How patient-, tooth- and restoration-level factors affect the longevity of amalgam and composite restorations in elderly patient populations is an important question given the current national debate on the quality, cost and appropriateness of health care. Currently available results on restoration longevity in the elderly are limited in several ways (few studies on the elderly;most existing studies conducted outside the US with geographically limited, often specialized populations;small sample sizes;and inherent biases in data sources). This project proposes analyzing restoration longevity using electronic dental record (EDR) data from two major health maintenance organizations, HealthPartners and the Marshfield Clinic. We will do so in two specific aims (SA): (1) generate a research dataset from the Virtual Data Warehouse (VDW) for statistical analysis using an ontology-based method;and (2) perform statistical analysis of the dataset to determine factors influencing restoration longevity in the elderly. In SA 1, we will first extend the Oral Health and Disease Ontology (OHDO) being developed collaboratively at the University of Pittsburgh and the University at Buffalo to include all study variables. Then, we will link the OHDO to the relational dataset of EDR data in the Virtual Data Warehouse (VDW), a project of the Health Maintenance Organization Research Network and transform the data in the OHDO into a dataset suitable for statistical analysis. SA 2 will answer two research questions for patients 65 years and older: (1) Does the longevity of amalgam restorations (AR) and composite (CR) restorations depend on a patient's caries history? (2) Does the longevity of AR and CR depend on whether the root surface is involved in the restoration? We will answer these research questions by analyzing EDR data from a population of elderly patients. Our project is innovative because it will use EDR data on a large scale for a broad population, produce a streamlined and reproducible process of generating statistical datasets from EDR data, and offer a methodological advance useful for similar studies. The project will address research questions important to patients and dentists;produce a generalizable method for extracting a variety of clinical data from multiple EDR systems and using them for clinical, comparative effectiveness, epidemiological and other research;and create future opportunities for the conjoint analysis of dental and medical data.
This project will address how patient-, tooth- and restoration-level factors affect the longevity of amalgam and composite restorations in an elderly patient population in two large health maintenance organizations, HealthPartners and the Marshfield Clinic. We will do so using an innovative methodology for converting data in electronic dental records into a dataset for statistical analysis using an ontology. The project will not only answer an important dental research question, but also produce a generalizable method for extracting a variety of clinical data from multiple EDR systems and using them for clinical, comparative effectiveness, epidemiological and other research.