Cover Image

HARDBACK
$54.00



View/Hide Left Panel

Pesticide Food Poisoning from Contaminated Watermelons in California, 1985

LYNN R.GOLDMAN, M.D.

DANIEL F.SMITH, Dr. P.H.

California Department of Health Services

Environmental Epidemiology and Toxicology Branch

Emeryville, California

RAYMOND R.NEUTRA, M.D.

California Department of Health Services

Berkeley, California

L.DUNCAN SAUNDERS, D.T.P.H.

University of Alberta

Edmonton, Canada

ESTHER M.POND, Ph.D.

JAMES STRATTON, M.D.

California Department of Health Services

Sacramento, California

KIM WALLER, M.P.H.

University of California, Berkeley

Berkeley, California

RICHARD J.JACKSON, M.D.

California Department of Health Services

Berkeley, California

KENNETH W.KIZER, M.D.

California Department of Health Services

Sacramento, California

ABSTRACT. Aldicarb, a carbamate pesticide, is the most potent pesticide in the market and has a LD50of 1 mg/kg. In the United States it is illegal to use aldicarb on certain crops, e.g., watermelons, because it is incorporated into the flesh of the fruit. Once an accidental or illegal use of such a potent pesticide occurs, there is no easy way for the agricultural or public health system to protect the populace. This paper describes the impact of one such event upon the health of individuals and the institutions of California. On July 4, 1985, California and other western states experienced the largest known outbreak of food-borne pesticide illness ever to occur in North America. This was attributed to watermelons contaminated through the illegal or accidental use of aldicarb by a few farmers in one part of the state. Within California, a total of 1 376 illnesses resulting from consumption of watermelons was reported to the California Department of Health Services (CDHS). Of the 1 376 illnesses, 77% were classified as being probable or possible carbamate illnesses. Many of the case reports involved multiple illnesses associated with the same melon among unrelated individuals. Seventeen individuals required hospitalization. There were 47 reports of illness

Archives of Environmental Health 45(4):229–36, July/August 1990. Reprinted with permission of the Helen Dwight Reid Educational Foundation. Published by Heldref Publications, 1319 Eighteenth Street, N.W., Washington, D.C. 20036–1802. Copyright 1990.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education Pesticide Food Poisoning from Contaminated Watermelons in California, 1985 LYNN R.GOLDMAN, M.D. DANIEL F.SMITH, Dr. P.H. California Department of Health Services Environmental Epidemiology and Toxicology Branch Emeryville, California RAYMOND R.NEUTRA, M.D. California Department of Health Services Berkeley, California L.DUNCAN SAUNDERS, D.T.P.H. University of Alberta Edmonton, Canada ESTHER M.POND, Ph.D. JAMES STRATTON, M.D. California Department of Health Services Sacramento, California KIM WALLER, M.P.H. University of California, Berkeley Berkeley, California RICHARD J.JACKSON, M.D. California Department of Health Services Berkeley, California KENNETH W.KIZER, M.D. California Department of Health Services Sacramento, California ABSTRACT. Aldicarb, a carbamate pesticide, is the most potent pesticide in the market and has a LD50 of 1 mg/kg. In the United States it is illegal to use aldicarb on certain crops, e.g., watermelons, because it is incorporated into the flesh of the fruit. Once an accidental or illegal use of such a potent pesticide occurs, there is no easy way for the agricultural or public health system to protect the populace. This paper describes the impact of one such event upon the health of individuals and the institutions of California. On July 4, 1985, California and other western states experienced the largest known outbreak of food-borne pesticide illness ever to occur in North America. This was attributed to watermelons contaminated through the illegal or accidental use of aldicarb by a few farmers in one part of the state. Within California, a total of 1 376 illnesses resulting from consumption of watermelons was reported to the California Department of Health Services (CDHS). Of the 1 376 illnesses, 77% were classified as being probable or possible carbamate illnesses. Many of the case reports involved multiple illnesses associated with the same melon among unrelated individuals. Seventeen individuals required hospitalization. There were 47 reports of illness Archives of Environmental Health 45(4):229–36, July/August 1990. Reprinted with permission of the Helen Dwight Reid Educational Foundation. Published by Heldref Publications, 1319 Eighteenth Street, N.W., Washington, D.C. 20036–1802. Copyright 1990.

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education involving pregnant women, two of whom reported having subsequent stillbirths. Thirty-five of the remaining pregnant women were followed—up 9 mo after the epidemic; no additional stillbirths were found. To control the epidemic, it was necessary to embargo on July 4 and to destroy all watermelons in the state on July 7 and to effect a field certification program. The epidemic and the costly resultant control measures illustrate the difficulties in assuring the safe use of the most potent pesticide. The use of pesticides is controlled by an elaborate set of crop specific regulations. State and federal regulators use laboratory tests of produce samples to insure that regulations are followed. When inadvertent or illegal applications of pesticide occur in a particular crop, there is no system that guarantees that the public will not be exposed. For most pesticides, the effects may not be dramatic, but when a potent pesticide appears in a widely eaten commodity, the impact on health and the institutions that are designed to protect it can be devastating. This paper describes the course of one such event. ON JULY 3, 1985, the Oregon Department of Health notified the California Department of Health Services (CDHS) of several cases of possible pesticide illness related to consumption of watermelons that were thought to have been grown in Arizona.1,2 At 4:00 A.M. on July 4, a 62-y-old woman on digoxin therapy was treated at a Lake County California, emergency department for hypotension, severe bradycardia (31 beats per minute [bpm]), atrial fibrillation, diaphoresis, vomiting, diarrhea, lacrimation, salivation, and muscle twitching. She had eaten watermelon about 30 min earlier. Her symptoms resolved following treatment with atropine. Two other family members who had consumed the same watermelon were also ill and had similar though milder symptoms. The treating physician notified the San Francisco Bay Area Regional Poison Control Center, which subsequently notified CDHS. Later on the morning of July 4, Oregon officials reported to CDHS that aldicarb sulfoxide (ASO), a toxic degradation product of aldicarb, had been detected in several of the melons related to illness episodes in that state and that the origin of the melons was, in fact, from California.1,2 Aldicarb, CAS No. 116–06–3, is a cholinesterase-inhibiting carbamate pesticide that is not registered for use on watermelons in the U.S. but commonly used on citrus, cotton, potatoes, peanuts, and soybeans. Within 2 h, calls to 10 California poison control centers, 20 selected emergency departments, and 1 county health department had identified an additional 12 presumed cases of pesticide illness related to consumption of watermelons. This included a group of 4 individuals in Bakersfield who had eaten a striped melon purchased at a roadside stand, a group of 6 individuals who had eaten a striped melon from a Los Angeles-area supermarket warehouse, and 2 individuals in the San Francisco Bay Area who had eaten green melons purchased at different retail stores. These illnesses were investigated by state and local health officials, and arrangements were made for obtaining watermelon samples. Just prior to noon on July 4, statewide media advisories were issued that warned against eating watermelons, and an embargo was placed on the sale of watermelons throughout California. Usual product recall mechanisms were inoperative because the day was a national holiday. By late afternoon on July 4, case investigations and tracking of sources of melons back through the distribution chains had implicated a single Kern County shipper in several, but not all of the episodes. Subsequently, in the melon from the first known California case, ASO was found at 2.7 parts per million (ppm). The embargo remained in effect for the next 3 d. On July 7, all watermelons in retail outlets or in the chains of distribution were destroyed because it was impossible to distinguish ASO-contaminated melons from melons free of ASO. A field certification program was implemented on July 10, and the embargo was lifted. Surveillance after that time identified only one further illness episode in California associated with a melon that tested positive for ASO. Product certification was conducted by the California Department of Food and Agriculture (CDFA) and involved testing composite samples of melons from fields for aldicarb and its metabolites. Melons from fields that tested negative were labeled by CDFA to certify that they had been cleared. Methods Commencing late on the morning of July 4, the public was advised through the mass media to report any watermelon-associated illness to their local health department. An active surveillance network set up by CDHS on July 5 involved (a) daily calls to California’s 10 regional poison control centers and selected emergency departments, (b) daily contact with all local health departments in California, and (c) periodic calls to several western states and the western provinces of Canada. Local health departments were asked to complete and return an illness report form (described below) to CDHS for all cases reported to them. They were also asked to periodically call selected hospital emergency departments within their jurisdiction so as not to miss illnesses severe enough to require emergency treatment or hospitalization. The CDHS illness report form and a case-definition algorithm were developed based on the expected cholinergic symptoms resulting from ingestion of aldicarb (Table 1). The case definition divided illness reports into three categories: (1) probable, (2) possible, or (3) unlikely, depending on timing of symptom onset, nature and severity of

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education symptoms, and number of people ill from the same melon. The CDHS illness report forms were distributed rapidly to local health department officials in an effort to speed collection of uniform case information. The forms included questions about symptoms, time and location of melon purchase, and others who ate the same melon. All reports of illness with date of onset after July 10 were telephoned to CDHS and promptly reviewed by a physician to identify probable poisoning cases from melons bearing certification labels. Additional information was sought from persons who reported illness, if necessary. Samples of melons from probable cases were collected and shipped by local health departments to the nearest participating CDFA or CDHS laboratory for analysis. Analyses for aldicarb, ASO and aldicarb sulfone (AS) in watermelons were performed by CDFA. In addition, several confirmatory analyses were performed by the U.S. Food and Drug Administration (FDA) regional laboratory in Los Angeles and CDHS’s Food and Drug Laboratory. Analyses by CDFA and FDA were performed using liquid chromatography. The minimum detection level was usually 0.2 ppm but ranged between 0.1 and 0.5 ppm ASO. Confirmatory analyses by CDHS were performed using gas chromatography and a method developed by Union Carbide for detecting aldicarb residues in water [method ALDICARB-FPD-WATER(a)].3 The detection level by this method for all aldicarb residues combined was 0.01 ppm. Selection of melons for testing was completed in two stages. During the first stage, i.e., prior to July 10, attempts were made to confirm the source and extent of the epidemic. The second stage, after July 10, involved sampling melons from fields that had passed the certification program. The theoretical ability of the field certification sampling plan to detect a single, highly contaminated field was quite good, but given the practical limit of detection of ASO, the necessary compositing of samples, and the large number of fields involved, it was still possible that some contaminated melons might have reached retail markets. Therefore, melons associated with “probable” illnesses that occurred after July 10 were assigned top priority for testing. Active surveillance continued until the end of August 1985. All case reports were reviewed later for completeness, and additional data were sought when needed. Data from individual case reports were then analyzed using the standardized case definition. In March 1986, an attempt was made to contact by mail and telephone the 47 women who reported being pregnant when they experienced their watermelon-associated illness. Information was obtained on the pregnancy outcome, birthing complications, birth defects, and any other relevant problems. Six of the 47 were lost to follow-up. Of the remaining 41, 2 denied having been pregnant, and 1 refused to participate. The other 38 women provided information on a standard questionnaire about the outcome of the pregnancy and the baby’s health. Case reports were tabulated in an attempt to identify the geographic source(s) of the epidemic. Illness rates and numbers of illness were mapped by county using SAS/GRAPH, 1980 U.S. Census population denominators, and Tektronix plotter.4 In an attempt to pinpoint store chains (and through them, wholesalers and farmers) who might have sold contaminated melons, we compared the frequency with which the various chains were identified by “probable” cases and by “unlikely” cases. Our reasoning was that “unlikely” cases probably approximated a random sample of the population as to their use of the various store chains so that we could analyze the data as one would a case-control study. We calculated odds ratios and 95% confidence limits. This measure of association divides the odds of using a particular store chain by “probable” cases by the odds of using that chain among “unlikely” cases. For rare diseases, it is an estimate of the rate ratio, i.e., the incidence of poisoning in patrons of that chain divided by the incidence in nonpatrons. Distributors that served counties or store chains with high odds ratios would be most suspect as sources for contaminated watermelons. Because of the difficulty in using the complete case definition given in Table 1, which required asking cases about the occurrence of multiple symptoms in several categories, simpler alternative case definitions were explored using data on symptom rates and onset times. Results Active surveillance. Case reports were received for dates as early as June 1, 1985. Table 2 shows the number of case reports received in California for the period of active surveillance (June–August 1985) by case classification. In all, 1 376 case reports were received; 78% were classified as probable or possible pesticide poisoning. The geographic distribution of illnesses was evaluated in an attempt to identify the origin of the contaminated melons, but mapping did not suggest a source or sources. Analysis of stores where melons associated with pre-July 10 illness were purchased showed that there were four major supermarket chains involved. Only one of these had a significantly elevated odds ratio, 1.89 (95% confidence limits 1.00 and 3.56), for “probable” vs. “unlikely” illness reports. However, the watermelon distribution systems were too intermingled to quickly determine the suppliers for this chain. The majority of incidents (61%) involved one person becoming ill after eating a melon. Twenty-two percent of the reports involved 2-person episodes; 10% were 3-person clusters, and 3% were 4-person clusters. Additional clusters involving 5, 6, 9, and 13 persons becoming ill after eating from the same melon also were reported.

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education Table 1.—Case Definitions for Watermelon-Associated Illness Outbreak—California, July 1985 Classification of Cholinergic Symptoms Group 1: Gastrointestinal Abdominal pain Nausea and/or vomiting Diarrhea Group 2: Other peripheral autonomic Blurred vision and/or watery eyes Pinpoint pupils Excess salivation Sweating or clamminess Group 3: Skeletal muscle Muscular weakness Twitching Group 4: Central nervous system Seizures Disorientation or confusion Excitation Classification of Illness Reports 1. Probable case: Melon positive for aldicarb or aldicarb metabolites; onset ≤2 h after consuming melon; AND ONE OF THE FOLLOWING: Multiple groups of cholinergic symptoms or a single group of symptoms and more than one person ill from the same melon; OR onset between 2 and 12 h after consuming melon, multiple symptom groups, and more than one person ill from the same melon. 2. Possible case: Onset less than 2 h after consuming melon, a single group of symptoms, and no other illnesses reported from the melon; OR onset within 2 to 12 h after consuming melon and multiple symptoms or symptoms from only one group. 3. Unlikely case: Some other cause of illness judged to be more likely; OR any illness with onset of symptoms more than 12 h after eating melon. Table 2.—Numbers and Percentages of Watermelon-Associated Illnesses Reported in California, June 1– August 31, 1985, by Onset Date and Case Definition Case definition Onset Onset Onset unknown Total 6/01–7/10 7/11–8/31 Probable 493 (49%) 197 (57%) 2 (8%) 692 (51%) Possible 269 (27%) 101 (29%) 6 (23%) 376 (27%) Unlikely 195 (19%) 40 (12%) 0 … 235 (17%) Incomplete 48 (5%) 7 (2%) 18 (69%) 73 (5%) Total 1 005 … 345 … 26 … 1 376   Note: See Table 1 for case definition. Figure 1 shows the epidemic curve of probable watermelon illness reports within California by date of purchase of melons. The first probable case was reported for a melon purchased June 16; reports rose sharply thereafter. Reports peaked for melons purchased on July 3. There was an abrupt decline in reports for melons purchased after July 4, which coincided with the melon embargo, media advisories, and other measures. Illness onsets for probable cases peaked July 4, and, as with onsets by purchase date, sharply declined after July 4. Severity of illness. Most people had relatively short-term minor illnesses that resolved quickly; however, some were severely ill. Several reports of cardiac arrhythmias, dehydration, seizures, and other severe illnesses were associated with watermelon consumption before and after July 10 (Table 3). Overall, 17 persons were reported to require hospital admission, 16 of whom were admitted prior to July 10. Of 6 reported deaths, all of which were autopsied, none could be attributed by the coroners to aldicarb/ASO ingestion. Pregnancy outcomes. Of the 38 women pregnant when they had watermelon-associated illness, 18 were classified as probable cases, 9 as possible, and 10 as unlikely. In one case, the information to classify the illness was inadequate. During the two months immediately after the incident, three pregnancies were investigated. Two near-term pregnancies resulted in stillbirths following acute illnesses associated with watermelon consumption. One pregnant woman had a “probable” illness, and the other had a “possible” illness. Fetal tissues from both stillbirths tested negative for aldicarb and its metabolites (personal communication, Union Carbide Corporation, 1985).

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education Fig. 1. Watermelon aldicarb illness reports by case definition and melon purchase date California, 1986. Nine months later an attempt was made to contact the other women who reported being pregnant when they had their watermelon-associated illness. Among the 35 women contacted, 2 neonatal deaths were reported. One was a premature infant born to a mother with “possible” illness, who reported headache and fever 1 wk prior to delivery, raising the possibility that the premature birth and death may have been due to an infection. The second death was due to hypoplastic left heart syndrome; this occurred to a mother with a “probable” illness during the 25th wk of gestation. Laboratory testing. Of 62 laboratory-tested melons purchased prior to July 10 and associated with illness, 9 (14.5%) were ASO positive. For illnesses associated with melons purchased after July 10, 188 melons were tested, and 1 (0.5%) was ASO positive. In no case was the parent compound aldicarb identified, but some melons contained AS. In addition to the 1 noted aldicarb-positive melon purchased in California after July 10, 2 other aldicarb-positive CDFA-labeled watermelons associated with illness after July 10 were reported in Canada (personal communication, 1985) and Oregon.1 One of the 3 positive melons found after July 10 could be traced to a particular California field. Case definition. The case definition algorithm was compared with symptom reports (Table 4). In general, the 28 with laboratory confirmation of watermelon contamina Table 3.—Severe Illness in California Associated With Watermelon Consumption, Summer 1985.   Number of cases reported Condition Before July 10 July 10 and after Seizures 3 0 Loss of consciousness 4 1 Cardiac arrhythmia 6 1 Hypotension 4 0 Dehydration 17 2 Anaphylaxis 3 0 Note: Some individuals had more than one of the above symptoms. tion with ASO were more likely to have had symptoms compatible with carbamate poisoning than those for whom melon tests were negative or not performed. Symptoms reported by at least 50% of those who consumed confirmed ASO-contaminated melons included abdominal pain, nausea, vomiting, diarrhea, blurred vision, salivation, sweating, muscle twitching and/or weakness, and disorientation. These symptoms were also found, but with less frequency, among cases classified as probable, possible, and unlikely. Symptom group 1 (gastrointestinal symptoms) showed the smallest differences in reporting between laboratory-confirmed melon

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education Table 4.—Cases With Various Symptoms, by Case Definition*: California Aldicarb in Watermelon Episode, 1985†   Illness report classification Symptom Melon positive‡ for ASO Probable Possible Unlikely (Total) 28 (100.0)§ 689 (100.0) 311 (100.0) 303 (100.0) Group 1   Abdominal pain 23 (82.1) 493 (71.6) 212 (68.2) 158 (52.2) Nausea/vomiting 24 (85.7) 563 (81.7) 250 (80.4) 200 (66.1) Diarrhea 24 (85.7) 466 (67.6) 179 (57.6) 181 (59.7) Group 2   Blurred vision 17 (60.7) 223 (32.4) 31 (10.0) 40 (13.2) Salivation# 14 (50.0) 128 (18.6) 22 (17.1) 21 (6.9) Sweating# 20 (71.4) 356 (51.7) 61 (19.6) 59 (19.5) Group 3   Muscle// 15 (53.6) 222 (32.2) 41 (13.2) 40 (13.2) Group 4   Disorientation 17 (60.7) 208 (30.2) 36 (11.6) 45 (14.9) Other symptoms   Breathing** 2 (7.1) 20 (2.9) 4 (1.3) 5 (1.6) Urination†† 5 (17.9) 150 (21.8) 21 (6.8) 22 (7.3) Fever‡‡ 4 (14.3) 151 (21.9) 44 (14.2) 52 (17.2) Hearing problem 0 (0) 20 (2.9) 9 (2.9) 5 (1.7) *See Table 1. †Excludes 45 cases that could not be classified and with untested melons. ‡Not mutually exclusive from other classifications. ASO is a metabolite of aldicarb. §Values are given as number and percentage, which appear in parentheses. #Excessive salivation or sweating. //Muscle weakness and/or twitching. **Difficulty breathing or shortness of breath. ††Excessive urination or incontinence. Was not included in case definition because of likelihood of urination associated with consumption of a large amount of watermelon. ‡‡As noted by respondent. cases and the other case groups, and therefore may be the least specific of the cholinesterase inhibitor symptoms. Fever was reported by 14.3% of those who consumed laboratory-positive melons and by 14% to 22% of those in the other groups. Fever was included to differentiate those persons with infectious illness (e.g., viral gastroenteritis), but it failed to do this (possibly because fever was self-reported). To screen for over-reporting, questions were asked about hearing problems; less than 3% of persons in any category reported same. Several simpler case definitions were developed for illness that occurred within 2 h of watermelon consumption. The following symptom patterns were compared to the more complex case definition used for this outbreak: diarrhea only, nausea and/or vomiting only, diarrhea and nausea/vomiting, and diarrhea or nausea/vomiting. For the four definitions, sensitivity and specificity were calculated. Diarrhea or nausea/vomiting within 2 hr of watermelon consumption had the highest sensitivity (79%) and specificity (82%). Hence, if cases with ASO-positive melons had been classified on the basis of these two symptoms alone, 79% of the cases defined as “probable” using the complete definition would have been identified. Cantaloupe-associated illness. In addition to the reports of watermelon-related illness, there were in this same period 77 illness reports associated with consumption of about 25 cantaloupes. Many of these cantaloupes were tested, and all tested negative for ASO. About half were tested for other pesticide residues (i.e., carbamates, organophosphates, and chlorinated pesticides); none were found. A few complaints about other types of fruit (e.g., honeydew melons) also were received, but none could be linked to any pesticides. Discussion Aldicarb is the most acutely toxic pesticide registered in the United States. It has two primary breakdown products: (1) ASO (for rats, LD50=0.9 mg/kg) and (2) AS (for rats, LD50=24 mg/kg).5 With well over 1 000 reports of probable pesticide illness from within and outside California, this episode ranks as the largest recorded North American outbreak of foodborne pesticide illness. In the past, intentional or inadvertent misapplication of aldicarb to cucumbers and mint was associated with similar, though more limited, outbreaks. The spectrum of illness reported in these outbreaks was similar to the current

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education one, ranging from mild to severe. No deaths have been reported from any of these food poisoning episodes.6–8 In these cases, as with the 1985 watermelon epidemic, identification of the epidemic was dependent on alert clinicians who quickly recognized the symptoms and signs of carbamate pesticide poisoning and on the abilities of laboratories to identify aldicarb metabolites (a test not routinely performed when testing for pesticide residues). Without careful surveillance, it would be easy to overlook such an epidemic because of the nonspecific nature of symptoms of early cholinesterase toxicity. Aldicarb has been implicated in at least two deaths in agricultural workers.9,10 Although no deaths in this epidemic were attributable to ASO, the spectrum of clinical illness seen in this episode included many severely ill people. Some of the more serious symptoms and signs reported, such as marked bradycardia and hypotension, could have been lethal, particularly in the very young, the elderly, and the chronically ill. The prompt embargo and widespread publicity almost certainly were responsible for preventing a much larger epidemic and saving lives. There are no known long-term or reproductive effects of aldicarb and its metabolites in the absence of maternal toxicity, and it is not a suspect carcinogen.5,11 One would expect that there would be a certain number of people in the state who had gastrointestinal illness onset coincidentally within 2 h of eating melon; hence, some of the sporadic cases were reported through September. However, under-reporting at the beginning of the outbreak may have been substantial, given the long Fourth of July weekend and that the active surveillance system required 1 wk to implement fully. For example, the poison control centers were initially so overwhelmed with calls that they often did not have time to record complete reports; thus, many cases may have been lost to follow-up during the first week of the outbreak. However, a greater proportion of “probable” cases occurred after July 11; this suggests that a reporting bias in favor of minor coincidental illness may have occurred when the epidemic was first reported by the media. It has been asserted that the entire epidemic was created by media coverage and reporting of illness coincidental with eating aldicarb-contaminated watermelons. However, the episode cannot be explained by coincidence. This is clear from the fact that those with laboratory-positive watermelons were likely to have a greater number of symptoms and more symptoms of severe acetyl cholinesterase inhibition than others. A study of the geographic case distribution revealed no single retail source for contaminated melons, even when confined to cases confirmed with ASO-Positive tests in melons. This is probably due to the prevailing methods of distributing watermelons, which involve mixing unlabeled melons from numerous different sources. This results in marked intermingling during the distribution process. Any future outbreaks of illness related to watermelon will likely be difficult to trace using epidemiological information alone. This certainly suggests a need for better labeling or tracking methods for watermelons. There were many illnesses clinically compatible with carbamate poisoning but associated with aldicarb-negative melons. Although, as mentioned above, some of these could have been coincidental occurrences, it is also possible that the laboratory analysis could not detect ASO at levels that can cause illness. This issue has implications for the regulation of pesticide residues in foods and deserves further study. An outbreak of this explosiveness and magnitude could never have been investigated and documented without the full support and participation of California’s local health departments, emergency departments, and poison control centers. The work-load generated by this event in these institutions and CDFA is hard to quantify. CDHS has time accounting records that suggest thousands of person hours were devoted by one agency alone. Since the 1985 epidemic, California has begun an integrated food surveillance program that involves local health and environmental health departments, CDFA, and CDHS. Monitoring for pesticide-related illness uses a report form similar to the one used for the 1985 outbreak, but with the simpler case definition for a probable case of carbamate poisoning of diarrhea or nausea/vomiting within 2 h of eating produce. This case definition is easier to use in the field and has sufficient sensitivity (79%) so that any future outbreaks of consequence should not be missed, even though it will overlook one of five individual illnesses. * * * * * * * * * * Management of this epidemic involved hundreds of individuals in government agencies at all levels and at numerous private institutions. The authors thank all of these persons. Special thanks go to Harvey F.Collins, Ph.D., for his editorial assistance; to Barbara Hopkins, David Epstein, and Martha Harnly, who assisted with data processing and analysis and illustrations; and to Carolyn Harris and Gette Meneses, who typed the manuscript. Submitted for publication June 27, 1989; revised; accepted for publication March 13, 1990. Requests for reprints should be sent to: Lynn R.Goldman, M.D., Environmental Epidemiology and Toxicology Branch 5900 Hollis Street, Suite E, Emeryville, CA 94608. * * * * * * * * * * References 1. Green MA, Heumann MA, Wehr HM, et al. An outbreak of watermelon-borne pesticide toxicity. Am J Pub Health 1987;77:1431–34. 2. U.S. Centers for Disease Control. Aldicarb food poisoning from contaminated melons—California. MMWR 1986;35:254– 58. 3. Union Carbide Corporation. ALDICARB-FPD-WATER(a). Agricultural Products Co., Inc., Research and Development Department, P.O. Box 8361, South Charleston, West Virginia 25303. March 1980.

OCR for page 588
Environmental Medicine: Integrating a Missing Element into Medical Education 4. SAS Institute, Inc. SAS/GRAPH user’s guide. Cary, NC: SAS Institute, Inc., 1981. 5. Risher JF, Mink FL, Stara JF. The toxicologic effects of the carbamate insecticide aldicarb in mammals: a review. Environ Health Perspectives 1987;72:267–81. 6. Marshall E. The rise and decline of Temik (TM). Science 1985;229:1369–71. 7. Goes AE, Savage EP, Gibbons G, et al. Suspected foodborne carbamate pesticide intoxications associated with ingestion of hydroponic cucumbers . Am J Epidemiol 1980;111:254–60. 8. U.S. Centers for Disease Control. Suspected carbamate intoxications—Nebraska. MMWR 1979;28:133–34. 9. U.S. Environmental Protection Agency. Aldicarb information sheet. July 1985. 10. Lee MH, Ransdell JF. A farmworker death due to pesticide toxicity: a case report. J Toxicol Environ Health 1985;14:239–46. 11. Cambon C, Declume C, Derache R. Effect of the insecticidal carbamate derivatives (carbofuran, primicarb, aldicarb) on the activity of acetylcholinesterase in tissues from pregnant rats and fetuses. Toxicol Appl Pharmacol 1979;49:203–08.