William Norman Pickles (1885–1969) was, for most of his life, a general practitioner in the village of Aysgarth, Wensleydale in the North of England, a relatively remote area in the early part of this century with many of the doctors' visits made on horseback or on an ancient motorcycle. The story of Pickles' life, how he became, among other honours, the first President of the Royal College of General Practitioners, is told in Pemberton's book Will Pickles of Wensleydale. It may seem curious that Pickles, a country GP, should rank with other great men in epidemiology: Jenner who worked on smallpox, Budd on typhoid fever, Snow on cholera, yet it was the remoteness of his practice combined with his own acute observation and accurate recording that enabled him to contribute so significantly to epidemiology.

Epidemiology in Country Practice by William N. Pickles, published in 1939, has been a source of continuing interest and challenge especially to general practitioners (Watson, 1982; Booth, 1987). Pickles worked for over 50 years as a general practitioner (GP) in rural Wensleydale where there were many isolated villages in which natural immunity against various infections was often lacking. And so the source of infection could usually be traced, and, with little or no immunity, spread was often rapid.

Many infections, especially viral infections, show a certain periodicity which can be wholly or partly predictable. It is important to be able to recognize these patterns because anticipating changes in incidence should lead to better diagnosis, treatment and control of infection. In this review patterns of some viral and other infections in England and Wales which have been observed from laboratory reports over a period of 21 years, from 1967 to 1987, are described.

The clinical application of the principles of epidemiological research has become fashionable. Three recent textbooks carry the words ‘Clinical Epidemiology’ in their titles (Feinstein, 1985; Sackett, Haynes & Tugwell, 1985; Weiss, 1986), and a venerable journal has recently changed its name to this effect (Feinstein & Spitzer, 1988). Deans of medical schools, Government Health Officials and editorialists pay tribute to the new science which is expected to steer the medical profession of the third millennium.

There is a significant role for the practice of epidemiology throughout the world if health professionals are to work effectively towards ‘Health for All’. The status quo leaves a great deal to be desired, as evidenced by recent signals from the premier international health agency, the World Health Organization (WHO). The forty-first World Health Assembly meeting in Geneva, Switzerland, in May 1988 approved a resolution stressing ‘the importance of epidemiology as a tool for the formulation of rational health policy’ (PAHO, 1988). In addition to its key role ‘in studying the causes and means of prevention of disease’, the Assembly noted epidemiology's valuable inputs in ‘health systems research, information support, technology assessment, and the management and evaluation of health service’.

To investigate cross-colonization with and persistence of Pseudomonas aeruginosa in cystic fibrosis (CF). 181 isolates from 76 CF patients were typed using a P. aeruginosa -specific DNA probe. Whereas sibling pairs predominantly harboured genotypically identical P. aeruginosa strains, all of the other patients harboured different strains. Seventy-nine per cent (22/31) of the infected CF patients harboured the same strains at the beginning and the end of a summer camp. A change of strains was seen in 10% (3/31) of the patients at the end of the camp. Forty-six per cent (6/13) of the patients who were apparently initially uninfected, acquired P. aeruginosa by the end of the period. Genotyping proved that strain change or acquisition was due to cross-colonization in four of nine cases. Very little P. aeruginosa was isolated from the inanimate environment. Persistence of P. aeruginosa after a temporary loss due to antibiotic therapy was seen in 12/16 paired patient strains before and after antibiotic therapy. Thus, suppression followed a flare-up seemed to occur in these patients rather than eradication and a new infection. When 35 patients were followed over a period of 6 months, 7 (20%) changed the strain in their sputum. Only one of 43 patients harboured two different P. aeruginosa strains simultaneously over a long period.

Campylobacter jejuni/coli strains from 164 chickens were serotyped by the methods previously described by Penner et al. and Lior et al. The chickens were sampled during breeding from hatching to the age of 42 weeks. The birds were housed, in two separate groups, under different environmental conditions, (for comparison of the effect of hygienic precautions on the transmission of the bacteria during breeding). In the group where the hygienic conditions could be controlled to a greater extent, the chickens became colonized later in the breeding chain and with only one single campylobacter strains.

The haemagglutinating ability of 141 isolates of Aeromonas spp. for human. horse and guinea-pig erythrocytes was examined. Although the majority of isolates (136/141) agglutinated human group O erythrocytes, all the eight possible patterns of agglutination were observed. Haemagglutination of human group O erythrocytes. but not horse or guinea-pig, was associated with the ability to agglutinate yeast cells (Saccharomyces) and with aggregation in a low concentration of ammonium sulphate. Haemagglutinating ability was further characterized by reactions in the presence of mannose, galactose or fucose. All the possible patterns of inhibition with individual sugars were observed, but haemagglutination of human group O erythrocytes not inhibited by mannose, galactose or fucose was more common among isolates from patients with diarrhoea, and isolates producing a Vero cell cytotoxin than would be expected by chance. This might represent a virulence factor.

Sinks in a new long-stay hospital (LSH) were cultured weekly during 4 consecutive months to evaluate the microbial profile before and after occupancy of the hospital. From the elderly patients admitted to the patient care rooms oral and rectal specimens were collected to examine the contribution of the patients' flora to the sink contamination. Isolates were typed biochemically, serologically and by susceptibility pattern. Before occupancy Gram-negative bacilli were not isolated. Once the elderly patients, who were highly colonized on admission, occupied their rooms identical strains gradually contaminated the sinks. Escherichia coli, Klebsiella, Pseudomonas and Acinetobacter species were the major correlating strains. The mean concentration of the correlating isolates was higher in throat and intestines compared to the mean concentration of the non- correlating strains. These strains seem to have a greater chance to be shed and then transferred via the hands of personnel to sinks. This resport shows that the major route of environmental contamination is from patient carriers to sinks, and not the reverse way.

Chickens which had been inoculated orally with a nalidixic acid-resistant strain of Salmonella typhimurium were reared on a diet containing different concentrations of avoparcin in an attempt to explain the variation in response to commercial levels of this antibiotic oberved by different workers. In one experiment small increases in faecal excretion of the inoculated salmonella occurred at 2·5 and 5·0 mg kg−1, with greater increases between 7·5 and 20 mg kg−1 In a second experiment there was a significant increase in excretion at 7'5 mg kg−1 and in a third experiment in which generally higher excretion rates were detected in all groups, significant increases were observed at 10 and 12·5 mg kg−1 only. In addition avoparcin significantly increased the faecal excretion of S. cholerae-suis, S. dublin and S. arizonae, serotypes not usually associated with poultry-derived food-poisoning in the United Kingdom. It did not increase faecal excretion of S. pullorum. Avoparcin at 10 mg kg−1 appeared to have little effect on the normal intestinal flora of S. typhimurium infected chickens despite the fact that when tested in vitro individual organisms were susceptible to this drug concentration. At 100 mg kg 1 viable counts of intestinal enterococci and Gramnegative anaerobic bacteria were considerably reduced while those of S. typhimurium and Escherichia coli increased. Antibiotic activity due to avoparcin was detectable in the alimentary tract and there was some increase in concentration of the antibiotic in the more distal regions.

A total of 1012 milk filters collected from 498 dairy farms in south-western Ontario during three study periods (December 1985–March 1986) were tested for the presence of verocytotoxin-producing Escherichia coli (VTEC). VTEC were detected and isolated using a Vero cell assay. Supernatants from 20 of the milk filter cultures had verocytotoxic activity and 7 VTEC strains were isolated from these positive samples. The prevalence of VTEC in the samples in each of the three study periods were 0·44, 0·65 and 0·99% respectively. All seven VTEC strains isolated were sensitive to commonly used antimicrobial agents. The serotypes of these strains were O 26.H11, O43.H2, O 153.H25, O ?. H8, O ?. H19, O ?.non-motile, and Orough. H19. Two of these serotypes (O 153. H25 and O 26.H11) have previously been associated with disease in humans.

The spread of two strains of Staphylococcus aureus with high level resistance to mupirocin is described. The resistance proved to be easily transferred to other S. aureus strains by filter mating experiments and on the skin of mice. No plasmid band corresponding to the resistance could be demonstrated by agarose gel electrophoresis or by caesium chloride gradient centrifugation but cleavage of ‘chromosomal’ DNA from resistant recipients showed bright bands of DNA absent from sensitive controls.

The degree of protection conferred by vaccinated dams on infant mice against colonization by Campylobacter jejuni depended on the bacterial strain, preparation, and route of administration of the vaccine. In some instances of homologous protection, serum bactericidal titres correlated well with protection. However, boiled C. jejuni vaccine, which was non-protective, also elicited a strong bactericidal antibody response. Conversely, bactericidal activity could not be demonstrated against strains capable of cross-protection. There was a good correlation between high campylobacter-specific IgG response and bactericidal activity.

Six hundred and thirteen fresh diarrhoeal faeces were inoculated on Skirrow blood agar (SK), on Preston blood free agar (PBF), and in Campy-thioglycolate broth (CT). After 24 h of storage at 4 °C, specimens were again inoculated on SK and PBF, and in Campylobacter enrichment broth (CEB). CT tubes were placed overnight at 4 °C. Plates and CEB tubes were incubated at 43 °C in microaerophilic conditions. A total of 68 specimens was positive for campylobacter on direct plating. Sixty-four of them were also recovered after subculturing from CT, and only 51 from CEB. Delayed inoculation of plates after storage of samples at 4 °C yielded 57 isolates. The storage of faeces at 4 °C for 24 h significantly reduces the number of campylobacter isolates. When samples are not plated immediately we recommend inoculating a CT tube maintained at 4 °C overnight as a holding medium.

The seroepidemiology of chlamydial infections in the Finnish population was studied by analysing the prevalence of chlamydial complement fixing (CF) antibodies in patients sera sent for virus serological screening tests over 17 years from 1971 to 1987. The total number of sera studied was over 160 000. In the early 1970s, the prevalence of chlamydial CF antibodies (CF titres ≥8) was low (less than 2%), but later the proportion of seropositive cases rose, and in 1976, 18% of sera contained antibodies. In 1984, the seropositivity rate was over 31%. The prevalence of high chlamydial CF titres (titres ≥64) also showed annual variation. In general, under 1% of sera contained chlamydial CF antibodies in high titre, but in 1979 and 1984, distinct peaks occurred when 1·3% and 1·4% of sera, respectively, had titres ≥64 The age-related antibody positivity rate showed a decline during early infancy, an increase in childhood and adolescence, and a stable level in adulthood when approximately 20% of the sera contained antibodies. The chlamydial antigen used in this survey was genus-specific, i.e. it detects antibodies against, all chlamydial species. Epidemiological data support the hypothesis that infections due to a novel chlamydial species, TWAR chlamydia, are the most likely explanation for the relatively frequent occurrence of chlamydial CF antibodies and for the variation in CF antibody prevalence.

Three enzyme immunoassays were used for the serodiagnosis of Trypanosoma evansi in camels in the Sudan in order to evaluate their ability to discriminate between infected and non-infected animals. Two assays were used for the detection of trypanosomal antibodies, one using specific anti-camel IgG conjugate and another using a non-specific Protein A conjugate. The third assay detected the presence of trypanosomal antigens using anti-T. evansi antibodies in a double antibody sandwich assay. Inspection of the frequency distribution of assay results suggested that the ELISA for circulating trypanosomal antibodies using specific antisera and the ELISA for circulating antigens can distinguish between non-infected camels and infected camels exhibiting patent infections or not. The ELISA using Protein A conjugate to bind non-specifically to camel immunoglobulin did not appear to discriminate between infected and non-infected animals.

The food and water hygiene in two Liberian communities was studied in a house-to-house diarrhoea survey. The level of contamination with enterobacteria of drinking water stored in the households was significantly higher than at the water sources. Food hygiene standards were low, particularly in the urban slum where storage of cooked food for long periods led to bacterial multiplication at high levels. Infant foods were particularly heavily contaminated. It is concluded that when water supply programmes are planned, the presence of other risk factors for water-related diseases should be investigated. To ensure maximum health benefits, water projects should as a rule be accompanied by other interventions.

The finding that there had been multiple isolations of Palyam serogroup orbiviruses from aborted cattle fetuses in neighbouring Zimbabwe, suggested that there was a need to investigate the possible occurrence of the viruses in South Africa. Unidentified viruses isolated in South Africa, which had been in storage, were examined. Four viruses which had been isolated from Culicoides midges collected at various sites in the years from 1969 to 1977, were identified as three strains of Gweru virus and one of the Nyabira virus (Palyam group serotypes originally described from Zimbabwe) A fifth virus, isolated in 1967 from the blood of a cow with mild fever, was found to be a distinct new member of the Vellore antigenic complex of the Palyam serogroup and was named Apies River virus. Sera from 476 cattle, 150 sheep, 24 goats and 78 humans from 10 farms were tested for neutralizing antibodies to the above three serotypes of virus plus Abadina and Marondera serotypes. Only 1 of 100 cattle sera from two farms in the southern coastal area had antibody, but elsewhere there was a high prevalence of antibody with 254 (53%) of all cattle exhibiting activity for one or more of the five serotypes of virus tested. Only 6 (4%) sheep, 3 (12–5%) goats and 11 (14%) humans had antibody.

A study of influenza in residential schools provided the opportunity to assess the significance of antibody as a predictor of immunity. Five hundred and fifty-six pupils from 8 schools were included in the investigations, and the outcome for these children in 27 naturally occurring outbreaks of influenza was analysed. The outbreaks comprised 5 caused by strains of influenza A H3N2, 10 caused by strains of influenza A H1N1, and 12 caused by strains of influenza B. On 8 occasions a second outbreak of the same serotype occurred in a school.

There was a general correlation between the presence of antibody to the outbreak strain and protection from infection. For each of the three influenza virus serotypes the infection rate in those with no detectable antibody was approximately 80%. Those with past experience of the virus but no antibody to the outbreak strain experienced lower infection rates (62% overall) but the infection rates were lowest in those with intermediate and high level antibody to the challenge strain (18% overall).

Vaccine was used by three of the schools. The effect of antibody derived from recent experience, either natural or vaccine-induced, on subsequent challenge with a drifted strain i.e. one showing antigenic drift away from the previous strain, was compared. Intermediate or high level antibody to the challenge strain in those who had experienced a recent natural infection was associated with a low infection rate (9%). A similar level of antibody produced in response to vaccination was associated with a significantly higher infection rate (23%:P 0-025). Among the vaccinees who had produced such antibody the infection rate was highest (32%) in those who had responded to vaccine in the presence of antibody to the vaccine strain.

The evidence from this study indicates that whilst antibody surveys of populations may provide some information about susceptibility to challenge with new strains of influenza viruses, the cirucumstances of the induction of the antibody affect its value as a predictor of immunity.

Isolates of capripoxvirus collected from sub-Saharan Africa were compared in sheep, goats and cattle and by restriction endonuclease digestion of their purified DNA. Biochemical techniques were used to precisely identify strains of capripoxvirus for epidemiological investigations. Strains of capripoxvirus infecting cattle have remained very stable over a 30-year period and are closely related to strains recovered from sheep in Africa.