The Impact of Reference-Based Pricing for Imaging Services on Choice of Provider and Spending in a National Health Plan

Health insurers are increasingly using benefit design to channel patients to lower-priced providers. With reference-based pricing (RBP), patients pay the difference between a defined contribution (called the reference-price) and the provider’s price. Because the patient is now responsible for the “last dollar” versus the “first dollar” (as is the case with deductibles), the hope is that patients will focus on where to receive care and not whether to receive any care at all.

In prior studies patients in RBP plans are more likely to switch to lower-priced providers, and this is associated with lower spending. Possibly most striking was that many providers lowered their prices in response to the program. However, the studies have been limited to a small number of large employers. Because these employers often have significant market power there is concern that the prior work may not be generalizable.

In this paper, we examine the effect of RBP for imaging services offered by a large national health plan across the U.S. The intervention population are employees from three self-insured firms of different sizes that introduced RBP for advanced imaging services (CT scans and MRI scans) beginning in January 2015. The firms did not give enrollees a choice between plans with and without RBP which avoids bias from enrollee selection. Our control population was a random sample of members from self-insured plans with similar benefit design but no RBP.

Using medical claims data from 2014-2016, we selected all claims for advanced imaging procedures in the outpatient setting (n=208,674 procedures). We analyzed the proportion of imaging per quarter at a provider with a price below the reference-price, and average total spending per procedure using linear and logistic regression models that control for enrollee age, gender, state and time (quarterly) fixed effects. Standard errors clustered on member in all models.

In the second year after the introduction of RBP for imaging services, enrollees in RBP plans were significantly more likely to choose a lower-priced provider than enrollees in plans without RBP (3.7 percentage points or 8% effect). Price per advanced imaging service for RBP enrollees was $65 lower than that in the control population, or 6% relative to the RBP group baseline price. After the introduction of the RBP program there was no relative change in utilization of high-priced imaging in the intervention and imaging control populations. We also report on whether there is differential patient response by health status, stratify the response by plan sponsor size within regions, and test the sensitivity of findings to different measures of geographic region (e.g., Hospital referral region).

Implementation of a RBP program among three self-insured firms was associated with a 6% reduction in per-imaging procedure price. This reduction is more modest than what has observed in prior studies of RBP.