INTRODUCTION

Seasonal influenza vaccination (SIV) is recommended for people with diabetes, but its effectiveness has not been demonstrated. All of the available studies are observational and marred with the healthy vaccine bias, that is, bias resulting from the generally better health behaviours practised by people who choose to be vaccinated against influenza, compared with those who do not. This protocol is intended to study the effectiveness of SIV in people with treated diabetes and simultaneously to control for bias.

METHODS AND ANALYSES

This case-control study is nested in a historical cohort and is designed to study vaccine effectiveness (VE) assessed by morbidity, mortality and anti-infective drug use. The cohort will comprise a representative sample of health insurance beneficiaries in France and will cover 10 consecutive epidemic seasons. It will include all patients reimbursed three separate times for drugs to treat diabetes. The first study of VE will use reasons for hospitalisation as the primary end point, and the second with the use of neuraminidase inhibitors and of antibiotics as the end points. A case will be defined as any person in the cohort reaching any end point at a given date. The case patient will be matched with the largest possible number of controls (individuals not reaching the end point by this date) according to the propensity score method with an optimal calliper width. A conditional logistic model will be used to estimate ORs to take into account both the matching and the repetition of measurements. The model will be applied separately during and outside of epidemic periods to estimate the residual confounding.

ETHICS AND DISSEMINATION

The study has been approved by the French Commission on Individual Data Protection and Public Liberties (ref: AT/CPZ/SVT/JB/DP/CR05222O). The study's findings will be published in peer-reviewed journals and disseminated at international conferences and through social media.