
Pull any workers’ compensation file of decent size and find the operative report. Then count how many times it appears.
Three is common. Eleven is not unusual on a file that has been through a hospital release, a subpoenaed provider production, the applicant attorney’s own copy and a defense production, each with its own fax header and cover sheet.
None of this is anybody’s misconduct. It is what happens when four parties independently gather records from overlapping sources and nobody deduplicates before serving.
The scale
daisyBill’s analysis of California medical-legal bills that reported record review found 40.6 million pages across 50,738 bills. That works out to roughly 800 pages per evaluation on files large enough to trigger MLPRR at all.
In 2023, physician evaluators billed $93.2 million under MLPRR for just over 31 million pages beyond the base allowances, across about 131,000 bills.
An earlier daisyBill analysis attributed 29% of the increase in medical-legal expense to attorneys serving over 54 million pages of records on evaluators.
Those numbers describe pages served. They do not describe distinct medical content. Nobody publishes that figure, which is a shame, because it is the one that determines how long the reading takes.
Why you cannot just skip them
Two reasons, and both matter.
You attest to the pages under penalty of perjury. The attestation is about what you reviewed. Deciding that pages 340 through 372 are a duplicate of 118 through 150 without checking is a judgment you are making about documents you did not read, and occasionally the “duplicate” contains a handwritten annotation that the original does not.
And the record is the evidence. If a page appears in the file and you did not read it, opposing counsel gets to ask you about it. “Duplicate” is a fine answer when you verified it. It is a poor answer when you assumed it.
The honest way to handle it
Deduplicate by comparison, not by assumption.
That means someone identifies candidate duplicates by document type, date and provider, then confirms them page against page, and records the confirmed duplicates with their page ranges. What you end up with is a file where the medical content is read once, and every repeat is accounted for by page number.
Your attested page count does not change. You received what you received. What changes is that you are reading 300 pages of distinct content instead of turning 740 pages of paper, and you can say precisely what the repeats were if anyone asks.
Doing this manually, on the physician’s time, is not sensible at $3 a page. Which is why almost nobody does it, and why almost every QME in California is reading the same documents multiple times at rates that assume they are not.
The related problem: what counts as a page
Blank separator sheets. Fax cover pages. Certification affidavits from record custodians. Color photographs of a document. A single-line letter confirming no records exist.
These arrive inside attested productions and they are pages. They are also not medical content, and if a substantial share of a 700-page production is administrative padding, that is worth knowing before you plan your week.
There is no rule requiring the serving party to strip them. There is nothing stopping you from tracking what proportion of your incoming files is padding, and raising it with the firms that consistently send the worst productions.
What this is really about
The medical-legal fee schedule pays $3 a page for record review beyond the base allowances. That rate was set on an assumption about how long a page takes to read.
The assumption breaks when a meaningful share of pages carry no new information but still have to be handled. It breaks further when the file is unordered, so you cannot read chronologically and have to hold four partial timelines in your head at once.
You cannot change the fee schedule. You can change how the file reaches your desk.
Where CUBEXLE fits
We index and deduplicate medical-legal files before they reach the evaluating physician. Confirmed duplicates identified by page range, providers indexed, records placed in date order, and a medical glossary for the terminology you would otherwise be looking up at 11pm.
The page count you attest to stays exactly what you received. The number of pages you have to read to form an opinion is considerably smaller.
Send us the page counts on your last three files and we will tell you, without charge, roughly how much of it is repetition. Write to connect@cubexle.com.