A review of perioperative practices did not reveal abnormality. Water samples from patient household water did not recover M fortuitum. All water samples collected from the hospital yielded growth of NTMs. Tap water collected from the operative room showed growth of M mucogenicum. M fortuitum was recovered from the shower water in the patient's room and from tap water in the nurse's office. These M fortuitum strains were submitted for genotyping ( Fig 1). The strains collected from the water in the shower in the patient's previous room and from the tap of the nurse's office appeared indistinguishable (>97% similarity; no band differences) and clustered with the clinical specimen, which exhibited >95% similarity (1 band difference) with the strains from the 2 water samples. Control strains were distinctly different from one another and from cluster G1.
The reproducibility of the automated rep-PCR was examined by comparing DNA profiles from the recovered M fortuitum strains. Two rounds of DNA extraction and 5 replicates of rep-PCR were performed, and the profiles obtained from these analyses were similar or undistinguishable. In addition, water samples were collected, and M fortuitum was recovered in water samples from 3 sites outside the department of gynecology. This supported the presence of M fortuitum in the water supply for the whole hospital.