Developing and Launching a Research Participant Registry.

BACKGROUND: A major challenge in clinical research today is the difficulty that studies have in meeting recruitment goals. Up to 48% of studies do not meet accrual goals within the specified timeframe, significantly delaying the progress of projects and the dissemination of findings. This pervasive problem is a recruitment crisis. We developed a representative, ethnically and racially diverse research participant registry in a predominantly rural state with high levels of health care disparities and minority populations. We sought input at each step of development from members of community advisory boards (CABs) across Arkansas. We report how community involvement in the development of the registry was implemented. METHODS: Members of CABs were partners in developing all aspects of the registry website, including the name, content, appearance, educational messages, and testimonials used. Constructs from the Health Belief Model informed the educational messages and supported the intense multimedia campaign used to launch and promote ongoing registrations. Using CAB guidance, community events were held throughout the state of Arkansas at venues with diverse racial and ethnic attendance. RESULTS: From April 1, 2016 to September 1, 2017, 4,002 people registered statewide who match the demographic profile of Arkansas. CONCLUSION: CAB involvement in the registry, multiple cues to action, and face-to-face contact with diverse lay audiences throughout the state were key components of the successful registry launch.

Development of non-invasive diabetes risk prediction models as decision support tools designed for application in the dental clinical environment.

The objective was to develop a predictive model using medical-dental data from an integrated electronic health record (iEHR) to identify individuals with undiagnosed diabetes mellitus (DM) in dental settings. Retrospective data retrieved from Marshfield Clinic Health System's data-warehouse was pre-processed prior to conducting analysis. A subset was extracted from the preprocessed dataset for external evaluation (Nvalidation) of derived predictive models. Further, subsets of 30%-70%, 40%-60% and 50%-50% case-to-control ratios were created for training/testing. Feature selection was performed on all datasets. Four machine learning (ML) classifiers were evaluated: logistic regression (LR), multilayer perceptron (MLP), support vector machines (SVM) and random forests (RF). Model performance was evaluated on Nvalidation. We retrieved a total of 5319 cases and 36,224 controls. From the initial 116 medical and dental features, 107 were used after performing feature selection. RF applied to the 50%-50% case-control ratio outperformed other predictive models over Nvalidation achieving a total accuracy (94.14%), sensitivity (0.941), specificity (0.943), F-measure (0.941), Mathews-correlation-coefficient (0.885) and area under the receiver operating curve (0.972). Future directions include incorporation of this predictive model into iEHR as a clinical decision support tool to screen and detect patients at risk for DM triggering follow-ups and referrals for integrated care delivery between dentists and physicians.