AIM: to develop and validate hemorrhoidal disease (HD) specific national questionnaire.MATERIALS AND METHODS: the Russian questionnaire, which included questions on the most typical HD symptoms, was developed by leading national specialists-proctologists. The regression analysis was used to evaluate linear and nonlinear relationships between the fact of the HD presence, objectively confirmed by a doctor, and the patients’ answers to the questions from the questionnaire. The final version of the questionnaire included only questions regarding clinical characteristics typical for hemorrhoids that were significantly associated with the presence of the disease. Validation of the questionnaire was performed on both training and test samples of patients to determine its sensitivity (Se) and specificity (Sp) in identifying patients with HD.RESULTS: initially, 7 main domains reflecting the presence and severity of symptoms typical for HD were included in the questionnaire. A training sample of patients was used to search for linear and nonlinear relationships between objectively confirmed hemorrhoids and responses to questions in the initial questionnaire. The final version of the questionnaire, consisting only of questions regarding symptoms that have demonstrated to be significant predictors of hemorrhoids, was validated on a training sample (Se = 86.5%, Sp = 73.4%, and the probability of identifying a patient with HD, according to Bayes’ theorem = 66.6%). After the developed version of the questionnaire was validated on a test sample, the final screening questionnaire for HD demonstrated Se 83.7%, Sp 74.1%, and the probability of identifying patients with HD with a positive questionnaire result, calculated according to Bayes’ theorem, was 67%.CONCLUSION: the first national questionnaire for screening HD has diagnostic characteristics indicating its high accuracy in identifying patients with HD and allowing it to be recommended for use in the general population.
Read full abstract