Predictive Analytics and Peer-Driven Intervention for Guideline-based Care for Sleep Apnea

OSA affects 7 to 12% of the US population and is an independent risk factor for several clinical consequences such as systemic hypertension, cardiovascular disease, stroke, reduced HR-QOL, increased all-cause mortality, and motor vehicle accidents due to sleepiness. However, OSA and other sleep disorders are woefully underdiagnosed because of time constraints at the PCP's office. Such poor CPAP adherence is associated with increased risk for fatal and non-fatal cardiovascular events. CPAP therapy has been associated with up to 3-fold reduction in fatal and non-fatal cardiovascular events in both on-treatment analysis of RCTs and observational studies. CPAP therapy is associated with 7-fold reduction in sleepiness-related accidents. The overarching aim of this proposal is to institute a multi-level strategy aimed at improving guideline based care for OSA in populations with health disparities. The overarching aim of the program of research is to eventually institute a multi-level strategy with interventions aimed at both providers and patients in order to improve guideline based care for OSA in populations with health disparities. Specifically, in order to address the under-diagnosis of OSA, an electronic health record (EHR) based predictive analytics system will be implemented. A similar EHR-based alert system with reflex order sets for initiating testing for OSA that the investigators have developed will enable PCPs to effectively diagnose and treat OSA. It was found that in a nationally representative dataset of \~220,000 patients, lower neighborhood income was associated with lower CPAP adherence. Considering the time and access-related barriers in the PCP office, it is believed that such socioeconomic status (SES) related differences in CPAP adherence needs to be addressed by culturally-competent peer educators (such as promotors) with OSA serving as "peer-buddies". In a recent multi-site RCT, it was found that CPAP adherence and patient satisfaction was improved by peer-driven intervention through an interactive voice response system (PDI-IVR; PCORI-IHS-1306-02505), whereby the peers shared their experiences with CPAP naïve patients and educated them about OSA and risks of CPAP nonadherence. Whether PDI-IVR can improve adherence in underserved population in low resource clinics is unclear. In this proposal, 110 CPAP naïve patients and 35 experienced peer-buddies will be recruited in from a large healthcare system that provides care for an underserved population with an excess burden of sleep disorders (Medicaid and dual-eligible Medicare beneficiaries). This study will address health disparities in the diagnosis and treatment of OSA using a patient-level educational intervention (PDI-IVR system) targeting patients and thereby promote health equity. This proposal will also inform future studies by gathering additional data on the effect of PDI-IVR on adherence to sleep study testing and CPAP adherence in an underserved population in low resource clinics. Specific Aim #1: To test a peer-driven intervention with interactive voice response (PDI-IVR) system to improve adherence to sleep study testing in a lower socioeconomic population. Hypothesis #1: A PDI-IVR support system will lead to greater adherence to sleep study testing in patients with OSA when compared to patients receiving conventional education. Specific Aim #2: To test a peer-driven intervention with interactive voice response (PDI-IVR) system to improve CPAP treatment adherence in a lower socioeconomic population. Hypothesis #2: A PDI-IVR support system will lead to greater adherence to CPAP therapy in patients with OSA when compared to patients receiving conventional education.