This study was initially conducted because of a request of the New Taipei City Health Bureau in response to the MK outbreak (8). The swimming pools at the resort had been filled with tap water. Unfortunately, we received information about the MK outbreak 1 day after cleaning and disinfection of the swimming pools had taken place (8). Before water samples were collected, the resort was temporarily closed by the health authorities, and the facility was cleaned and disinfected as recommended by the health authorities (i.e., treatment with 5 ppm free available chlorine for 3 hours). The cleaning and disinfection procedures included draining of all water reservoirs, pools, and tubs, followed by surface scrubbing to remove any potential biofilms and disinfection with sodium hypochlorite. The pretreatment methods of samples, PCR conditions, phylogenetic analysis, and all protocols were performed as described in our previous studies (2,6).

We collected 19 water samples and 8 soil samples from the swimming resort and its surrounding environment (Appendix Figure 1). We sequenced all 17 amplicons of the 15 test-positive samples, analyzed them by using BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi), and compared them with reference species from GenBank to determine the most closely related species. All the amplicons were homologous to microsporidia, with identity ranging from 89% to 99% (Table). Only (29.4%) amplicons showed a high degree of homology (>97% identity) to the reference strain.

In the initial survey (10 days after site disinfection), V. corneae was not detected. However, other Vittaforma-like microsporidia were identified in the standard swimming pool and foot washing pool. According to the Enforcement Rules for Swimming Pool Management in Taiwan, swimming pool water should be chlorinated (with a concentration of free available chlorine of ≈0.3–0.7 ppm) to prevent the spread of waterborne diseases. The presence of these other Vittaforma-like microsporidia indicates that some problems might have occurred during the disinfection process.

According to the New Taipei City Health Bureau, all pools, water source tanks, waterlines, and tubs in this facility were drained and scrubbed during the cleanup and disinfection process. The source of V. corneae infection remains debatable. Chlorine disinfection studies have shown that residual chlorine is capable of inactivating and reducing the number of microsporidians. However, microsporidian spores are relatively resistant to the typical concentrations of chlorine used for swimming pool disinfection (depending on the species, spores are inactivated after an exposure time of 10–120 min [i.e., inactivation of Encephalitozoon intestinalis spores does not reach 100% under 5 ppm of free available chlorine even after 120 minutes of treatment]) (9–11). Many clinical studies have indicated that soil or mud exposure, visits to hot springs, and outdoor activities, especially after rainfall, are all risk factors for ocular microsporidiosis (12,13). In addition, our previous study provided evidence for the presence of V. corneae in hot springs, thus indicating that pools in outdoor environments were associated with the presence of V. corneae (6). Therefore, we considered that the contaminating microsporidia in the swimming resort might have been brought to the resort from outside by human activities.

Our data show that many clade IV microsporidia were present in the soil and water samples from the resort site (Appendix Figure 2). Most of the microsporidia in clade IV are of terrestrial origin, according to small subunit rRNA gene phylogenetic analysis (14,15). Therefore, given that previous studies have shown that rainfall is an important risk factor for ocular microsporidiosis (5,12,13), we believe that water contamination might originate from soil environments after rainfall. Rainfall occurred near the sampling location during June 15–19 and again during July 1–4 (as recorded by the Taiwan Central Weather Bureau). We conducted a follow-up site survey after the July 1–4 rainfall and found Vittaforma-like microsporidians were found in pavement rainwater and the parking lot of the resort. Hence, we hypothesized that the swimming pool water was contaminated through soil or water brought in by human activity during the rainfall. Vittaforma-like microsporidians were found in the swimming pool and foot washing pool in the initial survey, which was conducted after a careful disinfection procedure, but microsporidia were not found in tap water or other pools. Therefore, these results suggest that the contamination was not from the waterlines or water sources, and the pools may have been contaminated from an outside source owing to human activities and poor facility configuration. In the follow-up survey, we found 100% identical amplicons in the parking lot and foot washing pool, suggesting a possible transmission pathway Vittaforma-like microsporidians from the outside environment to the swimming pool (Appendix Figure 1).

Our study demonstrated the presence of Vittaforma-like microsporidia in a swimming resort and nearby environments in Taiwan. Human activities, rainy weather, and soil-rich or park environments might have been possible sources of microsporidia in the waters at the facility. The foot washing pool and shoe cabinet area are possible contamination areas and might facilitate transmission of microsporidia throughout the swimming resort. We suggest several precautions, including improving the frequency and efficacy of disinfection procedures at the facility, using a continuous water flow facility in foot washing pools, and paying attention to the disinfection and cleaning of the shoe cabinet area, especially during the rainy season. In addition, for swimming resorts that are located in a park, enhanced monitoring of the environment surrounding the swimming pool is warranted.