1046
Hospital
Practice
HOSPITAL OUTBREAK OF CLOSTRIDIUM
PERFRINGENS FOOD-POISONING MAIR THOMAS
NORMAN D. NOAH
Epidemiological Research Laboratory, Central Public Health Laboratory, Colindale Avenue, London NW9 5HT G. E. MALE Environmental Health Department, St. Albans AL1 3TZ
to arise until Wednesday, by which time 379 patients (a third of the total) had been affected. Diarrhoea was the main clinical feature, but in several patients with chronic constipation it was the last symptom to appear. Many patients had pain and vomiting, but fever was not observed. 2 dehydrated patients required intravenous fluids. A man with severe hydrocephalus died; inhalation of vomit was found at necropsy with acute enteritis and
liquid facces, proximal
to
an
impacted megacolon.
Numerous Clostridium perfringens (C. welchii) were found in post-mortem cultures of faeces, intestinal tract, and spleen, but no other food-poisoning organisms were isolated.
Epidemiology M. F. STRINGER R.
J.
MARGARET KENDALL GILBERT
Food Hygiene Laboratory, Central Public Health Laboratory, London NW9 5HT
P.
H. JONES
K. D. PHILLIPS
Public Health Laboratory, Luton and Dunstable Hospital, Luton LU4ODZ
An outbreak of Clostridium
perfringens (C.welchii) food-poisoning affected a third of the patients in a large hospital, and one frail patient died. C.perfringens type A, serotype 1, was isolated from 46 (61%) of 76 patients examined and from food, and a new serotype (61) was isolated from 16. The attack-rate among patients who ate a minced-ham dish was 78%. The cooking and storage of this mince was faulty: cuts of meat were much too large, they were kept at room temperature too long before refrigeration, and after cooking they were put into mincers used also for raw meat. C.perfringens type A, serotype 1, was isolated from meat scraps in a mincer. Final reheating was inadequate to destroy even vegetative bacteria, and multiplication may have occurred during slow distribution to the wards. Outbreaks of C.perfringens food-poisoning are common in hospitals, and some underlying problems are discussed. A plea is made for the Food Hygiene Regulations to apply to National Health Service premises and for simple but effective reforms in institutional catering management.
Summary
THE OUTBREAK
AT about 6 P.M. on a Saturday some patients in a large hospital developed gastroenteritis. Cases continued
6. Hill, H. N. Am.J. med. Sci. 1950, 219, 394. 7. Raftery, E. B., Banks, D. C., Oram, S. Lancet, 1969, ii, 1147. 8. Stem, I. Am. Heart J. 1950, 40, 325. 9. Stein, I., Weinstein, J. Am. Practnr Digest Treat. 1955, 6, 64. 10. Wagner, M. L., Arbeit, S. R. Am. J. Cardiol. 1964, 13, 81. 11. Bean, W. B., Read, C. T. Am. Heart J. 1942, 23, 362. 12. Bean, W. B., Flamm, G. W., Sapadina, A. Am. J. Med. 1949, 7, 765. 13. Ryle, J. in The Natural History of Disease, p. 323. London, 1949. 14. Parry, C. H. An Inquiry into the Symptoms and Causes of the Syncope Anginosa commonly called Angina Pectoris; illustrated by Dissections. 1899. 15. Frank, N. J. Am. med. Ass. 1959, 170, 1147. 16. Harrison, C. E., Jr., Spittell, J. A., Jr., Mankin, H. T. Proc. Mayo Clin.
1962, 37, 293. Johnson, W. J., Achor, R. W. P., Burchell, H. B., et al. Archs. 1959, 103, 253. 18. Steinke, W. E., Curry, J. H. Am. J. Cardiol. 1968, 22, 436. 19. Rubler, H. B., Diefenbach, W. C. L. ibid. 1960, 6, 989. 17.
intern.
Patients in 19 of 34 wards were ill. The only common factor "minced diet" served in the affected wards but not sup to plied any of the 473 patients in 15 unaffected wards. Within the affected wards the attack-rate was 78% (200/256) among those prescribed the minced diet and 42% (179/422) of patients nominally having other diets (table I). The nature of the outbreak was characteristic of Cperfringens food-poisoning. The usual incubation period is 8-24 h. Foods prepared exclusively for the "minced diet" within this period were ham, served at 7.30 A.M., pureed butter beans and tapioca at noon, and grated cheese at 330 P.M. Only the butter beans were available for laboratory examination. was a
Med.
Preparation and Distribution of the Suspected Foodstuff On the day before the outbreak eleven 5.5kg hams were boned and rolled then cooked by steaming. They were stood to cool in a centrally heated room and were minced at some unrecorded time between 10 A.M. and 4 P.M. The mince was then put into the refrigerator. At 6 A.M. next day this mince was mixed with water and a soup powder and heated in large steam-jacketed cauldrons. Distribution Food from the central kitchen was distributed to 34 wards in three vehicles, each divided internally by two horizontal shelves. These were heated electrically before leaving the kitchen and cooled gradually en route for up to an hour. There was some further delay before serving from electric hotplates in the wards.
of Catering Premises These were visited on the fourth day of the outbreak. The kitchen was old and not well arranged. The staff was below establishment strength, particularly of trained workers. Mincers and choppers were shared between raw and cooked meats and contained raw-meat residues. There was an accumulation of unwashed pans. Refrigerators were not well ordered, and in one there was a cooked leg of beef, very much larger than the safe maximum weight of 3kg. A room open to the main kitchen and warmed by two radiators was used for the prerefrigeration "cooling" of cooked dishes. Staff toilets in Condition
outbuilding were unclean, and one was without toilet paper; although paper towels were available a dirty roller towel hung in a corner. The premises had been visited two months earlier by the environmental health officer, who had found widespread disrepair, obsolescence, and poor hygiene. an
His report included detailed recommendations for urgent upgrading of the hospital catering facilities.
Bacteriological Examination of Food Samples taken four days after the outbreak included scrap ings of raw meat from mincers, cooked ham, stock, dry butter beans, soup powder, grated cheese, and pease pudding. These were selected because they were relevant to the menus served before the outbreak.
1047 TABLE I-ATTACK-RATES BY DIET IN THE
C.perfringens type A, serotype 1, and Salmonella typhimurium phage type DT1 were isolated from meat scrapings in the mincer; a different strain of C.perfringens was isolated from butter beans. An anaerobic gram-positive bacillus cultured from the stock died before it could be studied. Bacteriological Examination of Patients and Staff Faeces were plated directly onto horse-blood agar containing 120 µg/ml neomycin sulphate, and a semiquantitative estimate of the growth of C.perfringens was made after 24 h anaerobic incubation. Colonies were then classified according to the &agr;-&bgr;-non-hæmolytic convention. Where possible, three colonies of each type were picked for serological typing. C.perfringens was isolated from 61 of 75 patients, other than the deceased, and from 2 of 5 symptom-free kitchen staff examined. Serology of C.perfringens Isolates Of the 238 human isolates of C.perfringens serotyped, 29 from the deceased patient, 207 from the 61 other affected patients, and 2 from kitchen staff. Further strains were obtained from the butter beans and the scrapings from the mincers. They were serotyped by slide agglutination with the 64 antisera available at the time of the outbreak. Among the 62 patients from whom C.perfringens was isolated, serotype 1 was common to 46 (74%), including the patient who died and the 2 dehydrated patients. Most of these cultures were x-hxmolytic. No other recognisable serotype was common to more than 3 of these 62 patients, but a large number of untypable strains was encountered, many of them &agr;-&bgr;-hæmolytic. To ascertain whether a second serotype was involved in the outbreak, antisera were prepared against 5 such strains from different patients. One of these, now designated serotype 61, matched strains from 16 of the 62 patients (26%), thus increasing the proportion identifiable from 57% to 66%. Serotypes 1 and 61 were found together in 8 patients, including the deceased. The isolate derived from the mincer was of serotype 1, and that from butter beans was untypable. An antiserum prepared from the latter did not agglutinate with any strain from patients. Multiple serotypes of C.perfringens were identified in many individuals; altogether 15 serotypes were recognised, and at least 8 types were isolated post mortem from the deceased patient. were
is
19
AFFECT I I:D WARDS
practicable provision should be made for rapid cooling after cooking, adequate refrigeration and speedy serving of the meal after removal from the refrigerator".’ This was written in 1954, yet C.perfringens foodpoisoning outbreaks remain common in hospitals, and the annual number reported appears to be increasing (table n). In the outbreak described, the preparation of a minced diet was faulty at successive stages. Joints of meat were cut to twice the 3 kg safe maximum size for thorough cooking and quick cooling. The delay between the end of cooking and the beginning of refrigeration not
exceeded the safe maximum of 1-½h .2—4 Cooked meat was minced in machines used for raw meat-a hazardous practice made even more dangerous by inadequate cleaning.5-’ Each of these three faults of size, delay, and contamination is easily avoidable. Premises and equipment were dilapidated. Because staffing was inadequate, supervision was incomplete, particularly at weekends and at 6 A.M., when breakfasts were prepared. The need for a competent supervisor in hospital kitchens during all working hours is evident. A bonus for inconvenient hours may be the price to pay for safety. Clinical Picture The diarrhoeal non-febrile illness which affected most of the victims was characteristic of C.perfringens foodpoisoning. In this large outbreak a few were severely ill, and one frail patient died. An unusual clinical feature was absent or delayed diarrhoea in a number of patients who presented with pain or vomiting, including the deceased patient, who had enteritis proximal to an impacted megacolon. Incontinence and chronic constipation are not uncommon in hospital populations and -may require management with both constipating agents and periodic enemas. It may be that colonic stasis contributed to the severity of the illness in some patients. Some of the toxic effects of intestinal obstruction may be related to the presence of C.perfringens.8 9 Experimentally, morphine has increased the pathogenicity of this species for the gut of sheep. 10
DISCUSSION
The Outbreak "Meat should be eaten
on
the
day
it is cooked. If this
TABLE 11---C.PERFRINGENS FOOD-POISONING IN ENGLAND AND
WALES (1971-76)*
Bacteriology The investigation was complex because multiple serotypes of C.perfringens were isolated from many patients, including at least 8 from the deceased. Food-poisoning with multiple serotypes of C.perfringens has been described.11 " The serotype 1 most frequently isolated from patients was cultured from a mincer in which the suspect food was prepared. The revised serotyping scheme developed at the Food Hygiene Laboratory" 12 enabled 15 serotypes to be distinguished, identifying two-thirds of the strains examined. Only types 1 and 61 were found in more than 3 people. Catering Services
’Data from the
Epidemiological Hygiene Laboratory, Colindale. 1 Provisional.
Research
Laboratory
and Food
The World Health Organisation has recognised the special problems of aviation catering and has set out
1048
guidelines for food safety in aircraft. 13 Hospital catering is another demanding service. Meals must be provided every day for large numbers of patients with differing dietary needs, who are unable to come to a central dining-room, so that difficulties of distribution are added to those of preparation. The D.H.s.s. Health Service Catering ManuaP advice is clear and appropriate. It could be more effectively applied and the value of the environmental health officer’s knowledge enhanced if the prevailing Crown Exclusion were to be waived and the .Food Hygiene Regulations became applicable to N.H.s. hospitals. 14- I, The value of safe catering relative to other aspects of patient care should be recognised. We thank Dr J. G. M. Mortimer and Dr T. M. Pollock for their support and constructive criticisms; Dr J. I. Pugh and Dr L. A. Nahkla for their reports on the post-mortem examination of the deceased patient; and the medical, nursing, and catering staff of the hospital for their cooperation in the investigation of the outbreak. Requests for reprints should be addressed to N. D. N. APPENDIX
Summary of Recommendations Made after the Outbreak 1. Hot food should not be held at ambient temperature for more than 1 z h before consumption or refrigeration, and a ventilated cooling-room should be available.
Points of View A MEASURE OF ACCEPTANCE
Patients’ Attitudes
Hæmophilia "HAEMOPHILIA should be stamped out," said one of our patients with mild haemophilia. He spoke with strong emotion even though, with an antihaemophiliaglobulin level of 6%, he has a bleed requiring treatment only three or four times a year. to
The half dozen or so local haemophiliacs who have had vasectomies are also in the "mildly affected"category, and a young woman with von Willebrand disease who asked for a termination of pregnancy and tubal ligation had had, to my knowledge, only one bleeding episode in her life requiring replacement therapy. Of the severely affected patients who bleed often and sometimes severely, one has asked for a vasectomy, three known carriers (mothers of severely affected children) and two wives are pregnant, and one wife and one carrier are undergoing exhaustive investigations for infertility. It may seem surprising, in these days of restricted families, prenatal diagnosis for a number of serious conditions, and selective abortion, that families of severely affected haemophiliacs have not eagerly followed these trends. This is not because of ignorance, since all are taught about the mode of inheritance of the disease; and when they are given opportunities to discuss their personal situation they appear to be well informed. The British Medical Journal, in an editorial referring the prenatal diagnosis of sickle-cell anaemia,’ asks, "Do we really know enough about sickling disorders to be able to predict the quality of life for any particular It is essential that our enthusiasm for individual? to
...
1. British
Medical Journal, 1977, i, 531.
2. Portions of meat should not exceed 3 kg. 3. If cooked food is reheated a boiling temperature must be attained throughout its bulk. 4. Mince should be made only from raw meat. 5. Utensils should be dismantled and washed in very hot water with detergent after every use. 6. Raw and cooked food processes should be separated. 7. Sample meals should be retained in the refrigerator for 48 h (as is done in many schools). 8. The D.H.S.S. manual2 should be in daily use, and its provisions should be explained to staff by the catering officer.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
A. W.
H., Turner, W. C. Br. med. J. 1954, i, 686. Health Service Catering Manual: Hygiene. D.H.S.S., 1974, vol. 2. Public Health Laboratory Service, Br. med. J. 1970, ii, 611. Hobbs, B. C. Food Poisoning and Food Hygiene. London, 1974. Gilbert, R. J. J. Hyg., Camb. 1969, 67, 249. Beck, A., Foxell,
Lancet, 1969, ii, 677.
Dempster, J. F. J. Hyg., Camb. 1973, 71, 369. Williams, B. W. Lancet, 1927, i, 907. Gleeson-White, M. H., Bullen, J. J. ibid. 1955, i, 384. Bennetts, H. W. Bull. Coun. sci. md. Res., Melb. no. 57, 1932. Sutton, R. G. A., Hobbs, B. C. J. Hyg., Camb. 1968, 66, 135. Hughes, J. A., Turnbull, P. C. B., Stringer, M. F. J. med. Microbiol. 1976 9, 475. 13. Wkly epidem. Rec. 1977, 52, 101. 14. Food Hygiene Regulations. s.i., 1970, no. 1172. H.M. Stationery Office. 15. Lancet, 1970, ii, 510. 16. Br. med. J. 1977, i, 851.
technology should not put pressure on families of this type". Much the same could be said of haemophilia. In our experience families at risk do not press for such studies, and interest in the investigation of carrier status is only lukewarm. new
*
*
«
A recent television programme dealing with some aspects of hxmophilia2 brought some interesting comments. The mildly affected patients thought that the programme was accurate and fair, although it portrayed haemophilia of a severity quite outside their own experience. Severely affected patients and their parents, however, thought it painted too black a picture and that it had undermined the confidence of employers and friends. One baby-sitter was quoted as saying, "I’ll never dare to look after your little boy again". Occasionally, mildly affected haemophiliacs appear with remarkably bad haemorrhages—e.g., a subungual haematoma 3 days after a hammer blow; a large cyst on the buttock 5 days after sitting on a cricket ball; a tense knee from a week-old haemarthrosis. Having few bleeds they expect to get away unscathed after most minor traumas. They seldom visit their doctors, and they are not always able to judge when treatment is necessary. The younger children take a long time to become accustomed to injections and to appreciate the benefits of treatment. As a result, haemorrhages, though rare, considerably disrupt their lives. Pinkerton,3referring to childhood asthma, points out that in a chronic condition acceptance of the diagnosis and its implications is essential if the patient’s family is to learn how to manage the episodes and make the best use of available treatment. A measure of acceptance is vital for haemophiliacs, too. Failure to come to terms 2. Genetic Chance with the Hæmophllic 1976. 3. Pinkerton, P. Br. med. J. 1972, iii, 462.
Family.
B.B.C. Television. Dec
Hospital
Practice
HOSPITAL OUTBREAK OF CLOSTRIDIUM
PERFRINGENS FOOD-POISONING MAIR THOMAS
NORMAN D. NOAH
Epidemiological Research Laboratory, Central Public Health Laboratory, Colindale Avenue, London NW9 5HT G. E. MALE Environmental Health Department, St. Albans AL1 3TZ
to arise until Wednesday, by which time 379 patients (a third of the total) had been affected. Diarrhoea was the main clinical feature, but in several patients with chronic constipation it was the last symptom to appear. Many patients had pain and vomiting, but fever was not observed. 2 dehydrated patients required intravenous fluids. A man with severe hydrocephalus died; inhalation of vomit was found at necropsy with acute enteritis and
liquid facces, proximal
to
an
impacted megacolon.
Numerous Clostridium perfringens (C. welchii) were found in post-mortem cultures of faeces, intestinal tract, and spleen, but no other food-poisoning organisms were isolated.
Epidemiology M. F. STRINGER R.
J.
MARGARET KENDALL GILBERT
Food Hygiene Laboratory, Central Public Health Laboratory, London NW9 5HT
P.
H. JONES
K. D. PHILLIPS
Public Health Laboratory, Luton and Dunstable Hospital, Luton LU4ODZ
An outbreak of Clostridium
perfringens (C.welchii) food-poisoning affected a third of the patients in a large hospital, and one frail patient died. C.perfringens type A, serotype 1, was isolated from 46 (61%) of 76 patients examined and from food, and a new serotype (61) was isolated from 16. The attack-rate among patients who ate a minced-ham dish was 78%. The cooking and storage of this mince was faulty: cuts of meat were much too large, they were kept at room temperature too long before refrigeration, and after cooking they were put into mincers used also for raw meat. C.perfringens type A, serotype 1, was isolated from meat scraps in a mincer. Final reheating was inadequate to destroy even vegetative bacteria, and multiplication may have occurred during slow distribution to the wards. Outbreaks of C.perfringens food-poisoning are common in hospitals, and some underlying problems are discussed. A plea is made for the Food Hygiene Regulations to apply to National Health Service premises and for simple but effective reforms in institutional catering management.
Summary
THE OUTBREAK
AT about 6 P.M. on a Saturday some patients in a large hospital developed gastroenteritis. Cases continued
6. Hill, H. N. Am.J. med. Sci. 1950, 219, 394. 7. Raftery, E. B., Banks, D. C., Oram, S. Lancet, 1969, ii, 1147. 8. Stem, I. Am. Heart J. 1950, 40, 325. 9. Stein, I., Weinstein, J. Am. Practnr Digest Treat. 1955, 6, 64. 10. Wagner, M. L., Arbeit, S. R. Am. J. Cardiol. 1964, 13, 81. 11. Bean, W. B., Read, C. T. Am. Heart J. 1942, 23, 362. 12. Bean, W. B., Flamm, G. W., Sapadina, A. Am. J. Med. 1949, 7, 765. 13. Ryle, J. in The Natural History of Disease, p. 323. London, 1949. 14. Parry, C. H. An Inquiry into the Symptoms and Causes of the Syncope Anginosa commonly called Angina Pectoris; illustrated by Dissections. 1899. 15. Frank, N. J. Am. med. Ass. 1959, 170, 1147. 16. Harrison, C. E., Jr., Spittell, J. A., Jr., Mankin, H. T. Proc. Mayo Clin.
1962, 37, 293. Johnson, W. J., Achor, R. W. P., Burchell, H. B., et al. Archs. 1959, 103, 253. 18. Steinke, W. E., Curry, J. H. Am. J. Cardiol. 1968, 22, 436. 19. Rubler, H. B., Diefenbach, W. C. L. ibid. 1960, 6, 989. 17.
intern.
Patients in 19 of 34 wards were ill. The only common factor "minced diet" served in the affected wards but not sup to plied any of the 473 patients in 15 unaffected wards. Within the affected wards the attack-rate was 78% (200/256) among those prescribed the minced diet and 42% (179/422) of patients nominally having other diets (table I). The nature of the outbreak was characteristic of Cperfringens food-poisoning. The usual incubation period is 8-24 h. Foods prepared exclusively for the "minced diet" within this period were ham, served at 7.30 A.M., pureed butter beans and tapioca at noon, and grated cheese at 330 P.M. Only the butter beans were available for laboratory examination. was a
Med.
Preparation and Distribution of the Suspected Foodstuff On the day before the outbreak eleven 5.5kg hams were boned and rolled then cooked by steaming. They were stood to cool in a centrally heated room and were minced at some unrecorded time between 10 A.M. and 4 P.M. The mince was then put into the refrigerator. At 6 A.M. next day this mince was mixed with water and a soup powder and heated in large steam-jacketed cauldrons. Distribution Food from the central kitchen was distributed to 34 wards in three vehicles, each divided internally by two horizontal shelves. These were heated electrically before leaving the kitchen and cooled gradually en route for up to an hour. There was some further delay before serving from electric hotplates in the wards.
of Catering Premises These were visited on the fourth day of the outbreak. The kitchen was old and not well arranged. The staff was below establishment strength, particularly of trained workers. Mincers and choppers were shared between raw and cooked meats and contained raw-meat residues. There was an accumulation of unwashed pans. Refrigerators were not well ordered, and in one there was a cooked leg of beef, very much larger than the safe maximum weight of 3kg. A room open to the main kitchen and warmed by two radiators was used for the prerefrigeration "cooling" of cooked dishes. Staff toilets in Condition
outbuilding were unclean, and one was without toilet paper; although paper towels were available a dirty roller towel hung in a corner. The premises had been visited two months earlier by the environmental health officer, who had found widespread disrepair, obsolescence, and poor hygiene. an
His report included detailed recommendations for urgent upgrading of the hospital catering facilities.
Bacteriological Examination of Food Samples taken four days after the outbreak included scrap ings of raw meat from mincers, cooked ham, stock, dry butter beans, soup powder, grated cheese, and pease pudding. These were selected because they were relevant to the menus served before the outbreak.
1047 TABLE I-ATTACK-RATES BY DIET IN THE
C.perfringens type A, serotype 1, and Salmonella typhimurium phage type DT1 were isolated from meat scrapings in the mincer; a different strain of C.perfringens was isolated from butter beans. An anaerobic gram-positive bacillus cultured from the stock died before it could be studied. Bacteriological Examination of Patients and Staff Faeces were plated directly onto horse-blood agar containing 120 µg/ml neomycin sulphate, and a semiquantitative estimate of the growth of C.perfringens was made after 24 h anaerobic incubation. Colonies were then classified according to the &agr;-&bgr;-non-hæmolytic convention. Where possible, three colonies of each type were picked for serological typing. C.perfringens was isolated from 61 of 75 patients, other than the deceased, and from 2 of 5 symptom-free kitchen staff examined. Serology of C.perfringens Isolates Of the 238 human isolates of C.perfringens serotyped, 29 from the deceased patient, 207 from the 61 other affected patients, and 2 from kitchen staff. Further strains were obtained from the butter beans and the scrapings from the mincers. They were serotyped by slide agglutination with the 64 antisera available at the time of the outbreak. Among the 62 patients from whom C.perfringens was isolated, serotype 1 was common to 46 (74%), including the patient who died and the 2 dehydrated patients. Most of these cultures were x-hxmolytic. No other recognisable serotype was common to more than 3 of these 62 patients, but a large number of untypable strains was encountered, many of them &agr;-&bgr;-hæmolytic. To ascertain whether a second serotype was involved in the outbreak, antisera were prepared against 5 such strains from different patients. One of these, now designated serotype 61, matched strains from 16 of the 62 patients (26%), thus increasing the proportion identifiable from 57% to 66%. Serotypes 1 and 61 were found together in 8 patients, including the deceased. The isolate derived from the mincer was of serotype 1, and that from butter beans was untypable. An antiserum prepared from the latter did not agglutinate with any strain from patients. Multiple serotypes of C.perfringens were identified in many individuals; altogether 15 serotypes were recognised, and at least 8 types were isolated post mortem from the deceased patient. were
is
19
AFFECT I I:D WARDS
practicable provision should be made for rapid cooling after cooking, adequate refrigeration and speedy serving of the meal after removal from the refrigerator".’ This was written in 1954, yet C.perfringens foodpoisoning outbreaks remain common in hospitals, and the annual number reported appears to be increasing (table n). In the outbreak described, the preparation of a minced diet was faulty at successive stages. Joints of meat were cut to twice the 3 kg safe maximum size for thorough cooking and quick cooling. The delay between the end of cooking and the beginning of refrigeration not
exceeded the safe maximum of 1-½h .2—4 Cooked meat was minced in machines used for raw meat-a hazardous practice made even more dangerous by inadequate cleaning.5-’ Each of these three faults of size, delay, and contamination is easily avoidable. Premises and equipment were dilapidated. Because staffing was inadequate, supervision was incomplete, particularly at weekends and at 6 A.M., when breakfasts were prepared. The need for a competent supervisor in hospital kitchens during all working hours is evident. A bonus for inconvenient hours may be the price to pay for safety. Clinical Picture The diarrhoeal non-febrile illness which affected most of the victims was characteristic of C.perfringens foodpoisoning. In this large outbreak a few were severely ill, and one frail patient died. An unusual clinical feature was absent or delayed diarrhoea in a number of patients who presented with pain or vomiting, including the deceased patient, who had enteritis proximal to an impacted megacolon. Incontinence and chronic constipation are not uncommon in hospital populations and -may require management with both constipating agents and periodic enemas. It may be that colonic stasis contributed to the severity of the illness in some patients. Some of the toxic effects of intestinal obstruction may be related to the presence of C.perfringens.8 9 Experimentally, morphine has increased the pathogenicity of this species for the gut of sheep. 10
DISCUSSION
The Outbreak "Meat should be eaten
on
the
day
it is cooked. If this
TABLE 11---C.PERFRINGENS FOOD-POISONING IN ENGLAND AND
WALES (1971-76)*
Bacteriology The investigation was complex because multiple serotypes of C.perfringens were isolated from many patients, including at least 8 from the deceased. Food-poisoning with multiple serotypes of C.perfringens has been described.11 " The serotype 1 most frequently isolated from patients was cultured from a mincer in which the suspect food was prepared. The revised serotyping scheme developed at the Food Hygiene Laboratory" 12 enabled 15 serotypes to be distinguished, identifying two-thirds of the strains examined. Only types 1 and 61 were found in more than 3 people. Catering Services
’Data from the
Epidemiological Hygiene Laboratory, Colindale. 1 Provisional.
Research
Laboratory
and Food
The World Health Organisation has recognised the special problems of aviation catering and has set out
1048
guidelines for food safety in aircraft. 13 Hospital catering is another demanding service. Meals must be provided every day for large numbers of patients with differing dietary needs, who are unable to come to a central dining-room, so that difficulties of distribution are added to those of preparation. The D.H.s.s. Health Service Catering ManuaP advice is clear and appropriate. It could be more effectively applied and the value of the environmental health officer’s knowledge enhanced if the prevailing Crown Exclusion were to be waived and the .Food Hygiene Regulations became applicable to N.H.s. hospitals. 14- I, The value of safe catering relative to other aspects of patient care should be recognised. We thank Dr J. G. M. Mortimer and Dr T. M. Pollock for their support and constructive criticisms; Dr J. I. Pugh and Dr L. A. Nahkla for their reports on the post-mortem examination of the deceased patient; and the medical, nursing, and catering staff of the hospital for their cooperation in the investigation of the outbreak. Requests for reprints should be addressed to N. D. N. APPENDIX
Summary of Recommendations Made after the Outbreak 1. Hot food should not be held at ambient temperature for more than 1 z h before consumption or refrigeration, and a ventilated cooling-room should be available.
Points of View A MEASURE OF ACCEPTANCE
Patients’ Attitudes
Hæmophilia "HAEMOPHILIA should be stamped out," said one of our patients with mild haemophilia. He spoke with strong emotion even though, with an antihaemophiliaglobulin level of 6%, he has a bleed requiring treatment only three or four times a year. to
The half dozen or so local haemophiliacs who have had vasectomies are also in the "mildly affected"category, and a young woman with von Willebrand disease who asked for a termination of pregnancy and tubal ligation had had, to my knowledge, only one bleeding episode in her life requiring replacement therapy. Of the severely affected patients who bleed often and sometimes severely, one has asked for a vasectomy, three known carriers (mothers of severely affected children) and two wives are pregnant, and one wife and one carrier are undergoing exhaustive investigations for infertility. It may seem surprising, in these days of restricted families, prenatal diagnosis for a number of serious conditions, and selective abortion, that families of severely affected haemophiliacs have not eagerly followed these trends. This is not because of ignorance, since all are taught about the mode of inheritance of the disease; and when they are given opportunities to discuss their personal situation they appear to be well informed. The British Medical Journal, in an editorial referring the prenatal diagnosis of sickle-cell anaemia,’ asks, "Do we really know enough about sickling disorders to be able to predict the quality of life for any particular It is essential that our enthusiasm for individual? to
...
1. British
Medical Journal, 1977, i, 531.
2. Portions of meat should not exceed 3 kg. 3. If cooked food is reheated a boiling temperature must be attained throughout its bulk. 4. Mince should be made only from raw meat. 5. Utensils should be dismantled and washed in very hot water with detergent after every use. 6. Raw and cooked food processes should be separated. 7. Sample meals should be retained in the refrigerator for 48 h (as is done in many schools). 8. The D.H.S.S. manual2 should be in daily use, and its provisions should be explained to staff by the catering officer.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
A. W.
H., Turner, W. C. Br. med. J. 1954, i, 686. Health Service Catering Manual: Hygiene. D.H.S.S., 1974, vol. 2. Public Health Laboratory Service, Br. med. J. 1970, ii, 611. Hobbs, B. C. Food Poisoning and Food Hygiene. London, 1974. Gilbert, R. J. J. Hyg., Camb. 1969, 67, 249. Beck, A., Foxell,
Lancet, 1969, ii, 677.
Dempster, J. F. J. Hyg., Camb. 1973, 71, 369. Williams, B. W. Lancet, 1927, i, 907. Gleeson-White, M. H., Bullen, J. J. ibid. 1955, i, 384. Bennetts, H. W. Bull. Coun. sci. md. Res., Melb. no. 57, 1932. Sutton, R. G. A., Hobbs, B. C. J. Hyg., Camb. 1968, 66, 135. Hughes, J. A., Turnbull, P. C. B., Stringer, M. F. J. med. Microbiol. 1976 9, 475. 13. Wkly epidem. Rec. 1977, 52, 101. 14. Food Hygiene Regulations. s.i., 1970, no. 1172. H.M. Stationery Office. 15. Lancet, 1970, ii, 510. 16. Br. med. J. 1977, i, 851.
technology should not put pressure on families of this type". Much the same could be said of haemophilia. In our experience families at risk do not press for such studies, and interest in the investigation of carrier status is only lukewarm. new
*
*
«
A recent television programme dealing with some aspects of hxmophilia2 brought some interesting comments. The mildly affected patients thought that the programme was accurate and fair, although it portrayed haemophilia of a severity quite outside their own experience. Severely affected patients and their parents, however, thought it painted too black a picture and that it had undermined the confidence of employers and friends. One baby-sitter was quoted as saying, "I’ll never dare to look after your little boy again". Occasionally, mildly affected haemophiliacs appear with remarkably bad haemorrhages—e.g., a subungual haematoma 3 days after a hammer blow; a large cyst on the buttock 5 days after sitting on a cricket ball; a tense knee from a week-old haemarthrosis. Having few bleeds they expect to get away unscathed after most minor traumas. They seldom visit their doctors, and they are not always able to judge when treatment is necessary. The younger children take a long time to become accustomed to injections and to appreciate the benefits of treatment. As a result, haemorrhages, though rare, considerably disrupt their lives. Pinkerton,3referring to childhood asthma, points out that in a chronic condition acceptance of the diagnosis and its implications is essential if the patient’s family is to learn how to manage the episodes and make the best use of available treatment. A measure of acceptance is vital for haemophiliacs, too. Failure to come to terms 2. Genetic Chance with the Hæmophllic 1976. 3. Pinkerton, P. Br. med. J. 1972, iii, 462.
Family.
B.B.C. Television. Dec
