What medical record review actually costs a plaintiff firm

Almost every plaintiff firm handles medical records one of two ways. Either a paralegal reads the set and builds the chronology by hand, or the file goes to a vendor who charges by the page and sends back a summary. Nearly half of firms outsource at least some of it. (All the figures on this page are collected and sourced here.)

What is unusual is how rarely the two get compared on the same case, using the same assumptions. They are priced in different units — one in dollars per page, the other in days of somebody’s time — and units that do not match are units that do not get compared.

Medical record review cost when you outsource it

Vendors quote per page. A routine treating file runs to roughly a thousand pages, and the invoice for one lands somewhere in the high hundreds to low thousands depending on turnaround and whether you want a narrative summary or just an index.

Two things distort that figure in practice.

The first is that page counts vary far more than page rates do. A soft tissue case with one provider and eight weeks of treatment is not the same document as a case with six providers across fourteen months, and the second one is not six times the price — it is closer to fifteen. Firms budget from the rate and get surprised by the total.

The second is rush work. The cases where you most need the records read quickly are the ones where a deadline moved, and that is exactly when the rush multiplier applies.

The in-house number

In-house review feels free, because the salary is already being paid. It is not free; it is just already spent.

The arithmetic that matters is time. For a routine soft tissue case, most firms put chronology work at a day to a day and a half. For a treating file with several providers and a year of records, three to four days is common. The variance comes from provider count and legibility rather than page count alone — thirty pages of handwriting cost more than three hundred pages of clean billing ledger.

Take a fully loaded paralegal cost, divide it into working days, and multiply by the days per case. Most firms doing this for the first time find the in-house number is not dramatically lower than the vendor number, and on complex files it is higher.

The real cost is rarely the invoice. It is the case you declined because nobody had four days.

The number nobody puts in the comparison

Both of the above are costs of doing the work. Neither is the cost of not doing it well.

A treatment gap is in the file from the day the records arrive. Whether it costs you anything depends on which side reads page six hundred first. If the adjuster raises an eleven week break in treatment before you have accounted for it, you are negotiating from a worse position than the facts require — not because the gap is fatal, but because you did not know it was coming.

The same is true of billing that does not reconcile, a prior injury in the history, or a missing authorization that surfaces in discovery. These are all findable. They are findable by a person with enough hours, which is precisely the resource under pressure.

What changes when software does the first pass

The useful framing is not automation versus people. It is which pages a person reads.

A system that reads the set can produce a dated chronology with every entry cited to the page it came from, and flag the things worth a second look: gaps, providers appearing once, billed amounts that do not match paid amounts, dates that contradict the narrative. A paralegal then reviews a chronology instead of building one, and reads the twenty pages that were flagged rather than the thousand that were not.

The honest limits are worth stating plainly. Typed records, ledgers and structured forms read reliably. Handwriting ranges from fine to genuinely illegible, and anything claiming to handle all of it is overselling. Accuracy on the hard material is the thing to test before trusting, which is why every entry citing its source page matters more than any accuracy percentage — it makes checking cheap. There is a ten-minute procedure for doing that checking.

How to actually decide

Run the comparison on files you have already closed, where you know the answer.

1. Take ten recent cases across your normal mix, not ten easy ones. (Or start from the calculator.)
2. For each, record what the records actually cost you: vendor invoices, or paralegal days at a fully loaded rate.
3. Add the cases you declined or delayed because the work was queued. This is the number most firms have never written down, and it is usually the largest one.
4. Compare that annual total against what a build or a subscription would cost over three years, including keeping it working.

If the total does not comfortably clear the cost of building something, buy an off-the-shelf product or carry on outsourcing. That is a real outcome and it is the right answer for plenty of firms. Volume is what carries a build, and a firm running a modest caseload usually does not have it.

What we would push back on is the third line of that list being left out. A records process that runs at capacity does not announce itself as a cost. It shows up as a partner deciding, quietly and reasonably, that the firm cannot take on another treating case this month.