OBJECTIVE: To perform a sensitivity analysis using a decision model for simulating opportunistic screening for oral cancer and precancer in order to obtain an estimate of health gain from screening high risk dental patients, identified with the aid of artificial intelligence. DESIGN: A hypothetical opportunistic screening programme based on preselection of high risk patients was compared with screening all eligible patients and with a 'do nothing' scenario. The basic assumptions and data used in the model have been described previously. SETTING: Screening would be carried out in dental practices. PARTICIPANTS: A notional population of 100,000 adults of average age 55 years and 20 years life expectancy. INTERVENTIONS: Systematic clinical examination of the oral mucosa of high risk individuals identified by artificial intelligence on the basis of personal characteristics and life style. OUTCOME MEASURES: QALYs and equivalent lives saved RESULTS: The following estimates of health gain were obtained: the preselected screened group (n = 25,000) would achieve 1,993,294 QALYs. Without preselection, screening an unselected population of 100,000 eligible adults would result in 1,993,094 QALYs. Under the 'do nothing' scenario, 100,000 unscreened individuals would achieve 1,992,982 QALYs. Selective screening could avoid the equivalent of 15 deaths per 100,000 subjects examined. Screening preselected subjects at heightened risk would save the equivalent of two to three times the number of healthy lives compared with non-selective screening. Only one quarter of the population would need to be examined. CONCLUSIONS: Opportunistic screening of preselected dental patients for oral cancer and precancer appears to be a promising health promotion strategy and should be subject to formal economic appraisal.
OBJECTIVE: To perform a sensitivity analysis using a decision model for simulating opportunistic screening for oral cancer and precancer in order to obtain an estimate of health gain from screening high risk dental patients, identified with the aid of artificial intelligence. DESIGN: A hypothetical opportunistic screening programme based on preselection of high risk patients was compared with screening all eligible patients and with a 'do nothing' scenario. The basic assumptions and data used in the model have been described previously. SETTING: Screening would be carried out in dental practices. PARTICIPANTS: A notional population of 100,000 adults of average age 55 years and 20 years life expectancy. INTERVENTIONS: Systematic clinical examination of the oral mucosa of high risk individuals identified by artificial intelligence on the basis of personal characteristics and life style. OUTCOME MEASURES: QALYs and equivalent lives saved RESULTS: The following estimates of health gain were obtained: the preselected screened group (n = 25,000) would achieve 1,993,294 QALYs. Without preselection, screening an unselected population of 100,000 eligible adults would result in 1,993,094 QALYs. Under the 'do nothing' scenario, 100,000 unscreened individuals would achieve 1,992,982 QALYs. Selective screening could avoid the equivalent of 15 deaths per 100,000 subjects examined. Screening preselected subjects at heightened risk would save the equivalent of two to three times the number of healthy lives compared with non-selective screening. Only one quarter of the population would need to be examined. CONCLUSIONS: Opportunistic screening of preselected dental patients for oral cancer and precancer appears to be a promising health promotion strategy and should be subject to formal economic appraisal.