Penn Medicine OnDemand
Penn Medicine Center for Connected Care
Office of the Chief Medical Information Officer
The Wharton School
UPHS Quality & Patient Safety Award, 2018
Milton W. Hamolsky Jr. Faculty Member Scientific Presentation Award Finalist, SGIM National Meeting
Seeing a primary care physician (PCP) urgently isn't always easy, especially if the practice is not close to where you work, you work long hours that are not conducive to daytime appointments, or you're not an established patient.
When we started work on this project, one in five employees at Penn Medicine was visiting the emergency department (ED) each year, with a significant percentage qualifying as avoidable visits.
Self-insured employers like Penn Medicine are interested in reducing avoidable ED visits. Directing employees to higher-value care can improve satisfaction and outcomes, decrease unnecessary utilization, and reduce benefits costs.
Penn Medicine OnDemand (PMOD) is a fully scaled virtual telemedicine practice operated by Penn Medicine’s Center for Connected Care. Providers are available 24 hours a day, seven days a week, 365 days a year.
PMOD makes it easy for users to obtain care when they need it, regardless of physical location, using video visits. When an in-person, office-based appointment with a care provider is required, PMOD providers help users schedule an in-person appointment at a time and location that is convenient for them.
In its first two years serving the 60,000 covered lives under the Penn Medicine insurance plan, PMOD conducted more than 7,200 clinical encounters, saved the health system over $1.5 million in avoidable ED or urgent care visits, saved employees over $425,000 in copays, and substantially increased the number of employees establishing primary care within Penn Medicine.
On January 1, 2020, PMOD began offering service to the general public. In March of that year, PMOD became a critical tool used by Penn Medicine to keep patients and providers safe during the COVID-19 pandemic. At the outset of the pandemic, the team adapted in real-time to handle the increased volume – with daily calls skyrocketing from 50-60 to more than 400.
Fake back end
Fake back end
Fake back end
It is essential to validate feasibility and understand user needs before investing in the design and development of a product or service.
A fake back end is a temporary, usually unsustainable, structure that presents as a real service to users but is not fully developed on the back end.
Fake back ends can help you answer the questions, "What happens if people use this?" and "Does this move the needle?"
As opposed to fake front ends, fake back ends can produce a real outcome for target users on a small scale. For example, suppose you pretend to be the automated back end of a two-way texting service during a pilot. In that case, the user will receive answers from the service, just ones generated by you instead of automation.
Fake back end
We launched and marketed a simple landing page offering Penn Medicine employees access to free virtual care and engaged a pool of clinical partners to triage appointments over three months. When requests came in, we manually routed phone calls to providers based on a preset schedule.
Piloting the service as a fake back end helped us learn about demand, feasibility, and viability. It also enabled us to collect insights about user needs, behaviors, and what kind of staff would be needed to support a permanent service. Finally, findings from this pilot helped us secure executive support for PMOD and informed the program's design at scale.
We ran a vapor test in the initial pilot to learn more about how people wanted to connect with care providers. Specifically, we wanted to test our hypothesis that patients would prefer video visits.
The pilot landing page offered Penn Medicine employees the option to call, video conference, or chat with providers. At the time, we were only equipped to treat patients over the phone. If a user requested a video conference or chat, a message popped up, notifying them that the service they selected was not yet available but that they could be connected immediately with a provider for a phone encounter.
During the pilot, 56 percent of patients chose the chat option compared to 26 percent for video and 18 percent for phone calls. However, follow-up surveys revealed that 67 percent of patients were satisfied with the phone encounter after speaking with a provider. During debriefs, pilot providers indicated that they felt that video visits would very rarely have changed treatment plans for patients.
These insights helped the team discern what functionality was necessary for the program's first iteration at scale.