9 Lessons Learned from UCSF’s Digital Transformation of Care Delivery
Financial and e-commerce industries like Amazon and Netflix have long since made the leap to digital. By leveraging user data, they are better informed about their consumers and able to delight their customers with rich, personalized, and intuitive experiences.
While healthcare made the transition from paper records to electronic health records several years ago, they are still lagging behind other industries in providing modern digital experiences for patients, families, caregivers, providers, and staff.
In the American Hospital Association’s webinar “Leveraging Modern Tools and Approaches to Transform Complex Care Delivery,” Aaron Neinstein, M.D., Vice-President of Digital Health, shared nine learnings from UCSF Health’s digital transformation to propel the health system to become a digital-first organization (link internal to UCSF) providing more empathetic, accessible care to patients wherever they are and whenever needed.
1. Reframe Problems with “How Might We”
Neinstein recommends stepping back, finding out what patients want, and it can be accomplished in new and different ways by changing workflows to create new experiences.
With thousands of kidney patients who might need a transplant, only a few hundred can be put on the transplant list. By reimagining and reframing the problem, UCSF Health launched a Kidney Transplant Pre-List Chat to allow patients who might one day need to get listed to enter the process early. By doing so, patients are engaged and can receive education and information such as the criteria needed to enter the formal listing process, while also allowing UCSF to better target matches.
Another example was during the pandemic. Following a lung transplant, patients need careful monitoring for signs of organ rejection, but COVID-19 made it too risky for UCSF’s lung transplant patients to travel to labs for onsite testing. By reframing the problem, approximately 500 patients were sent home spirometry kits and virtual chat tools. With the pivot, patients were able to safely monitor their health from home. And, we’ve also been able to replace six in-person visits per patient during the first two years after lung transplant with home spirometry and video visits.
2. Organize Cross-Functional Teams Around Areas of Value
By creating cross-functional teams and using collaboration, iterative experimentation, and process improvement, traditional organizational silos were broken down and teams were able to solve problems requiring the input and expertise of various departments and perspectives.
An example of what a cross-functional team can accomplish: this past summer, UCSF’s Cancer Center offered self-scheduling on the web (link internal to UCSF) for people newly diagnosed with cancer, which meant appointments could be made without a referral or a phone call in just a few minutes, saving time and reducing stress for patients and staff alike.
3. Create Analytics That Capture and Illuminate the Patient Journey
Through granular data measurement and instrumentation, we can understand the friction patients encounter in seeking care. In looking at every step of the patient’s journey and capturing and analyzing information such as search keywords, browser type, and website pathways we can make it easier for patients to find care and schedule themselves more quickly.
4. Think Outside-In, Not Inside-Out
Making the scheduling experience as frictionless and as easy as possible for patients is easier said than done to ensure patients are able to schedule at the right clinic and with the right provider for their condition. Neinstein says it is crucial to think outside-in rather than inside-out. The team worked on the backend to ensure the patient had all the information needed to get them the care they need, lessening the burden on the patient rather than solely thinking about internal operational needs.
5. Leverage Analytics to Enable Segmentation, Personalization, & Equity
Through granular data collection and understanding who is or is not using digital tools, we can learn what healthcare disparity gaps exist and take steps to make healthcare accessible to all.
By knowing our patient population, we can tailor product requirements to better meet the needs of the broader demographic served by UCSF, and to empower more people to control their care.
6. Measure Key Process Metrics
Neinstein talks about avoiding the trap of just implementing digital technologies but not really doing digital transformation. As an example, a clinic can implement a system to send a text message to patients after receiving their referral with instructions to call and schedule during business hours. By looking outside-in, the team took a step back to look at experience from a different perspective – and to start measuring themselves by the appointments scheduled without requiring a phone call.
He stresses the importance of a patient-first focus to understand patients and what they are trying to achieve, then to build the measurements around those outcomes versus simply measuring technology deployment milestones.
7. Run Small, Quick, Cheap Experiments to De-risk
In digitally-transformed organizations, the quickest and cheapest way to de-risk a concept to determine it has merit before committing significant resources to it is to deploy the product to a minimum number of users, evaluate, then iterate to make improvements.
The team tested two hypotheses in their experiment for cancer referrals:
8. Leverage AI for Your Boring Problems
A practicing endocrinologist, Neinstein understands the importance of patient interactions. While some are concerned that technology and Artificial Intelligence (AI) might fracture the doctor-patient relationship, Neinstein says that automation in fact allows physicians more time to interact due to lessened administrative overheard.
In the webinar, he talks about huge opportunities for AI to replace manual and labor-intensive processes, while allowing staff to focus more on the patients.
CDHI launched Referrals Automation several years ago to improve the efficiency of referral scheduling, and it is estimated that it could save UCSF administration staff more than 25,000 hours annually in faxed document processing with the addition of AI to determine fax type, triage priority, and more.
9. Enable a Difference in Kind, Not Degree
Finally, Neinstein talks about the impact of automation and the difference AI makes in kind, and asking ourselves if workflows should be done in an entirely new and different way.
With the addition of AI to Referrals Automation, faxes are prioritized with the highest level of urgency first, while without AI, faxes are processed in the order received regardless of the urgency of the case.
In leveraging AI for logistical and operational and workflow improvement, UCSF has partnered with Phillips to look at the patient flow capacity suite.
To watch the webinar on demand, click here To download the slides, click here