Initial Clinic Response and Laboratory Report to Health Department
The clinic employed one ophthalmologist and consisted of a single treatment room. Patient A had surgery on day 0 and experienced symptoms 2 days later (Figure 1). Patients B and C had surgery on day 7 and became symptomatic 3 days later, on day 10. After identifying infections in patients A and B (on days 5 and 11, respectively), the clinic paused surgeries. A third infection, in patient C, was identified the next day (day 12). Clinic staff members conducted a round of environmental cleaning, discarded or replaced some equipment, and retained other equipment for future environmental testing. Retained items included the procedure room refrigerator, and an epikeratome, a surgical device that uses suction to separate the corneal epithelium from the underlying corneal layers. The epikeratome consists of the machine that generates suction, suction tubing (Figure 2), and a handheld attachment that touches the eye during surgery which is sterilized after each use.
On day 13, the laboratory detected mold in the clinical specimens and notified the health department. The following day, the mold in patient A’s culture was identified as P. lilacinum; P. lilacinum was detected in a second patient’s culture shortly thereafter.
In accordance with standard practice for an outbreak investigation, the NYC Health Department’s institutional review board did not review this activity. This activity was reviewed by CDC, deemed not research, and was conducted consistent with applicable federal law and CDC policy.*
A case of P. lilacinum keratitis was defined by the NYC Health Department as eye pain or vision loss in a patient after laser eye surgery at the clinic on days 0-13 and 1) corneal cultures demonstrating P. lilacinum or 2) scrapings identifying fungal elements. Cases were identified through medical record review and clinician interviews. Among 11 patients who had laser eye surgery during the analysis period, three patients, who had surgery on day 0 (patient A) and day 7 (patients B and C), experienced illness meeting the case definition 2-3 days after surgery (Figure 1). Two of these three patients received a positive P. lilacinum culture result and one patient’s corneal scrapings identified fungal elements. All three ill patients experienced vision loss, and one of the two who received a positive P. lilacinum culture result required corneal transplantation. After identification of the fungal infection, the ill patients were started on topical voriconazole and natamycin. Eventually, all were transitioned to oral posaconazole. Patients sought care from multiple providers, and information on clinical outcomes is not available.
Based on the limited demographic and clinical information collected by the clinic, the health department reviewed information on the age, sex, and comorbidities of the 11 patients who underwent surgery during the analysis period. Surgeries at the clinic were conducted in a highly protocolized manner. All patients treated during the analysis period were in contact with the same surgical equipment in the same operative environment. The surgical procedures for all three ill patients, and at least four of the eight patients who did not become ill, involved use of the epikeratome. Of the eight patients who did not become ill, two had surgery on day 0 and six had surgery on day 7. Patients received the same perioperative medications, including three types of eye drops (steroid, nonsteroidal anti-inflammatory, and antibiotic). None of the patients was known to be immunocompromised, and no apparent demographic, clinical, or procedural differences were identified between patients who did and did not become ill.
On day 15, health department staff members visited the clinic to assess infection prevention and control (IPC) practices. The IPC assessment was conducted using CDC’s Infection Control Assessment and Response tool (4), which includes modules that can be selected on the basis of facility-specific concerns or applicability to a specific organism. The assessment focused primarily on the domains of instrument sterilization, environmental cleaning, medication safety, and water exposures.
Multiple IPC deficiencies were identified, including incomplete sterilization logs for surgical instruments, use of a disinfectant that is not approved by the Environmental Protection Agency (EPA) for surface cleaning in the procedure room, intraoperative use of expired topical ocular medications, and opportunities for patient exposure to nonsterile water during surgery. For example, clinic staff members reported that an undated bottle of saline solution was used to provide surgical irrigation for multiple patients over the course of weeks. In addition, among the equipment that clinic staff discarded and replaced as part of their initial response to the infections were multiple cool-mist humidifiers. During the IPC assessment, the new cool-mist humidifiers were present in the procedure room.
Whole genome sequencing analysis of the two P. lilacinum clinical isolates demonstrated that they differed by approximately 55 single nucleotide polymorphisms (SNPs) from each other and by approximately 200,000 SNPs from control isolates. This analysis suggested a common infection source. Antifungal susceptibility testing demonstrated good in vitro activity of azoles and supported resistance to amphotericin B.
On day 41, health department staff members collected the saline bottle, stored epikeratome tubing, and swabs from five refrigerator locations (gasket, interior door, interior bottom, interior upper, and medication compartment) for fungal cultures and amplicon sequencing. The humidifiers in use when the infections occurred had been discarded before the investigation and were not available for testing.
All environmental cultures were negative for P. lilacinum, but fungal amplicon sequencing detected P. lilacinum DNA in the epikeratome suction tubing (Figure 2). During surgery, the epikeratome draws air away from the eye through a handheld attachment that touches the eye and is connected to suction tubing that contains a filter to prevent flow of debris into the machine. The handheld attachment is sterilized after each use and was therefore not made available for sampling. The suction tubing is intended for single use and is provided as a sterile product. Clinic staff members reported that tubing was never shared between patients; it is not known whether the set of tubing collected for testing was the one used for any of the infected patients’ procedures. The epikeratome machine is a multiuse item. It was not sent for testing because no clear sampling approach could be identified.
Read more on Centers for Disease Control and Prevention
