It’s obvious that COVID-19 has radically reshaped many aspects of our society and our lives over the past two years. But perhaps unnoticed by most, the pandemic has also accelerated a quiet but powerful transformation in the way that we at UPMC approach how we care for patients. The result has been hundreds of lives saved and the promise that we can continue to improve care for years to come.
Prompted by the urgent need to help patients facing a deadly virus, UPMC committed itself at the start of the crisis to truly becoming a “learning health system,” with near real-time changes in how we manage patients based on the analysis of our own data, as well as the latest scientific findings and regulatory approvals.
For decades, clinicians and researchers nationwide have known that our processes for digesting credible, scientific knowledge and integrating that into care for patients are too slow and too cumbersome – sometimes taking years. That’s why the National Academy of Medicine called for development of a learning health system (LHS) in 2009 and set the goal that, by 2020, “90 percent of clinical decisions will be supported by accurate, timely, and up-to-date clinical information, and will reflect the best available evidence.”
Others have defined the LHS as an environment in which “science, informatics, incentives, and culture are aligned for continuous improvement and innovation, with best practices seamlessly embedded in the delivery process and new knowledge captured as an integral byproduct of the delivery experience.”
That’s a mouthful, and a heavy lift for any health system. At UPMC, we’d made some progress toward that goal in recent years, particularly in the science and informatics realms. Because we already had a robust data and analytics infrastructure, we could turn our attention quickly at the start of the pandemic to the incentives, culture and embedding of best practices needed to care for our patients.
That first involved creating a multidisciplinary COVID-19 Therapeutics Committee – composed of physicians, pharmacists, hospital leaders and others — to evaluate any possible treatment options and to rapidly share updated guidelines with all facilities across our system.
In addition to continuously evaluating UPMC’s internal patient data and controlled clinical trials, the committee had to weigh a surge of COVID-19-related publications from around the globe – some peer-reviewed and many not. This placed a premium on expertise in evaluating the merits of published information that appeared at unprecedented scale – and our researchers led or were involved in many trials that rapidly proved which treatments worked best for which patients.
Thus, while our therapeutics committee recognized the benefits of and implemented some treatments – like steroids, remdesivir and tocilizumab — it refuted the use of hydroxychloroquine, despite the government’s approval for its emergency use early on in the pandemic.
Information technology specialists at UPMC created alerts and orders in our electronic health record systems to reinforce guideline recommendations. They also collaborated with our research teams to integrate clinical practice with clinical trial enrollment across our vast system to ensure equitable care, regardless of geography or hospital type. Instead of separating “research” and “care,” all care became an opportunity to learn quickly about improving patient results.
Alongside the therapeutics committee, an intensive care unit management group made real-time recommendations regarding respiratory support strategies and other critical, supportive care, while a systemwide infection prevention taskforce guided testing, contact tracing, isolation and the use of personal protective equipment.
The result of this concerted effort was significant improvement in care in a stunningly short time and mostly prior to mass vaccination. Based on our data from more than 11,400 COVID-19 patients, the adjusted risk of in-hospital mortality decreased monthly by an average of 5% from March 2020 through early June 2021.
At the same time, we saw no appreciable variation in the type or volume of drugs and therapies used for patients with COVID-19 across 22 hospitals, achieving our goal of equity and access regardless of a patient’s ZIP code.
While this pandemic is not over, we’re already looking ahead to what’s next for our learning health system. Although we can’t determine the extent to which any single change improved outcomes for our COVID patients, we know that our accelerated learning saved lives.
Now we must continue to apply these lessons to the continuing hard work of health care long after this crisis passes. Our patients deserve nothing less.
Dr. Marroquin is Chief Health Care Data and Analytics Officer for UPMC.