Revenue measures how much a company sold last year; in medical testing it says very little about whether a hospital can keep buying from that company next year.
The reason is regulatory. An instrument that has not been cleared cannot be sold at all, and every additional assay, imaging modality or software function needs its own clearance. Siemens Healthineers, Roche Diagnostics and Abbott each hold thousands of individual registrations across the U.S. FDA 510(k) and De Novo pathways, the EU In Vitro Diagnostic Regulation (IVDR) and China's NMPA system — an asset accumulated over decades that no amount of capital can compress into a product cycle. A cleared assay menu is therefore closer to a licence than to a product.
The VerityRank Clinical Clearance & Installed-Base Index scores five weighted dimensions:
• Regulatory Clearance Depth (30%) — breadth and currency of FDA, CE-IVDR and NMPA registrations across imaging, immunodiagnostics, clinical chemistry, molecular testing, hematology and endoscopy.
• Installed-Base Lock-in (25%) — instruments actually running in clinical service and the cost to a laboratory of replacing them.
• Diagnostic Menu Breadth (20%) — how far a brand spans the major testing disciplines.
• Clinical Evidence & Guideline Standing (15%) — peer-reviewed validation, professional-society guideline inclusion and companion-diagnostics programmes.
• Global Service & Reimbursement Footprint (10%) — countries with direct installation, application support and reimbursement.
Each brand receives a Composite Brand Score (0-100), normalised against peers and refreshed annually from audited filings and regulatory registries. Quality and environmental management certification — ISO 13485 for medical device quality systems, ISO 14971 for risk management, ISO 15189 for medical laboratory competence, IEC 60601 for electrical safety and MDR 2017/745 / IVDR 2017/746 for European market access — is treated as a baseline qualification, not a differentiator, because no serious manufacturer can bid without it.
Disclaimer: This ranking is compiled from third-party authoritative sources including audited annual reports, regulatory registries and industry research. VerityRank is independent and receives no compensation from any company for inclusion, exclusion or position. Fiscal year ends differ between companies, and where divisions are not separately audited the parent group's disclosed segment figure is used and labelled as such.
The test is simple: strip away the box and ask what is left. A franchise leaves behind reagents, software, service contracts, a trained operator base and clinical evidence. An assembly business leaves behind a bill of materials.
Closed consumable ecosystems are the clearest marker. Roche Diagnostics pairs its cobas platforms with proprietary reagents and reported CHF 13.8 billion in division sales for 2025, with pathology alone growing 14% to CHF 1.7 billion. Sysmex holds more than 50% of the worldwide blood cell analysis market on the same model, which is why it can lose 40.8% of operating profit in a single year and still generate JPY 51.8 billion of it on JPY 500.0 billion of sales.
Vertical control of the hard components is the second marker. Siemens Healthineers builds its own superconducting magnets, detector chips and imaging software; Olympus grinds its own lenses and designs its own CMOS sensors, which is how it defends roughly 70% of the global gastrointestinal endoscope market; United Imaging claims full self-development of superconducting magnets, gradient amplifiers and PET crystals. GE HealthCare produces roughly two-thirds of its global CT output at its Beijing plant and over 40% of its ultrasound systems at Wuxi — vertical scale expressed as geography.
Service density is the third. A core laboratory or an endoscopy unit cannot tolerate a scanner that is down for a week. Roche and Sysmex both run direct service organisations across more than 150 countries, and Sysmex derives 88.3% of sales outside Japan — revenue that only exists because the service footprint exists. Abbott operates more than 100 manufacturing and R&D facilities on the same logic.
Conversely, a company that buys sub-assemblies, rebadges instruments and has no reagent chemistry of its own competes on price and lead time alone. That is a defensible position in low-acuity segments, but it cannot survive a hospital tender that weighs total cost of ownership across a seven-year instrument life.
A competitor can reverse-engineer a scanner in eighteen months. Replicating a validated, reimbursed assay menu takes a decade — which is why the menu, not the hardware, is the real barrier in medical testing.
Every individual test a laboratory can order must be separately cleared or approved, with its own analytical validation, reference ranges, interference studies and — critically — clinical utility evidence. Roche Diagnostics introduced two instrument platforms, six digital solutions and 53 new tests in 2025 alone, on top of an existing menu that runs into the hundreds. Abbott's diagnostics segment generated US$8.937 billion in 2025 across core laboratory, molecular, point-of-care and rapid diagnostics, with core laboratory growing 3.6% organically in the fourth quarter even as the segment faced a COVID base unwind.
Three compounding effects make menus self-reinforcing. First, hospital laboratories standardise on a single vendor to avoid running parallel quality-control programmes and duplicate training — so every added assay deepens the lock. Second, once a test enters a clinical guideline, switching suppliers means re-validating against that guideline, which laboratories are reluctant to do without a clinical reason. Third, companion diagnostics tie a test to a specific drug programme: Roche pairs its assays with pharmaceutical partners so the test and the therapy become commercially linked.
Molecular and pathology are where the menu barrier is highest. Roche's pathology franchise grew 14% to CHF 1.7 billion in 2025, supported by the US$1.05 billion acquisition of PathAI and the US$595 million purchase of SAGA Diagnostics for molecular residual disease monitoring — acquisitions that buy digital pathology algorithms and ultra-sensitive monitoring assays rather than capacity. Sysmex made the same calculation from the opposite direction, acquiring JEOL's clinical chemistry business in September 2025 to bolt a second discipline onto its hematology franchise.
For a hospital, the practical lesson is that instrument specifications are the least durable part of a purchasing decision. The menu, the reagent supply chain, the quality-control software and the service contract determine what the laboratory can actually report five years later.
For two decades China was the growth story in every medical testing earnings call. In 2025 it became the reason profit fell faster than revenue at almost every major brand.
Volume-based procurement (VBP) applies centralised, price-competitive tendering to reagents and consumables that previously carried premium margins in a closed-system model. DRG/DIP payment reform compounds the effect by paying hospitals a fixed amount per diagnosis, which turns every test into a cost line rather than a revenue line. Domestic-preference rules then steer public-hospital tenders toward local manufacturers. The result is a market that keeps buying instruments while paying materially less for the reagents attached to them.
The damage is visible in the numbers. Sysmex, whose hematology franchise is heavily exposed to Chinese hospital tenders, saw operating profit drop 40.8% to JPY 51.8 billion while revenue fell only 1.7% to JPY 500.0 billion; China accounted for JPY 89.5 billion, or 17.9%, of its sales. Roche reported its Diagnostics Division growing 2% at constant currency but declining 3% in Swiss francs, explicitly citing the impact of healthcare pricing reforms in China on its immunodiagnostics business. Olympus said Chinese fourth-quarter volumes were pulled down by heightened competition and policies favouring domestically manufactured products. GE HealthCare, Philips and Abbott all flagged similar Chinese tender pressure.
The Chinese leaders felt it worst because they sit inside the reform. Mindray, the country's largest device maker, reported revenue of RMB 33.28 billion, down 9.38%, and net profit down 30.28%, with domestic revenue falling 22.97% while overseas revenue grew 7.40% to RMB 17.65 billion — 53.03% of total sales and the first year international revenue exceeded China. United Imaging went the other way, growing total revenue 33.98% to RMB 13.80 billion with overseas revenue up 51.39% to RMB 3.43 billion, as domestic equipment-renewal subsidies offset tender delays.
The strategic response is now uniform across the industry: multinationals localise manufacturing in China for China, while Chinese champions export into Europe, the Middle East, Southeast Asia and Latin America. Both halves of that movement compress margins — one by duplicating supply chains, the other by competing on price in unfamiliar service territories.
AI in medical testing is real, but it is not yet a moat. It is currently a feature that protects an existing franchise rather than a franchise of its own — and the difference matters enormously for how these companies are valued.
Where AI already earns its keep is in the reading loop. Olympus positions ENDO-AID as an assistive lesion-detection system that highlights suspected early colorectal lesions during live colonoscopy — valuable precisely because it is bound to the EVIS imaging platform, so it strengthens the endoscope franchise instead of competing with it. Siemens Healthineers has pushed AI into scan acquisition and reconstruction, and deepened its clinical work through a ten-year strategic technology alliance with the Cleveland Clinic covering AI-enabled precision imaging and digital-twin applications. Roche has taken the same approach to pathology, buying PathAI for US$1.05 billion to embed image analysis into its tissue-diagnostics business.
The economics, however, still lead back to consumables. AI does not generate its own revenue at scale in diagnostics; it increases utilisation of the instrument, the reagent and the service contract that already exist. Philips illustrates both the opportunity and the trap: it invested heavily in AI-enabled imaging and reported €17.8 billion of group sales in 2025 with an adjusted EBITA margin of 12.3%, but its Diagnosis & Treatment businesses delivered 0% comparable growth for the full year, because software differentiation does not override hospital capital-budget cycles or Chinese procurement policy.
Regulatory reality further limits the moat. An AI diagnostic function needs its own clearance — the FDA's device software pathways in the United States, CE marking under MDR 2017/745 in Europe and NMPA approval in China — and once cleared, algorithmic performance must be demonstrated on the local patient population, not just the training cohort. That is a genuine barrier for small vendors, but it is a barrier incumbents clear far more easily than newcomers.
The honest read. AI raises switching costs modestly for brands that already own the installed base, and it is becoming table stakes for premium imaging and endoscopy platforms. What it has not yet done is create pricing power independent of the instrument, reagent and service revenue that medical testing brands have monetised for decades. Investors and hospital procurement teams should treat AI capability as a necessary feature of a modern platform — not as a separate reason to pay a premium.
