Shaded area indicates number of cholera cases. No surveillance was conducted between July 1998 and November 1998 because V cholerae is generally absent from water samples taken from Lima during these cooler months.4,5 Eight sites defined the surveillance area, which monitored the northern (1 site in Callao), central (1 site in Miraflores), and southern (1 site in Chorrillos and 5 sites in Pampas de San Juan de Miraflores) sections of Lima. Collection occurred in 4 raw sewage oxidation-purification ponds, and at 4 sea outlets for the city's major sewer lines.
Speelmon EC, Checkley W, Gilman RH, Patz J, Calderon M, Manga S. Cholera Incidence and El Niño–Related Higher Ambient Temperature. JAMA. 2000;283(23):3072-3074. doi:10.1001/jama.283.23.3068i
To the Editor: Since the 1991 epidemic, cholera continues to be an important health problem in Peru. More than 260,000 cases were reported in Lima alone by the end of 1993, when the epidemic was brought under control. Recent investigations suggest the existence of an environmental aquatic reservoir for Vibrio cholerae O1.1- 5 Furthermore, cholera seasonality in endemic areas suggests possible long-term survival of V cholerae in the environment. Therefore, extreme weather phenomena like the 1997-1998 El Niño may facilitate the growth of V cholerae in the environment, which may in turn trigger the onset of a cholera outbreak. We have previously suggested probing sewage water for the presence of V cholerae as an environmental surveillance measure against future cholera outbreaks.5
Sewage samples were probed weekly for V cholerae O1 from November 1997 to July 1998 and from November 1998 to May 1999 in Lima, Peru. Methods are described elsewhere.5 During this period we obtained the weekly number of new cholera cases in Lima from the Ministry of Health and the weekly average ambient temperature from the Peruvian Weather Service.
Between January 1997 and May 1999, 1175 clinical cholera cases were reported in Lima, but 1039 of these occurred between January and May of 1998 (Figure 1). From November 1997 to July 1998, V cholerae O1 was isolated in 10 (29%) of 35 surveillance weeks and in 11 (4%) of 280 sewage samples. However, V cholerae O1 was not isolated from any of the 208 sewage samples obtained between November 1998 and May 1999. On average, 62 cholera cases per week were reported when V cholerae O1 was isolated vs 11 cases per week when no V cholerae O1 was detected (P<.001).
Ambient temperature was positively correlated with the number of cholera cases (r=0.49; P<.001). This outbreak coincided with the exceptionally high 1998 summer ambient temperature increase due to the El Niño phenomenon, and the peak in ambient temperature also preceded the peak of the outbreak by 3 weeks. Regression analysis revealed that 19.3°C was the threshold ambient temperature that predicted a large increase in cholera cases. V cholerae O1 detection in sewage preceded the mid March 1998 outbreak peak by 14 weeks. After August 1998, no more than 2 new cases of cholera were reported per week, nor was V cholerae O1 isolated from sewage samples. As of March 2000, no cases of cholera have been reported.
As in our previous study,5 cholera was detected in sewage several weeks prior to the onset of the cholera outbreak, thus confirming our previous observation that sewage surveillance can therefore assist in preparedness measures for cholera outbreaks. Furthermore, the higher ambient temperatures associated with the 1997-1998 El Niño event appear to have led to an increase in environmental V cholerae O1 and in turn to the 1998 cholera outbreak, after the nearly complete absence of cholera cases in Lima since late 1995. During the 1998 summer (January-March 1998) the mean ambient temperature was 26°C, which is 3.4° higher than the previous 5 summers.6 Previous studies have demonstrated a positive relationship between water temperature and survival and reproduction of V cholerae.3 Our results support the use of sewage surveillance to predict cholera outbreaks and suggest a link between the 1998 cholera outbreak in Lima and the excess ambient temperature associated with the El Niño phenomenon.
Funding/Support: This study was supported in part by an Applied Research on Child Health Grant from the US Agency for International Development (Asociacion Benefica PRISMA, Lima, Peru); grants F31-HD08488 National Research Service Award from the National Institutes of Child Health and Development (Dr Checkly); CR-823130 Cooperative Agreement Grant and R-824995-O1-0 Research and Development Grant from the US Environmental Protection Agency (Johns Hopkins School of Public Health); a Fullbright grant (Ms Speelmon); and from the charitable RG-ER foundation for the advancement of climate and health research.
Acknowledgment: We thank Luis Suarez, MD, for his assistance; M. D. Chestnut, MD, for helpful comments; and J. B. Phu and D. Sara for their editorial support.