Investigation of a Cluster of Rapidly-Growing Mycobacteria Infections Associated with Joint Replacement Surgery in a Kentucky Hospital

BACKGROUND: Rapidly-growing mycobacteria are an infrequent cause of post-traumatic and post-surgical wound infections and have been associated with surgical implantation of devices, including joint replacement. At least one outbreak of post-surgical rapidly-growing mycobacteria infections associated with a colonized healthcare worker has been reported. Hospital A in Kentucky notified the Northern Kentucky Independent District Health Department (NKIDHD) of a cluster of five rapidly-growing mycobacteria infections that occurred between January -- September, 2013 among patients of an orthopedic medical practice, all of whom had hip or knee replacement surgery at Hospital A within nine months prior to mycobacteria isolation. The organisms involved, Mycobacterium wolinskyi and Mycobacterium goodii, had never before been isolated from a patient at Hospital A. NKIDHD and the Kentucky Department for Public Health investigated to identify the source of the infections.

METHODS: Case-patient hospital and clinic charts were reviewed to identify common exposures. Surgical techniques and processes were reviewed and environmental samples were collected from non-sterile water sources in the operating rooms and recovery ward. A case-control study was conducted with twenty control patients randomly selected from all joint replacement surgeries conducted at Hospital A between October 2012 and March 2013. Odds ratios were calculated for age, operating room used, day of surgery, time of surgery and each person present in the operating room.

RESULTS: No clinic location, physician, or visit date was common among the five cases. No improper surgical techniques or procedures were identified. Rapidly-growing mycobacteria grew out of environmental samples from operating room scrub sinks, recovery ward ice machines and a portable cold-therapy unit reservoir, but none were M. wolinskyi or M. goodii. Species recovered included M. sphagni, M. mucogenicum and M. abscessus. Case and control patients were similar in gender (3/5 vs. 14/20, respectively, were female) and age distribution (mean of 59 vs. 64 years, respectively). There was no significant association between case status and operating room, weekday of surgery, or time or day of surgery. One operating room healthcare worker—present in 5/5 case surgeries and in 6/20 control surgeries—was significantly associated with patient case status (OR: undefined, 95% CI: 2.1 – undefined, p<0.01).

CONCLUSIONS: The results of this investigation did not identify any environmental sources of infection but did find an association with an operating room healthcare worker. Prevention recommendations given to Hospital A included identification and management of surgical personnel potentially colonized with rapidly-growing mycobacteria and general reinforcement of sterile technique.