We discuss the use of the trichotomous logistic model to discriminate between patients with gastrointestinal (GI) cancer, patients with benign GI disease and 'normal' subjects, using symptoms and the concentrations of some serum proteins that are potentially indicative of malignancy as covariates. A parsimonious model can be obtained by invoking an indistinguishability hypothesis which is appropriate when a covariate is considered to have no predictive value between categories. It is shown that the polychotomous model can be re-parameterised under the null hypothesis to give a 'reduced form', which can be fitted by maximum likelihood. The validity of the use of the same methods for retrospective sampling is discussed. The approach is illustrated by the development of a logistic model to identify symptomatic and asymptomatic subjects with a high risk of GI cancer.

Self administered questionnaires are becoming popular investigative tools in medical research, yet few reports state the extent of methods used to validate these questionnaires before their general use. A pilot study was therefore carried out to validate a 41 item questionnaire for use in a population screening study for gastrointestinal disease. Participants in the study comprised 69 population controls, 40 patients with benign disease, and 35 patients with gastrointestinal cancer. Acceptability, ease of completion, reliability, and reproducibility of the questionnaire were all assessed. Only one subject refused to complete the questionnaire. Ninety six per cent of the questions were completed by each subject and only one response in 1440 was altered in the reproducibility study. The questionnaire disclosed symptoms similar to those elicited by a clinician and highlighted unreported gastrointestinal symptoms in the control group. Three questions were found to be unreliable and were altered before the questionnaire was put into general use. It is concluded that a pilot study to validate a new questionnaire is simple to perform and necessary to identify unreliable questions.

No abstract

No abstract

No abstract