BackgroundPostoperative use of oral prednisone to augment the effect of multimodal pain regimens after total knee arthroplasty (TKA) has increased in popularity. However, data on the risks of its utilization, especially as it relates to infection, have been lacking. We tested the null hypothesis that perioperative prednisone use is not associated with the incidence of surgical and medical complications after TKA. MethodsUsing a national administrative claims database, we identified 949,555 patients undergoing primary TKA. We excluded patients who filled oral prednisone prescriptions within 90 days prior to surgery or between 90 and 364 days after surgery. Patients who had acute prednisone use were defined as those who filled prednisone prescriptions only within 30 days after surgery. Outcomes consisted of surgical and medical complications after TKA. Multivariable logistic regression models were used to evaluate the association between acute prednisone use and complications, adjusting for age, sex, region, insurance plan, and Elixhauser comorbidities. ResultsPatients in the acute prednisone cohort had greater adjusted odds of subsequent manipulation under anesthesia (adjusted odds ratio [OR] = 1.23 [95% confidence interval {CI}: 1.09 to 1.38]; P < .001) and lysis of adhesions (adjusted OR = 1.58 [95% CI: 1.02 to 2.33]; P = .03) compared to patients who did not have acute prednisone use. Patients who had acute prednisone use also had greater adjusted odds of acute kidney injury (adjusted OR = 1.47 [95% CI: 1.25 to 1.71]; P < .001) and pneumonia (adjusted OR = 4.04 [95% CI: 3.53 to 4.59]; P < .001). There was no increased incidence of infection. ConclusionsPrednisone use shortly following TKA may be associated with a higher incidence of certain surgical and medical complications, but without increased risk for infection. However, given these risks, the optimal patient profile for postoperative prednisone use remains to be defined.
Read full abstract