When making healthcare decisions, patients, physicians and policy makers need unbiased information about the treatment effects on a variety of health outcomes. Nonetheless, more than half of medical treatments lack valid evidence of effectiveness, particularly for long-term and patient-centered outcomes. Therapies that demonstrate efficacy in randomized controlled trials may perform differently in general clinical practice, where there is a wider diversity of patients, providers, and healthcare delivery systems. The effects of these variations on treatment are often unknown but can influence the net benefits and risks of different therapy options in individual patients.

Population-based secondary data, i.e., the main sources for measuring real-world practice, have proliferated during the past few decades. Among these, healthcare utilization (HCU) databases that collect data for healthcare delivery system management are valuable resource for healthcare observational framework. However, as observational studies are particularly vulnerable to several sources of measurement and confounding biases, and because secondary HCU data especially suffer from the lack of diagnostic, therapeutic and clinical information, these fields should be investigated with robust methodologies.

We will focus on strengths, weaknesses, potentials of secondary computerized healthcare data for generating robust real world evidence and supporting policy decisions.