About MedaRep

Territory intelligence for independent reps and distributors.

MedaRep shows who does the cases in your territory, who is already in the room, and who signs off. It is built on the same federal data the enterprise platforms start from, sized to one territory and priced for one rep or distributor, month to month.

Why

The data was never the expensive part.

Most of what a five- or six-figure data platform shows a device company starts in public federal files: Medicare claims, industry payments and provider enrollment. Those platforms are built for a whole sales organisation. An independent rep or a distributor working a few states needs the same picture, for their own surgeons, without the enterprise contract.

MedaRep does that work once: it joins the files surgeon by surgeon, models what they leave out, and keeps it current, from $39 a month.

How it stays affordable

Modeled from public data, checked in the field.

Enterprise platforms license commercial claims, and that bill rides in every seat they sell. MedaRep does not buy them. It starts from the record that is public and counted surgeon by surgeon, Medicare, and models the rest.

  • Modeled, with a range. Medicare Advantage is scaled by each county’s share, and commercial and other payers by each procedure’s patient age mix, checked against New York’s all-payer discharge records. Every estimate carries its range, and the models are scored on years they never saw (the figures below).

  • Checked by the people who know the surgeons. On any surgeon’s case figure a rep can say too high, too low or about right, or give their own number. That feedback is pooled by area to find where a commercial mix may need adjusting. One answer about one surgeon never changes a figure, and Medicare counts stay as CMS published them.

  • Right where it decides the call. Who to see first, who is growing, who just moved, who is likely the incumbent: the order of the surgeons and the direction they are heading carry those decisions. What a licensed dataset adds is an exact commercial count, which MedaRep shows as an estimate with its range, never as a count.

  • No data bill in the price. With no licensed claims to pay for, plans start at $39 a month, month to month, where an enterprise platform runs five to six figures a year by published estimates.

What it does

From territory to target.

  • Target list. Surgeons ranked by case volume across 66 procedure families in 18 device segments, with assistants and facility billers filtered out.

  • Who’s in the room. Every manufacturer already engaged with a surgeon, from royalties to lunches, so you know who is likely the incumbent before the first call.

  • Who decides. Groups, hospitals and ASCs, and the connections between surgeons, so you know who gates the door.

  • What’s moving. Surgeons who change groups, volume turning up or down, and where the patient population is heading.

Method

How the numbers are made.

Every figure is labelled reported, estimated or forecast where it appears. Behind the estimates and forecasts is a machine-learning pipeline built the way a quantitative research team would build one: trained on everything public, scored only on years it never saw, and kept only when it beats the simpler way.

  • Reported. Counts as CMS and the other publishers release them, with the year they describe. Each procedure family is a fixed set of billing codes followed through CMS’s code changes: where a code set was rewritten (hernia repair in 2023, for one), the old and new codes are joined so a renumbering never reads as a surgeon’s decline, and the models never learn from the break.

  • Algorithmic estimates. Statistical models fill in what the public files leave out: the low volumes CMS withholds for privacy, placed with the surgeons most likely to hold them, and the Medicare Advantage and commercial patients Medicare claims never show, scaled by each county’s Medicare Advantage share and each procedure’s patient age mix. The payer assumptions behind every all-payer figure have been checked against a state’s hospital outpatient discharge records. Each estimate is shown with a range.

  • Machine-learned forecasts. CMS publishes about two years late. Gradient-boosted decision trees forecast each surgeon’s cases for the years since, and a second model scores the chance they stop billing Medicare here, reading the newest Medicare enrollment records as early signals of a move or a retirement. Both learn from every surgeon who ever billed a case in the loaded states, including the ones who have since left, so they learn what leaving looks like, not only what staying looks like.

  • Out of sample, every time. Each model is scored in rolling backtests that replay every year as if it were the present, trained only on what was known then. Candidates go head to head (linear models, random forests and other boosting libraries, other objectives, new data signals) and only what improves out-of-sample accuracy ships. The measured errors set the ranges you see.

  • Trends that mean something. A surgeon’s change in Medicare cases is read after removing their county’s shift from traditional Medicare to Medicare Advantage, which otherwise looks like a decline in every growing market.

  • Checked every release. Every data release is compared with the one before it and kept only if it passes, and every region is refitted before its forecasts are shown. Treat the range, not the point, as the claim.

No patient-level data is used, and no figure claims more than the files can show: shared groups and hospitals suggest who knows whom, but no public file records referrals.

Sources

Where the data comes from.

Everything is published free: by the federal government, by hospitals under federal price-transparency rules, and by OpenAlex.

Medicare claims and enrollment · CMS
Procedure volumes by surgeon, hospital and setting; group practices, affiliations and where clinicians bill.
Industry payments · CMS Open Payments
What device makers report about their engagement with physicians: the dollars, and what for.
Identity · NPPES and PECOS
Names, specialties, practice addresses and group membership.
Devices · FDA
510(k) clearances and PMA approvals, recalls, and adverse-event reports (MAUDE) by product code and manufacturer.
Companies · SEC EDGAR
Revenue, R&D and earnings of the public device makers, from their filings.
Facilities and prices · CMS and each hospital
Hospitals, ASCs, ownership and systems; negotiated rates from price-transparency files.
Population · U.S. Census Bureau
Residents by county and age, for demand and the outlook.
Publications · OpenAlex
Research output, for the surgeons who publish.

Every file, with attribution

Who built it

Built from a decade in medical device intelligence.

MedaRep was built and developed with a decade of medical device intelligence experience: turning claims, payments and market data into the territory plans, target lists and competitive reads that field teams sell from. That is also where it became clear what the enterprise platforms cost, and how much of what they show comes from public files that a rep working a few states could never afford to have joined for them.

MedaRep is that work, done once and kept current. The figures are built the way you would want them built if your own quota depended on them: labelled for what they are, checked against what was later published, and honest about what the public data cannot show.

Where it is going

From the surgeon to the whole market.

Today MedaRep is a synthesis of CMS data wired into the sales cycle: who to call, who is already there, and what is changing. The same database already holds the device side of the market, from FDA clearances, recalls and adverse-event reports to the public device makers’ filings, and the next releases connect the two: which products and companies are moving in a territory, where a recall may disrupt supply or service at a likely incumbent, and what that may mean for the surgeons it engages.

See your own territory.

Free to look. Paid plans start at $39 a month.