UCSF

Background

Hospital length of stay (LOS) and facility discharge are primary drivers of the cost of total knee arthroplasty (TKA). We sought to identify modifiable patient factors that were associated with increased LOS and facility discharge after TKA.

Methods

Prospective data were reviewed from 716 consecutive, primary TKA procedures performed by two arthroplasty surgeons between 2006 and 2012 at a single institution. Preoperative body mass index (BMI), Veterans RAND-12 (VR-12) physical component score (PCS), and hemoglobin level were collected in addition to other adjusters. Multivariate linear and logistic models were constructed to predict LOS and facility discharge, respectively.

Results

After adjustment, higher BMI was associated with increased LOS in a dose-response effect: Compared to normal weight (BMI <25) overweight (25-29.9) was associated with longer LOS by 0.32days (P=0.038), class-I obesity (30-34.9) by 0.33days (P=0.024), class-II obesity (35-39.9) by 0.67days (P=0.012) and class-III obesity (>40) by 1.15days (P<0.001). Class-III obesity was associated with facility discharge (odds ratio=2.08, P=0.008). Poor PCS was associated with increasing LOS: compared to PCS≥50, PCS 20-29 was associated with a LOS increase of 0.40days (P=0.014) and PCS<20 with a LOS increase of 0.64days (P=0.031).

Conclusion

Patient BMI has a dose-response effect in increasing LOS. Poor PCS was associated similarly with increased LOS. These associations for of BMI and PCS suggest that improvement preoperatively, by any amount, may potentially translate to decreased LOS and perhaps lower the cost associated with TKA.