Utilization of Drains and Association With Outcomes: A Population-Based Study Using National Data on Knee Arthroplasties.

Although surgical drains have been used routinely in total knee arthroplasties (TKAs), results from several large trials have led to recommendations against their use. Because national data are lacking, we aimed at assessing utilization patterns of drains and perioperative outcomes in TKA procedures. We included 1,130,124 TKA procedures from the national claims-based Premier Healthcare Database (2006 to 2016). Patients receiving a drain were compared with those who did not. Multivariable multilevel models measured associations between drain use and blood transfusions, postoperative infections, 30-day readmission, and length/cost of hospitalization. Odds ratios and 95% confidence intervals are reported. Propensity score analyses were performed to assess the robustness of results. Drain use decreased from 33.0% (n = 22,901 of 69,370) in 2006 to 15.6% (n = 19,418 of 124,440) in 2016 and was particularly higher in large (>500 beds; 27.1%) and nonteaching hospitals (26.9%). After adjustment for relevant covariates, the use of drains (compared with no use) was significantly associated with increases in particularly blood transfusions (odds ratio, 1.27; 95% confidence interval, 1.24 to 1.30 n = 138,306 total transfusions), whereas minimal effects were seen for other outcomes. Propensity score analyses confirmed these results. Although retrospective, the current study provides an important insight into real-world clinical practice regarding the use of drains. With current evidence not supporting their use in TKA, we found that national utilization is slowly decreasing. Moreover, because drain use is associated with negative outcomes, future studies should focus on drivers of their continued use. Level III, therapeutic study.