Medicare payments around episodes of inpatient surgery vary as much as 4070% across hospitals. Although such variation is attributable to multiple factors, differential use of physician services is one key component. In this context, we explored the use of medicine consultations after inpatient surgery, factors associated with increased utilization, and variations in practice patterns across hospitals.
We used national Medicare data to identify elderly patients undergoing colectomy or total hip replacement (THR) between January 2005 and November 2007. Hierarchical logistic regression was used to examine factors associated with medicine consult use. A random effect was incorporated to account for variation due to unmeasured hospital factors. To better understand reasons for variations in medical consultation rates among hospitals, we compared use in low risk patients (<75 yo with <= 1 comorbid condition) versus high risk patients (= 80 yo with = 2 comorbid conditions).
Our cohort included 105,810 (244,693) patients undergoing colectomy (THR) at 3,957 (3,447) hospitals. In this sample, 49.2% (47.8%) of colectomy (THR) patients received = 1 general medicine consult. Other common consults were cardiology (21.7%) and oncology (22.0%) after colectomy, and physical medicine and rehabilitation (8.6%) after THR. Factors associated with a greater likelihood of receiving a medicine consult included older age, more comorbidities, and being treated at a nonteaching hospital (Table). The use of medical consultations varied widely across hospitals (Figure). Although high risk patients received substantially more consultations, low risk patients were primarily responsible for wide variation in consultation rates across hospitals.
Postoperative medical consultations are common, but their use varies widely across hospitals. The greatest variation was seen among low risk patients. It is unclear whether this represents an opportunity for hospitals to improve efficiency, or whether these practice patterns benefit patients.
Table 1Factors Associated with Ordering at Least One Medicine Consult, Adjusted
|PATIENT AND HOSPITAL FACTORS||ODDS RATIO FOR ORDERING AT LEAST ONE MEDICINE CONSULT|
|Colectomy||Total Hip Replacement|
|% or Mean||Odds Ratio (95% CI)||% or Mean||Odds Ratio (95% CI)|
|Male||44.2||1.21 (1.181.25)||37.0||1.18 (1.151.20)|
|7074||20.2||1.16 (1.111.22)||26.7||1.16 (1.131.20)|
|7579||24.0||1.46 (1.391.53)||27.0||1.46 (1.411.50)|
|>=80||42.8||2.46 (2.342.58)||28.8||2.18 (2.112.25)|
|Black||7.8||1.20 (1.121.28)||3.9||1.09 (1.031.16)|
|Other||3.3||0.92 (0.841.00)||1.7||0.98 (0.901.07)|
|1||21.2||1.94 (1.832.06)||32.2||1.81 (1.751.87)|
|>=2||71.4||6.05 (5.726.40)||53.1||4.79 (4.644.94)|
|No||83.6||1.89 (1.602.23)||81.9||1.45 (1.131.86)|
|NUMBER OF BEDS|
|200349||31.1||1.43 (1.291.58)||30.7||1.73 (1.502.01)|
|350499||17.6||1.42 (1.231.64)||17.5||1.78 (1.442.21)|
|>=500||18.9||1.21 (1.021.45)||19.8||1.92 (1.472.50)|
|NURSETOCENSUS RATIO||1.51||0.92 (0.900.93)||1.57||0.96 (0.950.98)|
|% MEDICAID DAYS||16||0.22 (0.150.31)||15||0.32 (0.190.53)|
|Urban||90.4||1.14 (0.991.31)||90.9||1.21 (0.991.49)|
|Forprofit||11.2||1.44 (1.281.62)||11.4||1.12 (0.951.32)|
|Other||9.9||0.79 (0.710.89)||8.7||0.81 (0.680.97)|
Figure 1Variation in Hospital Use of Medicine Consults after Colectomy (top panel) and Total Hip Replacement (bottom panel).
To cite this abstract:Birkmeyer J, Chen L, Banerjee M. Hospital Variation in Use of Postoperative Medicine Consultations. Abstract published at Hospital Medicine 2012, April 1-4, San Diego, Calif. Abstract 97650. Journal of Hospital Medicine. 2012; 7 (suppl 2). https://www.shmabstracts.com/abstract/hospital-variation-in-use-of-postoperative-medicine-consultations/. Accessed April 4, 2020.