{ Penn Medicine Center for Health Care Innovation

We are excited to announce the winners for our second round of Innovation Grants.

The Innovation Grant Program was designed to advance promising ideas and promote exploration in the areas of connected health and employee health. Winners receive funding and support from the Center for Health Care Innovation to facilitate the rapid translation of ideas into action and measurable outcomes over six months.

We were impressed by the quality and scope of the more than 50 proposals we received, and look forward to starting work with the teams in June.

Cloud-based platform for ICU EEG monitoring and visualizing results
A team led by Brian Litt, MD, professor in Neurology & Bioengineering, will build an automated, cloud-based platform for Intensive Care Unit (ICU) electroencephalogram (EEG) interpretation. Patients are monitored continuously with EEGs in ICUs worldwide. Recent studies show a large percentage of ICU patients have seizures, brain ischemia, encephalopathy, or other conditions that can be detected early on an EEG, allowing therapy to be initiated promptly. However, continuous long-term EEG monitoring currently presents two major problems: it must be interpreted manually by physicians, delaying the delivery of results to the caregivers, and those caregivers rely on written reports from these studies, thus inhibiting the ability to view trends over time or forecast when a patient’s condition may deteriorate. The project aims to build an automated, cloud-based system for interpreting long-term ICU EEG data to speed response to changes in patients’ conditions and improve patient outcomes.

Telemedicine to improve access to genetic services
Angela R. Bradbury, MD, assistant professor of Hematology-Oncology in the Abramson Cancer Center, will use telemedicine to increase access to genetic testing and counseling services. Genetic testing for cancer susceptibility is now an essential component of oncology care, increasing the need for genetic counseling specialists to assist in care of patients and their families. Testing is typically available only at large, academic facilities, leaving many providers and patients without access to genetic counseling locally. Genetic testing should always be conducted in conjunction with proper pre- and post- test counseling to contextualize the test and outline what the results may mean. As genomic applications in oncology expand, the demand for genetic expertise will increase and gaps in delivery will worsen. Through an NIH-funded study, Bradbury and her team showed telemedicine can be an effective way to expand genetic services to populations with limited or no access to care. The new project seeks to transition the team’s research-supported telemedicine program to a sustainable clinical model.

Technology to Improve Prenatal Services
Spearheaded by Ian Bennett, MD, PhD, associate professor of Family Medicine & Community Health, this initiative uses text messages to engage and educate patients, enabling early interventions to reduce poor pregnancy outcomes. Low income women have high rates of poor pregnancy outcomes, including low birth weight, preterm birth, and preeclampsia. While signs of these conditions and associated risk factors can be identified in the course of prenatal care and targeted by interventions, the effectiveness of prenatal visits can be limited by patient literacy and engagement, as well as limited time to educate them. Delays in the identification of these disorders can result in poor perinatal outcomes. Penn Medicine’s Helen O. Dickens Center for Women serves more than 3,000 low income patients each year, primarily African American women who are at increased risk for these outcomes. The project will create an application to deliver information regarding signs and symptoms of adverse pregnancy conditions to at risk women via text message. Fundamental to this project is the belief that an informed and engaged patient will increase the effectiveness of monitoring for pregnancy disorders.

Innovation Grant Round One Winners

Shreya Kangovi: Dissemination of IMPaCT
Dr. Kangovi has developed an effective model for deploying community health workers (CHWs) to improve health outcomes for low socioeconomic status patients. A recent RCT demonstrated fewer recurrent hospital readmissions, increased access to primary care, improved communication (measured by HCAHPS), higher patient activation and improved mental health for this vulnerable population. Based on these findings, Penn Medicine established a Center to foster her model within our hospitals and practices. That Center now receives requests from other health systems and payers seeking to implement the IMPaCT model, but we currently don’t have the infrastructure or capacity to support dissemination. As changing health policies drive health systems to seek such programs (and pay for help setting them up), their investment in models not based on evidence is likely to be wasteful (most prior programs have been ineffective) and even have the potential for adverse outcomes. The goal of this project is to establish scalable tools, including an interactive resource for hiring, training and guiding CHWs, and processes to efficiently adapt the model to local needs, while also testing for a business model enabling these programs to be self-sustaining while making a positive impact.

Jules Lipoff, Carrie Kovarik, Misha Rosenbach, William James and Brendan Carr: Teledermatology Most community hospitals do not have sufficient access to dermatologists. Having developed an effective technology platform and shown the potential for teledermatology in both Botswana and Philadelphia health clinics, the team’s goal for this project involves establishing a sustainable business model while also experimenting to explore processes around response times and scalability for broader impact. Telehealth appears poised to play an important role in the future of healthcare delivery, especially in areas where access to specialists is limited (most regions). With rapidly evolving laws and regulatory guidelines, this project gives us an opportunity to learn what works while making a difference for patients.

Lynn Washington, Tricia Shustock, Nancy Findley and Christina Zalewski: faster testing, reporting and anticoagulation adjustments (starting with VAD patients, but looking to expand) to reduce adverse outcomes
Conventional anticoagulation management methods continue to result in readmissions, patient dissatisfaction and rising costs. Frequent trips to labs, variations in care obtaining blood samples and staff capacity constraints (among other issues) impact this high risk population. Delays in identifying and correcting INR results too often has severe consequences. One aspect of this project will involve understanding how to effectively incorporate greater self-management with point of care (POC) testing kits, providing immediate INR readings to enable faster interventions. It is understood, though, that self-testing isn’t new and that alone will not be a magic bullet resolving the complexity of caring for these patients. We’re engaging in this project with the belief that connected health approaches and process innovation have the potential to improve patient outcomes.

Keith Hamilton, Kevin Haynes and Jimish Mehta: generation of near real-time antibiograms leveraging the Penn Data Store (PDS)
Funded by Dr. PJ Brennan. Antibiogram reports, ensuring patients receive correct antibiotics and controlling the spread of drug resistant bacteria, often require months to create. This effort makes the process prohibitive in many facilities, as well as delaying identification of changes in susceptibility patterns. Without accurate and timely information on both national guidelines and local susceptibility patterns (which can vary even by unit), patient care and outcomes are adversely affected. This team has demonstrated the potential to generate the reports from PDS in less than 15 minutes, while also improving accuracy. Their pilot work, eliminating effort, inaccuracies and delays, has attracted demand from providers to inform appropriate treatment. Additional programming is required to develop web and mobile interfaces that can be used by providers, infection control, antibiotic stewardship and administration, as well as to better capture outpatient susceptibility patterns and unit-specific data. Bringing current and local data quickly and directly to prescribers is expected to reduce variations in care, improve outcomes, limit the spread of antibiotic resistance and enable unit level surveillance to allow the design of more effective interventions.

Hanna Zafar and Tessa Cook: automated monitoring of recommended follow-up for indeterminate or suspicious lesions in abdominal imaging reports
Funded by Dr. PJ Brennan. Currently UPHS can’t accurately monitor follow-up recommendations within abdominal imaging reports and determine whether appropriate follow-up is scheduled, completed or missed. Misses can result in serious implications for patients, such as delayed cancer diagnosis, and legal risk. In prior research, less than half of patients with incidental findings obtained recommended follow-ups. Manual review is not feasible given the volume of studies and variability in the structure of recommendations. As of July 1, the Department of Radiology implemented standardized categorization for lesion reporting. This team will leverage these diagnostic categories, adding requirements for specifying modality and timing of follow-up for indeterminate and suspicious lesions, with an automated system determining the status of recommended follow-ups and generating messages to providers in the case of patients with no scheduled or completed follow-up within a specified time frame. Approximately 400 abdominal imaging studies are performed daily across Penn Medicine. Once developed, automated monitoring of standardized imaging follow-up recommendations could be implemented throughout UPHS to improve patient and health system outcomes.