Burden Hours Explanation

OMB Question: Please explain why the burden hours is very different in this package. The burden for 2016 was 13,640 and now it is 1,150.

CMMI Response:

The LAN’s 2016 survey was the first effort of its kind, so we submitted a much higher number of burden hours than actually incurred due to the uncertainty about (1) the number of health plans that would voluntarily participate and (2) the length of time it would take for plans to complete the survey. For 2016, we estimated that as many as 220 health plans might participate in the survey; in fact, only 72 payers were surveyed by the LAN in 2016. In conducting the 2016 survey, the LAN was also able to calculate a more precise estimate of 23 burden-hours per survey, 8 hours less than the original 31 burden-hour estimate.

In addition, the LAN’s 2016 PRA application collected two measures from each health plan—a 2015 look-back measure and a 2016 point-in-time measure. These two measures required two responses from each plan, so we doubled the number of expected responses to 440 (220 health plans x 2 responses). When we executed the survey, however, we learned that both America’s Health Insurance Plan (AHIP) and Blue Cross Blue Shield Association (BCBSA) were conducting their own APM surveys and were only collecting the 2015 look-back measure, not the 2016 point-in-time measure.

These two organizations, AHIP and BCBSA, represent approximately200 health plans across the country. Because of the parallel surveys, the LAN’s 2016 survey collected responses from 70 plans and 2 Medicaid FFS states. In 2016, the LAN did include the AHIP and BCBSA collected information in an aggregated reporting of results rather than ask health plans to complete multiple surveys. This streamlined the data collection the LAN undertook, but ensured a high level of representation of the results through coordination with external partners.

For the LAN’s 2017 survey, we anticipate a similar process to the one that transpired in 2016: AHIP and BCBSA will conduct their own surveys, all surveys will contain just the one look-back measure, and the results will be aggregated and reported through the LAN.

In sum, there are three reasons that we have submitted a much lower burden-hour estimate this year than for 2016:

1. We did not include the 200 AHIP and BCBSA health plans in this year’s burden-hour estimate. We now anticipate a maximum of 50 responses (50 plans x 1 response), not 440 (220 plans x 2 responses).

   1. Our rationale is that AHIP and BCBSA are voluntarily conducting their own 2016 look-back measures, among other information, using their own resources and for their own purposes. The LAN will receive an aggregate number that will be aggregated with LAN results.

   2. The LAN conducted a market analysis to identify a smaller number of target plans that would ensure representation of the diversity of health plans as well as covered lives. As a result of this analysis and the trade association partnership, the LAN is targeting only 38 health plans to participate directly, plus approximately 12 Medicaid FFS states. We know not all plans invited will commit, thus we estimate a maximum of 50 direct participants.

2. We are collecting only one measure in 2017 (look-back measure).

3. We have a more accurate estimate of burden-hours per survey based on our 2016 experience: each survey should require an average of 23 burden-hours, not 31.