NeoGraph Analytics
Healthcare IT

Why two forecasts of the same market differ by ten times

Across ten markets we sized this month, base-year estimates spread up to 24x while growth rates spread barely at all. The gap is almost never a disagreement about the future — it is a disagreement about what was counted.

NARD
NeoGraph Analytics Research Desk
••8 min read•Pune, India

Ask four research houses what a healthcare market is worth this year and you can get four answers an order of magnitude apart. The instinct is to average them, or to pick the one from the most familiar brand. Both are mistakes, and the reason is visible in the numbers themselves.

The pattern: bases disagree, rates agree

From ten markets we sized this month, each measured at the same base year across publishers:

MarketSpread at the base yearSpread in published CAGR
GLP-1 receptor agonists24.2x8.47% - 17.5%
Spatial biology16.9x—
Health data interoperability14.6x—
Point-of-care diagnostics11.0x1.37x (5.1% - 7.0%)
Real-world evidence10.0x2.32x
Liquid biopsy4.36x2.55x
Bispecific antibodies3.49x—
Radiopharmaceuticals1.78x3.35x (4.9% - 16.4%)
Neuromodulation1.52x2.77x (4.4% - 12.17%)
CAR-T1.19x2.97x (14.6% - 43.3%)

Read the first and fourth rows together. In point-of-care diagnostics, published base-year values differ by eleven times while the growth rates differ by 1.37 times. The houses broadly agree about how fast the thing grows and disagree violently about what the thing is. That is a definitional disagreement wearing a numerical costume.

Rows six through ten invert it, and that inversion is its own finding. In radiopharmaceuticals the bases agree to 1.78x while the rates spread 3.35x — because the rate is where a forecaster encodes an assumption about product mix, in this case how fast therapy displaces diagnostics. In neuromodulation, a 2.77x spread in rates over a 1.52x spread in bases is a disagreement about whether external stimulation counts, expressed as growth.

How wide the definitions actually are

Three cases we met this month, each checkable:

One publisher at both extremes. In health data interoperability, Credence Research publishes a market reaching roughly $10.63 billion by 2032 in one report and roughly $349.9 billion by 2032 in another — a 33-fold gap inside a single catalogue. Neither number is a lie; they are answers to different questions, sold under near-identical titles.

A subset larger than its superset. In spatial biology, one house's narrow-scope figure exceeds its own wider-scope figure by 5.4x. Another publishes a $1.48 billion base at a 19.3% rate, which compounds to $8.64 billion — not the $10.2 billion endpoint printed alongside it, which implies 21.3%. The series does not close on itself.

Agreement that is not corroboration. In decentralised clinical trials, four houses land within 1.07x of each other at the base year. That reads as consensus until you notice there is no independent way to measure the category: only 37 of 607,160 studies registered on ClinicalTrials.gov text-match "decentralized clinical trial", and the FDA retitled its own guidance in September 2024 to drop the phrase in favour of "decentralized elements". Four houses agreeing closely about an unmeasurable category is more likely a shared derivation than four independent measurements.

What to do instead of averaging

Divide. Every market has something countable underneath it, and the per-unit number tells you which scope you are actually looking at:

  • Point-of-care diagnostics. The CMS CLIA database recorded 233,448 Certificate of Waiver and 21,736 provider-performed microscopy sites in September 2026. The highest published market size over those 255,184 sites is $208,087 per site per year — about $570 a day of waived testing per site. That rules out a professional-sites-only reading immediately.
  • Health data interoperability. The Sequoia Project lists 11 Designated QHINs. The highest published figure works out at $9.4 billion per QHIN per year, which is more than Epic's entire reported company revenue (~$6.7 billion for 2025) — and Epic's Nexus is one QHIN among eleven.
  • Liquid biopsy. Guardant Health's FY2025 oncology revenue of $683.6 million across roughly 276,000 tests is $2,477 per test. The top published 2034 forecast divided by that price implies 43 million tests a year, against roughly 20 million annual cancer cases worldwide. The forecast is not about therapy selection; it is an unstated bet on population screening.
  • Hospital-at-home. CMS has approved 373 hospitals across 140 systems, and the programme has served about 31,000 patients in four years. Published market sizes imply between 202,000 and 811,000 episodes annually.

None of these divisions produces a market estimate. Each one rules something out, which is more useful than a fifth estimate: it tells you which of the published figures cannot be describing the market you care about.

The headline is a scope decision

The conclusion we keep arriving at is uncomfortable for the format: a report whose headline is a single number is reporting the analyst's scope decision, not a finding about the market. The useful output is the range, the reason for the range, and the arithmetic that eliminates part of it.

So that is what our reports lead with. Every base year is a named publisher's own figure advanced at that publisher's own rate, said so on the page. Every headline is divided through by something a reader can count. Where a split is not disclosed — regional shares, country shares, vendor shares — the section is absent rather than apportioned, which is why our competitive sections name players and describe position in words.

How we produce our figures · Browse the research

Sources

  • CMS, CLIA laboratory statistics release, September 2026 (301,284 registered laboratories; 233,448 Certificate of Waiver; 21,736 provider-performed microscopy).
  • The Sequoia Project, Designated QHIN listing, retrieved 16 October 2026 (11 QHINs).
  • ClinicalTrials.gov, study counts retrieved 16 October 2026.
  • FDA, decentralized-elements guidance retitling, September 2024.
  • Guardant Health, FY2025 results, 19 February 2026.
  • CMS Acute Hospital Care at Home approval list; Chartis and American Medical Association reporting on the five-year extension, February and March 2026.
  • Published market estimates as cited in each report, with publisher, date and stated scope.

Spreads above are computed from the published estimates tabulated in the individual reports; the full source lists, and the figures we examined and chose not to publish, are in the research notes behind each one.

Get Expert Insights

Fill out the form and our team will get back to you within 24 hours.

0 + 0 =

Please solve this simple math problem to verify you are human

Quick Contact

24×7 Sales & Support Available