Background: In recent years, there has been a move toward value-based health care. Value is generally defined as outcome divided by cost; however, it is not clear exactly how to define and measure outcomes. In this study, we utilized the Nationwide Inpatient Sample (NIS) to determine how hospital volume and other factors affect quality for patients undergoing total hip and knee arthroplasty.
Methods: Using the NIS of the Healthcare Cost and Utilization Project (HCUP) of the Agency for Healthcare Research and Quality (AHRQ), we conducted a retrospective study of all total hip and total knee arthroplasties performed from 2001 to 2011. We identified all procedure and outcome variables using the International Classification of Diseases, Ninth Revision (ICD-9) billing codes. Patients were grouped into quartiles based on the corresponding hospital's procedure volume. The quality measurement for each hospitalization was binary, with perfect inpatient care reflecting a favorable result for all of the following outcomes of interest: death, sepsis, postoperative infection, thromboembolic events, venous thrombosis, hematoma, blood transfusion, and length of stay below average. The Perfect Inpatient Care Index (PICI) was then calculated for each hospital. The PICI was defined as the number of hospitalizations with no unfavorable outcomes divided by total volume of arthroplasty. Value was measured as the PICI divided by the mean total charges. Multivariable nested regression was used to determine variables that predict perfect inpatient care.
Results: From 2001 to 2011, the NIS database reported 1,651,354 total hip or total knee arthroplasties. Hospital arthroplasty volume ranged from 0 to 11,758 procedures. Overall, hospital PICI scores increased as arthroplasty volume increased. In multivariable nested regression analysis, procedure volume (odds ratio [OR] for the highest quartile compared with the lowest quartile, 2.116 [95% confidence interval (CI), 1.883 to 2.378]) and lower patient acuity (OR, 2.450 [95% CI, 2.429 to 2.472]) were independently associated with better PICI scores. Value increased as hospital procedure volume increased.
Conclusions: Hospital procedure volume varied widely. Although small differences were seen in individual outcome measures, composite scores (PICI) and value were substantially better at hospitals that had higher procedure volume and in lower-acuity patients.
Level of evidence: Economic Level IV. See Instructions for Authors for a complete description of levels of evidence.