Abstract
Aim
This was a multicenter, randomized, clinical trial to compare the 90-day 1) incidence of surgical site complications (SSC); 2) health care utilization (the number of dressing changes, readmission, and reoperation); and 3) the patient-reported outcomes (PRO) in high-risk patients undergoing revision total knee arthroplasty (rTKA) with postoperative closed incision negative pressure wound therapy (ciNPT) versus a standard of care (SOC) silver-impregnated occlusive dressing.
Method
A total of 294 rTKA patients (15 centers) at high-risk for wound complications were prospectively randomized to receive either SOC or ciNPT (n = 147 each). The ciNPT system was adjusted at 125 mmHg of suction. Investigated outcomes were assessed weekly up to 90 days after surgery. A preset interim analysis was conducted at 50% of the intended sample size, with planned discontinuation for clear efficacy/harm if a significance of p < 0.005 was attained.
Results
A total of 242 patients completed the required follow-up (ciNPT: n = 124 (84.4%); SOC: n = 118 (80.3%)). Demographics, baseline comorbidities, causes of revision (prosthetic joint infection, aseptic loosening, implant-related, and periprosthetic fractures), and duration of treatment were similar in both cohorts (p > 0.05). Intention to treat analysis demonstrated lower rates of SSC with ciNPT (3.4%) compared to SOC (14.3%) (p = 0.0013) (Table 1 – not included in the proceeding). Similar outcomes were obtained with the modified intention to treat analysis (p = 0.0013). The ciNPT cohort exhibited lower readmission rates (p = 0.0208), and number of dressing changes (p = 0.0003). Conversely, differences in the 90-day incidence of SSI and measured patient-reported outcomes did not reach statistical significance (p > 0.05).
Conclusions
ciNPT mitigates the risk of SSC and readmission among high-risk rTKA patients. The lower frequency of dressing changes within the ciNPT cohort may provide added value for healthcare utilization without compromising pain and function.
For the table, please contact authors directly.