A New Model for Patient Payments
Justin Nicols, Founder & CEO of Sift Healthcare, breaks down how data science holds the key to improving patient payment models
In this weekly “Think Tank” series, our version of a digital op-ed column, StartUp Health is giving voice to the hundreds of entrepreneurs and innovators working to achieve health moonshots. In this week’s edition of Think Tank, Justin Nicols, Founder & CEO of Sift Healthcare unpacks how data science and propensity to pay scoring are the biggest missed opportunities when it comes to improving patient payment strategies.
By: Justin Nicols, Founder & CEO, Sift Healthcare
I talk to providers every day who ignore the fastest-growing revenue opportunity in healthcare: improved patient payments. Ignore might be too strong of a word. Providers are aware, but many accept the status quo for managing patient payments. The status quo — descriptive analytics and simple scoring — leaves money on the table. This is money that providers have already earned. It can also alienate patients and walks a fine line around regulatory compliance. Healthcare needs a new model, built on data science, for patient payments management.
The ongoing adoption of high deductible healthcare plans continues to increase patient financial responsibility. Patient responsibility after insurance increased by 88% between 2012 and 2017. As of 2017, the average patient deductible was $1,820 and the out-of-pocket maximum was $4,400. In the same period, the Federal Reserve Board reported that 40% of Americans don’t have enough in savings to cover a $400 emergency. Patients are responsible for more but are not equipped to pay.
Healthcare is the fifth largest consumer expenditure category in the US. Although bucketed as a consumer expenditure, medical expenses are not a part of the average American’s monthly budget. They’re lumped together, but healthcare “consumers” are quite different than retail consumers. Buying a new car is a different prospect than paying for a hospital stay. It’s the difference between disposable income and emergency savings, planned purchases and surprise bills, a lifestyle choice and a confusing obligation. When it comes to payment strategies, looking at patients as retail consumers doesn’t work. We need a new, healthcare-specific approach.
Propensity to pay scores have become the gold standard in patient payment strategies. But a propensity to pay score is a point in time recommendation that relies on consumer data. A FICO credit score segments patients based on their ability to pay their car loan, their mortgage, their Macy’s credit card bill. This doesn’t match their ability and willingness to pay a medical bill. Most consumers with medical debts show no other signs of financial distress. Fifty percent of consumers who have medical trade lines, had previously “clean” credit reports with no past delinquencies. Most propensity to pay scores ignore basic attributes that impact whether the bill or at least a portion of the bill, will get paid. These excluded attributes include bill size, procedure and insurance type. Further, this basic scoring fails to provide recommendations for how to best collect from the patient.
A better method to build propensity to pay scores is with historical data. This type of modeling takes into account the uniqueness of provider facilities, procedures, payer mix, specialty and region. Scores and recommendations are based on how a specific provider’s patient payments perform. When you move away from current scores and look at historical patient payment data, propensity to pay accuracy increases. You also unlock the ability to determine how much each individual is likely to pay, which patients need options like payment plans and at what amounts, as well as the best method and cadence for patient contact. This new approach maximizes the dollars collected from each patient while enabling providers to offer patients increased access and flexible payment options.
Better propensity to pay scoring is only one part of an improved process for increased patient payments. Applying data science is part two.
Many likelihood to pay scores stop with the score. There is no reporting on payment outcomes and no continuous learning from the data. An effective patient payments management system should integrate a feedback loop. Who actually paid their bill? When? How much? This feedback loop drives *real* data science, giving providers a new level of insight into their payments and the means to improve propensity to pay scoring. A feedback loop enables future predictions that are more accurate, helps providers better understand patient populations and pinpoints workflow optimizations.
At a macro level, applying data science enables a provider to optimize their patient payments portfolio. Patients are segmented into risk groups and the key drivers that optimize payment outcomes for each group are identified. These types of data science tools for managing risk across consumer groups are table stakes in other industries. They’ve yet to be adopted at scale in healthcare. It’s time to move in that direction.
Patient responsibility will continue to grow. Maximizing patient payments will only become more critical. Providers who fail to develop targeted insights will be at a disadvantage. Healthcare needs a new model for patient payments, one that gives patients the flexibility they need to pay their medical bills and that helps providers maximize revenue. Data science delivers on both.
Message Justin Nicols on StartUp Health HQ or email siftmd@startuphealth.com. Learn more about Sift Healthcare.
Learn more about StartUp Health Moonshot Academy and our Cost to Zero Moonshot.
Sign up for StartUp Health Insider to receive industry news, funding insights, special access to events, and more delivered to your inbox every Wednesday.
