Virtualized Homes: Tools for Better Discharge Planning

Hospital readmissions are difficult for patients and place institutions at increased financial risk for uncompensated care. Creating post-hospital care plans that leverage the home environment and compensate for its deficiencies may help to reduce readmissions. Better understanding of a patient’s residence can enhance tailoring of home-based self-care and disease management interventions. Currently, information about the home environment, such as adequate utilities, come from patients or from home visits. These visits can be intrusive and costly, and at best may yield text descriptions or simple images that lack precision and do not provide adequate detail. Patient self-reports rely on recall and can reflect selective perception, understating hazards or overstating space and affordances.

Technology makes it possible to create full-scale 3-dimensional (3D) models of the interiors of patient homes for provider review. Displaying these models and making them accessible through the electronic health record (EHR) could improve discharge planning by creating a better understanding of the patient’s home situation and enabling plans to be more tailored to the patient’s life. This research demonstrated the technical feasibility and clinical value of making full 3D models of actual patient homes available for exploration and note taking, from within the EHR.

The specific aims of the research were as follows:

• Develop a new research technology, Home3D, that will enable visualization, exploration, manipulation, and annotation of a full-3D model of a patient’s home through a web interface, and storage of those annotated images in the EHR. 
• Develop the Home3D viewer and the data interface with the EHR. 
• Demonstrate an interoperable data storage strategy. 
• Determine the usefulness for patients and for clinicians of integrating a full rich description of the home into clinical care, using the discharge planning process as a clinical use case. 

The research focused on finding an efficient way to securely store the large amount of data generated from the 3D images of each home. The storage solution needed to be accessible from within the EHR, without storing the actual data there. In addition, the project worked to resolve issues to make the solution functional, including issues around identity management, privacy, efficiency when rendering images, and transmission speed.

A web-based viewer was developed that supported the creation of 3D models of actual patient homes and was accessible via browsers or mobile devices. Access to the viewer was integrated into the EHR, allowing clinicians to navigate around the 3D image and annotate it. Still images from the viewer were able to be exported and stored back in the EHR. Once implementation, integration, and display of the house data was validated, clinician stakeholder focus groups were held to explore their opinions of the type of patients for whom this resource would be of assistance, the tools’ usefulness, utility, and acceptability. Focus groups reacted positively to the tool and its potential usefulness for planning and preparing patients for discharge. In particular they felt that not all patient homes would require scanning, but that it would be useful for patients with complex discharges, with greater illness burden, those requiring rehabilitation, and those whose self-reports on their homes required greater detail for better understanding of their environment.

Future research will aim to better understand the experience of the end user, develop ways to measure the quality of the system during motion, and explore alternative strategies to improve speeds.