Personalized Clinical Decision Support to Improve Participation in Hospital at Home

PROJECT SUMMARY Inpatient hospitalization is costly – accounting for $1.1 trillion in health care spending annually in the United States – and is associated with high morbidity and mortality risks. Hospital at Home (HaH) is an alternative care model where care teams provide acute hospital-level care in a patient’s home. Despite decades of data

that support HaH efficacy relevant to key patient-centered outcomes, barriers to HaH participation limit its widespread adoption and population health impact. Our central hypothesis is that providers and patients require Clinical Decision Support (CDS) integrating data from disparate sources and a Shared Decision Making

(SDM) framework to help inform point-of-care decisions regarding HaH and surmount low participation rates. The overarching goal of our work is to improve value-driven care by helping patients engage in the decision of which acute-level care option best meets their needs. The objective of this study is to evaluate whether a

Health Information Technology (IT)-enabled SDM solution incorporating expected patient outcomes and preferences and deployed at the point-of-care improves patient and provider participation in HaH as a care model. To achieve this objective, we will: 1) characterize patient, caregiver, and provider perceptions of the risk

tradeoffs, needs, and care preferences for HaH; 2) partner with patients, caregivers, and providers to iteratively design Hospital-level Outpatient Management Evaluation and Decision Support (HOME-DS), a Health IT- enabled SDM solution that incorporates risk-model probabilities and patient and caregiver preferences; and 3)

evaluate the feasibility of implementing HOME-DS in acute care and establish the acceptance rate of HaH. We will focus HOME-DS on adults aged 18 and older hospitalized with suspected pneumonia, a prevelant condition that has been commonly included in HaH models. We will apply the previously validated and broadly

accepted Pneumonia Severity Index (PSI) as the quantitative risk score input for HOME-DS. User personas and needs for the initial HOME-DS prototype will be defined through key informant interviews with patients hospitalized with pneumonia, their caregivers, and providers (Aim1); user-centric design principles will further

guide iterative development of the HOME-DS prototype (Aim 2); and we will test the feasibility of implementing HOME-DS in acute care to guide patient and caregiver decision making in selecting hospital level care in the home or traditional hospital (Aim 3). We hypothesize HOME-DS is feasible to implement within the provider

workflow for hospital admission and can yield participation rates in HaH of 50%. The proposed project will engage a heterogeneous population with pneumonia, as this is a population with substantial acute care utilization costs and a large gap in understanding implementation challenges to explain why alternatives to

traditional hospitalization are not used more widely. Results will demonstrate the value of Health IT that integrates clinical data with patient preferences to promote effective SDM enhancing patient-centered acute care options.