Olivier Saint-Lary1, Jonathan Sicsic2. 1. Collège National des Généralistes Enseignants (CNGE), University Versailles Saint-Quentin en Yvelines, Department of Family Medicine, 78180 Montigny le Bretonneux, France. 2. CERMES3, UMR8211, INSERM U988, Site CNRS, 7, rue Guy Môquet, 94801 Villejuif Cedex, France. Electronic address: jonathan.sicsic@inserm.fr.
Abstract
BACKGROUND: In 2009, a voluntary-based pay for performance scheme targeting general practitioners (GPs) was introduced in France through the 'Contract for Improving Individual Practices' (CAPI). OBJECTIVE: To study the impact of the CAPI on French GPs' consultation length. METHODS: Univariate analysis, and multilevel regression analyses were performed to disentangle the different sources of the consultation length variability (intra and inter physician). The dependant variable was the logarithm of the consultation length. Independent variables included patient's sociodemographics as well as the characteristics of GPs and their medical activity. RESULTS: Between November 2011 and April 2012, 128 physicians were recruited throughout France and generated 20,779 consultations timed by residents. The average consultation length in the sample was 16.8 min. After adjusting for patients' characteristics only, the consultation length of CAPI signatories was 14.1% lower than that observed for non signatories (p<0.001). After adjusting for GPs' characteristics and the case mix, the CAPI was no longer a significant predictor of the consultation length. The results did not change significantly from one type of consultation to another. CONCLUSION: Although the CAPI was extended to all GPs in 2012, our results provide a cautionary message to regulators about its ability to generate higher quality of care.
BACKGROUND: In 2009, a voluntary-based pay for performance scheme targeting general practitioners (GPs) was introduced in France through the 'Contract for Improving Individual Practices' (CAPI). OBJECTIVE: To study the impact of the CAPI on French GPs' consultation length. METHODS: Univariate analysis, and multilevel regression analyses were performed to disentangle the different sources of the consultation length variability (intra and inter physician). The dependant variable was the logarithm of the consultation length. Independent variables included patient's sociodemographics as well as the characteristics of GPs and their medical activity. RESULTS: Between November 2011 and April 2012, 128 physicians were recruited throughout France and generated 20,779 consultations timed by residents. The average consultation length in the sample was 16.8 min. After adjusting for patients' characteristics only, the consultation length of CAPI signatories was 14.1% lower than that observed for non signatories (p<0.001). After adjusting for GPs' characteristics and the case mix, the CAPI was no longer a significant predictor of the consultation length. The results did not change significantly from one type of consultation to another. CONCLUSION: Although the CAPI was extended to all GPs in 2012, our results provide a cautionary message to regulators about its ability to generate higher quality of care.
Authors: Louise Devillers; Jonathan Sicsic; Angelique Delbarre; Josselin Le Bel; Emilie Ferrat; Olivier Saint Lary Journal: PLoS One Date: 2018-01-25 Impact factor: 3.240