Background: Preventive activities carried out in primary care have important variability that makes necessary to know which factors have an impact in order to establish future strategies for improvement. The present study has three objectives: 1) To describe the variability in the implementation of 7 preventive services (screening for smoking status, alcohol abuse, hypertension, hypercholesterolemia, obesity, influenza and tetanus immunization) and to determine their related factors; 2) To describe the degree of control of 5 identified health problems (smoking, alcohol abuse, hypertension, hypercholesterolemia and obesity); 3) To calculate intraclass correlation coefficients.
Design: Multi-centered cross-sectional study of a randomised sample of primary health care teams from 3 regions of Spain designed to analyse variability and related factors of 7 selected preventive services in years 2006 and 2007. At the end of 2008, we will perform a cross-sectional study of a cohort of patients attended in 2006 or 2007 to asses the degree of control of 5 identified health problems. All subjects older than 16 years assigned to a randomised sample of 22 computerized primary health care teams and attended during the study period are included in each region providing a sample with more than 850.000 subjects. The main outcome measures will be implementation of 7 preventive services and control of 5 identified health problems. Furthermore, there will be 3 levels of data collection: 1) Patient level (age, gender, morbidity, preventive services, attendance); 2) Health-care professional level (professional characteristics, years working at the team, workload); 3) Team level (characteristics, electronic clinical record system). Data will be transferred from electronic clinical records to a central database with prior encryption and dissociation of subject, professional and team identity. Global and regional analysis will be performed including standard analysis for primary health care teams and health-care professional level. Linear and logistic regression multilevel analysis adjusted for individual and cluster variables will also be performed. Variability in the number of preventive services implemented will be calculated with Poisson multilevel models. Team and health-care professional will be considered random effects. Intraclass correlation coefficients, standard error and variance components for the different outcome measures will be calculated.