One key theme that emerged out of last month’s Medical Group Management Association (MGMA) Annual Conference and detailed by Jeff Lagasse in Healthcare IT News, was that “Predictive Analytics is now a Business Necessity.”
Predictive analytics will bring enormous value to population health initiatives across healthcare. Most health systems, however, are currently focused on the wrong use cases. Vendors with big promises and sophisticated tools are inviting health systems to invest in expensive predictive analytics initiatives that come with high risks and uncertain rewards. That is NOT the right place to start. There are many easy wins available for predictive analytics. Systems should take advantage of these opportunities before pursuing the more difficult ones.
High Reward + Low Risk approach
Health systems have a treasure trove of data. The depth and breadth of their customer profiles would be the envy of any other industry. Predictive analytics should start by targeting high reward, low risk opportunities.
- High reward opportunities are ones that drive health for patients AND return for health systems.
- Low risk opportunities are ones that health systems can execute against without big implementation or change management initiatives.
In particular, systems should be targeting initiatives that can deliver payback within 12 weeks, and 2-4x monthly returns going forward. There are many of these to be had.
Value of a Patient Relationship
The average lifetime value of patient relationship to a health system is $50,000-100,000, with huge variation by patient. More importantly, the average lifetime value of that relationship to the patient is even higher. What would it be worth to a family if we were able to: (a) prevent a daughter from contracting cervical cancer, (b) prevent a pre-diabetic father from becoming diabetic, (c) catch and treat a mother’s breast cancer before it progressed…? How would that family value a relationship with a health system that was proactively thinking about them and reaching out to engage them in their health?
Proactively reaching out to patients is something most health systems have already begun to do via care coordinators, case managers, marketing teams, and other groups. Most systems, however haven’t reached the second stage of proactive engagement: prioritizing and coordinating who to reach out to and what to say.
A great place for predictive analytics to start, therefore, is by focusing on predicting the “expected value” of reaching out to patients. We might want to reach out to a patient about cancer screening, or a inpatient discharge follow-up visit, or a diabetic wellness program. In any case, predictive analytics should be used to determine the expected value of proactive engagement with every single patient in the health system’s sphere of influence. The output of this exercise could easily be used to feed existing outreach efforts and, if done correctly, multiply their value to both patients and systems.
Annual Wellness Visit Example
A typical Annual Wellness Visit has a value of around $250. That value represents more than just the margin on the bill for the visit. It includes the value of potential referral opportunities, risk factor scoring opportunities (for ACO patients), readmission avoidance, ER usage avoidance, vaccinations, and more. Clearly, some visits will be worth a lot more than $250. Some will be worth less.
To size the opportunity, let’s assume a health system with 100 primary care providers. Let’s further assume that each primary care provider could support four additional patient visits per day. If each of these patient visits was an Annual Wellness Visit worth $250, each primary care provider could create an incremental $1,000 of value per day. At 20 work days in a month, that is $2 million of incremental value those 100 providers could be generating for the system each month. $24 million per year.
When we reach out to patients to invite them to schedule Annual Wellness Visits, we obviously want to reach out to the ones for whom the visit has the most potential value AND the ones who are most likely to be influenced to action by our reach out. (There are many other variables at work here that are left out for this illustration.) Defined in this way, the problem is custom made from predictive analytics.
Predictive analytics can generate a prioritized list of patients targeted for proactive engagement. That list can be immediately fed to care coordinators, case managers, digital marketing channels, and phone-based outreach resources to be converted to appointments. Because most health systems already have patient outreach efforts in place, this approach does not require workflow changes.
Low Hanging Fruit
There is a lot of low-hanging fruit with short time-to-value and big ROI in healthcare predictive analytics. These opportunities create value for patients and health systems alike. Executing on these opportunities generates money for future investments, and also allows the growth of the organizational process and momentum to ensure success of more complex initiatives in the future.
The health system with 100 primary care providers envisioned above would typically have $20-30 million per year in easy-to-mine use cases. These typically require weeks to launch and weeks to payback. Systems who take this approach end up seeing quick success and rapid growth behind early wins. Ready to mine some use cases?