- Legal
The IMR Record Package: Why California Utilization Review Appeals Fail on Documentation, Not Medicine – Copy
Table of Contents
Ineligibility is climbing, and 2026 is the worst year since 2016
Through June 2026, the IMRO received 89,725 unique applications and found 79,683 eligible, according to the DWC IMR Update for the second quarter of 2026. That puts ineligibility at 11.2% on a half-year basis. If the pace holds, 2026 will be the worst year for screened-out applications since 2016.
| Reason for ineligibility | H1 2026 | Full-year 2025 |
|---|---|---|
| Conditionally non-certified (CNC) UR decision | 6,181 | 7,492 |
| Untimely filing | 1,328 | 2,481 |
| No UR report attached | 1,318 | 2,616 |
| No signature | 618 | 1,022 |
| Other | 546 | 1,161 |
| No signature and no UR report | 51 | 88 |
| Total | 10,042 | 14,860 |
CNC is a records failure that starts at the RFA
The April 2026 amendments to 8 CCR 9792.9.5 help here. A denial letter must now identify the missing information and state that the request will be reconsidered on receipt of a new RFA containing it. Read that paragraph as a work order. Our breakdown of the AB 1293 changes to QME timelines covers a related set of documentation deadlines that shifted in the same period.
The honest counterargument
The California Workers’ Compensation Institute examined 1.57 million IMR decision letters issued from 2015 through March 2025. In the twelve months ending March 2025, CWCI found that the top 1% of requesting physicians, 81 doctors, accounted for 42.2% of all disputed service requests that went through IMR. The top ten individual physicians accounted for 10.9%. Seven of those ten appeared on the same list the previous year.
Two readings fit. One is that a small group of outlier prescribers requests treatment the guidelines do not support, and IMR is working as designed. The other is that a small group of practices has a systematic process problem in how requests are documented and supported, and repeats it every year.
Roughly three weeks of every IMR is spent waiting for records
The six-month window is the most misread rule in the process
If the clinical basis sits in a report seven months old, in an operative note from a prior surgeon, or in a QME report predating the window, it will not appear in the administrator’s submission. It reaches the reviewer only if you send it. This is the same gap that shows up in medical records review across a workers’ compensation file generally.
Your clock is the same as theirs
Two numbers most practitioners have wrong
The overturn rate understates available relief.
The headline figure for 2025 was 10.2% of treatment requests overturned, down from 12.7% in 2024. Measured at the case level in DWC's Q4 2025 regional sample of 37,872 cases, dispositions ran 87.3% upheld, 6.9% overturned, and 5.8% partially overturned. Close to one case in eight produced some authorization. Partial overturns rarely appear in the commentary, and they change how a package should be built, because carving a disputed request into separable components creates more than one opportunity for a favorable finding.
Expedited review is almost entirely unused.
Of 152,351 final determinations issued in 2025, DWC reports that 108 were expedited. That is under one in a thousand. Expedited cases averaged two days from receipt of records to decision. Where a worker faces an imminent and serious threat to health under 8 CCR 9792.6.1(j), and the treating physician will certify it, the process exists and virtually nobody uses it.
For claims professionals tracking reversal rates
The CNC figures are the ones worth watching. A conditional non-certification represents a UR cycle that produced no necessity determination, a treatment decision deferred rather than resolved, and in many cases an IMR application filed and screened out. Rising CNC volume is a cycle-time and administrative-cost signal well before it becomes a reserve question, a dynamic we cover further in our analysis of how record quality affects TPA reserve accuracy.
The WCAB is not the backstop
A build checklist
- Confirm the UR decision is not CNC. If it is, work the RFA rather than the appeal.
- Diary from the RFA date. The six-month window runs backward from there.
- Calendar the deadline: 30 days from the mailing date of the UR decision, or 10 days for formulary-drug-only disputes. Filing by mail extends these to 35 and 15 days, and mailing from outside California extends them to 40 and 20 days.
- Check the signature and attach the UR letter. Those two defects accounted for 1,987 ineligible applications in H1 2026.
- Read the administrator’s document list against your file. Anything omitted is yours to supply.
- Pull the material outside the window: prior treater reports, operative notes, QME and AME reports, imaging reads, functional findings.
- Confirm the treating physician expressly addressed guideline prerequisites. Under amended 9792.9.5, where the requesting physician opines that a prerequisite should be disregarded, the reviewing physician must explain why that reasoning is insufficient.
- Assess whether the request separates into components that could draw a partial overturn.
- Ask whether the case qualifies for expedited review.
- Submit under 9792.10.5(b) rather than assuming the file is complete.
The capacity problem behind all of this
Rapid Care MRR’s medical record review platform was built for that volume: organizing large record sets into usable chronologies, surfacing what falls inside and outside a date window, and flagging gaps while there is still time to fill them.
Outcomes in this system are close to uniform. The file is the variable.