T. Andriotti1, E. Goralnick1, M. Jarman1, M. A. Chaudhary1, L. Nguyen3, P. Learn2, A. Haider1, A. Schoenfeld1 1Harvard Medical School,Surgery,Boston, Massachusetts, MASSACHUSETTS, USA 2Uniformed Services University Of The Health Sciences,Surgery,Bethesda, MD, USA 3Harvard Medical School,Vascular And Endovascular Surgery,Boston, Massachusetts, MASSACHUSETTS, USA
Introduction:
Healthcare performance evaluators have prioritized reduction of length of stay (LOS) and readmissions as important measures of quality in health care. However, these two measures represent competing demands as decreased LOS may result in increased unplanned readmissions. Our objective was to assess the optimal LOS that leads to the lowest readmission risk after discharge for knee replacement arthroplasty.
Methods:
A retrospective, open cohort study design was performed using Tricare claims, the Department of Defense’s Health Insurance product, to identify all eligible adult patients (18 – 64 years) who were discharged from elective total knee arthroplasty from 2006-2014. To estimate the optimal timepoint for LOS related to lowest 90-day readmissions, a generalized additive model with spline regression was generated to assess for the predicted risks of readmission (graph 1) adjusted for age, sex, gender, military rank as a proxy of socioeconomic status, any complications during hospital stay and Charlson comorbidity score. Readmissions included stays for all unplanned causes, reported by the principal diagnosis at the index (i.e., initial) inpatient stay within 90 days after discharge from elective total knee arthroplasty.
Results:
11,517 patients (6,910 women and 4,607 men) with a mean [SD] age of 56.94[6.34] years underwent the procedure within the study frame. 50.14% were white, 1.37% were Asian, 9.31% were black, 0.66% were American Indian, 3.12% were other and 35.39% were unknown. The median LOS was 3 days (IQR: 2-3 days) and the main causes of 90-day readmissions were post-operative infection (0.81%), mechanical complication of other internal orthopedic devices (0.41%) and knee lymphedema (0.31%). The lowest risk of being readmitted in 90 days was observed in patients discharged on the 1st day of post-operatory discharge (POD-1) (risk = 4.8%). Moreover, as LOS increases, the risk of readmissions significantly increases up to 9.73% for patients discharged on the 8th day (POD-8) (p=.0004).
Conclusion:
LOS reduction up to one day may not culminate in increased risk of readmissions in patients whose clinical conditions allow them to be discharged on POD-1 after elective total knee arthroplasty. Hence, orthopedists may consider discharge patients with good post-surgical conditions as soon as one day after elective total knee arthroplasty.
