Evaluating clinical guidelines for chronic disease management: Do they enable the personalization of care?

Objective: To described how general practitioners (GPs) personalize interventions for patients with chronic diseases and compare practice with the corresponding guidelines.

Study design: Scoping review followed by a multicentre cross-sectional study in French general practices.

Methods: We identified elements of personalization described in guidelines related to diabetes, hypertension, dyslipidaemia, insomnia and depression. Then, GPs completed questionnaires for pharmacological (PI) and non-pharmacological interventions (NPI) after any consultation for these diseases to collect: when, on which the variables (clinical, biological characteristics, etc.), how and by whom the interventions were personalized, and what was personalized in the interventions. Agreement between GPs' practices and guidelines was analyzed using Cohen's Kappa.

Results: We extracted 204 elements of personalization in 10 guidelines, and GPs described 1512 elements of personalization in 161 PI and 1313 elements in 131 NPI. Personalization was mainly based on patients' general characteristics (20.6 % of PT; 24.8 % of NPI) and treatments characteristics (14.5 % of PI; 9.8 % of NPI). GPs accounted for patients' preferences in 64.6 % of PI and 79.4 % of NPI. For PI, the agreement between GPs and guidelines was globally low (kappa = 0.21[0.11; 0.31]) but moderate for treatment characteristics (kappa = 0.48 [0.09; 0.87]) and high for disease characteristics (kappa = 1.00[1.00; 1.00]). For NPI, agreement was globally very low (kappa = 0.16[0.10; 0.25]) but moderate for treatment characteristics (kappa = 0.59[0.19; 1.00]) and disease characteristics (kappa = 0.48[0.12; 0.87]).

Conclusions: Guidelines insufficiently described the tailoring variables and the subsequent modifications of the interventions. They need to be better described to promote a medicine that is both personalized to each patient and homogeneous between physicians.