Length of stay, pain and opioid consumption following manual versus robotic-assisted total hip arthroplasty

Abstract

Background

Robotic assistance in total hip arthroplasty (THA) has increased, but the influence on outcomes compared to manual THA remains uncertain. With the growing emphasis on reducing opioid consumption after arthroplasty, we studied whether robotic assistance was associated with length of stay (LOS), pain, and opioid use after THA.

Materials and methods

We included 14,501 opioid-naïve patients who underwent THA at a single institution between 2019 and 2023 (8900 manual and 5601 robotic). In-hospital pain scores (NRS), LOS, and opioid consumption patterns were collected. Opioid dosages were converted to morphine milligram equivalents (MMEs). After preliminary bivariate analysis, multivariable linear regression analyses were performed adjusting for age, sex, race, BMI, ASA-class, smoking status, cement use, marital status, year of surgery, surgeon experience, approach and periarticular injection.

Results

Robotic THA was associated with significantly shorter LOS (Estimate: 6.8 h, 95 %CI: 8.0, −5.6, p < 0.0001). Robotic THA patients had higher minimal and mean pain scores (Estimate: 0.03, 95 %CI: 0.02–0.05, p < 0.001; Estimate: 0.08, 95 %CI: 0.03, 0.14, p = 0.0042). Robotic THA patients used less MMEs per hour of hospitalization (Estimate −0.11 MMEs, 95 %CI -0.174, −0.039, p = 0.0021), but were prescribed more MMEs at discharge (Estimate: 3.59 MMEs, 95 %CI: 0.323, 6.856, p = 0.0312). The differences in MMEs refilled after discharge and total 90-day opioid prescription patterns were not significant.

Conclusion

Robotic assistance in THA was independently associated with a slightly shorter LOS. The significantly higher pain scores (0.08 points of NRS) and lower in-hospital opioid consumption (0.11 MMEs/hour) suggest that while some statistically significant differences exist between robotic-assisted and manual THA, these differences may not be clinically meaningful.