Using administrative data for epidemiological research: case study to identify persons with periodontitis

The availability of automated administrative databases in medical and dental healthcare, combined with modern computing power, allows sophisticated multi-layered analyses of the causes and consequences of care delivered to large populations. These automated databases are now also commonly linked to data extracted from electronic medical records, and are used to measure health system performance, adjust payments, assess providers for quality improvement purposes, and inform policy decisions. Use of health insurance databases can be particularly powerful when linked through common unique identifiers to allow examination of the influences of dental care on medical outcomes and costs, and vice versa. This review describes the growing use of these databases for epidemiological and health services research, their advantages and limitations, and presents a case study in which dental claims data are used to identify persons with periodontitis in order to enable critical assessment of the relationship between periodontal interventions and systemic health outcomes/medical costs.