ASSOCIATION BETWEEN POSTOPERATIVE WOUND LEAKAGE AND INFECTION AFTER ARTHROPLASTY: RESULTS OF A NATIONAL WOUND CARE APP IMPLEMENTATION STUDY

Background

The duration and extent of postoperative wound leakage after joint arthroplasty in patients with or without a complicated course, like a prosthetic joint infection (PJI), is currently unknown. Adequate differentiation between normal postoperative wound leakage and wound leakage due to a postoperative PJI is important and prevents unnecessary surgical procedures. We investigated the association between postoperative wound leakage and development of PJI in patients who used a previously developed mobile wound care app.

Methods

A multicenter, prospective cohort study with patients aged 18 years or older after primary implantation or revision of a total joint arthroplasty. During 30 post-operative days after arthroplasty, patients recorded their wound status in the woundcare app. An algorithm calculated a daily score from imputed data. If the daily score exceeded a predefined threshold, the patients received an alert that advised them to contact their physician.

Results

Of 1020 included patients from 11 centers, 14 patients developed a PJI. Of 1006 patients without PJI, any form of postoperative wound leakage occurred in 51%, 12%, 7% and 3% during the 1^st, 2^nd, 3^rd and 4^th postoperative week, respectively. Median duration of wound leakage was eight days (IQR 3.5–12.5) for patients with PJI and two days (IQR 0–4) for patients without PJI (p <0.001). In total, 498 (49%) patients received 2589 alerts. Receiving an alert was not predictive for the development of a PJI. The odds ratio for a PJI was higher in patients with wound leakage compared to patients without wound leakage: OR 1.76 (0.59–5.29), OR 45.42 (10.04–205.53), OR 16.76 (3.68–76.28) and OR 18.09 (1.59–205.66) in the 1^st, 2^nd, 3^rd and 4^th postoperative week, respectively.

Conclusion

In patients who received a knee or hip arthroplasty, the odds ratios for PJI were significantly increased in patients with postoperative wound leakage. However, the high absolute number of patients with wound leakage and no PJI showed that wound leakage alone is not a sensitive, i.e. discriminative, indicator to guide the decision whether to reoperate patients for a suspected PJI.

The abstract was submitted on behalf of the Woundcare app study group: H.M.J. van der Linden, Leiden University Medical Centre; D. Broekhuis, Leiden University Medical Centre; M. de Jong, Leiden University Medical Centre; M.R. Benard, Alrijne Hospital; A.P. Wassenaar, Alrijne hospital; A.S.B. Mol, Alrijne Hospital; M. Rutgers, Reinier Haga Orthopedic Center; J.Pasma, Reinier Haga Orthopedic Center; R. Bazuin, Reinier Haga Orthopedic Center; N. Mathijssen, Reinier Haga Orthopedic Center; C.E. Van Der Wijngaart, Onze Lieve Vrouwe Gasthuis; N.W. Willigenburg, Onze Lieve Vrouwe Gasthuis; M.E. Van Der Hoorn, Onze Lieve Vrouwe Gasthuis; B. Dijkstra, Medical Center Leeuwarden; L.D. De Jong, Rijnstate Hospital; H. Haan, University Medical Centre Groningen; M. Stevens, University Medical Centre Groningen; M. Reijman, Erasmus Medical Centre Rotterdam; H. Hoogeboom, Nijsmellinghe Medical Centre; C. Meijer, Park Medical Centre + all authors listed above this article