Raw Milk Causes 1927 Epidemic

Copied from the Journal of the American Medical Association Magazine, November 5, 1927

AN OUTBREAK OF GASTRO-ENTERITIS MILK-BORNE EPIDEMIC AT DYERSBURG, TENN., CAUSED BY SALMONELLA SUIPESTIFER H. C. STEWART, M.D., C.P.H. Director, Division of Epidemiology AND WILLIAM LITTERER, Ph.C., M.D. Director, Division of Laboratories, Tennessee State Department of Health NASHVILLE, TENN.

Medical literature of recent years has contained an increasing number of articles dealing with outbreaks of gastro-enteritis. The articles deal with the etiologic agents, the symptomatology, the epidemiology, and the conveyers of the infectious agent. A variety of names have been given these outbreaks, and many organisms have been accused of being the cause of these conditions. This epidemic is described as an example of an outbreak of this kind, the vehicle being raw milk. Most of the outbreaks reported have shown the causative agent to be carried in some meat food. Nearly all epidemics which involve a large number of people show that milk has been the vector of the infectious agent.

SANITARY CONDITIONS IN DYERSBURG Dyersburg is the county seat town of Dyer County. The Mississippi River borders the county on the west. Dyersburg is 75 miles northeast of Memphis. The census of 1920 gave the city a population of 6,444, and local estimates made at the time of the investigation gave a population of about 9,000 persons, with 6,000 of these white and 3,000 colored.' The water supply is owned by the municipality and comes from wells from 600 to 800 feet deep ; the water is settled, and -rapid sand filtration is employed. The water is not chlorinated. The analyses of the previous six months had shown a water of good sanitary quality. There is a fairly complete system of sewerage, but there is an estimated number of between 500 and 600 privies in the city which have tubs or cans, with scavenger service. This service is not maintained with any great degree of completeness. Six dairies supply the bulk of the trade of the city. There are a number of small producers, owners of two or three cows, who supply from 25 to 35 gallons of milk daily. All milk used in Dyersburg is raw milk. Some of the ice cream used is made locally, and a good deal is shipped in from Memphis. Ice is manufactured and distributed by two local companies. Raw fruits and vegetables are secured from a variety of sources, and distributed through markets and pcddlers. Dyer County has a full-time county health department, and this service is cooperated in by the city of Dyersburg. The county health officer also serves as city health officer, and the sanitary inspector serves both county and city. A public health nurse and an office clerk complete the personnel of the health department. The work of the department has been well planned and well supervised. The need of some sort of milk control had previously been brought to the attention of the city officials. A privy ordinance was in the process of passage at the time of this outbreak, and has since been adopted.

CLINICAL SYMPTOMS After a short incubation period, ranging from eight to thirty-six hours after the entrance into the body of the causative agent, symptoms came on rather suddenly. Headache, followed by severe backache, nausea and vomiting, abdominal pains and diarrhea were almost constant symptoms. More than half the patients gave a history of one or more severe chills followed by a rapid rise in temperature to a height of 104 or 105 F., in some cases. A convulsion was the initial symptom in some of the young children affected. A few patients suffered a very marked collapse, with small pulse and cold, wet skin. Coma or slight delirium was noted in a few cases. As a rule, the acute symptoms subsided within two or three days, nearly all persons affected being up and around within a week after the onset of symptoms. A remarkable degree of muscular weakness was shown after the patient had passed through the acute stage of the sickness. No deaths occurred.

EPIDEMIOLOGIC OBSERVATIONS From Saturday night, September 4, until Monday night, September 6, fully 98 per cent of the cases had their onset. Prior to Saturday there had been no cases of this kind so far as the health department could ascertain. One hundred and fifty persons with this trouble were seen by physicians during the outbreak, and there were in all likelihood others with symptoms not severe enough to warrant the calling in of a medical attendant. The presence of this number of persons with practically identical symptoms, in such a short period of time, established the presence of an epidemic of a very explosive type, and gave rise to the natural assumption of a common vehicle of infection. Of the seventy-six patients studied, four had the onset of the disorder on Saturday, fifty-three on Sunday, eighteen on Monday, and one on Tuesday. Seventy-one of these were white and five colored, four of the colored patients being cooks in white households, and the other a boy working about the yard of a white family and taking his meals in the kitchen of that family. Thirty per cent of the cases occurred in children under 10 years of age, this being nearly twice the estimated population percentage of that age. The cases occurred in the well-to-do section of the city. Eighty-six per cent of the persons affected used milk as a beverage. An additional 11 per cent used milk in some other way. Eighty-four per cent of those affected used milk from dairy 1, which was the second largest distributor. No connection could be established between the occurrence of the disease and the use of any ice, ice cream, raw fruit or vegetable supply, nor could the sanitary conditions of the premises be shown to be involved. The city water supply was used in all cases investigated, either exclusively or nearly so. It is very difficult to conceive of a water-borne infection causing such severe symptoms in such a short period of time. Should it be possible for such a contamination to occur, it is quite beyond the possibility of chance for so many cases of the disease to have developed in one group of persons. It would have been impossible for a contact infection to develop so explosively. Thus the milk supply remained as the remaining vehicle of infection. Sixty-four, or 84 per cent, of the patients investigated obtained milk either wholly or partly from dairy 1. Three of the patients supplied with milk from dairy 2 partook of home-made ice cream in which milk and cream from dairy 1 had been used. The two other patients supplied by dairy 2 had milk drinks Saturday evening at the one drug store which was supplied with milk from dairy 1. The two users of milk from dairy 3 had been on a pleasure trip on Sunday and had eaten beef sandwiches at a roadside camp. They attributed their illness to these sand¬ wiches, as they said "they did not taste right" when they ate them. Two other members of this family used no milk at all and yet suffered violently. The two users of milk from dairy 4 also had had milk drinks at the drug store mentioned Saturday night. One boy using milk from the family cow had worked in the drug store Sunday as extra help, and had eaten whipped cream from the supply of cream from dairy 1. It can be seen, then, that, with the- exception of the cases in the family previously mentioned, all cases investigated can be either directly or indirectly traced to the milk supply of dairy 1. Sixty of the patients were exposed to milk from the morning's delivery only. Six were exposed to both morning and evening deliveries, and in six cases the time of delivery was not noted. It is reasonable, then, to assume that the causative agent was present in only one delivery, that of Saturday morning, September 4. Dairy 1 was visited and found to be quite well equipped for a dairy of such size, but not with all the equipment necessary for the production of clean, safe milk. The pasture used by the dairy had at one end a ditch, which was, at the time of the investigation, nearly dry. There had been no rains since August 21, but before that there had been rains every day since August 1. About 200 yards above the pasture the effluent from a large septic tank emptied into a small stream. At no time had the pasture been overflowed by water from this stream, but it was possible and entirely probable that some of the effluent from this tank flowed with the stream into this ditch, which under usual conditions was dry. No history of previous cases of a similar disease or of possible carriers was reported from the homes contributing sewage to this tank. The dairy owner had been away on a trip for three days, arriving home Saturday noon, September 4. He was not in a position to say definitely that all his usual requirements as to the handling of milk had been complied with during his absence. Saturday evening, he noticed that three of his cows came to the barn with flanks and udders considerably soiled with mud from the ditch in question. The possibility of human carriers was considered. All stool specimens from workers of the dairy were negative.

ISOLATION OF ORGANISM Bactériologie and serologie Studies.—Dr. R. L. Motley, pathologist to the Baird-Brewer Hospital of Dyersburg, Tenn., selected two cases for study as being representative of the entire group of patients with the disease. Within twelve hours of the onset of the symptoms, cultures had been made on endo-agar plates of the feces of the two patients. Several plates from each specimen of feces were made. On every plate from the two samples there appeared quite a number of colorless typhoid-like colonies of a gram-negative motile bacillus that conformed culturally to the bacillus paratyphoid organism. Several of these micro-organisms were studied from each plate and all presented the following characteristics : On agar plates they showed a uniform grayish translucent growth of a fairly motile gram-negative bacillus. They fermented dextrose, mannite, maltose and Russell's double sugar (alkaline slant), but did not react to lactose or saccharose. There was no liquefaction of gelatin. Agglutinative Tests : The organisms were strongly agglutinated by the serums of five patients who had had the disease under study, none of whom gave any history of any typhoid-like disease at any time. One of them had been inoculated with triple typhoid vaccine in 1918, but none of the others had ever had vaccine. All but one of the patients had recovered from their illness when the tests were made, and this patient had a continuous fever for three weeks. The serums of these patients also agglutinated strongly stock paratyphoid bacilli, but did not agglutinate stock typhoid or paratyphoid A. The organisms under study were not agglutinated by normal control serums. These results were also confirmed by Dr. William Krauss of Memphis, director of the West Tennessee Branch Laboratory of the State Department of Public Health. Their conclusions were that the micro-organism under investigation belonged to the bacillus paratyphoid group or Salmonella paratyphi group. Further Investigations.—A culture was sent to one of us (W. L.) to carry on a detailed study of the micro-organism. Fermentative Reactions : The micro-organism fermented, in addition, the following sugars : rhamnose, sorbite, galactose, xylose and dulcite (eight days). No fermentation was recorded on arabinose, trehalose and inosite. In testing inosite, Jordan's method in determining the pn values of the mediums at stated intervals was performed; it was very satisfactory with the control organisms used, as well as with the one under discussion. Indole was not found, and there was no Voges-Proskauer reaction. Agglutinin and Absorption Tests : Rabbits were immunized with two strains of Salmonella enteritidis (Gärtner) ; two strains of Salmonella paratyphi (Schottmüller) ; two strains of Salmonella aertrycke (Mutton-Newport), and two strains of Salmonella suipestifer. Also rabbits were immunized with the micro-organism under investigation. Very satisfactory titer was obtained from these tests, ranging from 1: 3,000 to 1: 10,000. Without going into the details of the well known absorption method, suffice it to say that the two stock strains of Salmonella suipestifer com¬ pletely absorbed the serum of the micro-organism under investigation, and likewise that the latter microorganism completely absorbed the serums of the two stock strains of Salmonella suipestifer. The serum of the organism under investigation and the two serums from stock strains of Salmonella suipestifer showed only partial absorptive powers to all of the other serums mentioned, and vice versa. A study of all the immune serums made it quite evident that the agglutinative tests alone could not be wholly relied on. Group agglutination with unabsorbed serum is often so pronounced as to give little definite information unless additional tests are made with properly absorbed serum. Agglutination Tests on Patients' Serums : The serums from three patients were obtained about two months after convalescence. The accompanying table shows their agglutinative powers. Agglutination Tests on Patients' Serums Isolated S. Sui- S. Aer- S. Para- S. EnterPatient Organism pestifer trycke typhi B. itidis 1 1:640 1:640 1:80 1:80 1:10 2 1: 320 1: 320 1: 60 1: 60 1: 10 3 1: 160 1: 160 1: 40 1: 40 1: 10 Application of the absorption tests with these serums shows that the isolated organism and Salmonella suipestifer are identical. The agglutinine of the other micro-organisms are practically absorbed by Salmonella suipestifer and by the strain under investigation. The other micro-organisms do not materially affect the agglutinative power of the serums to Salmonella suipestifer or to the isolated micro-organism under investigation. Study of the Feces After Convalescence : About one month after convalescence an attempt was made to recover the micro-organism. Five patients were selected and their stools examined every day for three days with negative results. Animal Experiments : The micro-organism under investigation was not markedly virulent or toxic for guinea-pigs or mice. Four rabbits were also inoculated with 0.01 cc. of an eighteen hour broth culture. These rabbits died in from five to seven days. The strain therefore has a virulence for rabbits which has been noted by Theobald Smith in his work with Bacillus cholerae-suis or Salmonella suipestifer. Autopsy showed pale spots on the spleen ; a mottled liver ; small hemorrhages in the lungs ; congestion of the vessels in the small intestine, and hemorrhagic areas in the peritoneum. The micro-organisms were recov¬ ered from the heart's blood, liver and spleen.

SUMMARY AND CONCLUSIONS 1. The outbreak of gastro-enteritis here described involved more than 150 people. 2. Symptoms of a toxic nature were noted, such as severe headache, nausea and vomiting, diarrhea and profound muscular weakness. A few cases showed delirium. The incubation period ranged from eight to thirty-six hours. Recovery was practically complete within one week. 3. The vehicle of infection was a raw milk supply. 4. It seems probable that the infection gained entrance to the milk through lax measures in cleansing the udders and flanks of the cows. 5. The organism isolated conforms in its morphology, tinctorial and serologie reactions, and virulence to rabbits to Salmonella suipestifer {Bacillus cholcrae-suis). 6. In a review of the literature we do not find any report of an outbreak caused by milk contamination with Salmonella suipestifer.