Temporal Relations of Unplanned Readmissions Following Total Knee Arthroplasty: A Study of Large State Inpatient Databases.

BACKGROUND: Centers for Medicare & Medicaid Services stipulate a 90-day global period for hospitals for unplanned readmissions after primary total knee arthroplasty (TKA). However, not all readmissions are directly attributable to index surgery, and reasons for readmissions vary during this time period. This study identifies causes and temporal relations of readmissions using large state inpatient databases.

METHODS: State inpatient databases of New York and California were queried for all primary TKAs performed from 2005 to 2011 and frequencies of all causes of unplanned readmission were identified from 0 to 90 days after index surgery using the International Classification of Diseases, Ninth Revision, codes. Temporal differences in proportions of readmission diagnoses were tested using the Pearson chi-square test.

RESULTS: The query identified 419,805 cases of primary TKA from 2005 to 2011. There were 26,924 readmissions during the 90-day recovery period, with 15,547 (57.7%) at 0-30 days, 6593 (24.5%) at 31-60 days, and 4784 (17.8%) at 61-90 days. Primary diagnoses at readmission that were identified to be directly attributable to surgery comprised 38.3% readmissions at 0-30 days, 24.0% at 31-60 days, and 16.3% at 60-90 days. Proportion of readmissions directly attributable to surgery decreased over the 90-day period after index surgery.

CONCLUSION: From this analysis of 2 large state inpatient databases, primary diagnoses at readmission vary with time, and majority of these may not be directly attributable to index surgery or postoperative state up to 90 days. These findings suggest that the current 90-day global period policy for this procedure should be reformed to better reflect the profile of unplanned readmissions after TKA.