Background:

Internists play an increasingly important role in the medical management of orthopedic surgery cases. A randomized controlled trial of patients undergoing total hip and knee replacement found that those who were comanaged by internal medicine hospitalists had fewer minor complications than patients managed by orthopedists only. Length of stay (LOS) was unchanged in this study. At our specialized orthopedic facility, a hospitalist/internal medicine service began in September, 2010 to consult on orthopedic and rehabilitation patients if they developed medical complications. There was concern that the new service might be associated with increased costs and LOS. This report describes the effect of our hospitalist consult service on costs, readmissions, and LOS among patients undergoing elective total hip and knee replacement.

Methods:

In a retrospective observational study, we reviewed the electronic medical records (EMR) of all patients hospitalized for elective total hip or knee arthroplasty in calendar 2011. Data came both from the institution's EMR and from the University Hospitals Consortium (UHC) database. To reduce bias because of differential selection of sicker patients for consultation, the study used the UHC's calculation of individual patient severity of illness and ICD‐9 codes for hospital‐acquired conditions to derive a propensity score. This score measured likelihood that a patient would be selected for consultation. Study outcomes, adjusted for propensity score, were the UHC‐reported ratios of observed‐to‐expected LOS (LOS O:E.) and direct cost (direct cost O:E.). In addition, we compared propensity‐score‐adjusted rates of 30‐day readmission among consultation and nonconsultation patients.

Results:

There were 1757 patients who underwent a total hip or knee arthroplasty according to both the UHC database and the EMR during calendar year 2011. Of these, 320 were seen in consultation by a hospitalist and 1437 were not. After adjustment for propensity score, patients seen by a hospitalist had an average increase in LOS O:E of 0.07 (P = 0.14). For direct cost O:E the increase was 0.02 (P = 0.55). Readmissions were lower among consulted patients by 0.4% (P = 0.72).

Conclusions:

We found no important or statistically significant effect of hospitalist consultation on LOS O:E, direct cost O:E, and readmissions within 30 days. These results do not support an adverse effect of internist consultation on LOS or cost, but neither do they demonstrate improved care. It is possible that the team model reported in the above‐referenced trial may be more effective than reactive consultation; early and ongoing involvement of medical hospitalists on orthopedic teams may reduce the number of complications and improve their management.

Patient Summary



Total (n = 1757) Not Seen by Hospitalist (n = 1437) Seen by Hospitalist (n = 320) Estimate of Effect of Being Seen by Hospitalist
Average age 63 62 63
Female 1142 (65%) 924 (64%) 218 (67%)
Mortality risk: minor 1569 (89%) 1306 (90%) 263 (81%)
Mortality risk: moderate 182 (10%) 126 (8%) 56 (17%)
Mortality risk: major 6 (0%) 5 (0%) 1 (0%)
Severity of Illness: minor 992 (57%) 853 (59%) 139 (42%)
Severity of Illness: moderate 697 (40%) 541 (37%) 156 (48%)
Severity of Illness: major 68 (4%) 43 (2%) 25 (8%)
LOS, mean (median) 4.11 (4) 3.90 (3) 5.13 (4)
LOS O:E, mean (median) 1.23 (1.09) 1.18 (1.06) 1.46 (1.29) 0.066 (P = 0.142)
Direct cost $14,346 ($13,390) $14,024 ($13,147) $16,030 ($14,623)
Direct cost O:E 1.19 (1.13) 1.17 (1.10) 1.28 (1.21) 0.017 (P = 0.553)
Readmissions within 30 days 39 (2.3%) 30 (2.1%) 9 (2.8%) −0.004 (P = 0.72)