Global IVD Reagents and Consumables Market Report, Size, Share and Forecast 2026-2033: The Instrument Is a Financing Arrangement

Contents

1. Market forecast snapshot, 2025–2033
2. Research scope and definitions
3. The thesis: the reagent is the product, the instrument is a financing arrangement
4. Market size construction and attribution
5. Product segmentation: nine classes of consumable
6. Regional structure
7. End-market and channel structure
8. Price formation: three regimes, one P&L
9. Volume, test economics and realised revenue per test
10. Competitive landscape
11. Value chain and margin distribution
12. Market risk and sensitivity
13. Methodology and editorial sourcing

Market Forecast Snapshot, 2025–2033

Research scope and definitions

This report measures the reagent and consumable layer of in vitro diagnostics: assay reagents and their kits, calibrators and controls, quality-control materials, single-use cartridges and test strips, blood-culture bottles and AST panels, culture media, histological stains and antibodies, slides and cassettes, cuvettes, sample cups, pipette tips, tubes and specimen-collection consumables.

It excludes instruments, analysers, track systems, middleware and laboratory software, and it excludes service, maintenance and reference-laboratory testing services. Those are the instrument and service layer, which this report sizes only as a remainder: USD 29.6 billion of the USD 108.0 billion IVD base in 2025, or 27.41%.

How this report relates to the rest of the cluster. The layer is additive with the instrument and service layer and not additive with the four discipline reports already published under the IVD hub. Global Immunoassay measures a USD 26.4 billion discipline including its analysers; this report measures the USD 21.6 billion of immunoassay reagents inside it. Summing the two would count the reagent twice. The correct arithmetic is horizontal: 78.4 + 29.6 = 108.0, the IVD base.

The thesis: the reagent is the product, the instrument is a financing arrangement

Two companies published their revenue by category last year and disagreed by 21.8 percentage points about what business they are in.

Sysmex, in the year to March 2026: reagents JPY 307,923 million, instruments JPY 103,232 million, service JPY 69,275 million, other — which it defines as including consumables — JPY 19,575 million, total JPY 500,006 million. Reagents are 61.58% of net sales. Reagents, service and consumables together are 79.35%. Its own twenty-year series makes the point better than any table: ongoing sales rose from 66.4% to 79.5% of revenue, a 13.1-point drift in the direction of the annuity.

bioMérieux, calendar 2025: reagents EUR 3,393.2 million, equipment EUR 282.3 million, services EUR 254.1 million, equipment rentals EUR 61.9 million, other revenue EUR 78.3 million, total EUR 4,069.8 million. Reagents are 83.38% of revenue. Equipment is 6.94%.

Both companies sell a closed system. Both are paid per test. Both put the same physical product in the same kind of laboratory. The difference is that bioMérieux rents its instruments and books the reagent share embedded in that rental as rent — its own note says the equipment-rental line includes revenue reclassified from reagents — while Sysmex sells analysers and tracks outright and books them as instruments. The razor-and-blade ratio is not a property of the industry. It is a property of the contract.

That matters because the ratio decides what a price cut does to a P&L. A vendor at 83% reagent content takes an administered price reduction on 83% of revenue with its instrument cost already committed. A vendor at 62% reagent content takes it on 62%, and has an instrument line it can still price, bundle or place at a margin. The same tender, two different outcomes, determined before the tender is announced.

Market size construction and attribution

The USD 78,400 million level is built bottom-up. Nine vendors disclose a reagent, consumable or recurring-revenue figure, or disclose a divisional revenue to which a stated reagent intensity is applied. The intensities are printed rather than buried, because four of the five largest diagnostics companies in the world do not publish a reagent split at all.

A 51.24% residual is the largest in this cluster, and it is the mirror image of the finding in haematology, where 87.72% attributed to named vendors against a 12.28% residual. Haematology attributes cleanly because a six-part differential cannot be assembled from third-party reagents: no unlisted middle can exist. A reagent, by contrast, is a formulation. The physical fit is proprietary; the chemistry is not. So thousands of firms — Chinese provincial producers, Indian generic houses, Korean and Brazilian suppliers, every manufacturer of a cuvette, a tip and a sample cup — sit in the middle of this market with no disclosure obligation and no reason to file one.

This is not a research gap. It is the structural difference between a market whose barrier is an instrument and a market whose barrier is a recipe.

Product segmentation: nine classes of consumable

The layer splits into nine product classes. Three of them — immunoassay reagents, molecular consumables and clinical chemistry reagents — are 63.0% of it, and they are trending in three different directions.

The three leading classes are derived from, and reconcile to, the discipline reports already in this cluster. Immunoassay reagents are 82% of a USD 26.4 billion immunoassay market. Clinical chemistry reagents are 70% of a USD 16.8 billion market — the lowest intensity of the three, because chemistry is the one discipline where the analyser is a real product rather than a placement. Haematology and coagulation reagents are 77.2% of a USD 7.62 billion field.

The spread in growth rates is the story. Immunoassay reagents grow at 4.55% and chemistry reagents at 3.00%, both below the layer, because both are where the administered price lands. Molecular consumables grow at 8.00% and become the second-largest class in the layer by 2033 — flow cells, library-preparation kits, amplification chemistry — because a sequencing consumable is priced against a menu that is still expanding and a buyer who is not yet being tendered. Point-of-care cartridges grow at 5.98% for a related reason: the unit of purchase is a single test in a sealed housing, and its price is embedded in a device contract, not in a provincial award.

Regional structure

China is 15.8% of the layer and, on the evidence of 2025, sets the marginal price for the other 84.2%. That is not a rhetorical claim: Roche's China diagnostics revenue fell 24% while its ex-China business grew 7%, and the language both Roche and Abbott used to explain the drag was identical — volume-based procurement. A price formed in one province of one country propagates into every analyst model, every discount negotiation and every framework agreement outside it, because the buyer in those negotiations knows what the reagent sold for in Anhui.

End-market and channel structure

Two-thirds of this market is bought by organisations that are measured on cost per reportable result rather than on revenue. That is the single most important structural fact in the report, because a cost-per-result buyer does not care whether the analyser or the reagent is where the vendor makes its margin. The buyer cares about the combined cost of producing a validated number. The vendor therefore decides which half of the pair to price sharply, and the razor-and-blade ratio is that decision, written down.

Price formation: three regimes, one P&L

A reagent reaches a laboratory through one of three price mechanisms, and the three act on different lines of the same P&L.

1. The reagent is tendered (China). The buyer sets the price of the consumable directly. Anhui's alliance, led by the province's healthcare security administration and joined by 28 provinces, purchased tumour-marker reagents across 16 items and thyroid-function reagents across 9, producing 25 sub-categories and 5,978 winning records. The methods in scope were chemiluminescence, flow fluorescence and time-resolved immunofluorescence. Average reduction: 52.6%. Term: two years, executing in Yunnan from 1 March 2026. Combined tumour-marker and thyroid panels were excluded. Roche, Abbott and Siemens all appear among the winners — a reminder that a vendor can win a tender and still lose revenue.

2. The test is administered, the reagent is not (United States). Medicare prices the CPT code. The laboratory buys its reagents privately. So deflation arrives at the output price while the input price is negotiated, which means a US laboratory facing a cut absorbs it as margin compression rather than passing it upstream. The preliminary CY 2027 Clinical Laboratory Fee Schedule cut chemistry 16% across 387 codes and capped annual reductions at 15% for 2027–2029, which tells you the underlying medians sit further below current rates than one year can reach.

3. Neither is published (most of Europe, most of Asia-Pacific). Framework agreements, distributor margins, and local-manufacture criteria. Prices form in negotiation and are not observable from outside.

The consequence is that the same 4.55% growth in immunoassay reagents is produced by three completely different mechanisms in three regions, and only the Chinese one starts with the reagent.

Volume, test economics and realised revenue per test

Global IVD reportable tests are modelled at 47,300 million in 2025 rising to 75,400 million in 2033, a 6.00% CAGR. The reagent layer grows 5.31%. The difference is the price effect:

A −0.65% annual drift in realised revenue per test is mild compared with the −1.3% a year in clinical chemistry, and the reason is mix: as molecular consumables and point-of-care cartridges take share, the average basket moves toward products whose price is not yet administered. The average is therefore improving while the individual lines inside it are deteriorating. That is the most dangerous kind of stability in a forecast, and it is why the sensitivity table below carries a five-point band.

Methodology and editorial sourcing

This report is built at a 70% secondary / 30% primary split. The secondary layer is audited results, statutory filings and government procurement notices read from the issuing body. The primary layer — 52 depth interviews across 14 markets, 31 open-ended tender-cycle interviews, a 191-respondent laboratory survey and observation at four conferences — establishes mechanism and sequence and carries no figure. No market-research competitor material is cited, quoted or used to set any level.

Editorial sources

• Sysmex — Net Sales by Destination, Business and Product type (instruments, reagents, service, other; year to March 2026)
• Sysmex Report 2025 (sources of revenue; ongoing sales 66.4% → 79.5%)
• bioMérieux — FY2025 consolidated financial statements (IFRS 15 revenue by category)
• Danaher — Fourth Quarter and Full Year 2025 Results (Diagnostics USD 9,941m; segment operating margin 26.7%)
• Danaher — Form 10-K, FY2025 (revenue by type: recurring versus non-recurring)
• Bio-Rad — Fourth Quarter and Full Year 2025 Results (Clinical Diagnostics USD 1,562.1m; quality controls as driver)
• Bio-Rad — Form 10-K, FY2025 (segment description; revenue largely recurring)
• Snibe — 2025 Annual Report (domestic and overseas reagent versus instrument revenue; closed-system model)
• Yunnan Provincial Healthcare Security Administration — execution notice for the Anhui-led 28-province IVD reagent alliance (25 sub-categories; 5,978 winning records; average 52.6%)
• 2025 IVD volume-based procurement outlook (Gansu routine reagents; Anhui cardiac markers; Jiangsu NIPT services)

Company and entity links

bioMérieux Q2904021 · Danaher Q1159256 · Bio-Rad Laboratories Q2904022 · Sysmex Q7663566 · Roche Diagnostics Q1592110 · Abbott Q306764

Related research in this cluster

• Global In Vitro Diagnostics (IVD) Market — the USD 108.0 billion hub this layer sits inside
• Global Immunoassay Market — the 27.60% of this layer that is bought on a chemiluminescence analyser
• Global Clinical Chemistry and Urinalysis Market — the discipline where the reagent is most nearly a commodity
• Global Haematology and Coagulation Diagnostics Market — the discipline where the reagent cannot be substituted at all
• The first broad reset since 2018 cuts chemistry sixteen per cent
• Sysmex took 1.7 per cent off sales and 40.8 per cent off operating profit

Articles on this market

Two companion analyses are published against this report. Two Companies Sell the Same Instrument and Disagree by Twenty-Two Points About What Business They Are In examines the reagent-intensity ratio and why it decides what an administered price does to a vendor's P&L. Anhui Bought Twenty-Five Reagent Categories at Fifty-Two Point Six Per Cent Off works through the procurement mechanism behind the price formation section above.

Paysage concurrentiel

Competitive Landscape

There is no leader of this market, because it is not one market. It is nine product classes with nine different barriers, and the vendor that leads each one leads it for a different reason. The only useful way to read the competitive structure is to ask, class by class, what actually stops a buyer switching.

Three moats, and only three

The instrument lock. The reagent fits one analyser. This is the strongest barrier in the layer and it is why immunoassay and haematology reagents attribute cleanly to named vendors while chemistry reagents do not.

Menu breadth. A closed system is only as valuable as the number of items it can report. Snibe’s 2025 annual report states the position precisely: 250 chemiluminescence reagents sold globally, 207 with EU access, 201 registered domestically across 288 registration certificates, covering tumour markers, thyroid, infectious disease, gonadal and cardiac markers, inflammation, glucose metabolism, bone metabolism, liver fibrosis and autoantibodies — plus 80 clinical chemistry reagents and seven coagulation reagents. Menu count is not a marketing statistic. Each item is a shelf in the laboratory, and a shelf the competitor cannot stock.

The compliance obligation. Quality-control materials are the one consumable a laboratory cannot decline. A laboratory can defer an analyser, renegotiate a reagent price, or send a test out. It cannot stop running the controls that validate the result it reports. That is why Bio-Rad’s Clinical Diagnostics segment grew 0.8% at constant currency to USD 1,562.1 million in 2025 with quality controls named as one of two drivers — in the same year and the same segment where lower reimbursement rates for diabetes testing in China were named as the offset. One product line is a purchase; the other is a licence to operate.

The margin tells you where the value sits

Danaher’s Diagnostics segment reported USD 9,941 million of revenue and USD 2,650 million of segment operating profit — a 26.7% margin, against 26.8% in 2024. Core revenue grew 1.5%. A business that is roughly four-fifths recurring consumables held its margin flat while growing 1.5%, in a year when its Chinese reagent prices were reset by procurement. Sysmex, by contrast, took sales down 1.7% and operating profit down 40.8%, margin 17.2% to 10.4% — a company with a 61.58% reagent ratio and a large Japanese and Chinese installed base, absorbing the same reset through a smaller annuity and a heavier instrument commitment.

Two vendors, the same tender cycle, a 16-point gap in operating margin. The difference is not product quality. It is what share of the P&L is protected by an installed base that must buy, and what share is exposed to a price that can be administered.

Chaîne de valeur

Value Chain and Margin Distribution

A reagent travels through five stages, and the money is not distributed evenly across them.

The pinch is the bead, not the vial

A reagent is largely water, protein, salt and plastic. The value is in the antibody pair, the calibration and the lot-to-lot consistency — and upstream of all three, in the raw material that captures the analyte. Snibe’s annual report lists its technology platforms in order, and the first two are magnetic microspheres and key reagent raw materials, ahead of instrument development and reagent development. That ordering is a strategy statement. The raw material is the only input in the chain that can be simultaneously scarce and cheap, which makes it the only place where vertical integration buys both cost and defensibility.

It also explains the shape of the residual. The 51.24% of this layer that attributes to nobody is not an instrument problem — thousands of firms can formulate a chemistry. It is a supply-chain fact: whoever controls the bead, the antigen or the enzyme controls who can enter, and those suppliers are largely unlisted.

Which leg of the pair carries the asset

The instrument is a balance-sheet decision before it is a product. A vendor that sells the analyser books revenue at placement and transfers the depreciation to the laboratory. A vendor that rents it keeps the asset and books the reagent share embedded in the rental as rent — bioMérieux’s EUR 61.9 million equipment-rental line is exactly that, and its own note says the line includes reagent revenue reclassified as rent.

The consequence is a difference in what a price cut touches. Where the instrument is sold, a reagent cut reduces revenue on a cost base that has already been recovered. Where the instrument is rented, the same cut has to cover a lease the vendor still owns. Neither model is better in the abstract. But the model determines whether a vendor can answer an administered reagent price by placing more instruments — which is precisely what Snibe did in 2025, growing domestic instrument revenue 6.88% in the same year domestic reagent revenue fell 13.77%.

Local fill-finish as a strategic asset

A reagent is short-dated, temperature-sensitive and customs-sensitive. That combination makes local fill-finish the difference between competing for a provincial award and being excluded from it, and it is why local manufacture has become an explicit criterion in procurement outside China as well as inside it. It is also why the Chinese domestic reagent industry could fill 2,314 biochemical specifications and 5,978 reagent award records: the capacity to fill a vial locally existed before the tender did.

Where the margin concentrates

Across the chain, the margin concentrates at the two ends. Upstream at the raw material, where the supplier of a bead or an antibody pair prices against a switching cost measured in revalidation. Downstream at the compliance consumable, where the buyer has no discretion. The middle — formulation, fill-finish, distribution — is where the price competition lives, and it is where the residual sits.

Activité d'investissement

Investment Analysis

The increment is the argument. Of the USD 40,200 million this layer adds between 2025 and 2033, a third comes from one product class that is a fifth of the base, and less than a tenth comes from the classes that carry the administered price.

The four positions worth underwriting

1. Molecular consumables capacity. 20.40% of the base supplies 33.85% of the increment. Flow cells, library-preparation kits and amplification chemistry are priced against a menu that is still expanding and a buyer who is not yet being tendered. The risk is not price; it is a sequencing-platform transition that makes a consumable format obsolete.

2. The compliance consumable. Quality-control materials and calibrators are 3.40% of the layer and grow at 4.87%, but they are the only product class whose demand does not fall when a laboratory cuts cost, because internal QC is a condition of reporting a result at all. Bio-Rad’s Clinical Diagnostics segment grew 0.8% at constant currency with quality controls named as a driver while Chinese diabetes reimbursement was cutting the same segment from the other side. This is the lowest-beta asset in the layer and should be valued as such, not as a diagnostics growth asset.

3. Surviving domestic Chinese reagent manufacturers. The 51.24% residual is where consolidation is happening, and administered price is the mechanism. A 52.6% average reduction across 5,978 award records does not eliminate the category; it eliminates the marginal bidder. Vendors holding menu breadth, raw-material control and local fill-finish come out with a larger share of a smaller price — which is why Mindray still names doubling domestic coagulation share among its three-year objectives after a year in which its domestic revenue fell 22.97%.

4. Instrument placement as a financial product. If the reagent price is administered, the only growth lever left is installed base. That converts the instrument from a product into a financing decision, and it rewards vendors with balance sheet and channel reach over vendors with the best assay. It is also the reason to discount a reagent-only business with no placement engine.

What the transactions of 2025 were buying

Nobody bought instrument capacity. QIAGEN completed the acquisition of Parse in December 2025 while discontinuing NeuMoDx and Dialunox at a cost of roughly USD 20 million of 2025 revenue — a menu purchase and a platform exit in the same year, which is the logic of this layer stated as a capital allocation decision. Danaher’s position is the other form of the same argument: a cartridge installed base above 60,000 instruments carrying more than 30 cleared assays and roughly USD 1 billion of sustainable incremental annual revenue on the company’s own framing. The scarcity in this market is not innovation. It is the right to be on the bench.

What not to underwrite

Do not underwrite price. Realised reagent revenue per test falls from USD 1.657 to USD 1.573 across the window, and every scenario in the sensitivity table is a variation on how fast that happens. A model that grows this layer by growing revenue per test will overstate 2033 by an amount equal to the entire molecular consumables increment.

Do not underwrite instrument revenue as a proxy for the franchise. Danaher grew Diagnostics core revenue 1.5% and held a 26.7% segment operating margin because four-fifths of the revenue was recurring. Sysmex’s sales fell 1.7% and its operating margin halved. The two numbers measure the same industry from opposite sides of the razor-and-blade ratio, and only one of them is a valuation input.

Three disclosures to watch

First, whether any of Roche, Abbott, Siemens or Danaher begins publishing a reagent line — a disclosure that would shrink the residual and signal that management sees the consumable ratio as a valuation argument. Second, whether Chinese domestic vendors keep reporting instrument growth alongside reagent decline; the quarter that inverts is the quarter the installed-base strategy stops paying. Third, whether any western vendor follows bioMérieux into renting more of its installed base, which would move the razor-and-blade ratio without changing a single price.

Technologie et innovation

Technology and Innovation

Nothing in this layer changes what a test measures. Everything in it changes who can be charged for the measurement, and how much. Four transitions are doing that work in the forecast window.

1. The cartridge is a pricing device before it is a technology

A single-use cartridge converts a reagent sale into a device-locked consumable. That matters commercially because a provincial procurement can specify a reagent — analyte, method, specification, price — but it cannot easily specify a cartridge without naming the reader it fits. The buyer’s lever is blunted by geometry rather than by patent.

This is the mechanism behind the divergence in the product table. Point-of-care cartridges and strips grow at 5.98% against 4.55% for immunoassay reagents and 3.00% for chemistry reagents, and the gap is not clinical demand. It is that the cartridge’s price is set inside a device contract, while the reagent’s price is set in a room with 28 provinces in it.

2. Lyophilisation removes the cold chain, and with it a local advantage

A room-temperature-stable reagent can be shipped, stored and used where there is no cold chain. That is what makes decentralised testing reachable outside hospital laboratories: physician offices, pharmacies, and settings with intermittent power. It also removes one of the reasons for local fill-finish — a freeze-dried reagent crosses a border more easily than a frozen one. The strategic effect is double-edged: larger addressable geography for the vendor, a weaker local-manufacture argument at the tender.

3. Menu chemistry is where reagent innovation still earns a premium

The clearest counter-example to this report’s deflation thesis is a chemistry problem. Small molecules carry a single antigenic epitope, so the standard two-antibody sandwich cannot be used and the field has relied on competitive immunoassay, which trades sensitivity for convenience. Snibe states it spent from 2017 to 2021 building a sandwich route for small molecules and now sells 16 products on that platform across six clinical domains — 25-hydroxyvitamin D, oestradiol, progesterone, aldosterone and tacrolimus among them — protected by 25 patents, 13 on assay process and 12 on antibody sequence. The line generated RMB 211.18 million in 2025, up 75.36%, with correlation coefficients against mass spectrometry of 0.994 for aldosterone, 0.990 for oestradiol and 0.993 for tacrolimus, and 0.97 for vitamin D across 1,314 samples.

Two things follow. First, a reagent innovation can grow 75% in a year inside a market whose average price is falling 52.6% at tender, because the item is not in the tender. Second, each such assay moves a test from send-out to in-house, which shifts volume between end markets rather than adding it.

4. The instrument engineer is paid to sell fewer reagents

There is a structural tension inside every vendor in this layer that no strategy document resolves. The analyser is designed to reduce reagent consumption per test: micro-volume reaction systems, concentrated reagents diluted on board, smaller sample volumes, longer calibration intervals. Snibe’s flagship chemistry analyser is specified at 2,000 tests per hour, 3,000 with the electrolyte module, on what the company describes as an ultra-micro reaction system with on-board dilution of concentrated reagent — and the stated benefit is a significant reduction in reagent and sample consumption.

That is the correct engineering answer and the wrong revenue answer. Every generation of instrument reduces the vendor’s own reagent revenue per installed analyser, which is why instrument placement has to keep growing for a reagent-led business to grow at all. The forecast in this report assumes exactly that: test volume compounding at 6.00% against reagent revenue at 5.31%.

A note on what is not yet a technology transition: digital and computational pathology are reshaping the read step, but they change stain and slide volume only at the margin. Histology and pathology consumables grow at 4.98% and remain 5.10% of the layer, and no assumption in this report depends on that changing.

Risque de marché

Market Risk and Sensitivity

Four risks matter, and they are not the four that are usually listed. None of them is a demand risk in the ordinary sense: tests per patient are not falling in any market in this forecast. Every one of them is a price, volume-per-patient or cost-of-compliance risk.

Risk 1 — the administered price extends into the volume base

Until 2025, Chinese reagent procurement concentrated on high-value immunoassay menus. In 2025 it moved to routine products: Gansu led the first volume-based procurement of routine blood, stool and urine reagents and strips, while Anhui extended to cardiac markers after the 25-province alliance of 2023 and the 28-province tumour-marker and thyroid procurement of 2024. Routine chemistry and haematology reagents are 22.50% of this layer. A 52.6%-style reduction applied to a fifth of the layer is the single largest downside in the report.

Risk 2 — panel unbundling cuts volume, not price

This is the risk most often missed, and it is disclosed in plain language. Snibe attributes its domestic decline to three policies acting together: provincial reagent procurement, the deepening of DRG/DIP per-case payment, and test-panel unbundling together with mutual recognition of results across institutions. The first is a price risk. The second makes the hospital a cost centre that manages the test as an input. The third and fourth reduce the number of tests performed per patient, because a panel that was ordered as a set is ordered item by item, and a result produced at one institution does not have to be produced again at another.

A volume risk is worse than a price risk for this layer, because the instrument commitment is fixed. A vendor that has placed an analyser on a five-year reagent commitment has already paid for volume that may not arrive.

Risk 3 — regulatory cost rationalises the menu

Registration is granted per reagent, not per instrument. Under IVDR in Europe the transition runs to May 2025 for Class D, May 2026 for Class C and May 2027 for the remaining classes, and each classification step adds notified-body review, technical documentation and post-market surveillance to every individual item. Under NMPA in China, Snibe holds 288 certificates to sell 201 chemiluminescence reagents — roughly 1.4 certificates per item. That arithmetic is a barrier for a large vendor and a wall for a small one, which means the residual will consolidate. The effect on price is the opposite of the effect on competition: fewer suppliers, less pressure.

Risk 4 — raw material and currency concentration

The upstream supply of antibodies, enzymes, magnetic beads and membranes is concentrated in a small number of suppliers, several of them unlisted. A single-source raw material can halt a reagent line regardless of demand. Separately, the layer is reported in USD on a converted basis: JPY 148, EUR 1.13 and CHF 1.15 to the dollar. A 10% move in the yen changes Sysmex’s converted reagent revenue by roughly USD 200 million without a single test changing hands.

Sensitivity

The band is 112,900 to 124,300, or 4.66% to 5.93% — a 127-basis-point range around a 5.31% base. Almost all of it is two governments and one hospital-payment reform. Nothing in the band is a clinical or demographic assumption, which is the point: this is the one layer of in vitro diagnostics where the forecast is mostly a policy forecast.

Paysage réglementaire

Regulatory Landscape

Regulation enters this layer at the level of the individual reagent, and that single fact shapes the competitive structure more than any other rule in the sector.

Europe — IVDR classifies the reagent, and the classification is the cost

Regulation (EU) 2017/746 replaced the old directive with a risk-class system, and the transition runs by class: Class D to May 2025, Class C to May 2026 and the remaining classes to May 2027. Most infectious-disease and blood-safety reagents sit in the highest classes. The practical effect is that each item carries its own conformity assessment, technical file, performance evaluation and post-market surveillance obligation, and each requires a notified body with capacity to review it.

Three consequences follow. First, the fixed cost per SKU rises, which hits vendors whose menu is small: a portfolio of eleven reagents carries almost the same regulatory overhead as a portfolio of two hundred. Second, menus get rationalised — items with modest volume are withdrawn rather than re-certified, which narrows the menu breadth that is supposed to be the competitive moat. Third, the 51.24% residual consolidates, with the effect on price described in the risk section.

United States — the reagent is a device, the test is a fee

An IVD reagent is a medical device. It reaches market through a 510(k) with a predicate, a De Novo classification or a premarket approval, and its risk class follows the analyte and the clinical use rather than the chemistry. Separately, CLIA categorisation — waived, moderate or high complexity — decides who may perform the test, which determines whether a given reagent can be sold to a physician office or only to a laboratory with the right certificate. Menu strategy and channel strategy are the same decision in the United States.

The price, however, is set somewhere else entirely. Medicare pays for the test under the Clinical Laboratory Fee Schedule, and the reagent is a private input cost. The preliminary CY 2027 schedule cut chemistry 16% across 387 codes and capped annual reductions at 15% for 2027–2029. That is a reduction in the output price with no corresponding change in the input price, which means a US laboratory facing it has one place to go for relief: the reagent contract.

China — three instruments acting on the same P&L

NMPA registration governs the item, and the certificate count is the barrier: Snibe holds 288 certificates covering 201 chemiluminescence reagents in the domestic market. Provincial healthcare-security administrations then set the price through alliance procurement — 25 provinces in 2023, 28 in 2024, routine reagents added through Gansu in 2025. And DRG/DIP per-case payment reform changes the buyer’s incentive from revenue to cost, which is what makes a hospital willing to enforce test-panel unbundling and to accept a result produced at another institution.

The sequence matters. Registration creates the menu, procurement sets its price, and payment reform decides how much of it gets ordered. A vendor can be registered, be awarded, and be paid less in the same year.

Japan and the rest

Japan revises the reimbursement schedule on a biennial cycle, which produces synchronised, predictable, modest reductions rather than tender cliffs. Most other markets combine device registration with framework agreements in which local manufacture is an increasing criterion — the same mechanism China used earlier, at a different point in the cycle.

What regulation does to the razor-and-blade ratio

Registration is per reagent. Instruments are certified once. So a vendor’s regulatory cost scales with menu breadth while its commercial value also scales with menu breadth — and the two curves cross at a menu size only large vendors reach. That is the regulatory reason the layer’s residual is 51.24% and falling, and the reason the strongest positions in this report belong to vendors with hundreds of certificates rather than the best single assay.