Five large outbreaks of food poisoning are described in which clinical, epidemiological or laboratory data indicated Clostridium perfringens as the causative organism. The foodstuff common to all incidents was boiled salmon served cold as an hors d'oeuvre. In all cases the fish had been subject to a long period of cooling or storage between boiling and consumption. It is thought that multiplication of the organism occurred during this time.

Recommendations are made for the avoidance of further similar incidents.

One thousand pharyngeal swab specimens were processed for aerobic culture to determine the carriage rate of Gram-negative bacilli (GNB). The isolates were identified and their sensitivity determined to 11 antibacterial drugs following standard techniques.

Similar pharyngeal carriage rates of GNB were found among the various groups of healthy subjects. Patients had higher colonization rates (27%) than healthy subjects (16%). The increase in prevalence of GNB seemed to be associated with underlying diseases and duration of hospitalization.

Klebsiella (36%) was the most frequent genus amongst the 215 isolates of GNB followed by pseudomonas (13%), enterobacter (13%) and acinetobacter (10%). Others were less frequently isolated.

Over 70% of all isolates were resistant to ampicillin (79%) and carbenicillin (72%); 55, 45 and 43% were resistant to cephalothin, tetracycline and streptomycin, respectively. The great majority of the strains were sensitive to the remaining six drugs.

The hospital isolates were more resistant than the non-hospital isolates to most drugs tested. The hospital strains were also more often multiply resistant (89%) than the non-hospital strains (60%). Sixty-five different resistance antibiograms of 1—10 drugs were observed among 191 strains. More varied types of antibiograms were observed among hospital strains.

The high frequency of multiple drug resistance of the isolates is an indication of the extensive use of antibacterial drugs, indicating the need for a policy for judicious use of drugs.

In a prospective survey carried out over 9 months in 1984 in the Department of Bacteriology, Belfast City Hospital, Campylobacter jejuni was isolated from 24 out of 1200 faecal specimens (2%) from patients with acute diarrhoea. This isolation rate is much lower than that from other parts of mainland Britain which report an isolation rate of between 8 and 15%. It is difficult to explain this large discrepancy but the limited availability of unpasteurized milk and the generally worse summer here (fewer barbecues, picnics) may be some reasons to explain this low incidence in N. Ireland.

Three groups of premises (butchers' shops, supermarkets and general dealers) which sell raw and cooked meats were compared. Salmonellas were not detected, but Escherichia coli, and to a lesser degree Staphylococcus aureus and Streptococcus faecalis, were widely distributed in all three groups of premises. Contamination of hands, towels and nail brushes was related to poor working practices. The presence of E. coli or Str. faecalis on slicing machines was associated with contamination of meat samples. A number of wiping cloths were heavily contaminated with E. coli, and many also contained Clostridium perjringens. Few premises provided written cleaning plans, and in many cases staff did not receive an adequate training in food hygiene. The use of disinfectants as part of the cleaning process did not necessarily reduce the level of bacterial contamination. In general there was poor correlation between microbiological results and a visual inspection made by an environmental health officer. The possible reasons for this finding arc discussed.

Amounts of Mycobacterium tuberculosis antibodies were determined in sera from patients with either active or inactive tuberculosis and in healthy subjects by an immunoenzymatic assay in which whole BCG cells attached covalently to polystyrene disks were used as antigen. Statistically significant differences (P < 0·005) were found both between the active and inactive tuberculosis groups and between the active group and healthy controls. No significant differences were found between the inactive group and controls. Since this procedure is efficient (91%) and can be used in areas which lack laboratory equipment, it appears promising for individual serodiagnosis and for epidemiological surveys.

Investigation of 39 JK-type coryneform isolates from patients at a single hospital revealed that 23 possessed plasmids, which formed six groups on restriction endonuclease analysis. Four of the groups were associated with production of similar bacteriocin-like substances, and shared a minimum of 6·4 kilobase pairs of DNA. These plasmids, found in isolates from different patients, provide strong direct evidence that person-to-person transmission of JK bacteria had occurred within the hospital.

The microporous polyethylene hollow fibre membrane is capable of adsorbing small-sized lipopolysaccharides (LPS) prepared by sonication dispersion, column chromatography on Sephadex G-75 and filtration through a filter membrane with a nominal pore size of 0·025 μg. Small-sized LPS had a one-thousandth of endotoxin activity as compared to intact LPS, when determined by the Synthetic Chromogenic Substrate method of LAL with a specific endotoxin activity of 73·7 ng/μg LPS. Fluorescent microscopy of fluorescein conjugated LPS on a microporous polyethylene hollow fibre showed that fiuorescein-LPS was adsorbed through the entire depth of the membrane texture. Accordingly the adsorption capacity of the filter for small-sized LPS was determined as 1·65 mg LPS/3·68 m2 surface/116 mg fibre/module.

A new adsorbent, microporous polyethylene chopped fibre, was produced from a high density polyethylene. This can adsorb lipopolysaccharides (LPS) linearly up to 2 h, and showed the highest capacity to adsorb LPS when compared with two other polyethylene-based adsorbents and a polystyrene-based adsorbent. More than twice as much orange II and 4-nitroquinoline N oxide were adsorbed in the new adsorbent as was LPS. The adsorption isotherm of the new adsorbent for LPS was of Ln type, the correlation between adsorption and concentration of solute was proportional; whereas orange II and 4-nitroquinoline N oxide were of L type (greater adsorption than Ln type); tetrachloroethylene adsorption was of S type, less than Ln type. Adsorption of LPS to the new adsorbent increased when temperature rose, whereas adsorption of orange II and 4-nitroquinoline N oxide decreased. These data suggest that the binding of LPS to the new adsorbent is a hydrophobic interaction, whereas the binding of both orange II and 4-nitroquinoline N oxide is not. The new adsorbent has a greater potential for the removal of endotoxin from tap water than other commercially available adsorbents such as charcoal and Amberlite XAD-2.

The prevalence of nasal colonization and infection with methicillin-resistant Staphylococcus aureus (MRSA) among patients and staff was studied in a section of a Paediatric Surgical Unit in Lisbon between February and July 1985. Nasal colonization was demonstrated in 41% of burned patients, 5% of non-burned patients and 35 % of the nurses. Infection by MRSA occurred in 30% of the burns.

The isolates had identical serological patterns, slight differences on phage typing and were resistant to methicillin, cephalosporins, tetracycline, erythromycin and aminoglycosides. A chloramphenicol resistance plasmid of 3 Md was present in those isolates which were chloramphenicol resistant and a small plasmid of 1·7 Md which coded for constitutive erythromycin resistance was present in many isolates. Gentamicin, tetracycline and inducible erythromycin resistance were chromosomal.

Several reasons for the apparent low virulence of the isolates are discussed. Attempts to control the outbreak by the discharge of colonized or infected patients, improvement of nursing practices and treatment with temporary removal from work of the colonized nurses did not eliminate the organism from the unit.

The survival of Escherichia coli in an aerosol was studied at several temperatures and over a range of relative humidities using a Henderson apparatus.

Death occurred in two phases, the first lasting approximately 1 min; in the second the number of viable microorganisms declined exponentially. E. coli was robust and remained viable for many hours. Death was most rapid at low humidities (< 50% r.h.) at 15 and 30 °C, with half-lives of 14 and 3 min respectively. In humid conditions the half-lives were much longer, approximately 83 and 14 min respectively.

Based on this work, preliminary recommendations for the climate of livestock buildings can now be given to control the airborne spread of E. coli.

A multiple-tube technique based on peptone water incubated at 44 °C for 24 h followed by detection of indole was found to be sensitive and specific for the detection of Escherichia coli in oysters and mussels. The method has the advantage of providing rapid results and is both less expensive and less time-consuming than other MPN techniques.

Four distinctive plasmid profile types have been identified in strains of Salmonella gold-coast isolated in Britain. Strains of one type, designated plasmid profile type 4, caused an extensive outbreak of food-poisoning in 1984, and it has been confirmed that the vehicle of infection was imported French pâté.

A survey on the bacteriological quality of both drinking water and flavoured drinks from coin-operated vending machines is reported. Forty-four per cent of 25 drinking water samples examined contained coliforms and 84% had viable counts of greater than 1000 organisms ml at 30 °C. Thirty-one flavoured drinks were examined; 6% contained coliforms and 39% had total counts greater than 1000 organisms ml. It is suggested that the D.H.S.S. code of practice on coin-operated vending machines is not being followed. It is also suggested that drinking water alone should not be dispensed from such machines.

The apparent reduction in the incidence of subsequent joint sepsis and of re-operation without evidence of infection during the course of a prospective study was an artifact of the analysis method.

Using the cross agglutinin absorption test 12 British leptospiral isolates of the Sejroe serogroup were identified to serovar level. Six strains isolated from cattle, two from pigs and one from a human were identified as Leptospira interrogans serovar hardjo. Two isolates from wildlife were identified as Leptospira interrogans serovar saxkoebing. One further strain isolated from wildlife closely resembled serovar saxkoebing, but specific identification was not possible. These are the first reported isolations of serovar saxkoebing in the United Kingdom. The problems associated with the cross agglutinin absorption test, and possible alternative typing procedures are discussed.

A total of 1203 unselected routine faecal samples from 1006 patients were cultured for Yersinia species by a cold enrichment technique. Seventy-five specimens (6·1%) from 63 patients were culture-positive for Yersinia spp. Fifty-two were Yersinia enterocolitica, 22 Yersinia frederiksenii and 1 Yersinia intermedia. The predominant Y. enterocolitica isolates belonged to biotype 1 – serotype 0:6, 30 or serotype 0:5, 27. Y. frederiksenii strains were non-typable. Forty isolates were recovered from 33 patients with gastroenteritis. During the study period 83 Salmonella spp. from 33 patients, 17 Shigella sonnei from 13 patients and 13 Campylobacter jejuni from 12 patients were cultured. Yersinia spp. was isolated in association with salmonella on three occasions, twice with rotavirus and once each with Shigella sonnei, Campylobacter jejuni and Trichuris trichiura.

A two-minute strep A direct swab test (SADST) was used to detect the presence of Lancefield group A streptococci (GAS) from the throats of 207 patients with pharyngitis at a primary-care clinic. The results were compared with a standard culture method. Fifty-one specimens were positive and 156 specimens were negative for GAS by culture. The SADST had a sensitivity of 96% (49 of 51) and specificity of 98·7% (154 of 156). The predictive values of a positive and negative SADST, for GAS, were 96% and 98·7 % respectively. The SADST showed negative reactions with five specimens containing beta-haemotytic streptococci other than GAS and 34 known stock cultures other than GAS. Our results indicate that SADST is a rapid, simple, convenient and reliable test to use for diagnosis of GAS pharyngitis at primary care clinics, physicians' offices and clinical laboratories.

Application of a skin lotion to the body after showering greatly reduced the number of bacteria and skin scales dispersed from 10 men and 10 women. This effect lasted for at least 4 h when surgical clothing was worn. The use of a skin lotion to reduce bacterial dispersal could provide a simple and inexpensive alternative to an ultraclean air system or uncomfortable operating clothing during surgery requiring these procedures.

A total of 571 swine sera collected at an abattoir in the city of Obihiro, Hokkaido during the period February–November 1984 were tested for antibody against human (H1N1) influenza virus strains. A high prevalence of antibody was observed for only 3 months from April to June in that year, in 81/180 sera (45·0%) to A/USSR/92/77 strain and in 50/180 sera (27·8%) to a current epidemic strain (A/Hokkaido/1/84). Some cross-reactions were observed between the A/USSR/92/77 and A/Hokkaido/1/84 antibodies (r = 0·75). Only minor relationships were noted between the A/New Jersey/8/76 (swine type H1N1) and A/USSR/92/77 (r = 0·35) or A/Hokkaido/1/84 (r = 0·51) antibodies. Absorption of sera positive for antibody to the A/Hokkaido/1/84 strain with the homologous virus strain removed all detectable antibodies, while the absorption of the sera with the A/New Jersey/8/76 strain produced incomplete absorption in one half of the sera tested. These resultsstrongly suggest that the swine became infected with a human H1N1 virus as piglets during an epidemic of influenza which occurred in the human population during January and February 1984.

Suspension tests were conducted on 69 commercial and 7 non-commercial disinfectant formulations to determine which classes of chemicals were most active against human rotavirus (HRV). Virus samples, in the presence of varying levels of organic matter, were exposed to the disinfectants for 1 min. The levels of remaining infectious virus were determined by plaque assay. Products were rated by their ability to reduce the levels of infectious virus by more than 3 log10 in the presence or absence of tryptose phosphate broth (peptides and inorganic salts) or fecal matter.

Of the commercially-available products tested, only 25% were rated as highly and 7% as moderately effective. The remaining 68% were either effective only in the absence of any additional organic matter (48%) or were completely ineffective (20%). The majority (64%) of the moderately and highly effective products were further examined for their ability to inactivate > 6 log10 of infectious HRV in the presence of fecal matter or tryptose phosphate broth. With one exception, all these products were still effective. Products potentially suitable as topical antiseptics, hard surface disinfectants and instrument soaks were identified. The results emphasize the care that should be exercised in the selection of disinfectants for the control and prevention of rotaviral infections.