A serious outbreak of staphylococcal infections in the maternity units in Blackburn was investigated. There were considerably more than one hundred cases of breast abscesses altogether, well over half of which occurred in primiparae. Staphylococcus aureus, phage type 80, was the predominating organism throughout the outbreak and at the peak period during the early part of the outbreak this type was responsible for nearly 80% of the infections.

After the introduction of a number of procedures for the general reduction of cross-infection the incidence of breast abscess fell markedly and a controlled trial of an antibacterial cream, containing neomycin and hibitane, which was applied to the nasal mucosa of all infants and mothers in the test group of patients, was undertaken. The conditions obtaining in the test and control groups were identical in every way except that the control patients did not receive the neomycin-hibitane cream. There were about 1250 mothers and infants each in the test and control groups; the incidence of breast abscesses in the test group was 0·8% and in the control group it was 2·7%. The method adopted for the detection and treatment of carriers among the nursing staff broke down on two occasions; this fact and the emergence of an unforeseen source resulted in a larger number of infections than should have occurred. Had it not been for these incidents there is little doubt that the trial would have shown more conclusively the effectiveness of the neomycinhibitane cream, by the method laid down, in reducing cross-infection.

Investigation of the bacterial flora on the nasal mucosa of over 1000 infants in the control group yielded results of considerable interest. Of 300 cases where there was early colonization by Staph. albus, this organism established its dominating position in 70% of the cases and it was not subsequently displaced by Staph. aureus. The significance of this observation and the evidence favouring nasal dissemination of Staph.aureus as the most important cause of hospital cross-infection are discussed.

Long-chain fatty acids are not appreciably bactericidal to tubercle bacilli at neutral pH, but become so even in slightly acid solution. In this way tubercle bacilli, normally resistant to 0·1 N hydrochloric acid (pH 1·0), are rapidly killed at this pH if traces of fatty acid are present, such as can be found inside test tubes that have been plugged with cotton wool and sterilized by dry heat.

The lethal effect is largely prevented by non-ionic, cationic, and to a lesser extent anionic detergents, which probably take up the fatty acid into the micelles so that it can no longer attack the bacilli.

The various fatty acids identified in the deposit from heated cotton wool were tested individually and found to be bactericidal. They comprised the usual C12 to C18 saturated acids and also oleic acid. The latter, however, showed no particular lethal toxicity in excess of that shown by, for example, palmitic acid. Shorter chained fatty acids than those represented in the deposit were less lethal.

We wish to thank Dr A. T. James for the chromatographic analyses and for advice; Dr R. J. W. Rees for carrying out counts of total acid-fast bacilli in two experiments; Messrs. Robinson & Sons Ltd, Chesterfield, for information on the industrial processes involved in the preparation of cotton wool; Mr E. J. Latham for advice on preparing glassware; and Dr J. Marks for helpful criticism of the manuscript.

1. An attempt has been made to follow up a consecutive series of seventy-two patients for the presence of Phase 1 C.F. antibody approximately 2 years or more after the acute illness. Fifty-one of the series were tested.

2. Fifteen of the fifty-one patients had detectable amounts of antibody, generally in low titre.

3. The presence of Phase 1 antibody correlated well, in older age-groups, with the duration of convalescence following the acute illness. There was no correlation with the duration of fever.

4. All cases with Phase 1 antibody also had Phase 2 antibody, usually in slightly higher titre. In no case in which Phase 2 antibody was absent or present in low titre was Phase 1 antibody found.

5. It is suggested that the presence of Phase 1 C.F. antibody is an indication of past persistent infection. It cannot necessarily be concluded that it is an indication of present persisting infection.

We wish to thank Dr H. Silverstone, Senior Lecturer in Medical Statistics, University of Queensland, for statistical advice.

When killed radioactive bacterial spores are inhaled by guinea-pigs and retained in the lung, it is found that the rate of loss up to 5 weeks later is sufficiently constant for the initial retention of a group of animals to be calculated from the measured retention at a later time (say 3 weeks after exposure). If, therefore, the inhaled cloud contained also a pathogen, in known proportion to the spores, the initial retention of the pathogen could be calculated. The experiments apply to groups of animals, but it is believed that the conclusions are valid also for single animals. In experiments with Br. suis it was found that the fate, as regards infection, of individual animals within a group exposed to a common dosage was not significantly related to the estimated retained dose, despite the wide variation of individual doses about the group mean.

A simple micro-method of measuring ‘D-’ and ‘C-’ antigen contents of poliovirus preparations is described. The D-antigen of poliovirus is almost certainly the component of the virus involved in eliciting neutralizing antibodies, and the D-antigen assay thus becomes a simple in vitro method for assessing the antigenic potency of poliovaccines and for investigating the influence of various factors upon the antigen. The value of the technique in following the concentration and purification of poliovirus preparations is illustrated in the studies reported in the accompanying paper by Fantes (1962). The results of experiments designed to test the validity of the idea that D-antigen is the antigen eliciting neutralizing antibodies are discussed by Beale (1961).

1. A bacteriological laboratory was provided in the medical departments of two passenger ships—a modern luxury liner cruising in the Mediterranean for 1 month and an older vessel, now out of commission, on a 3 months round voyage through the tropics. Galley hygiene in relation to gastro-enteritis was investigated on both ships.

2. 125 samples of water and iced water were examined. Samples of water chlorinated on the ships were usually satisfactory. Many port waters sampled from the barge or hosepipe were contaminated—some mildly, some profusely. A fault in an ice-making machine led to pollution of ice used for various purposes. Swimming-bath water was usually mildly polluted only; occasionally in the tropics when the baths were very popular the count rose to 1600 Esch. coli per 100 ml.

3. Thirty-five samples of milk, ice-cream and churn washings were examined. English liquid pasteurized milk stored in the cold gave satisfactory results, but the rehydrated dried milk used exclusively on the round voyage through the tropics and swabs from the apparatus used in its preparation were contaminated with Esch. coli. Ice-cream samples from ship I gave satisfactory results, and of two samples of ice-cream powder examined on ship II one was satisfactory and the other gave a poor result. Occasional churns which had been superficially cleaned but not sterilized had a high general and Esch. coli count.

4. 189 samples of food were examined, twenty-two for general and coliform counts only. Results were variable according to foodstuff and atmospheric temperature. Some cold cooked meatstuffs gave low counts and absence of coliform bacilli in 1/10 g.; others had plate counts of 300,000 to 25 million per gram with Esch. coli in 1/1000 dilution, e.g. crayfish ready for the table. Salmonellae were not found in any samples but occasionally small numbers of coagulase-positive staphylococci and non-haemolytic Cl. welchii were isolated.

5. Sixteen samples of imitation cream and washings from savoy bags were examined. Samples from freshly opened cans of cream gave satisfactory results. In whipped cream and in cream from cakes and sweets, Esch. coli was found in 1/1000 dilutions, and rinsings from savoy bags in use gave probable numbers of Esch. coli of 18,000 + / 100 ml. Counts were still high after the bags had been washed and dried and they were far from sterile even after ‘boiling’ and drying.

6. Twenty-three samples of wash water from salad vegetables and fruit had high counts of Esch. coli. After these articles had been washed in water containing 80 p.p.m. sodium hypochiorite, coliform organisms were not found in 100 ml. of water. Potassium permanganate was considered to be of doubtful value. The results of laboratory experiments to confirm the concentration of hypochiorite and time necessary to destroy Esch. coli on lettuce and watercress are shown in Tables 3 and 4.

7. Results from forty-four samples of wash and rinse waters from dish-washing machines and one bowl showed that the temperatures were often too low; with temperatures above 50° C. coliform bacilli were absent in 1 ml. quantities, but waters at temperatures below 42° C. gave probable Esch. coli counts of 1800 + / 100 ml. The bacteriological condition of the crockery and cutlery varied according to the wash and rinse waters. Rinses from metal dish covers gave poor results, with 1800+ Esch. coli per 100 ml. of rinse and occasionally Cl. welchii and coagulase-positive staphylococci. Wash and rinse waters, with temperatures of 35°ndash;42° C. used for pots and pans gave probable Esch. coli counts of 1800 + /100 ml.

8. Rinse waters from twenty cloths used in the galleys, pantries and dining rooms were nearly all heavily contaminated with Esch. coli, with probable numbers of 18,000 + /100 ml.; these counts were reduced to negligible figures by soaking the cloths in chloride of lime.

9. Swabs were taken from thirty-eight cutting-boards and twenty-six surfaces of sinks, shelves and miscellaneous articles. Many boards were heavily contaminated, the probable number of Esch. coli being 18,000 + /100 ml. of diluent; lecithinase-positive clostridia were found also. The boards could be sterilized by brushing with a stout wire brush followed by chlorination.

Other surfaces, and swabs from various knives in the kitchen, including the blade of the slicing machine, frequently gave high counts of Esch. coli.

10. 329 swabs from nostrils and fingers of saloon and galley staffs yielded one cutlure of coagulase-positive staphylococci from 111 nasal swabs and three cultures of coagulase-positive staphylococci and seven of Esch. coli from 218 finger swabs from twenty-seven men.

11. Twenty-eight faecal specimens were examined from fifty-four cases of gastro-enteritis reported on the Mediterranean cruises; salmonellae were not found, but heat-resistant Cl. welchii was isolated from six samples (21%), and the clinical picture was that of Cl. welchii food poisoning.

12. 121 faeces samples from 137 patients with gastro-enteritis were examined chiefly for salmonellae and coagulase-positive staphylococci on the outward leg of the round voyage through the tropics. Salm. typhimurium was isolated from a clinical case and Cl. welchii from one typical case of food poisoning. On the return journey all samples were examined also for Cl. welchii; of thirty-five stool samples from thirty-five patients, six (17%) yielded lecithinase-positive clostridia; salmonellae were not found.

13. Heat-resistant strains of Cl. welchii were also isolated from two chopping-boards, a knife, cold beef, brawn, two wiping cloths, a dish cover, two samples of pie and corned beef.

For a period of a year the total count, Coli-aerogenes count and turbidity were determined on samples of sewage and effluent taken every week from pilot-scale percolating filters treating domestic sewage. The eight filters contained four types of filtering medium in two sizes.

Numbers of coliform bacteria in the settled sewage were much higher during the warmer months but the total count showed no regular seasonal variation. Treatment in all the filters reduced the numbers of bacteria in both groups by over 95%. In general the reductions in numbers of bacteria and in numbers of all particles in suspension were proportional to the total area of filtering medium with which the sewage came into contact. The probability of discharging more than 10,000 coli-aerogenes per ml. ranged from 10% in the most efficient to 90 % in the least efficient filter.

Experiments in which Serratia indica was added to the influent sewage showed that only 0·4 % were recovered in the settled effluent after 24 hr. Treatment in the filters had little selective action on the different types of coliform bacteria.

This paper is published by permission of the Department of Scientific and Industrial Research.

Biological assays based on a quantitative response in particular (e.g. hormone assays) typically involve the estimation of (log) relative potency (= M) with respect to some standard preparation on the basis of one or more symmetrical determinations usually at equally spaced (log) dose intervals. It is the purpose of this paper to consider further statistical aspects of estimates of M, and especially confidence limits for M based on a combination of separate assays. These are given as examples of tests of linear hypotheses. Two applications are given on the combining of quantitative bio-assay data.

1. Conditions are described under which the three types of live or formalin-inactivated poliomyelitis viruses can be concentrated 10–100 times in good yield, with elimination of approximately 90% of the extraneous protein normally present in poliovaccines.

2. It was shown that 0·1–0·2 mg./ml. AlPO4 almost quantitatively adsorbed the antigens, provided the adsorptions were carried out at a pH round about 5. Re-elution of the antigens occurred readily at pH 7–pH 8.

3. No noticeable conversion of the ‘D-antigen’ to the ‘C-antigen’ occurred during the concentration.

The environment of two contiguous surgical wards was examined over a period of twelve months by means of a slit sampler, settle plates and blanket sweep plates. At the same time, nasal swabs were taken each week from the patients and all cases of sepsis examined bacteriologically.

Phage typing of more than 3600 isolations of Staph. aureus showed that there was one predominant strain in the air, bedding, patients' noses and infected wounds.

There was no relationship between the total number of bacteria in the ward air and the numbers of Staph. aureus.

The recovery of large numbers of Staph. aureus from the air at certain periods was associated with a high contamination rate in the blankets and with an increased incidence of staphylococcal sepsis.

Not all nasal carriers of Staph. aureus contaminated their bedding. There was evidence that some patients became nasal carriers of strains of staphylococci previously isolated from their bedding.

Some evidence was obtained that blankets may play a role in the transmission of staphylococci from patient to patient.

This work was supported by a grant from the Australian National Health and Medical Research Council. Our thanks are due to Mrs Elisabeth Bradshaw for her technical assistance and to Prof. John Loewenthal for his interest and for permission to study his wards.

A planned experiment to gauge the effect of regular blanket disinfection on bacterial, especially staphylococcal, contamination and infection in an acute geriatric unit is described.

With formalinization of blankets, the total colony counts of sweep-plates of blankets in use fell to 43 and 34% of the control period counts on two test wards respectively. The staphylococcal counts fell to 20 and 25% of their previous levels on the two test wards. Ordinary laundering with soap and water did not significantly alter the degree of bacterial contamination of the blankets.

Colony counts of slit-sampler plates taken during bed-making fell, with formalinization, to about one-third and one-half of the control period counts on the two test wards respectively. Staphylococcal colony counts fell to about one-fifth and one-third of their previous levels.

Staphylococcal contamination of patients was measured by the nasal carrier rate, there being very little sepsis on the unit. The gross figures for nasal carriage of Staph. aureus showed little variation attributable to blanket disinfection. The number of patients acquiring a new strain of staphylococcus after admission tended to increase as the investigation continued. This trend was reversed on both test wards coincident with the introduction of blanket disinfection, and it would seem that this measure may have the effect of lessening the chances of the hospitalized patient becoming a nasal carrier of Staph. aureus, but the results are not formally statistically significant.

There appears to be a lack of parallelism between the degree of environmental contamination and cross-infection rates. It is considered unlikely that contaminated blankets may cause infection except in two special circumstances: (a) in the wards, by direct contact with the patient's skin or in-dwelling catheter, and (b) in the operating theatre suite, either directly by contact with the patient's skin, or indirectly via the air.

I am grateful to Professor R. E. O. Williams for the loan of a slit-sampler; to Dr Martin Hynes for advice and encouragement; to Mr M. P. Curwen for the analysis of statistical significance; to Mr F. Seage for much technical help; to Dr J. Sharkey, geriatrician; to Sister J. Cleary and the nurses of the geriatric unit; and to Mrs H. M. Page and Miss M. Hanworth for secretarial assistance.

The work was supported by a grant from the free funds of the Northern Group Hospital Management Committee.