We conducted retrospective cohort studies among guests of hotels X and Y. Hotel X staff routinely recorded all guests in the hotel registry; only one guest per room was recorded in the hotel Y registry. Study participants included registered persons at hotel X and both registered and unregistered persons at hotel Y who stayed at the hotels during the night before or after the date the respective outbreak began. Adult participants were interviewed by telephone with a standardized questionnaire following informed consent; children were interviewed with the assistance of a parent following assent. Information about underlying medical conditions, symptoms of illness, and exposure to all common areas in the respective hotel was collected. Participants were asked to estimate the timing of their exposures to the nearest minute. Interviews were conducted within 2.5 weeks of the outbreak at hotel X and within 1 week of the outbreak at hotel Y.

We defined an ocular case as any individual who developed at least one of the following after arrival at the hotel: burning eyes, watery eyes, blurry vision, difficulty opening eyes, or photophobia. A respiratory case was defined as an individual who developed at least one of the following symptoms after arrival at the hotel: cough, wheezing, shortness of breath, chest tightness, or sore throat. We performed bivariate and stratified analyses using chi-square, 2-tailed Fisher’s exact, and Cochran-Mantel-Haenszel tests; p-values < 0.05 were considered statistically significant. We also calculated the relative risks (RR) and 95% confidence intervals (CIs). The effect of duration of pool exposure was modeled for either syndrome at both hotels using a smoothing spline fit with λ = 100,000 (JMP 5.1.2; SAS Institute Inc.; Cary, NC).

Environmental health investigations were conducted at both hotels using an extensive environmental health outbreak investigation report for swimming pools [Centers for Disease Control and Prevention (CDC) 2006]. We examined swimming pool and spa dimensions, and disinfection, filtration, and ventilation systems; we also identified chemicals in the mechanical room of each pool. We collected pesticide application and pool maintenance records for the 2 weeks preceding the outbreak and interviewed swimming pool staff about recent pool maintenance practices and changes in equipment or building structure. No air samples were collected during the outbreak. Water samples were collected by the state health department for microbial analysis at hotel X but were discarded before chemical analysis. A water sample collected in a plastic drinking water bottle by a hotel guest approximately 12 hr after the outbreak at hotel Y was tested with gas chromatography and mass spectrometry for four trihalomethanes (chloroform, bromoform, bromodichloromethane, and dibromochloromethane) using a method similar to that of Cardinali et al. (2004); it was not tested for chloramines. No clinical specimens were available.

A total of 72 people became ill with ocular and respiratory illness during outbreaks associated with indoor swimming pools at two hotels. We observed symptoms similar to those previously reported following chloramine exposure, including burning eyes, tearing, and cough (Goyder 2000; Massin et al. 1998; Momas et al. 1993; Thickett et al. 2002; Varraso et al. 2002). Symptoms occurred within minutes of exposure to indoor swimming pool environments, generally were more likely to develop with increased duration of exposure, and were sometimes prolonged and severe. We were able to document illnesses lasting up to 14 days; four bathers also sought medical care as a result of these outbreaks. As in a previous outbreak, children were predominantly affected, and attack rates were extremely high (Goyder 2000).

In the outbreaks at hotel X and hotel Y, exposure to the pool area was associated with symptoms; in at least some cases, contact with the pool or whirlpool water was not required. The ocular syndrome developed only among persons exposed to the pool area; the respiratory syndrome developed in 1 of 15 persons without pool-area exposure at hotel X but in no one without pool-area exposure at hotel Y. Activities within the pool, such as swimming under water or swallowing water, were not additional risk factors for illness, and use of swimming goggles within the pool area was not associated with lower risk of the ocular syndrome. This suggests that contact with or inhalation of contaminated air alone is associated with symptoms—as is expected with a mucosal irritant—and that dermal absorption or ingestion may not be necessary to produce symptoms. Generally, longer exposure to the pool area correlated with illness. However, those who remained longest in the pool area of hotel Y developed ocular, but not respiratory, symptoms; the reasons for this are unclear.

Although the etiologic agent in these outbreaks could not be determined definitively, chloramines likely contributed to the illness. Levels of chlorine at the hotel X whirlpool likely exceeded chlorine-to-nitrogen mass ratios required for breakpoint chlorination and would have increased production of trichloramine (Shang and Blatchley 1999; White 1999). This, when combined with a malfunctioning ventilation system in the pool area and chloraminated fill water, could have led to elevated chloramine levels in the air around the pool. Similarly, although pool maintenance records did not indicate excess chlorine at the hotel Y pool, the bather load and combined chlorine levels exceeded recommended levels during the outbreak period. A sample of swimming pool water from hotel Y was analyzed by the CDC but did not reveal potential causal agents. Chloramines are highly volatile and would not be expected to persist under the given sample storage conditions; therefore, the water sample was not analyzed for chloramines. However, trihalomethanes, which are potential ocular and respiratory irritants that may be found in swimming pools (Shaw and Eng 1987), are chemically stable and therefore might persist at high levels if initial concentrations were substantial enough to offset loss from nonstandard sample collection and storage. This sample did not contain abnormally high levels of trihalomethanes compared with previously published levels for U.S. pools (Armstrong and Golden 1986). Thus, no other potential causal agents, such as microbial pathogens or elevated levels of trihalomethanes, were found in pool water from hotel Y. Additionally, the observed incubation period and symptoms at both hotels were typical of those observed following exposure to chloramines (Goyder 2000; Massin et al. 1998; Momas et al. 1993; Thickett et al. 2002; Varraso et al. 2002). Together, these findings suggest that chloramine exposure played a role in the outbreaks. To appropriately quantify environmental concentrations of chloramines, air or water sampling must take place within hours of identification of a possible outbreak and before changes have been made to pool chemistry or ventilation. Water samples should be collected in amber glass bottles, which might not be readily available. Air sampling and testing methods for chloramines are also complex and beyond the routine capability of most public health departments (Hery et al. 1995). It is likely that these factors will contribute to future underreporting of outbreaks of chloramine intoxication.

To decrease exposure to chloramines in swimming pool environments, proper ventilation is also required. Because of limited air turnover and the potential for continuous generation of chloramines, air quality in indoor swimming pool enclosures is likely to differ substantially from that of outdoor swimming pool areas. It is not surprising, therefore, that differing levels of chloramines between indoor and outdoor swimming pool environments have been reported (Massin et al. 1998). Special attention to pool chemistry and ventilation may be required in cold weather climates, where a greater proportion of air may be recycled to prevent heat losses, and in indoor recreational pools and parks where wave action, sprayers, and other turbulent water activities have been documented to increase the aerosolization and volatilization of chloramines to a greater degree than standard leisure pools (Hery et al. 1995). Because little or no data exist on airborne concentrations of chloramines for pools, the American Society of Heating, Refrigerating and Air-Conditioning Engineers (2006) is investigating whether the amount of ventilation air currently required in pool enclosures is sufficient to remove these pollutants.

To ensure adequate pool operation and ventilation, it is critical to improve the training of pool operators. Currently, pool staff commonly lack adequate training in swimming pool chemistry and maintenance (CDC 2003). Maintenance staff also may be required to divide their time between oversight of the pools and many other activities, as in these two outbreaks. Violations of pool codes, especially swimming pool chemistry regulations, are a common result. In a study of > 22,000 pool inspections in 2002, more than half of inspections reported at least one violation. Child wading pools (18.4%), therapy pools (14.3%), and hotel/motel pools (14.0%) yielded the greatest percentage of violations (CDC 2003). In both outbreaks reported in the present study, pool chemistry abnormalities included elevated free and combined chlorine, as well as elevated cyanuric acid levels. Cyanuric acid is intended to protect chlorine from degradation due to ultraviolet light in outdoor swimming pools (Moody et al. 1993; Yamashita et al. 1988). As a result, higher free chlorine residuals are required in swimming pools treated with cyanuric acid (Yamashita et al. 1988), and it should be used only in outdoor swimming pools and spas. These problems suggest that standardized education of public pool operators should be mandatory. Additionally, trained personnel should be available for monitoring and maintenance duties at all times, especially at times when pool usage is highest. Both of these outbreaks began on weekends when the bather load was high; in one outbreak (hotel X), pool maintenance was not performed by the usual staff. Clearly labeling pool products containing cyanuric acid for outdoor use may also decrease inappropriate use of these products.

There are several limitations to these studies. Although we attempted to reach every guest registered at the hotel during the outbreak period, participation was incomplete and reported attack rates may be artificially elevated. Participants also may not have been able to recall precisely their symptoms or exposures. Finally, the lack of appropriate air and water samples from the outbreak period limited our ability to establish a definitive etiology for the illnesses associated with these outbreaks.

Clinicians and public health practitioners should watch vigilantly for possible outbreaks of illness due to swimming pool exposure, especially during the indoor swimming pool season. When an outbreak is suspected, the local health department should be notified immediately and environmental health staff should investigate promptly. To help establish outbreak etiology, swimming pool or spa water chemistry (free and combined chlorine levels) should be immediately documented before alteration of the water chemistry, and appropriate sample dilutions should be made if readings exceed the limits of test kits. Development of simple, rapid, field-based methods for measurement of chloramines and indoor air quality for use by public health departments would assist outbreak investigations.

These outbreaks involved extremely high attack rates and short incubation periods following exposure to indoor swimming pool enclosures. Similar outbreaks are likely to be common, although they have been reported rarely. As indoor pool swimming continues to gain popularity, careful attention to pool maintenance should help prevent illness from infectious and chemical agents. Health departments should be prepared to investigate outbreaks within hours of their identification and collect appropriate environmental samples. The public should be alerted to the potential health risks of poor hygiene practices such as urinating in or not bathing before entering swimming pools.