Waterborne disease surveillance.

Findings of a Survey of State and Territorial Epidemiology Programs

Introduction

Watershed protection, improved sewage disposal and the filtration and disinfection of drinking water have, since the early part of this century, reduced the frequency of infectious waterborne disease transmission (1). Despite this progress, in 1993, the largest waterborne disease outbreak in U.S. history occurred in Milwaukee, Wisconsin. This outbreak of waterborne cryptosporidiosis affected about 370,000 people, or about 25% of the exposed population, and was implicated in at least one hundred deaths of AIDS patients (2). Despite the size of the outbreak, detection occurred only when increased sales of anti-diarrheal medicines were observed and reported to the local public health agency. The detection of several waterborne cryptosporidiosis outbreaks (3-5) and routine recovery of Giardia cysts and Cryptosporidium oocysts from drinking water supplies (6-8), have raised concerns that most waterborne disease outbreaks are undetected or detected too late for effective intervention.

In the United States, infectious disease surveillance is conducted by state and local health departments. These health departments have primary responsibility for outbreak detection, investigation and control. States and/or territories differ in regard to which diseases are legally reportable, who is responsible for reporting cases, and the fraction of reported cases who are interviewed to obtain additional exposure information and algorithms for detecting outbreaks from case reports. Individual cases of more than 40 infectious diseases are reported by state and territorial health departments to the federal Centers for Disease Control and Prevention (CDC). Information on waterborne disease outbreaks is voluntarily reported to the CDC and is compiled and published biennially in conjunction with the U.S. Environmental Protection Agency (EPA) (9).

Efforts to improve state-based waterborne disease surveillance were sponsored by the EPA in the late 1970s and early 1980s. Three state waterborne disease surveillance projects were conducted in Colorado, Vermont and Washington State (10). In 1988, the EPA sponsored a "Workshop on Methods for Investigation of Waterborne Disease Outbreaks" to discuss techniques and make recommendations for improving waterborne disease surveillance. The workshop involved 32 panel members and 96 local, state and federal public health officials. A series of workshop recommendations was published by the EPA in 1990 [TABULAR DATA FOR TABLE 1 OMITTED] (11). Because of budget restrictions, state and local support for infectious disease surveillance has declined in the past decade (12). This has most likely reduced the resources available to implement the workshop recommendations. To ascertain the degree of implementation, a survey of state epidemiology programs was conducted in the summer of 1993.

Methods

The state epidemiologist from each state, U.S. territory and the District of Columbia was asked to identify a designated waterborne surveillance coordinator or a staff person familiar with the agency's waterborne disease surveillance systems and outbreak investigation procedures. A questionnaire was mailed to the designated individual. A second questionnaire was mailed if the first was not returned within four weeks. If a response from the individual was not received within eight weeks, the information was collected by telephone.

Questions, based on the 1988 workshop recommendations, were asked concerning the number of professional epidemiology staff, responsibility for infectious disease surveillance oversight (state versus local health agencies), disease reporting requirements, whether foodborne or waterborne disease coordinators are specifically designated, and the frequency and nature of outreach efforts to encourage health care providers and local public health agencies to participate in disease reporting. Additional questions asked whether inter-agency agreements were in place with health maintenance organizations (HMO) or the Indian Health Service (IHS) for obtaining reports of infectious diseases. Agreements to use HMO and IHS data as a surveillance tool were of interest since both provide health care to defined populations. The relationships between the state epidemiology program, the state drinking water regulatory program, the EPA, and the CDC were also documented. If a foodborne or waterborne disease coordinator had been designated, information on his/her training and work experience was collected.

Data analysis used SPSS/PC+[TM] Version 4.0. Statistical testing was done using a Chi squared test.

Results

Responses were obtained from the District of Columbia, Puerto Rico and Guam, and all of the states except Illinois. Half of the epidemiology programs had primary responsibility for infectious disease surveillance, while the other half shared that responsibility with local health agencies. For example, in some states the larger local health departments operate their own infectious disease surveillance system, such as Los Angeles, Seattle-King County and New York City, while in other states all infectious disease surveillance is done by the state epidemiology program, such as New Mexico and Delaware. In states without local public health agencies, regional centers operated by the state public health agency serve [TABULAR DATA FOR TABLE 2 OMITTED] many of the functions of a local health department, such as New Mexico.

All respondents indicated that monitoring of reportable diseases is computerized to some extent by their epidemiology programs. Most used EPI INFO (13), a CDC-developed software program for data entry, database management and statistical analysis. No state or territorial health department indicated that a formal agreement existed with any health maintenance organization for using nurse hotline records, computerized patient records or billing data to improve infectious disease reporting. Only 5 health departments had an agreement of any type with the Indian Health Service (IHS) for disease reporting, and most of these agreements are only for specific diseases such as tuberculosis or AIDS reporting. Even states with large American Indian populations, such as New Mexico, Oklahoma, North and South Dakota, had no procedure for using computerized IHS health care data to improve disease reporting and follow-up.

Although 47 of 52 (90%) states and territories require laboratory reporting of confirmed cases of reportable diseases, four states (Arizona, Delaware, Florida and North Dakota) and the District of Columbia do not. All but eight state or territorial epidemiology programs (Delaware, District of Columbia, Kansas, Massachusetts, Michigan, New Jersey, Washington, Guam) mail newsletters to physicians and other health care providers in the state to inform them of reported diseases and/or results of recent investigations. Of the 44 programs which mail newsletters, 20 (43%) mail them at least biweekly, 17 (39%) monthly, while the remaining 7 (18%) mail them quarterly or yearly. Twenty-two agencies hold yearly or periodic meetings to discuss infectious disease issues with local health agencies and health care providers (Table 1). Eighteen states have laboratory reporting requirements, a yearly infectious disease conference, and mail newsletters. In 23 states, laboratory reporting is required and newsletters are mailed but no yearly conference is held. In the District of Columbia and Delaware, laboratory reporting is not required, no conference is held, and no newsletters are mailed [ILLUSTRATION FOR FIGURE 1 OMITTED].

Twenty-five epidemiology programs had designated a waterborne disease coordinator other than the state epidemiologist; 26 had designated a foodborne disease coordinator [ILLUSTRATION FOR FIGURE 2 OMITTED]. In 12 additional states, the state epidemiologist was designated as the coordinator. Three states had a waterborne disease coordinator but no foodborne coordinator while four states had the reverse. In all but two cases where there was both a foodborne and waterborne disease coordinator, the same person served both functions. The majority of the 25 waterborne disease coordinators (90%) had a Masters in Public Health (MPH) degree, over 75% had more than five years of work experience in public health, and 80% had training or work experience in microbiology. However, only six of the 25 (24%) waterborne disease coordinators had either training or work experience in drinking water engineering, disinfection or treatment.

Fewer than half of the epidemiology programs had a designated waterborne disease contact in the state drinking water regulatory program and only seven waterborne disease coordinators (28%) held regular meetings with drinking water regulatory staff. For 22 (42%) state and territorial health departments, the drinking water regulatory authority was in another governmental agency. Four respondents, none of whom were designated waterborne disease coordinators, did not know where the authority for drinking water regulation resided. Twenty-two epidemiology programs (42%) received some or all reports of drinking water coliform violations from the drinking water regulatory program.

For the 26 epidemiology programs which are located in the same agency as the drinking water regulatory program, 15 (58%) had a designated person in the drinking water program to coordinate engineering and regulatory aspects of the outbreak investigation whereas only 7 (32%) of 22 programs located in different agencies had such a coordinator (p = .07). Being located in the same agency was not predictive of having a drinking water coordinator for outbreaks, receiving coliform reports or holding regular meetings with drinking water regulatory staff (Table 1). However, of those 22 epidemiology programs which had a designated drinking water outbreak coordinator, 14 (64%) received some or all coliform reports compared to 8 (27%) for the 30 programs without a designated drinking water contact (p [less than] .01).

It was suspected that programs with few infectious disease epidemiologists may be less able to designate a foodborne/waterborne disease coordinator, due to the wide range of responsibilities for a state epidemiology program. Six epidemiology programs have only one infectious disease epidemiologist and another 21 have two to four epidemiologists. Therefore, answers to several questions were compared between small (1 epidemiologist), medium (2-4 epidemiologists) and large (5 epidemiologists) programs. The frequency of designating a foodborne or waterborne disease coordinator was ascertained. The questions examined were: whether newsletters were mailed to health care providers, how frequently newsletters were mailed, whether a yearly infectious disease conference was held, whether regular meetings were held with water regulatory staff, and whether coliform violation reports were received (Table 2). Although programs with only one infectious disease epidemiologist were - in comparison to programs with greater than 2 - significantly less likely to have a designated waterborne disease contact, the departments with the largest number of epidemiologists (5 or more) were not more likely than intermediate programs (2-4 epidemiologists) to have designated a foodborne or waterborne disease coordinator. Five of six departments (83%) with only one infectious disease epidemiologist sent out a newsletter at least monthly. The same was true for 19 of 25 departments (76%) with 5 or more epidemiologists. All seven departments which hold regular meetings with drinking water regulators had two to four epidemiologists.

Figure 3 illustrates the number of waterborne disease outbreaks during the period 1986 to 1992. The proportion of states having regular meetings with the drinking water program or the proportion with a designated waterborne disease outbreak coordinator in the drinking water program did not differ significantly between states without waterborne disease outbreaks and states with three or more outbreaks.

Discussion

Waterborne disease surveillance and outbreak investigations have identified and described sources of contamination as well as deficiencies in drinking water disinfection and filtration. These findings have contributed to the improved safety of U.S. drinking water. But despite prior efforts to improve outbreak detection and estimate the underlying rate of waterborne outbreaks, the frequency and characteristics of undetected waterborne outbreaks remain unknown. Estimates have suggested that only 10-33% of waterborne outbreaks are reported (1). Even when outbreaks are detected, identification of an etiologic agent for waterborne outbreaks often eludes investigators. In 1991, of 15 waterborne outbreaks reported, etiologic agents were not identified for 12 (9). In one outbreak, despite almost 10,000 estimated cases, no etiologic agent could be identified.

Reporting of the occurrence of infectious disease by health care providers and the public is the basis of waterborne disease surveillance and outbreak detection. Even with limited surveillance resources, many state epidemiology programs employ a variety of tools to encourage disease reporting. Feedback to health care providers is furnished by newsletters and infectious disease conferences. In 37 of 52 states and territories, newsletters are mailed at least monthly. Yearly infectious disease conferences are held by 22 of the 52 agencies (42%). Use of a number of other unique and innovative methods for improving disease surveillance was also reported by state public health agencies (e.g. sentinel physician reporting, toll-free telephone numbers for disease reporting, mailing of reminder letters and instructions to health care providers, press releases, formal liaisons with hospital infection control personnel, presentations at medical conferences, and threats of prosecution for physicians who fail to make mandated reports).

The 1988 EPA workshop recommended that epidemiology programs promote the electronic transmission of computerized surveillance data as a long-term goal. Since the EPA workshop, large health maintenance organizations have grown rapidly in number and individual size. With that growth has come the increased availability of computerized health care data from populations served by these organizations. This survey indicates that few health departments have taken advantage of this source of information to improve infectious disease surveillance. Given the potential benefits of using these data to passively monitor diseases which come to the attention of health care providers, pilot projects for using these databases for disease surveillance should be considered. Existing cooperative agreements between the Centers for Disease Control and Prevention and several health maintenance organizations may provide an opportunity for testing the feasibility and potential of using these data to enhance state and local disease surveillance programs.

The 1988 EPA workshop also recommended that state epidemiology and state drinking water personnel regularly meet to review general trends and recent outbreaks of waterborne disease. In only seven instances did epidemiology staff hold such regular meetings. In only 22 cases was a drinking water regulatory staff person designated to coordinate waterborne disease outbreaks, and in seven instances no specific name was mentioned by the epidemiology program respondent. The combination of epidemiology coordinators having little contact with water treatment experts prior to an outbreak, little prior planning for how to coordinate a waterborne disease outbreak investigation and most of them lacking drinking water treatment training or work experience may delay the formation of an effective investigative team of epidemiologists, engineers and sanitarians. Even states with three or more waterborne disease outbreaks in the past seven years were no more likely to have developed the recommended relationships with the drinking water regulatory program.

Difficulties in maintaining routine contacts between the epidemiology and drinking water regulatory programs may result, in part, from either physical or administrative separation of the epidemiology and drinking water programs. Among comments reported on the questionnaire, 10 respondents reported coordination problems which resulted from administrative separation of the drinking water regulatory authority from the epidemiology program. In cases where the two programs were in the same department, a water regulatory program outbreak coordinator was more likely to be named and, when named, the epidemiology program was more likely to receive reports of some or all coliform violations. Twenty-two of 52 states and territories had administratively separated epidemiology and drinking water regulatory programs.

One of the recommendations of the 1988 EPA "Workshop on Methods for Investigation of Waterborne Disease Outbreaks" was that all states designate a waterborne disease coordinator who has a Masters in Public Health (MPH) and five years of work experience in public health. Fewer than half of epidemiology programs (48%) had a designated waterborne disease coordinator. Only 25% of the coordinators had drinking water treatment training or work experience. In light of the significant costs of recent federal drinking water treatment requirements designed to reduce the risks of waterborne disease transmission (e.g. Surface Water Treatment Rule, Coliform Regulations, Enhanced Surface Water Treatment Rule), the relatively small investments needed for a designated epidemiologist and improved communication between epidemiology and state drinking water regulatory staff would seem to be extremely cost-effective. The costs of federal drinking water treatment rules, however, may be passed along to water consumers while additional resources for disease surveillance and outbreak investigations must be obtained from the restricted budgets of state and territorial health departments.

At the federal level, little or no consideration is given to the need for increased resources to support the surveillance activities needed to detect and investigate waterborne outbreaks. Dedicated funding should be provided to epidemiology and drinking water regulatory programs to accomplish this task. Without additional dedicated resources, the workloads of the designated waterborne disease coordinators, many of whom are also the designated foodborne disease coordinator and have multiple other designated responsibilities, will likely continue to increase while the priority placed on waterborne disease surveillance decreases. Adequately preparing for infrequent waterborne disease outbreaks is often viewed as a low priority by state epidemiologists, given the competing demands on epidemiology staff and resources. Infrequent waterborne disease outbreaks can, however, occasionally cause large numbers of illnesses as recently demonstrated in Milwaukee. Health agencies must be adequately prepared to detect waterborne outbreaks in a timely manner so that effective action can be taken to prevent further illness. With dedicated resources for waterborne disease surveillance, investigation, and outbreak control - perhaps derived from the drinking water utility budgets - state and territorial health departments would be better able to provide EPA and other regulatory agencies with the information needed to more rationally direct regulatory efforts and optimally allocate scarce resources to reduce the risk and burden of waterborne disease.

REFERENCES

1. Craun G.F., (1986), Waterborne Disease in the United States, CRC Press ISBN 0-8439-5937-6.

2. MacKenzie W.R., N.J. Hoxie, M.E. Proctor, et al., (1994), "A massive outbreak in Milwaukee of Cryptosporidium infection transmitted through the public water supply," New England Journal of Medicine, 331 (3): 161-167.

3. Hayes E.B., T.D. Matte, T.R. O'Brien, et al., (1989), "Large community outbreak of cryptosporidiosis due to contamination of a filtered public water supply," The New England Journal of Medicine, 320(2):1372-1376.

4. D'Antonio R.G., R.E. Winn, J.P. Taylor, et al., (1985), "A waterborne outbreak of cryptosporidiosis in normal hosts," Annals of Internal Medicine, 103:886-888.

5. Smith H.V., R.W.A. Girdwood, W.J. Patterson, et al., (1988), "Waterborne outbreak of cryptosporidiosis," Lancet, 2:1484.

6. LeChevallier M.W., W.D. Norton, R.G. Lee, (1991), "Occurrence of Giardia and Cryptosporidium spp. in water supplies," Applied and Environmental Microbiology, 57:2610-2616.

7. LeChevallier M.W., W.D. Norton, R.G. Lee, (1991), "Giardia and Cryptosporidium spp. in filtered drinking water supplies," Applied and Environmental Microbiology, 57:2617-2621.

8. Ongerth J.E., H.H. Stibbs, (1987), "Identification of Cryptosporidium oocysts in river water," Applied and Environmental Microbiology, 53:672-676.

9. Moore A.C., B.L. Herwalt, G.F. Craun, R.L. Calderon, A.K. Highsmith, D.D. Juranek, (1993), CDC: Surveillance for waterborne outbreaks - United States, 1991-1992, 42:SS5:1-22.

10. Harter L., F. Frost, R. Vogt, et al., (1985), "A three-state study of waterborne disease surveillance techniques," American Journal of Public Health, 75 (11): 1327-1328.

11. Environmental Protection Agency, (Sept. 1990), "Workshop on Methods for Investigation of Waterborne Disease Outbreaks: Summary of recommendations," EPA, Office of Research and Development, Washington D.C., EPA/600/9-90/021.

12. Berkelman R.L., R.T. Bryan, M.T. Osterholm, et al., (1994), "Infectious disease surveillance: A crumbling foundation," Science, 264:368-370.

13. Dean A.G., J.A. Dean, D. Coulombier, et al., (1994), Epi Info, Version 6: A word processing, database, and statistics program for epidemiology on microcomputers, Centers for Disease Control and Prevention, Atlanta, Georgia.

Floyd Frost, Ph.D., The Lovelace Institutes, 2425 Ridgecrest Drive S.E., Albuquerque, New Mexico 87108