Few studies have assessed risk factors for sporadic cryptosporidiosis in industrialized countries, even though it may be numerically more common than outbreaks of disease. We carried out case-control studies assessing risk factors for sporadic disease in Melbourne and Adelaide, which have water supplies from different ends of the raw water spectrum. In addition to examining drinking water, we assessed several other exposures. 201 cases and 795 controls were recruited for Melbourne and 134 cases and 536 controls were recruited for Adelaide. Risk factors were similar for the two cities, with swimming in public pools and contact with a person with diarrhoea being most important. The consumption of plain tap water was not found to be associated with disease. This study emphasizes the need for regular public health messages to the public and swimming pool managers in an attempt to prevent sporadic cryptosporidiosis, as well as outbreaks of disease.

C ontaminated drinking water may expose large populations to the risk of disease, as shown by the Milwaukee incident, where approximately 400,000 people were affected.1 Epidemiological investigations can identify the causes of such outbreaks and reduce their impact. 2In the investigation of sporadic waterborne disease, the validity and reliability of drinking water intake is important to establish disease causation. 3,4 We only found two studies that addressed this issue. A British study examined fluid intake, comparing a questionnaire to a diary, in women with urinary tract infections. 5The questionnaire mean was 2.0 litres/day and the diary mean 1.9 litres/day. A United States study examined reports of home fluid intake, comparing a questionnaire to a diary, by pregnant women and their partners. The questionnaire mean was 0.75 litres/day and the diary mean 0.40 litres/day. Pearson's correlation coefficient was 0.78.Our aim was to determine the validity and reliability of telephone questionnaire drink-

Telephone interviewing is becoming increasingly common with both research and commercial organisations. The validity of research using telephone recruitment is dependent on high participation rates. Following low participation rates of controls in a pilot case-control study using telephone recruitment, we sought to improve the participation rate for the main study. We did this by sending an introductory letter to prospective controls one to two weeks before making any telephone contact. We report the effect of this letter on participation rates.We undertook a pilot case-control study of risk factors for cryptosporidiosis in Melbourne over May and June 1998. The main Melbourne case-control study commenced immediately after this. Cases were recruited by telephone, using persons notified to the Victorian Department of Human Services with laboratory-confirmed cryptosporidiosis. Population-based controls were recruited by telephone, using an electronic White Pages residential telephone directory, 1 which generated random telephone numbers and corresponding addresses. Controls were either interviewed at the time if there was an appropriate age-and sex-matched case, or asked to go on to a 'bank' for subsequent interviewing. A person agreeing to either the interview or the 'bank', was considered to be a consenting participant. To recruit controls for the pilot study we made telephone contact without an introductory letter. For the main study we sent an introductory letter to prospective households one to two weeks prior to their first telephone call. The introductory letter contained information about the general nature of our study, time commitments of participants and names of the organisations conducting the research. No factors, other than the introductory letter, were different between the pilot and main case-control study. We compared the participation rate of controls from the pilot study with the participation rate from the first three months of the main study.For the pilot study, the participation rate of controls was 48% (156/327, 95% CI 42-53). For the main study, the participation rate of controls was 77% (230/298, 95% CI 72-82). The difference in participation rates of controls between the pilot and main study was a highly statistically significant increase of 29% (95% CI 22-37, p<0.001). The participation rate for cases was 98% (62/ 63, 95% CI 92-100) for the pilot study and 100% (12/12, 95% CI 74-100.0) for the main study.The participation rate of controls increased considerably with the use of an introductory letter sent prior to the initial telephone call. It was a relatively simple and low-cost process, with batches of 100 letters sent out one to two weeks prior to the anticipated telephone call being made. Low participation rates threaten the validity of comparisons between case and control groups. 2 Feedback from our interviewers suggested that the high level of refusal in the pilot study was due to 'cold calling' households without any advance warning. The effect of 'cold calling' and the usefulness...

No abstract