Abstract
Background
Conflicting results about the impact of blood transfusions on outcomes after total knee arthroplasty (TKA) have been reported. We hypothesized that transfusions would be associated with greater readmission and complication rates after primary TKA.
Methods
We conducted a retrospective cohort study of the 100% 2008 Medicare Provider Analysis and Review database, and identified primary THA patients by ICD9 codes and excluded fractures/ER admissions to select for elective cases. Patients who received a perioperative blood transfusion (6,951 patients) were compared to a control group who did not receive transfusion (332,762 patients). Descriptive statistics of age, sex, race, diagnosis for surgery, Elixhauser comorbidities, mortality (inpatient, 30, 60, and 90 days and 2 years), readmissions (30, 60, and 90 days), complications (medical and surgical at 30 and 90 days), and revision at 2 years were assessed for both groups. Continuous variables were compared with Student's T-test and categorical variables with chi-square test. Multivariate logistic regression models were constructed to assess the association of transfusion with readmissions, complications, and revisions. Statistical significance was set at p < 0.01.
Results
Patients who received a transfusion were older (mean 74.4 vs. 72.6, p<0.0001), more likely to be male (75% vs. 66%, p<0.0001), and had a higher Elixhauser comorbidity count (2.0 vs. 1.85, p<0.0001). Transfused patients had significantly greater readmission rates at 30-days (8.2% vs 5.7%, p<0.0001), 60-days (11.7% vs 8.2%, p<0.0001), and 90-days (14.4% vs 10.4%, p<0.0001). Their overall complication rates at 30 days (2.1% vs 1.4%, p<0.0001) and 90 days (3.2% vs 2.1%, p<0.0001) were greater mainly due to greater surgical complications at 30 days (1.4% vs 0.9%, p<0.0001) and 90 days (2.1% vs 1.3%, p<0.0001). Patients who received a transfusion had significantly higher mortality rates at 2-years (2.2% vs. 1.4%, p<0.0001). Two years after discharge, transfused patients had no difference in revision rates (2.4% vs 2.4%, p=0.8805). Multivariate regression found that transfusion was the third most important factor for surgical complications at 30-days (OR: 1.6, 95% CI: 1.3 to 2.0) and an independent risk factor for readmissions at 60-days (OR: 1.4, 95% CI: 1.3 to 1.5). At 2-years, transfusion was independently associated with mortality (OR: 1.4, 95% CI: 1.2 to 1.7), but not revision (OR: 1.1, 95% CI: 0.9 to 1.3).
Conclusion
Primary TKA patients who were transfused had an independently higher risk for surgical complications and readmissions at 30 and 90 days, and mortality within 2 years. Transfusion, however, did not impact revision rates. A restrictive transfusion threshold should be considered for patients undergoing TKA.
