PND44 EPIDEMIOLOGY AND DETERMINANTS OF CHRONIC MIGRAINE: A REAL-WORLD STUDY IN PRIMARY CARE IN ITALY

Abstract

Chronic Migraine (CM) is a relevant public health concern. However, there are several clinical and sociodemographic features which make difficult the diagnosis of CM. Given the “gatekeeper” role of GPs in the public health system in Italy, the epidemiology of CM in primary care is pivotal. This study therefore aims to assess the epidemiology of CM in primary care in Italy by testing five operational algorithms defining CM. These algorithms adopted different combinations (i.e. increasing or decreasing completeness and correctness) between migraine-related contacts coded via ICD9CM and medications use with related durations among migraine sufferers. Using the Health Search Database we performed population-based cohort study. A set of algorithms defining CM were developed to estimate the prevalence and incidence rate of CM in a cohort of patients active in Health Search Database (HSD) on January 1, 2016. In addition, we conduct a nested case-control analysis in an attempt to quantify the level of association between candidate determinants of CM according to each algorithm. The prevalence rate increased from the first to the fifth algorithm ranging 0.03% - 0.28%. There was no 95% Confidence Intervals (CI) which overlapped the others, and every CI reliably maintained 2% precision. Incidence rates showed a growing trend (i.e., 0.008-0.056 per 100000 person-years) as well. Overall, all algorithms were able to capture sex (females show a greater risk than males), and NSAIDs overuse as statistically significantly determinants of incident cases of CM. The presence of depression was associated with a statistically significant increase of incidence rate of CM only for two algorithms. Our findings show that prevalence and incidence rate of CM are underestimated in HSD when compared with current literature. On the other hand, we found acceptable correctness of CM definition given the consistency across the defining algorithms in capturing significant associations with well-known determinants.