The past year and a half has certainly tested the resilience of healthcare organizations both nationally and globally, across a number of dimensions. At the same time, the pandemic has stimulated innovation progress that could pave the way for a promising future. In that context, we at Healthcare Innovation were once again thrilled with the outstanding quality of submissions we received from healthcare stakeholders across the U.S., as part of our 2021 Innovator Awards program. In addition to the four winning teams this year, our editorial staff also selected six organizations as semifinalists. Below are descriptions of these teams’ initiatives, written by the project leaders themselves.
Penn Medicine: Discrete Genetic Data in the EHR
Penn Medicine created a discrete genetics orders and results interface to operate with outside genetics labs. The objective of the project was to allow clinicians and genetic counselors to enter orders within the health system electronic health record (EHR), and electronically interface them with an outside lab while having the lab interface discrete results back in to the EHR. This interface then transmits discrete results into the EHR, segregating genetic variants which allows clinicians to make informed clinical decisions in order to provide better patient care. Penn Medicine’s EHR precision medicine tab also allows its providers to more efficiently locate genetic orders, results and reports on a patient’s chart. So far, 22 different departments and 25 providers have placed orders on close to 500 patients for genetic/genomic testing.
Landmark Health: Leveraging Machine Learning for Palliative Care Patients
The goal for Landmark Health—which provides interdisciplinary care and 24/7 home visits for high-risk, medically complex patients—was to ensure they provide the best standard of care to seriously ill patients by aligning medical care with patient goals and wishes. As many as 15 percent of Landmark’s patients will pass away each year, meaning it is critical for the organization to accurately and consistently identify patients with supportive care needs. To do so, Landmark tapped into machine learning and developed an end-of-life (EOL) prediction model that enables the care team to prepare EOL conversations with patients at the right time. Landmark’s data science team specifically developed a predictive algorithm to assess each patient’s risk of death, hospice, or palliative encounter in the next year, testing over 200 clinical, demographic, and social factors associated with EOL risk mined from clinical literature, expert interviews, and Landmark experience.
Texas Children’s Hospital: A Readmission-Prevention Strategy for Pediatrics
Texas Children’s Hospital (TCH) leaders were interested in using risk scores to help focus their efforts on the highest-risk patients. Readmission prevention programs that utilize risk scoring tools are common in adult medicine, but many available models don’t perform as well in a pediatric environment. Building on its foundation of 10+ years of EHR data, TCH developed its own predictive model using machine learning technology to create an algorithm based on historical pediatric data at the enterprise. TCH chose three acute care units plus the Heart Center to pilot its risk-based interventions; those highest-risk patients received a high-quality after visit summary; meds-to-beds for discharge medications and the scheduling of any necessary subspecialty follow-up visits before they left the hospital; and a follow-up phone call from a nurse within 72 hours of discharge.
Seattle Children's Hospital: In Pursuit of an Opioid-Free Pediatric Ambulatory Surgery Center
Approximately 130 people die a day because of the opioid epidemic in the U.S. Surgery is now known to be a gateway to long term opioid use; 5 percent of adolescents and 7 percent of adults develop persistent opioid use following even minor surgery. Within 18 months, the perioperative team at Seattle Children's Bellevue Surgery Center were able to reduce perioperative opioid exposure for outpatient surgery to near zero. They did so by deploying an AI solution that enabled to team to learn and adapt their protocols from real-world data captured by their EHRs. Previously, it took years to complete even one cycle of change when adjusting clinical protocols, but now it takes weeks as this solution allows clinicians to see across patients in the EHR, rather than just doing a deep dive on one patient. This means when the team changed from opioids to alternative medications, they were instantly able to understand if it was an improvement.
Stanford Children’s Health: Comprehensive Care from the Fetal Center Team
The Johnson Center for Pregnancy and Newborn Services at Stanford Children’s Health is exclusively focused on discovering ways to address the full range of disorders that may occur in the mother and fetus during pregnancy and gestation. The Fetal Center team experienced challenges with using a legacy application to track fetal status, and the unique functionality required does not currently exist in the industry standard EHR systems. Therefore, multiple clinical, business and IT teams at the health system partnered to develop their own application in-house, with the goal of replacing the legacy application, improving user experience, and optimizing clinical workflows. The application, named Puffin, has generated strong results, enabling the Stanford Medicine Fetal Center program to track critical information of complex pregnancies for maternal, fetal, and neonatal cases to provide comprehensive care. Puffin has also led to significant improvements in clinical workflows, including reducing duplicative documentation and improving patient specific outcome reviews.
The Harris Center for Mental Health and IDD: Connecting Law Enforcement with Mental Health Clinicians
Officers routinely respond to mental health calls with very limited expertise, which can often lead to negative outcomes associated with these interactions. As a result, a partnership between the Houston-based Harris County Sheriff Office (HCSO) and The Harris Center was implemented to improve response to calls involving a person with mental illness. The CORE program connects a law enforcement first responder with a mental health clinician using a tablet and HIPAA-compliant technology. There are 200 tablets distributed throughout HCSO, and direct impacts from the initiative include: 42 percent of such incidents were resolved on the scene with the patient remaining in the community; 58 percent resulted in transportation to hospitals; and in less than 2 percent of the time, the person was transported to jail or juvenile detention. What’s more, 86 percent of law enforcement partners themselves reported the clinician helped the officer de-escalate the consumer and 93 percent indicated the clinician helped them decide what course of action to take with the consumer.
