Surveillance of diarrheal disease in Houston, Texas: 1983.




Lemp, G.F.

Journal Title

Journal ISSN

Volume Title


University of Texas, Health Science Center at Houston, School of Public Health.


Surveillance for salmonellosis, shigellosis, and giardiasis was carried out in 10 randomly selected medical laboratories in Houston, Texas during the period January 1 to December 31, 1983. Prospective abstraction of data from laboratory records was performed during scheduled monthly or semi-monthly visits to each laboratory. A total of 11,803 persons submitted specimens for stool culture and/or ova and parasite examination to the 10 study laboratories during 1983. Salmonella was isolated from the stools of 305 persons, and Shigella from 243 persons. Giardia was identified in specimens from 282 persons. Positivity proportions (cases per 100 persons submitting specimens) for three diseases were 3.44, 2.74, and 4.16, respectively. The three diseases showed a seasonal pattern, with the largest number of cases occurring in the period July to October. Most of the cases occurred among young children under 6 years of age. Salmonellosis, shigellosis, and giardiasis age- specific positivity proportions peaked in the age intervals less than one year, 5 to 9 years, and 1 to 4 years, respectively. Positively proportions were significantly (p<0.001) higher among males as compared to females. Four serotypes showed evidence of significant (p<0.05) clustering in space and time: S. oranienburg, S. muenchin, Sh. flexneri 3a, and Sh. flexneri 3. Three additional serotypes exhibited space-time clustering of borderline significance (0.05<p<0.10): S. typhi, S. newport, and Sh. flexneri 2. The positivity proportions for shigellosis and giardiasis were 2.5 and 1.6-fold higher, respectively, among those census tracts with a median income under $20,000, as compared with those greater than or equal to $20,000. The positivity proportion for salmonellosis was highest for census tracts with median incomes in the middle range ($16,000-$19,999). Census-tract-specific positivity proportions for giardiasis were significantly (p<0.005) associated with the percentage of housing units in a census tract lacking complete plumbing. Persons in census tracts supplied by surface water had a 1.6-fold higher giardiasis positivity proportion as compared with persons in census tracts supplied by well (ground) water. Salmonellosis and shigellosis were not significantly associated with water source or plumbing. After control (stratification) for median income, the relationship was probably due to confounding by income. The results suggest that a medical laboratory-based surveillance system could be an effective means for detecting potential outbreaks of diarrheal disease and identifying, geographic, demographic, and seasonal factors related to incidence.


139 p., Dissertation


public health, human diseases, diseases, food poisoning, water quality