Continuity of care is a central feature of primary care. It represents the idea of a long lasting relationship between the patient and his doctor. It has been associated with a positive impact on the quality of care and on health outcomes such as improving patient satisfaction with health care or fewer hospitalizations and emergency room visits. However, continuity has been somewhat neglected in the French primary care reforms in contrast to issues such as access to care. Furthermore, it has been suggested that the efforts to improve access could have reduced continuity. Thus, the aim of the thesis is to evaluate continuity of care in France and assess its recent evolution and its impact on health. The final objective is to validate a measure of continuity that could be used as a management tool for Regional Agencies of Health.This work is based on data from a 1/97e, representative of the general population, sample of the national health insurance database that contains every health care reimbursements made to French beneficiaries. Continuity was assessed using the continuity index of Bice and Boxerman that assesses the concentration of medical consultations with the same general practitioner. This measure is an indirect reflection longitudinality. The measure was used to assess the level of continuity in France and assess the association between continuity and all-cause mortality using a Cox model and introducing the measure of continuity as a time-dependent variable. Furthermore, the measure was also used to assess the role of individual and contextual factors, including provision of care, on the level of continuity using a multilevel model. Finally, the metrological characteristics of the measure were tested to validate the measurement as an indicator based on the Agency for Health Research and Quality.The analyses showed that continuity of care in France has remained constant in recent years at a relatively high level. However, interregional variation with regions with better continuity of care and others with lower continuity were seen. Lower continuity of care values were associated with a four percent increase in likelihood of death in the general population. Continuity levels were also associated with individual factors, such as sex, age and social status of the beneficiaries, as well as factors at the physician level. Provision of care factors had a moderate role in the variation of continuity of care levels. However, the results confirm the negative association between continuity and access. Finally, the measure was validated as a quality indicator.In conclusion, continuity of care is associated with the level of health of the population. The evaluation of the continuity in the French context shows a relatively high level in France, no doubt due to the attachment in France to the general practitioner model. However, it also revealed interregional variations that suggest that there is room for improvement and thus an opportunity to improve the level of health of the population. The use of tools such as a continuity indicator could allow regulatory agencies to monitor continuity and measure the impact of implementing actions, such as care pathways. The elaboration of the national healthcare strategies is a tremendous opportunity to include continuity of care in primary care organization planning.
