[Validation of identification criteria for patients having surgery for rhegmatogenous retinal detachment at Toulouse University Hospital]

Introduction: Seasonal variations seem to modify the incidence of rhegmatogenous retinal detachment (RRD), with a summer peak, although the meteorological parameters involved have not been studied in France. In order to conduct a national study evaluating the association between the occurrence of RRD and various climatological variables (METEO-POC study), we need to establish a national cohort of patients having undergone surgery for RRD. The data of the National Health Data System (SNDS) allow the realization of epidemiological studies concerning various pathologies. However, as these databases were initially intended for medical administrative use, the identification of the pathologies coded in these databases must be validated before any use for research purposes. In order to carry out a cohort study using SNDS data, the objective of this study is to validate the identification criteria for patients having undergone RRD surgery at Toulouse University Hospital.

Methods: We compared the cohort of RRD surgery patients at Toulouse University Hospital from January to December 2017 assembled from SNDS data with the cohort of patients meeting the same selection criteria assembled from Softalmo software data.

Results: With a positive predictive value of 82.0%, a sensitivity of 83.8%, a specificity of 69.9%, and a negative predictive value of 72.5%, we can consider that our eligibility criteria are performing well.

Conclusion: Since the selection of patients via SNDS data is reliable at Toulouse University Hospital, we can use it at the national level for the METEO-POC study.