~130 Days from ASM Starting, do you know what's in your cost measure?
Published on APM Connect | August 2026
CMS has not yet posted the final list of clinicians who will be in the Ambulatory Specialty Model on January 1, 2027, 133 days from now, the ASM Participants data site still holds only the February preliminary file. But the two cost measures the model runs on are not a mystery, and they are not new. ASM's cost category uses the existing MIPS episode-based cost measures for heart failure and low back pain, and the 2026 performance-period Measure Information Forms for both have been public on the QPP site since December 2025. They are attached to this post. Those same two measures also decide who is in the model at all: ASM selection turns on having at least 20 attributed heart failure episodes or 20 attributed low back pain episodes in a twelve-month period, as measured by these EBCMs. The list is late. The rulebook that produces it has been sitting in public for eight months.
Read the forms and one thing reshapes how you should think about ASM: these measures track a relationship, not a procedure. Two Part B claims from the same NPI open an attribution window, within 180 days of each other for heart failure, within 60 days for low back pain and from then on a single reaffirming claim extends your accountability, by a full year for heart failure and by 120 days for low back pain. It takes two claims to start the clock and one ordinary follow-up visit to restart it. Episode windows accordingly run from 365 up to 729 days for heart failure and from 120 up to 484 days for low back pain. The denominator is not episodes or patients but days, so a patient you have managed for two years carries far more weight in your score than one you saw twice.
Both measures cap SNF costs at 30 days and IRF and LTCH costs at the 90th percentile, and both risk-adjust off a specific condition list. Review the measure specifications prior to 2027.
Access the most updated ASM measure specifications here: Heart Failure Measure &
Comments