The clock is part of the decision: timeliness in California utilization review

Every year the California Division of Workers' Compensation publishes what amounts to a report card on utilization review. The most recent one is reassuring on the question the industry argues about most. In 2025, the independent medical review organization overturned 10.2% of the UR denials it reviewed, down from 12.7% the year before.[1] Read the other way: roughly nine in ten disputed decisions were upheld on their clinical merits.

That statistic hides a different exposure. A decision can be clinically correct and still be worth nothing, because in California the calendar is part of the decision.

Why timeliness outranks the merits

In its 2014 en banc decision in Dubon v. World Restoration (Dubon II), the Workers' Compensation Appeals Board set a bright-line rule: a UR decision is invalid, and outside the IMR process, only when it is untimely. Timeliness is a legal question for the Board, and if a decision is late, the Board itself may decide medical necessity on substantial medical evidence.[2]

The practical consequence is severe. A timely denial is evaluated inside a structured framework: MTUS guidelines, a specialist reviewer, a defined appeal path. A late one moves the question to a judge, where the careful clinical reasoning inside the report may not be relied on at all. In one post-Dubon panel decision, the Board held that because UR was untimely, no part of the determination could be used to deny a requested lumbar fusion.[3]

Judges have looked past the decision date as well. A 2024 panel decision records the trial judge's observation that a determination made on time but not served on time can still be treated as untimely, for example when the injured worker or their attorney is never served.[4] "On time" means the entire chain, not the reviewer's signature.

How the clock actually runs

The rules sit in Labor Code section 4610 and section 9792.9.1 of Title 8 of the California Code of Regulations. In operational terms:

• The clock starts at first receipt. It begins when the request for authorization is first received by anyone in the chain, whether the employer, the claims administrator or the utilization review organization. A request received before 5:30 p.m. counts as day zero, and the next day is day one.[5]
• Standard prospective and concurrent reviews get five business days from receipt of a complete DWC Form RFA.[6]
• Expedited reviews get 72 hours from receipt of the information reasonably necessary to decide.[6]
• Incomplete requests carry their own deadline. Within five business days the reviewer must either treat the request as complete or return it marked "not complete" with the reasons stated.[6]
• More information buys time, not unlimited time. The request for records must go out inside the first five business days, and the outer limit is 14 days from the treatment recommendation.[7]
• Notice is part of the deadline. After the initial telephone, fax or email communication, written notice to the requesting physician is due within 24 hours for concurrent review and two business days for prospective review.[6]

The rules keep tightening. Regulatory updates effective April 1, 2026 make explicit that a request accepted as complete is subject to investigation and administrative penalties, which raises the cost of waving a defective RFA through rather than returning it.[8]

The clock is a clinical variable, not just an administrative one

It is tempting to treat timeliness as paperwork hygiene. The occupational health literature suggests otherwise: in musculoskeletal injury, what happens in the first weeks shapes the next year.

In the Disability Risk Identification Study Cohort, workers with acute back injuries who received more than seven days of opioids in the first six weeks had roughly double the risk of work disability at one year, after adjustment for pain and injury severity.[9] In a national sample of workers with acute, disabling occupational low back pain, early MRI that was not clinically indicated was associated with far longer disability and average medical costs $12,948 to $13,816 higher than matched cases without early imaging.[10] A systematic review of that literature reaches the same conclusion: early MRI without red flags is associated with longer disability.[11]

Read those findings next to the UR clock and a different picture emerges. The early window is exactly when the system is deciding whether a worker gets prompt conservative care, or an unnecessary scan and a cascade of downstream intervention. A system that takes nine days to authorize physical therapy, or that reflexively approves imaging because the record was too thin to argue, is making a clinical decision whether it intends to or not.

What utilization review actually contributes

The oldest empirical study of UM in workers' compensation is still one of the most clarifying. Reviewing 9,319 workers' compensation cases managed by a large UM program, Wickizer, Lessler and Franklin found denial rates of roughly 2% to 3%, with many denials later reversed. The measurable effect came from length-of-stay management, averaging 1.9 fewer days.[12]

The lesson for modern operations: the value of review does not sit mostly in the denials. It sits in what gets requested, how quickly appropriate care is released, and how disciplined the process is. Which means the failure modes that matter are process failures, and process failures are timing failures.

Five practices that protect the clock

1. Timestamp at first touch, on every channel. Fax, email, portal and mail should all generate a receipt record the moment a request arrives, not when someone opens it. If you cannot prove day zero, you cannot prove day five.
2. Triage completeness on day zero. A same-day completeness check leaves room to either accept the request or return it with specific reasons well inside the five-day window. The 2026 amendments make a sloppy acceptance a penalty exposure.
3. Sort the queue by time remaining, not by arrival. Every open request should display its remaining business days, with expedited reviews visually separated. A queue sorted by arrival time hides risk; a queue sorted by time remaining exposes it.
4. Ask for missing information once, and precisely. A blanket request for "all records" invites a second round trip and burns the extension. Name the specific report, test or visit note the determination depends on, and send it in the first five business days.
5. Treat service as part of the decision. Track delivery to the physician, the injured worker and any attorney as its own deadline, with proof of service stored next to the determination. Dubon jurisprudence turns on the whole chain.

The takeaway

The IMR data suggests California's reviewers are applying the guidelines competently. The larger and quieter exposure sits in the handoffs around them: intake, records, queue discipline and service. Getting the medicine right is necessary. Getting it right on time, with proof, is what makes the decision hold, and what determines whether the worker gets the right care during the window when it still changes the outcome.

This article is general information about workers' compensation rules and practice, not legal advice. Check current statutes, regulations and case law for your jurisdiction.

Back to all articles