Influence of epidemiological methods on the estimates of respiratory disease prevalence in Italy

<bold>Background:</bold> A large variability is often observed in respiratory disease prevalence, estimated by surveys using different methods of disease assessment. <bold>Aim:</bold> To evaluate the influence of the methods used to detect the disease on the estimates of respiratory disease prevalence in the Italian adult/elderly population. <bold>Methods:</bold> Epidemiological studies dealing with the prevalence of asthma, allergic rhinitis and COPD in the Italian population during the last two decades were retrieved and prevalence estimates were related to the method of disease assessment: questionnaire postal or phone surveys, surveys combined with spirometry, primary care records, and Health Care Utilization (HCU) databases. <bold>Results:</bold> The estimates of asthma prevalence were rather consistent, ranging from 6.1% to 8.5% in postal surveys, primary care database, and administrative databases. The only exception was a phone national survey, which reported a very low prevalence (3.4%). The prevalence of allergic rhinitis was much lower in primary care records (4.5%) compared to questionnaire surveys (25.8%-37.4%). However, differences nearly disappeared when considering only drug-treated cases. Prevalence estimates for COPD were also strongly affected by the method of assessment. Prevalence estimates were the lowest in primary care records and in phone national survey (around 3%), increased to 6.8% in HCU databases, and to about 10% in questionnaire surveys enquiring about doctor-diagnosed COPD or in surveys using spirometry. <bold>Conclusion:</bold> While prevalence estimates for asthma are rather consistent in the Italian adult/elderly population, the estimates of allergic rhinitis and COPD are strongly affected by the method of disease assessment.