PEPPER relaunch wraps in September, and staff coders can now download the report themselves

The MLN Connects edition dated July 23, 2026 tells providers the Program for Evaluating Payment Patterns Electronic Report is relaunching "in the coming months for all Medicare facility types, including hospitals, post-acute care providers, and specialty facilities." The report itself is not new. What changed is who can pull it and when your facility type is up.

What actually changed since the December relaunch

CMS relaunched PEPPER for short-term acute care hospitals in December 2025. At that point, in CMS's own words, "only authorized officials (AOs) and access managers (AMs) with active Identity & Access Management System accounts can download their reports."

Staff end users can now download their organization's PEPPER through the portal once an AO or AM approves the PEPPER business function in the Identity & Access system. That is the practical change. The person who reads the report no longer has to route every request through the person who signs the enrollment paperwork.

The login is the same one you already use for NPPES and PECOS, and the Comparative Billing Report business function can be requested at the same time.

The release schedule runs through September 2026

CMS says it is relaunching PEPPER via ongoing releases through September 2026. The schedule it published:

After the relaunch, short-term acute care hospitals get a PEPPER quarterly. Every other facility type on that list gets one annually. So if you are a skilled nursing facility or a partial hospitalization program, September is the only release scheduled for you in this cycle, and the access approval has to clear before then.

The coding target areas flag under-coding too

This is the part worth reading twice. PEPPER draws a control limit at the 80th percentile for every target area, so anything at or above it is a high outlier. Coding-focused target areas carry a second control limit at the 20th percentile, and anything at or below that is a low outlier.

That lower limit is the under-coding signal, and it only exists on the coding target areas. Outlier status gets evaluated in priority order: nation first, then MAC jurisdiction, then state, because the state comparison group is the smallest.

CMS describes the target areas as discharges and services vulnerable to improper payments, including diagnosis-related groups at risk for under-coding or over-coding.

What PEPPER will not do for you

PEPPER does not identify the presence of payment errors. It tells you where your statistics sit against a comparison group, and the chart review is still yours. A target area above the upper control limit is a place to look, not a finding.

The report also will not hand you a patient list. CMS suppresses any numerator or denominator under 11, so small-volume target areas can come back blank, and you have to rebuild the population yourself from the numerator definition in the user guide.

Each report covers the most recent 12 quarters available, so you can watch a target area move instead of reacting to one period. Index Analytics LLC develops and distributes PEPPER under contract with CMS.

What to do with it on the coding side

Treat the outlier list as a work queue for a self-audit, not as a scorecard. Pull the target areas where you sit outside a control limit in either direction, sample charts, and check whether the documentation in the record supports what was billed. That is the same question a contractor asks when it pulls your charts, which is what turns up in reviews like the Novitas nursing-home Part B audit and the insufficient-documentation denial patterns on the fee-for-service side.

For a risk-adjustment team, PEPPER sits on the fee-for-service side of the house and reports nothing about your Medicare Advantage HCC capture. It is still worth a look if your organization bills both, because the sampling method is the same one you already run on a chart chase. Line the supporting notes up in the evidence builder before you sign off that a sampled chart was coded correctly.