National Public Health Surveillance

Georgia’s National Center for Disease Control (NCDC) conducts passive surveillance for botulism. Physicians are required by law to report all suspected cases of botulism to local epidemiologists who, in turn, are required to report to region-level epidemiologists and to NCDC. Administratively, Georgia is divided into 10 regions, 2 major cities, and 2 autonomous republics. The two autonomous republics, South Ossetia and Abkhazia, are not under public health surveillance.

Medical Records Review

We visited hospitals in cities and regions of Georgia that reported at least one botulism case from 1980 to 2002. At each hospital, medical records were sorted by discharge diagnosis. A patient was considered to have botulism if medical records indicated that this was the final diagnosis. For each patient, a trained epidemiologist completed a standardized data abstraction form that included patient demographics, illness history, and clinical characteristics. A random sample of patient records (13%) was audited by a staff physician at the Tbilisi Infectious Pathology Center, which supplied most of the patients for this study, to confirm the accuracy of abstracted data; all paper abstraction forms and electronic records were compared to assess accuracy of data entry.

Outbreaks were not documented separately in national surveillance. To account for clustering of cases from common food sources, we defined an outbreak as two or more patients who were documented in the medical chart as being part of an outbreak; who had hospital admission dates no more than 3 days apart; and who had identical suspect food sources, town of residence, and hospital of admission. A botulism event was defined as an outbreak or as a sporadic case (i.e., an individual case not associated with other cases). When performed, diagnostic testing on food or human specimens was conducted at NCDC by using the standard mouse bioassay for detecting botulinum toxin (9).

Surveillance System Evaluation

Because our analysis demonstrated marked geographic variation in incidence, we performed retrospective case finding and active surveillance for botulism in two parts of Georgia that reported no cases from 1980 to 2001: the city of Poti, estimated 2002 population 46,000, and the region of Samegrelo, estimated 2001 population 405,500. For retrospective case finding in these two parts of Georgia, we interviewed local epidemiologists to identify botulism cases that were not reported to NCDC from 1996 to 2002. We also interviewed physicians working at the main outpatient clinics and on inpatient infectious diseases, critical care, and neurology wards. At hospitals, we reviewed medical records to identify suspected botulism cases among patients diagnosed with neurologic syndromes from 1996 to 2002. We also reviewed pharmacy records at these hospitals to determine whether botulinum antitoxin, which is widely available, was ever administered to a patient. For active surveillance, we telephoned epidemiologists and hospital-based physicians every month from April through December 2002 to ascertain whether any new cases of botulism occurred in these two parts of Georgia.

Home Visits

To learn more about practices in the home, we visited 14 homes in different regions of Georgia. We interviewed families about food preservation practices. We purchased tomatoes, peppers, and cucumbers at local markets, and asked persons who routinely preserve such vegetables to demonstrate this process for us.

Ethics Review

The studies reported in this manuscript were reviewed by the human subjects committees at the Centers for Disease Control and Prevention (Atlanta, Georgia, USA) and at NCDC (Tbilisi, Georgia).

National Surveillance Data

Figure 1

Figure 1. Number of botulism cases and cases per 100,000 persons in Georgia, 1980–2002. Data are derived from routine, passive national surveillance. Data are presented as one trend line because the incidence and...

From 1980 to 2002, we found 879 cases of botulism reported in Georgia. In 2002, a total of 39 cases were ascertained for an incidence of 0.9 per 100,000 persons (Figure 1). The median annual incidence increased from 0.3 per 100,000 persons (median case count 15), during 1980 to 1990, to 0.9 per 100,000 persons (median case count 41), during 1991 to 2002. The incidence was highest in 1994 (3.6 per 100,000 persons) when 173 persons became ill at a wedding from eating contaminated fish. Fifty-eight deaths were attributed to botulism, for an average death rate of 7%. The case-fatality rate ranged from 0% (several years) to 18% (in 1981). The median case-fatality rate increased from 3% (1980–1990) to 7% (1991–2002).

Medical Records Review

We identified medical records for 706 patients with a diagnosis of botulism from eight hospitals in five regions of Georgia. Though patients resided in eight regions of Georgia, 90% of patients were initially hospitalized at or transferred to one Tbilisi hospital, which serves as the national referral hospital for botulism. Because national surveillance data consisted of only summary statistics and lacked personal identifiers, we could not determine whether the 706 patients identified through hospitals represent a subset of the total reported in surveillance (879) or all hospitalized, diagnosed botulism case-patients in Georgia.

The median age of patients was 34 years (range 1–90). Three hundred fifty-five (50%) were female. Patient ethnicities included Georgian (73%), Azerian (11%), Armenian (10%), and Russian (5%). The median age, sex, and ethnicity of patients was similar to that of the general population, except for the proportion of Azerian patients, which was almost twice as large as the proportion of Azerians living in Georgia in 1989 and 2002 (10). Fifty-four (8%) patients died during hospitalization.

Figure 2

Figure 2. Cumulative incidence of botulism events by region of residence, per 100,000 persons, in Georgia, 1980–2002. Outbreaks and sporadic cases are counted as one event. Data are derived from review of medical...

We identified 329 botulism events: 154 outbreaks involving 531 persons (median 3 persons per outbreak; range 2–83) and 175 sporadic cases. Medical records were available for 83 of the reported 173 patients from the 1994 outbreak. The eastern part of the country, particularly the area surrounding Tbilisi, had the highest incidence rate (Figure 2). No botulism events were identified in three areas of the country: the city of Poti and the regions of Samagrelo and Racha-Lechkhumi.

All botulism events were suspected or confirmed to be foodborne. The most commonly implicated food was home-preserved vegetables, accounting for 261 (80%) events. Other implicated foods included smoked fish (12%) and smoked meat (2%). Among events involving home-preserved vegetables, the most commonly implicated vegetables were tomatoes (15%), peppers (15%), celery leaves (13%), eggplant (13%), and combinations of vegetables (13%). The proportion of cases attributed to vegetables did not vary significantly across years or regions. (Home-preserved vegetables were implicated more frequently among patients of non-Georgian ethnicity (83%) than among patients of Georgian ethnicity (64%) (p < 0.01). Botulism events varied markedly by season; most occurred in the winter (55%) or spring (23%), compared to summer (10%) or fall (12%). The proportion of events attributable to home-preserved vegetables was almost two times greater in the winter (86%) than in the summer (44%) (p < 0.01).

A specimen from either a patient or food was collected in 116 (35%) botulism events. Of events tested, 20 (17%), involving 75 persons, had botulinum toxin detected. Among these 20 events, type B toxin was identified in at least one specimen from 17 events (85%) (3 from blood, 13 from food, 1 from vomitus), type E in 2 events (10%) (1 from blood, 1 from food), and type A in 1 event (5%) (food). Home-preserved vegetables were implicated in all 17 events in which toxin type B was identified. Smoked fish was implicated in the one toxin type A event and smoked fish and home-preserved tomatoes in the two toxin type E events.

Surveillance System Evaluation

Retrospective case finding in the city of Poti and region of Samegrelo, both of which previously reported no cases, identified three unreported, possible cases of botulism from 1996 through 2002: one involved incorrect reporting of the site of residence, the second involved failure of a physician to report, and the third involved failure of an epidemiologist to report. In the last two cases, neither patient reported eating high-risk food, neither was treated with antitoxin, and neither was given a final diagnosis of botulism by the treating physician. Review of antitoxin distribution records did not identify any patients treated with antitoxin from 1996 to 2002. Active surveillance for botulism cases did not identify any new cases from April to December 2002 in Poti and Samegrelo.

Home Visits

We observed that the home preservation process varied between homes, but the sequence of events was common. We also found that many families inadequately sterilize equipment, conserve vegetables uncooked, fail to use pressure cookers, store vegetables for months to years, eat preserved vegetables uncooked, and, out of necessity, occasionally eat food that looks and smells spoiled. Salt, sugar, vinegar, and spices are frequently added but usually in small amounts.