Relevance of using healthcare reimbursement data to monitor syphilis epidemic in France

Abstract

Introduction In France, surveillance of early syphilis (primary, secondary and early latent) relies on the clinician-based ResIST sentinel network. Although ResIST enables the monitoring of trends, an insight of the complete picture of the syphilis epidemic is not possible. More specifically, cases reported by this network are mostly diagnosed in free STI clinics and hospitals (respectively, 75% and 24%). This study aims to estimate the number and rate of diagnoses made outside these health facilities by exploring health insurance data. Methods An algorithm combining healthcare reimbursements for specific diagnostic tests and recommended treatment was fitted to identify syphilis cases. Sensitivity analyses were used to validate the algorithm. Age and gender standardized diagnosis rates were estimated using census data. The study period was restricted to three years, from January 2011 to December 2013, before the stock depletion of benzathine penicillin G (BPG) in 2014 and 2015. Results Between 2011 and 2013, 12,644 (7.5 cases per 100,000 inhabitants) reimbursements were made for syphilis-related diagnoses. The annual number of cases increased by 22% from 2011 (n = 3771, rate = 6.7/100,000) to 2013 (n = 4589, rate = 8.2/100,000). The rate of syphilis diagnosis increased in men from 12.9/100 000 to 16.0/100,000, while it remained steady in women at approximately 1.8/100,000. The disease burden was greatest in French overseas territories (18.1/100,000) and in the Paris area (11.7 cases/100,000). Conclusion Despite the lack of data on the number of confirmed diagnoses and information about sexual behaviour, these findings demonstrate the relevance of analysing insurance data to help monitor the syphilis epidemic in patients who visit general practitioners and non-hospital-based specialists. With BPG availability, annual rate could be estimated again to carry on epidemiological dynamic analysis and guide prevention actions.