S1E3: Can Prior Authorization Stop Ambulance Taxis? (Riley League)

Listen on Apple Podcasts

In today’s episode, we discuss the impacts of litigation and prior authorization on the use of ambulances as taxis in the dialysis industry.

My guest is Riley League, a Ph.D. candidate in Duke’s Economics department. Riley works with Professor James Roberts, the Department Chair of the Economics Department at Duke and has worked on several projects related to the dialysis industry. I’m excited to learn from him about the convoluted world of the dialysis industry and how he and his coauthors are untangling its effect on patients and healthcare costs.

The paper we discuss can be found here.

In-depth show notes

00:29 – 03:10 Introduction to ambulance taxis and our guest Riley League.

03:11 – 12:19. Improper payments by the Centers for Medicare and Medicaid Services (CMS) take up a substantial part of the CMS budget but are difficult to study because of the fragmented nature of their enforcement across different districts. We discuss the dataset of Medicare fraud cases that Riley and coauthors compiled and hone in on exactly how bad actors in these cases used ambulance rides to defraud CMS.

Improper payments by CMS totaled 28.9 billion dollars in 2019, the equivalent of 7.3% of all CMS spending that year
Riley and coauthors compiled a dataset of dialysis using Public Access to Court Electronic Records (PACER) data and looking for court records and fraud news on search engines
Medicare policy specifies that ambulances should only be paid for by Medicare if it is the only safe way for a patient to travel, e.g. patient is bedridden or needs medical attention en route
How bad actors systematically and intentionally pocketed improper Medicare payments by providing ambulance rides for patients that did not require them
Strategies for recruiting patients to receive these ambulance rides: waiving copays, encouraging patients to spread the service through word-of-mouth
How large dialysis providers may use similar strategies to steer patients towards private insurance plans, which in turn pay significantly more for dialysis than Medicare

12:20 – 31:19. Riley and coauthors used their dataset and a two-way fixed effects model to answer the following questions: what is the effect of criminal and civil lawsuits on the use of non-emergent ambulance rides, and how does that compare to the effect of prior authorization.

How prior authorizations can delay care but may help combat wasteful or improper healthcare spending
Discussion of the two-way fixed effects model, a variation of difference-in-differences that allows for multiple points of treatment to begin
Fixed effects are ways to capture natural variability in space and time, enabling researchers to more clearly see additional changes attributable to interventions
Applying the analysis plan to a case study of payments for ambulance rides in the Pennsylvania East District
Addressing the assumptions of parallel trends and no exogenous shocks
Using placebo tests to demonstrate that the results are not due to random chance

31:20 – 34:08. Ultimately, they found that civil litigation tended to have very little impact on payment for non-emergent ambulance rides. Criminal litigation had some impact, but prior authorization had the most impact on reducing payments for non-emergent ambulance rides.

These conclusions may not have been reached without the two-way fixed effects model, highlight potential pitfalls of simpler models or studies that only sample one site
How their two-way fixed effects model allowed Riley and coauthors to get at causal inference in a situation where a randomized controlled trial would not be possible

34:09 – 39:52. We conclude with a discussion of whether reducing improper Medicare payments for ambulance taxis led to worse patient outcomes.

Importance of ensuring prior authorizations only reduce the use of services in fraudulent or improper cases, not among patients who really need them
In this case, patient data on dialysis use, hospitalizations, and mortality allowed Riley and coauthors to conclude that patient outcomes were not worse following reductions in payments for non-emergent ambulance rides

References