OBJECTIVE: To identify risk factors that might explain a sudden increase in the rate of surgical-site complications following laminectomy.
DESIGN: Retrospective cohort study.
PATIENTS AND SETTING: Patients who underwent laminectomy at a 120-bed hospital from August 1 through October 15, 1996 (the epidemic period). A case-patient was defined as a patient with postoperative surgical-site complications (surgical-site drainage, edema, or swelling) requiring surgical debridement.
RESULTS: Of the 148 patients who underwent a laminectomy during the epidemic period, 17 (11%) met our case definition. The rate of postoperative surgical debridement was 7.6-fold higher during the epidemic period than the preceding 19-month period (17/148 vs 15/995, P<.001). Development of surgical-site complications was associated with intraoperative receipt of morphine nerve paste (relative risk [RR], 11; P<.001), preoperative shaving by nurses rather than surgeons (RR, 6.6; P=.006), procedures done by a certain surgeon (RR, 3.1; P=.022), or receipt of iodine rather than povidone-iodine for preoperative skin antisepsis (RR, 5.1; P=.002). In multivariate analysis, only receipt of morphine nerve paste remained as a risk factor (RR, 18; P=.011). The paste was used to control postoperative pain and was applied directly to exposed dura and surrounding tissues. At the time of surgical debridement (median, 24 days postsurgery), the original surgical sites showed residual paste and a lack of healing. Ten of 16 cultures from surgical sites were positive; all but three grew skin commensals. Histological examination of surgical specimens showed a foreign-body reaction, but no marked acute inflammation.
CONCLUSIONS: The intraoperative use of morphine nerve paste may delay wound healing and increase postoperative morbidity. When new products are introduced, standardized protocols should be developed for their use, and systematic surveillance should be done to monitor for potential adverse outcomes.
Kramer MH, Mangram AJ, Pearson ML, Jarvis WR
Infect Control Hosp Epidemiol 1999 Mar;20(3):183-6