directly, and 30-day read-mission rates did not change si
gnificantly in the postintervention period. Satisfaction among primary care
physicians was high, with 90% of those answering a survey responding that
they would recommend a similar program to other primary care groups. Medica
l housestaff satisfaction with their educational experience also increased.
Conclusions: implementation of an inpatient physician program at this insti
tution significantly decreased resource utilization while maintaining or im
proving quality of care. Satisfaction with the program was high among prima
ry care internists and housestaff.