Objective: Health care has variable features, specific determinants, and consequences on people's health that should be investigated in a population perspective. We present a population-based analytical approach to assessing patterns, determinants, and outcomes of health care.
Methods: The approach uses standard epidemiological methods for sample selection and analysis, as well as statistical methods concerning model fitting and validation strategy. It also uses psychometric methods allowing pattern identification: factor analysis, cluster analysis, and polytomous logistic regression to investigate the factors associated with identified patterns of care. The approach is illustrated with an application to ectopic pregnancy (EP).
Results: EP diagnostic and therapeutic procedures appeared to be closely related, suggesting 3 differentiated patterns of care. Predictors of these patterns were not only the clinical features and the woman's prior gynecological history but also center characteristics and location. These patterns were found to influence immediate and middle-term outcomes and risk of recurrence of EP.
Conclusions: We illustrate the feasibility and profitability of a population-based analytical approach to identify patterns, determinants, and consequences of care. This approach could be used to analyze other pathological conditions and health care systems.