Section 232 and the Medical Technology Industry

1. Executive Summary

The Section 232 national security investigation into imports of personal protective equipment, medical consumables, and medical equipment, including devices was self-initiated by the Secretary of Commerce on 2 September 2025 and published in the Federal Register on 26 September 2025 at 90 FR 46383. As of 5 August 2026 the Bureau of Industry and Security still lists the investigation as ongoing. No Commerce report has been publicly confirmed and no presidential proclamation has issued. The statutory 270-day report deadline fell in late May or June 2026 depending on how the clock is counted, and practitioners tracking the docket — Covington in an April 2026 status survey — project presidential action in a window around mid-September 2026.

That status is the first and most important correction to the way this subject is usually framed. A great deal of commentary treats the medtech Section 232 investigation as the pending event that will determine the industry’s tariff exposure. It is not. The industry has already absorbed roughly eighteen months of tariff cost under four successive legal authorities, and the measures currently doing the most damage to medical device economics are not the medtech 232 at all. They are the Section 301 China tariffs on gloves, syringes and respirators, which reached 100 percent on gloves and syringes on 1 January 2026; the restructured Section 232 metals tariffs of 6 April 2026, which now apply duties to the full customs value of derivative products rather than to metal content alone; and the Section 301 “forced labour” tariffs that took effect on 24 July 2026 across the sixty economies investigated, a group accounting for the large majority of US imports. A manufacturer whose tariff planning is organised around waiting for the medtech 232 proclamation is mis-allocating attention.

The second correction concerns who is actually exposed. The intuitive framing — commodity consumables most exposed, complex capital equipment most insulated — is directionally right about ability to absorb and substantially wrong about magnitude of exposure. The largest disclosed tariff bills in the sector belonged to capital equipment makers: GE HealthCare initially guided to approximately $500 million of gross 2025 exposure, of which around $375 million was bilateral China-related, and Siemens Healthineers forecast €400 million for fiscal 2026 after €200 million in fiscal 2025. Stryker guided to approximately $400 million for 2026. These are not insulated numbers. What distinguishes the large caps is not low exposure but high absorptive capacity: gross margins in the high-sixties to high-seventies, pricing and productivity levers, and the balance sheet to carry a multi-hundred-million-dollar cost while mitigation catches up. The genuinely fragile position belongs to consumables specialists — ICU Medical disclosed a first-quarter 2026 tariff expense equal to roughly 2 percent of adjusted revenue — and to the small and mid-size manufacturers and importers who have neither the margin nor the customs infrastructure to respond.

The third correction is that the industry’s revealed response has been absorption, not pass-through, and not reshoring. Across the 2025–2026 reporting cycle, large medtech companies repeatedly halved their initial tariff estimates, offset the residual through pricing and productivity, and declined to raise hospital prices. Vizient’s July 2025 projection for 2026 healthcare supply chain inflation was 2.41 percent overall and 3.28 percent for surgical supplies — against a Premier survey finding that 82 percent of healthcare executives expected tariffs to raise costs by 15 percent within six months. The gap between the 15 percent fear and the 3 percent outcome is the measure of how much manufacturers absorbed. On reshoring, MedTech Dive’s one-year review of 9 April 2026 concluded that unlike pharmaceuticals, medtech did not pursue large-scale relocation to the United States; domestic expansion happened only where regulatory familiarity or technical collaboration already provided an advantage. Several of the largest capacity announcements in the period went to Malaysia and Mexico rather than the United States.

The fourth correction is that the upstream input exposure the industry should be worried about is not the one usually named. Tungsten is a real vulnerability but not in the form commonly stated: the United States has not mined tungsten commercially since 2015 and net import reliance exceeds 50 percent of apparent consumption, but China accounted for 26 percent of US tungsten imports over 2021–2024, not roughly half. China’s leverage runs through its 79 percent share of world mine production and its February 2025 export controls on selected tungsten items, which coincided with a concentrate price move from $252 to an estimated $380 per metric tonne unit between 2024 and 2025. The larger and more under-discussed input exposure is the 6 April 2026 metals proclamation’s shift to full-customs-value assessment on derivative products. A device with 8 percent steel content by value that previously paid 50 percent on that 8 percent may now pay 25 or 50 percent on 100 percent of its value. That change, applied to surgical instruments, orthopaedic implants, sterilisation trays, hospital beds, wheelchairs, cannulae and imaging housings, is capable of doing more damage to device economics than any plausible medtech 232 headline rate — and the public process for petitioning products in or out of the derivative lists was terminated by the same proclamation.

Finally, the structural reason medtech cannot behave like other tariffed sectors deserves to be stated plainly, because it is the fact that should anchor any manufacturer’s planning. Devices cannot re-source quickly because supplier, material and site changes are regulated events. A manufacturing site change for a Class III device generally requires a premarket approval supplement with a 180-day statutory review clock and a FY2027 user fee of $95,510, with possible pre-approval inspection; a panel-track supplement costs $509,386. And devices cannot pass cost through to payers because device prices are embedded in bundled procedure reimbursement set administratively. Needham’s Mike Matson put the asymmetry precisely: pharmaceutical companies could offer price concessions in exchange for tariff relief, but a device maker “can’t just say, ‘Well, I’m going to cut my knee replacement price by 20%’.” Medtech is squeezed between a supply chain it cannot legally reconfigure at speed and a demand side that cannot be repriced.

Key judgments

The most probable outcome of the medtech 232, on the evidence in the docket and the administration’s revealed preferences, is a narrow, high-rate remedy aimed at the two categories where domestic capacity has genuinely collapsed — medical gloves and syringes — combined with exclusions or reduced rates for allied and free-trade-agreement partners. That is the shape of the only well-developed pro-tariff filing in the docket, and it is consistent with an administration that has already deployed 100 percent Section 301 rates on exactly those two product lines. A broad ad valorem tariff across HTS chapters 90 and 30 is possible but poorly supported by the record, because medtech is a net-exporting sector with a domestic production share of roughly 70 percent, which undercuts the import-dependence premise of a national security finding.

A second judgment: the probability that the investigation is allowed to lapse without action is meaningfully above zero. The commercial aircraft Section 232 statutory deadline passed in 2026 without presidential action, establishing that this administration will let a 232 docket sit. RBC’s Shagun Singh has observed that medical devices are not a sector the administration appears to be targeting adversarially. Manufacturers should plan for a proclamation but should not treat inaction as an implausible scenario, because the two cases call for different inventory and contracting behaviour in the September window.

A third judgment: the near-term financial risk to most large-cap manufacturers in the second half of 2026 is not a new tariff. It is the reversal of a non-recurring benefit. Second-quarter 2026 results across the sector were inflated by refunds of duties collected under the International Emergency Economic Powers Act after the Supreme Court invalidated that authority on 20 February 2026 — €186 million at Philips, approximately €200 million at Siemens Healthineers, $129 million at GE HealthCare, approximately $80 million at Boston Scientific, $36 million at Intuitive Surgical. Philips’ refund alone contributed around 4.2 percentage points to second-quarter adjusted EBITA margin. None of it recurs. Any 2027 planning model built off second-quarter 2026 margins is wrong by several hundred basis points.

2. The investigation: scope, clock, and what the record actually contains

2.1 Scope

The Federal Register notice of 26 September 2025 (90 FR 46383, document 2025-18729, Commerce docket 250924-0160, regulations.gov docket BIS-2025-0258) defines three product buckets. The first is personal protective equipment: surgical masks, N95 and other respirators, gloves, gowns, and related medical parts and components. The second is medical consumables, defined as single-use or short-term-use items, and enumerating syringes, needles, infusion and intravenous pumps, IV bags, catheters and tracheostomy tubes — and, importantly, diagnostic and laboratory reagents. The third is durable medical equipment and devices: wheelchairs, crutches, hospital beds, pacemakers, insulin pumps, ventilators, magnetic resonance imaging machines and x-ray apparatus.

Two features of the scope deserve emphasis because they are routinely missed. First, pharmaceuticals are expressly excluded, having been addressed by the separate Section 232 pharmaceutical investigation initiated 1 April 2025 and concluded by proclamation on 2 April 2026. Second, the inclusion of diagnostic and laboratory reagents pulls the in-vitro diagnostics sector into the investigation. Most category-based commentary on this investigation omits diagnostics entirely, which is an error: IVD is a distinct segment with a distinct dependency profile, and the College of American Pathologists filed in the docket specifically seeking exemption for laboratory supplies and IVD products.

The 21-day comment window closed on 17 October 2025.

2.2 The clock

Section 232 gives the Secretary of Commerce 270 days from initiation to report to the President, and the President 90 days from receipt to determine whether to act. Counting 270 days from 2 September 2025 produces a report deadline around 30 May 2026 and an outside presidential action date around 28 August 2026. Covington’s April 2026 tracker projects action in a mid-September 2026 window, implying a report deadline nearer mid-to-late June. Either way, both dates are within weeks of this report’s publication.

Neither the report nor its transmission is required to be public, which means the absence of news is genuinely uninformative about whether Commerce has filed. What is informative is precedent: Holland & Knight noted in June 2026 that the administration had not acted on the commercial aircraft Section 232 despite its statutory deadline having passed. Deadlines in this programme are being allowed to lapse.

2.3 What the docket says

The comment record is lopsided, and reading it is the best available guide to the shape of any remedy.

Against tariffs, or seeking carve-outs: AdvaMed filed on 17 October 2025 asking for reciprocal tariff-free trade with allied nations, aggressive action on foreign market-access barriers, domestic regulatory reform, and procurement incentives favouring US production. Notably, AdvaMed has publicly preferred the Section 232 route to broader instruments — chief executive Scott Whitaker described it on 24 February 2026 as “sector-specific and evidence-based,” permitting tailored remedies rather than blanket duties. That is an unusual industry posture and a signal that the industry expects a narrow or favourable outcome.

The Medical Device Manufacturers Association argued tariffs would “needlessly disrupt a globally integrated industry that already meets domestic demand,” and asked, if tariffs proceed, for reciprocal relief with preferred partners, preservation of USMCA treatment, phased implementation, and continuation of duty drawback. It also asked for public-private stockpiling partnerships leveraging the Strategic National Stockpile.

The US Chamber of Commerce filed the most granular submission, with ten asks including no 232 tariffs on allies and FTA partners, narrower scope definitions differentiated by product category, non-application to USMCA-qualifying goods, availability of duty drawback, restoration of status quo rates with the EU, Japan, Korea and the UK, CMS payment adjustments favouring domestically produced components, expedited FDA pathways for US-manufactured devices, and streamlined domestic minerals permitting. Its central empirical argument was that onshoring some product categories would take a decade to reach equivalent productivity and quality, and that tariffs could push production out of the US market rather than into it.

The American Hospital Association sought exceptions for vital medications, devices, equipment, supplies and PPE, “especially critical for low-margin, high-volume goods,” and asked that Nairobi Protocol duty-free treatment for devices such as pacemakers, insulin pumps and prosthetics be preserved. Premier filed five recommendations including distinct treatment for products deemed essential by FDA and the Department of Defense, a “slow and steady glidepath” to implementation, clear definitions of which components are dutiable, and impact studies to confirm that tariffs achieve onshoring rather than shortages. Business Roundtable, the Health Industry Distributors Association, the National Association of Manufacturers, the Association of American Medical Colleges, the College of American Pathologists and AAHomecare also filed.

For tariffs, the record contains essentially one developed proposal. The Coalition for a Prosperous America asked for a specific duty of $0.045 per glove plus 150 percent ad valorem on nitrile gloves; $15 per kilogram on imported nitrile butadiene rubber feedstock; a per-unit duty on Chinese-origin syringes equal to 100 percent of average landed cost, which it put at approximately $0.58 per unit; trusted-partner exclusions for Canada, Mexico, Japan and the EU; anti-circumvention safeguards against transshipment; long-term federal purchase commitments; stockpiling mandates; and a formalised product-by-product inclusion process. No labour union filing was located in this docket.

The asymmetry is the analytically useful part. Where the record supports a remedy, it supports one aimed narrowly at gloves and syringes, structured as a specific per-unit duty as much as an ad valorem rate, with allied-country exclusions. Manufacturers should note the design implication: a per-unit specific duty defeats first-sale valuation and transfer-pricing mitigation entirely, because there is no ad valorem base to reduce.

3. The tariff regime that already exists

Any assessment of Section 232 exposure that does not begin with the layers already in force will misprice the marginal impact. The following is the state of play on 5 August 2026.

3.1 The four-authority sequence

The reciprocal tariff regime introduced in April 2025 under the International Emergency Economic Powers Act was struck down by the Supreme Court on 20 February 2026 in Learning Resources, Inc. v. Trump, No. 24-1287, by 6–3, on the ground that IEEPA does not authorise the President to impose tariffs. The Court declined to order refunds, leaving remedies to the Court of International Trade and Customs and Border Protection.

Within hours of the ruling the IEEPA orders were rescinded and replaced with a flat 10 percent surcharge on all imports regardless of origin under Section 122 of the Trade Act of 1974, which carries a 150-day statutory maximum. That surcharge ran from 24 February to 24 July 2026. The Court of International Trade held in May 2026 that Section 122 was also unlawful; the Federal Circuit stayed that ruling in June 2026.

Section 122 expired at midnight on 23 July 2026 and was replaced, effective 12:01 a.m. on 24 July 2026, by Section 301 “forced labour” tariffs covering the sixty economies investigated. The structure is two-tier: 10 percent for the nineteen economies that maintain forced-labour import prohibitions, 12.5 percent for the forty-one that do not, with rates capped net of most-favoured-nation duty for the EU, Japan, Korea, Switzerland and Taiwan — a carve-out more favourable than the rates originally proposed. Goods qualifying for preferential treatment under USMCA are exempt.

For device manufacturers, one feature of the forced-labour action matters more than its rate: goods already subject to Section 232 duties are exempt from it. That creates a genuine planning lever, discussed in Section 9.

3.2 The refund position

Approximately $166 billion in IEEPA duties was collected. At a Court of International Trade hearing on 9 June 2026, CBP reported roughly $90 billion in claims accepted and approximately $23 billion approved. The Department of Justice filed notices of appeal on 3 June 2026 challenging the CIT’s universal refund orders, arguing they conflict with the anti-nationwide-injunction doctrine of Trump v. CASA. The CBP refund mechanism has proceeded in phases, with the phase covering finally liquidated entries available, on the government’s position, only to importers who filed at the CIT.

The operational point for manufacturers is that protective filing at the CIT is the gating requirement for recovering the largest tranche. PwC estimated that up to $2.6 billion in IEEPA refunds is available to the medtech industry. Importers that paid the 24 February–24 July Section 122 surcharge should also be filing protests on liquidated entries within the 180-day window.

3.3 Section 232 metals: the change that matters most

The steel, aluminium and copper Section 232 programme was restructured by proclamation effective 6 April 2026. Rates are now 50 percent on articles made entirely or almost entirely of the metal, 25 percent on derivative articles that are not almost entirely metal, 10 percent for products using US-smelted or cast inputs, 200 percent on Russian aluminium, and 25 and 15 percent respectively for UK products.

The mechanical change is the consequential one: duties now apply to the full customs value of derivative products rather than to the value of the metal content. This materially raises effective duty on any metal-containing medical device. Surgical instruments, orthopaedic implants, needle and cannula stock, sterilisation trays, hospital beds, wheelchairs, infusion stands and imaging enclosures are all candidates.

The same proclamation terminated the public inclusions process for derivative products, replacing it with an internal Commerce and USTR process. There has never been a product-exclusion process under the current metals programme. The practical consequence is that a manufacturer that finds its product newly captured as a derivative has no formal petition channel and must advocate informally. This is a significant loss of process rights, and it is the strongest reason to read Annexes I-A and I-B of the April 2026 proclamation line by line against one’s own HTS classifications rather than relying on secondary summaries.

3.4 Section 301 China: the rates already at 100 percent

The four-year review concluded in 2024 set a phased escalation that has now fully taken effect. As of 1 January 2026: syringes and needles at 100 percent, with the temporary waiver for enteral syringes (9018.31.0080) expired; medical and surgical gloves at 100 percent, having gone from 7.5 percent to 50 percent on 1 January 2025 and then doubled; surgical and non-surgical respirators at 50 percent; disposable textile face masks at 50 percent. Baseline List 1 through 3 duties remain at 25 percent and List 4A at 7.5 percent.

There is no blanket medical exclusion. The 178 exclusions currently active were extended following the November 2025 US-China understanding through 10 November 2026 — a cliff worth diarising.

3.5 Country treatment as of August 2026

The critical structural point is buried in the phrase “capped net of MFN.” For high-MFN goods, a cap of that kind produces little incremental burden. But a great many medical device HTS lines carry MFN rates at or near zero, which means the cap provides no relief at all and the full 10 to 12.5 percent lands. Manufacturers assuming that the EU or Japanese framework ceilings protect them should check the MFN rate on their own lines before relying on it.

3.6 USMCA and the Canadian shock

The mandatory six-year USMCA joint review took place on 1 July 2026. The United States declined to agree to the optional sixteen-year extension; Mexico and Canada both supported it. Mechanically the agreement remains fully operative through 1 July 2036, with preferential tariffs, rules of origin, investment protections and dispute settlement continuing unchanged, and the extension can still be confirmed in writing at any time. The substantive change is that joint reviews are now mandatory annually through 2036, converting a single distant cliff into a recurring annual political risk.

Separately and more sharply, three proclamations issued 20 July 2026 imposed an additional 50 percent ad valorem duty on Canadian goods under Section 338 of the Tariff Act of 1930 — the first use of that authority in history — effective 19 August 2026. The scope runs across 554 HTSUS subheadings in three annexes, the largest covering 439 subheadings and approximately $19.3 billion of 2024 imports described as motor vehicles and agricultural and manufactured products. USMCA preference does not exempt. Goods already subject to Section 232 are exempt. Whether any medical device lines appear in those annexes was not resolvable from secondary sources, and any manufacturer importing from Canada should screen the annexes directly before 19 August.

3.7 The pharmaceutical 232 and the classification boundary

The pharmaceutical proclamation of 2 April 2026 imposes 100 percent on patented pharmaceutical articles and APIs listed in Annex I, with 20 percent for companies under a Commerce-approved onshoring agreement (escalating to 100 percent on 2 April 2030), 15 percent for the EU, Japan, Korea and Switzerland-Liechtenstein, 10 percent for the UK, and 0 percent for companies with both MFN-pricing and onshoring agreements (expiring 20 January 2029). Effective dates are 31 July 2026 for companies named in Annex III and 29 September 2026 for all others. Annex I modifies more than 130 HTSUS subheadings across chapters 29 and 30. Unusually, the pharmaceutical 232 rate operates as a ceiling rather than an adder: the higher of the 232 rate or the HTSUS Column 1 rate applies, and where two 232 rates apply, the lower governs.

The sleeper issue for device makers is classification. No reviewed source addresses how the pharmaceutical proclamation treats drug-device combination products, prefilled syringes, drug-eluting stents, convenience kits, or diagnostic reagents under HS 3822. Chapter 30 headings of live interest — 3002 for blood fractions and immunological reagents, 3005 for dressings and wadding, 3006 for sutures and diagnostic reagents — sit within the chapter universe from which Annex I draws. Meanwhile the medtech 232 scope separately claims diagnostic and laboratory reagents. A product could plausibly be claimed by either regime, at radically different rates. Reading Annex I and Annex IV of the pharmaceutical proclamation against one’s own classifications is not optional diligence.

4. What actually happened to the industry’s income statement, 2025–2026

The most valuable evidence about tariff impact on medtech is not modelling. It is eighteen months of disclosure. The pattern across the sector is consistent enough to name: initial estimate, halved estimate, absorption, refund.

4.1 The disclosure record

At the sector level, IQVIA estimated approximately $4 billion in tariff costs to medtech in 2025, and cited AdvaMed worst-case modelling roughly four times higher.

4.2 What the record actually demonstrates

Three findings follow, and each cuts against a common assumption.

First, the initial estimates were systematically too high, by roughly half. That is partly genuine mitigation and partly the collapse of the bilateral China escalation that drove the largest single component of the early numbers — $375 million of GE HealthCare’s $500 million. Manufacturers modelling a medtech 232 should be cautious about anchoring on first-pass gross exposure numbers, including their own.

Second, the cost was absorbed rather than passed on. PwC’s Glenn Hunzinger described companies as “not passing prices on to customers, bearing the brunt and finding efficiency.” Stryker’s Kevin Lobo was blunt that the company “would have raised our earnings without question, had the tariffs not appeared.” The offsets were pricing at the margin, productivity programmes, network optimisation and discretionary cost restraint — Medtronic reported roughly 30 basis points of price and 40 basis points of cost-of-goods efficiency in a single quarter. Research and development was protected: RBC’s Shagun Singh noted that “R&D is absolutely the last thing our companies want to cut.”

Third, disclosure quality degraded sharply in 2026. Abbott and Johnson & Johnson both stopped mentioning tariffs on their second-quarter 2026 calls; Dexcom never quantified. Absence of disclosure is not absence of cost, but it does indicate immateriality relative to other drivers. Boston Scientific cut 2026 guidance twice in 2026 — on WATCHMAN and electrophysiology operational issues, not tariffs. That is a useful corrective to the assumption that tariffs are the dominant variable in medtech earnings. They are not; procedure volumes are.

4.3 The refund distortion

Second-quarter 2026 results across the sector are inflated by non-recurring IEEPA refunds and must not be annualised. Confirmed: GE HealthCare $129 million, of which $106 million related to 2025 and was excluded from guidance; Siemens Healthineers approximately €200 million; Philips €186 million, worth around 4.2 percentage points of quarterly adjusted EBITA margin and roughly a full point on the year; Boston Scientific approximately $80 million; Intuitive Surgical $36 million; Zimmer Biomet $0.20 per share in the first quarter; Stryker a net benefit in the second quarter.

Several companies have refund optionality still outside guidance — Medtronic has applied but booked nothing, Steris’ FY27 guide assumes none, ICU Medical and Teleflex both exclude it. For a competitor-benchmarking exercise, the distinction between companies that have taken refunds into reported margin and those that have not is the single largest source of apparent margin divergence in the sector right now.

4.4 The reshoring question, answered by revealed behaviour

The policy objective behind these tariffs is domestic capacity. The record on whether medtech delivered it is unflattering.

The clearest tariff-attributed relocation in the sector is Siemens Healthineers’ move of Varian radiotherapy manufacturing from Baja California, Mexico to Palo Alto, California — a $150 million programme adding around 50 jobs, announced 16 May 2025, one week after the company’s €300 million tariff warning. Note that this is a capital equipment relocation, which is directly contrary to the assumption that complex equipment is the segment least likely to move.

Elsewhere: Abbott committed $500 million across Illinois and Texas for transfusion and diagnostics capacity; Roche Diagnostics $550 million in Indianapolis; Philips more than $150 million across Reedsville, Pennsylvania and Plymouth, Minnesota — but explicitly not attributed to tariffs. GE HealthCare moved PET/CT production from the Middle East to the US and a surgery line from Asia to the US, and enhanced USMCA duty-free qualification, duty drawback programmes and bonded logistics routes. Boston Scientific expanded in Minnesota — and in Malaysia. Intuitive Surgical is expanding instrument manufacturing in Mexico. Medtronic announced no reshoring programme at all, only “manufacturing network optimisation over time.” Zimmer Biomet had little to do, its chief financial officer noting that most of its production has been in the United States for some time.

MedTech Dive’s conclusion on 9 April 2026 was that medtech did not follow pharma into large-scale reshoring; domestic expansion occurred only where regulatory familiarity or technical collaboration already provided advantage, “not at scale replacing global networks.” KPMG’s Christopher Young described the pattern as “targeted shifts, not sweeping relocations.”

5. Category-by-category exposure, rebuilt

The four-category framework below replaces the conventional three-tier version. The addition of diagnostics and laboratory products is not cosmetic: the category is explicitly in scope, has a distinct dependency profile, and is systematically omitted from public analysis.

5.1 Commodity consumables and PPE

This is the category where the conventional analysis is right about vulnerability and wrong about almost every specific.

Import dependence is near-total, but the direction of dependence has shifted. Roughly 99 percent of medical gloves used in the United States are imported, and domestic production is approximately 1 percent of the roughly 120 billion nitrile gloves the country consumes annually. But the widely repeated claim that Malaysia supplies about 60 percent understates the position and, more importantly, obscures a split that matters enormously for tariff planning. Import data for January to July 2025, published 12 September 2025, shows medical nitrile and latex glove imports of 44.06 billion units, up 10 percent year on year, sourced 67.5 percent from Malaysia, 16.9 percent from Thailand, 9.8 percent from Vietnam — and only 2.8 percent from China. Chinese direct share of the medical nitrile segment has effectively collapsed under the Section 301 escalation. Vinyl gloves are the opposite case: 30.14 billion units imported in the same period, 90.8 percent from China. Industrial nitrile and latex, at 19.18 billion units and growing 43 percent year on year, is 39.9 percent Malaysian, 27.5 percent Thai and 11.6 percent Chinese.

The planning implication is precise. A buyer or manufacturer whose glove exposure is medical nitrile has already substantially exited China and faces a Malaysia and Thailand concentration risk instead. A buyer whose exposure includes vinyl examination gloves is almost wholly China-dependent and is carrying the 100 percent Section 301 rate. Treating “gloves” as one exposure is the mistake.

The cost escalation is well beyond the 10 to 25 percent range usually cited. Chinese medical and surgical gloves have carried 100 percent Section 301 duty since 1 January 2026, having stepped from 7.5 percent to 50 percent a year earlier. Syringes and needles have been at 100 percent since September 2024, and the enteral syringe waiver has expired. Respirators and disposable face masks sit at 50 percent. On pricing, average selling prices reported in February 2025 put Chinese gloves at roughly $15 per thousand pieces before the 50 percent step, $17 after, and an estimated $25.60 with a further 10 percent layer, against Malaysian product moving from $17–18 to $20–21. Reporting from March 2026 describes Malaysian and Thai prices rising 15 to 25 percent on demand shift, with distributor margins calculated on landed cost inclusive of tariff — a compounding effect on the buyer that is invisible in the headline duty rate.

Margin capacity in this category is genuinely absent. Top Glove’s gross margins ran 15 to 20 percent pre-pandemic, collapsed to low single digits after it, and have recovered only to the low teens. Global demand is roughly 300 billion pieces a year and broadly stagnant. There is no cushion in the supply chain; a duty increase in this category is a price increase, with a lag set by contract cycles.

And the reshoring alternative has been tested and has failed. This is the most important correction in this report to the optimistic view that commodity consumables can simply be re-sourced or re-domesticated. A federal programme committed $850 million across six companies to build domestic medical glove manufacturing. As reported on 7 July 2026, not a single medical glove has been produced under it. The flagship project, Blue Star NBR’s Virginia facility — which received $123 million in federal financing to become the first domestic producer of medical-grade nitrile butadiene rubber in more than thirty years — was operational by spring 2023, stalled that December requiring tens of millions more for utilities and an adjacent factory, and remains idle after roughly four and a half years, with its chief executive suggesting it could be sold for parts. Domestically made gloves cost approximately twice imported product. Nine domestic PPE manufacturers formed the American Medical Manufacturers Association in February 2023 to lobby for support.

The honest reading is that a tariff on gloves and syringes, absent long-term federal offtake commitments of the kind the Coalition for a Prosperous America requested, raises US healthcare costs without creating US capacity. That is not an argument against the policy — it is an argument that the policy as usually conceived is incomplete, and it is exactly what Premier asked Commerce to test with impact studies.

Geographic substitution is faster here than anywhere else, and this remains true. A dressing or a gauze pad does not carry a Class III premarket approval, so a sourcing shift to Vietnam, Malaysia, India or Thailand can be executed on commercial rather than regulatory timescales. That advantage is real. It is also the reason this category will see the fastest reshuffling and the highest pass-through — and, given the concentration data above, the reason the next fragility to worry about is Malaysian and Thai single-region dependence rather than Chinese dependence.

5.2 Durable medical equipment and mid-complexity devices

Wheelchairs, hospital beds, patient monitors, insulin pumps and continuous glucose monitors occupy the genuinely uncertain middle, and the uncertainty is mostly about geography rather than product.

The manufacturing base for this tier is heavily concentrated in Mexico and, for the diabetes segment, in a small number of large players with real supplier leverage. Mexico exported $19.3 billion in medical devices and instruments in 2024, supplies roughly one-fifth of all US device imports, and has displaced Germany as the leading origin of devices sold in the United States. The sector employs more than 140,000 workers directly across more than 2,500 manufacturers, with more than 600 holding active export registrations. The clusters are Baja California with 75-plus companies and over 74,000 workers, Chihuahua with 30-plus and over 40,000, Sonora with 26 and over 17,000, Jalisco and Nuevo León smaller. Product mix runs to cardiovascular devices, surgical instruments, orthopaedic implants, neurostimulation, diabetes management systems and blood processing equipment.

The article-level claim that Mexico “currently carries a 25 percent tariff rate” is not correct as of August 2026 and needs replacing with something more useful. USMCA-originating goods are exempt from the 10 percent Section 301 forced-labour layer altogether; non-originating goods bear the duty on non-US content subject to a 15 percent minimum effective rate. One shelter-manufacturing operator reports that 85 percent of its clients’ exports qualify under rules of origin. So for a compliant Mexican operation the forced-labour layer is close to a non-event. What USMCA does not do is shield against Section 232 or Section 338, and that is where the residual Mexican risk sits — a metals-derivative capture on a device with meaningful steel or aluminium content is not waived by origin.

For the diabetes technology names — Medtronic, Abbott, Dexcom, Insulet — the record supports the view that they have absorbed comfortably. Insulet reported roughly 50 basis points of 2025 gross margin impact and raised guidance anyway, citing an exemption for certain medical devices and a decade of roughly $1 billion in automation and facility investment; its Malaysian site came online in the third quarter of 2025 and was accretive in year one. Dexcom did not mention tariffs on its second-quarter 2026 call and reported gross margin up 400 basis points to 64.1 percent. This tier’s large players have leverage; its smaller ones do not.

The expected behaviour here remains selective supplier diversification rather than wholesale relocation, plus continued advocacy for country-specific treatment. What has changed since the conventional analysis was written is that the USMCA question has been partly answered: the agreement survived the July 2026 joint review intact through 2036, but now carries mandatory annual reviews, which converts a single planning horizon into a recurring one.

5.3 Complex capital equipment and implants

This is where the conventional framing most needs correcting, in two opposite directions at once.

On exposure magnitude, the “most insulated” characterisation is wrong. The largest disclosed tariff numbers in medtech belonged to imaging and therapy equipment makers. GE HealthCare’s initial 2025 gross exposure of approximately $500 million was the largest in the sector; $375 million of it was bilateral China. Siemens Healthineers guided to €400 million for fiscal 2026 after €200 million in fiscal 2025, and separately carried a €250 million currency headwind. Stryker guided to approximately $400 million for 2026. Intuitive Surgical’s initial 2026 gross margin guidance of 67.5 to 68.5 percent reflected 100 basis points of tariff impact — a surgical robot maker taking a full point of margin. These are the sector’s biggest numbers, not its smallest.

The reason is straightforward and worth stating because it is not intuitive: a single MRI or CT system is a globally sourced assembly, and the more countries a bill of materials touches, the more tariff surfaces it presents. Complexity is not insulation; it is exposure area. What complexity does provide is inelasticity of demand and pricing power, which is a different thing.

On absorptive capacity, the conventional view is right, and the numbers support it. Gross margins in this tier run from roughly 68 to 69 percent at Intuitive Surgical to 78 to 79 percent at Edwards Lifesciences, with Abbott at 57.1 percent and Zimmer Biomet around 71 percent. A company with a 78 percent gross margin and pricing power in a concentrated market can absorb a great deal. GE HealthCare’s chief financial officer described tariffs as “neutral to positive to our financial performance” by February 2026, and by mid-2026 tariffs were a minor component of a roughly $250 million total input-cost inflation bucket that also included memory chips, oil and freight. Siemens Healthineers expects to fully offset tariff effects over the medium term.

On substitution speed, the conventional view is right in principle but was contradicted in practice. The regulatory constraint is real: swapping a supplier for a critical implant component can trigger revalidation, and Siemens Healthineers’ chief financial officer noted that specialised sites — superconducting magnet production in Oxford, photon-counting centres — constrain what can move, and that the company would act “only once there is sufficient planning certainty and it makes economic sense.” Yet the single clearest tariff-driven relocation in the entire sector was in this tier: Varian radiotherapy production from Baja California to Palo Alto. Capital equipment did move. It moved because the assembly step was relocatable even where component qualification was not.

On the political read, the conventional view holds. J.P. Morgan’s Robbie Marcus has argued the industry may be largely insulated, citing bipartisan support and trade frameworks with the EU and Japan. RBC’s Shagun Singh has said medical devices are not a sector the administration appears to view negatively. Needham’s Mike Matson expects tariffs eventually but on a gradual escalation model similar to pharmaceuticals. That is a reasonable central case for this tier: gradual, absorbed, and not the source of a price shock.

One risk this tier carries that the others do not: retaliation. US medtech runs a trade surplus with nearly all major partners on roughly $75 to $80 billion of exports. High-value capital equipment is precisely the kind of politically visible export that foreign governments target, and they do not need tariffs to do it. China banned EU medical device procurement in July 2025 in retaliation for the EU’s own International Procurement Instrument measure of 19 June 2025, which excludes Chinese tenderers from device procurements above €5 million and caps Chinese content at 50 percent for winning bidders, with penalties of 10 to 30 percent of contract value. Procurement exclusion, domestic-substitution policy and volume-based procurement are China’s instruments of choice against Western imaging and device vendors, and they bite harder than duties. Canada is the acute near-term risk: the United States supplied approximately $1.4 billion of medical devices to Canada in 2024, nearly 40 percent of Canadian imports into a $9.8 billion market, and the Section 338 action taking effect 19 August 2026 is a formal discrimination finding against Canada in the middle of live USMCA negotiations.

5.4 Diagnostics and laboratory products: the missing category

Diagnostic and laboratory reagents are named in the Section 232 scope, yet almost never appear in category analyses of it. They deserve separate treatment for three reasons.

First, the dependency profile is different. IVD supply chains run on reagents, enzymes, antibodies, lab plastics, pipette tips and consumables, with meaningful Chinese content in the commodity end and concentrated Western and Japanese supply in the instrument end. Laboratory budget analyses through 2025 and 2026 have consistently flagged tariff-driven price increases on research and diagnostic supplies.

Second, the classification boundary risk is highest here. HS 3822 diagnostic reagents and chapter 30 headings 3002, 3005 and 3006 sit at the seam between the pharmaceutical 232 and the medtech 232. A reagent kit could be captured by either. The pharmaceutical proclamation’s Annex I and Annex IV must be read directly against IVD classifications.

Third, the sector organised separately in the docket. The College of American Pathologists filed specifically seeking exemption for laboratory supplies and IVD products, which is a signal that the segment does not consider itself covered by the broader industry asks.

Diagnostics also carries a distinct commercial exposure: Siemens Healthineers cut its fiscal 2026 revenue growth guidance in July 2026 on China diagnostics weakness, unrelated to tariffs. For diagnostics manufacturers, Chinese domestic substitution in the reagent and analyser market is a larger structural threat than US tariff policy.

5.5 Upstream inputs: the exposure that is mispriced

The general point — that inputs which are not themselves devices may matter more to total cost than finished-goods duty rates — is correct and under-appreciated. The specifics usually offered are not.

Tungsten. The claim that the United States sources roughly half its tungsten from China is incorrect. The US Geological Survey’s Mineral Commodity Summaries 2026 reports net import reliance exceeding 50 percent of apparent consumption over 2021–2025, with import sources over 2021–2024 of China 26 percent, Germany 14 percent, Bolivia 8 percent, Vietnam 8 percent and other sources 44 percent. Tungsten has not been mined commercially in the United States since 2015, and seven US companies process concentrates and scrap into finished products. China’s leverage is upstream of the import statistics: it produced 67,000 of an estimated 85,000 metric tonnes of world mine output in 2025, roughly 79 percent, and holds 2.5 million tonnes of reserves, first in the world. In February 2025 China implemented export controls on selected tungsten items, following US tariff increases to 50 percent on Chinese tungsten products. Rotterdam concentrate prices moved from $252 per metric tonne unit in 2024 to an estimated $380 in 2025, a rise of roughly 51 percent. So the vulnerability is real and arguably worse than the original framing suggests — it is a global chokepoint plus an export-control instrument, not a bilateral import share. On the application, tungsten’s dominant use is cemented carbide, around 60 percent of consumption; USGS notes depleted uranium as a substitute in applications requiring high density or radiation shielding, which is the mechanism by which tungsten matters to linear accelerator and radiation-therapy construction.

Metals derivatives. As set out in Section 3.3, the 6 April 2026 shift to full-customs-value assessment on Section 232 derivative products is the largest under-discussed cost event in this space. Combined with the termination of the public inclusions process, it means a device manufacturer can find its effective duty rate multiply without any medtech-specific action being taken and without a formal channel to contest the classification. Manufacturers should be running a full HTS screen of their portfolio against Annexes I-A and I-B, and modelling the difference between metal-content and full-value assessment on every metal-containing line.

Polymers and resins. Becton Dickinson’s disclosure is the most useful public datapoint in the sector on input hedging: approximately 50 percent of its North American resin is hedged, with multi-source resin suppliers and cap-and-roll mechanisms. Resin, not metal, is the dominant input for the consumables tier, and hedging coverage is a competitive variable that receives almost no external analysis.

Semiconductors and memory. Both GE HealthCare and Intuitive Surgical identified memory and semiconductor cost inflation as a live 2026 input pressure alongside tariffs — in GE HealthCare’s case as part of a roughly $250 million total input-cost bucket in which tariffs had become the minor component. Any 2027 cost model that treats tariffs as the dominant input variable is likely to be wrong for reasons that have nothing to do with trade policy.

6. Why substitution is slow: the regulatory arithmetic

The claim that a box of gauze can change factories faster than a Class III implant is correct, but it is usually asserted rather than quantified. The quantification is what makes it actionable.

For devices cleared under section 510(k), the governing question is whether a change to a supplier, material or manufacturing site could significantly affect safety or effectiveness. FDA’s guidance on deciding when to submit a 510(k) for a change to an existing device, finalised in October 2017, sets out the decision logic; where the answer is no, the change is documented in a letter to file, and where it is yes, a new premarket notification is required. A traditional 510(k) carries a FY2027 user fee of $28,653, reduced to $7,163 for qualifying small businesses, and runs against MDUFA performance goals of a 15-day acceptance review, substantive interaction by day 60, and a decision within 90 FDA review days — with the calendar-time total materially longer where an additional information request stops the clock.

For devices approved under a premarket approval application, the arithmetic is harsher. FDA’s December 2018 guidance on manufacturing site change supplements provides that a site change generally requires a PMA supplement with a 180-day statutory review period, and specifies thirteen categories of required information including process flow diagrams, validation master plans and protocols, environmental and contamination controls, calibration procedures, supplier and contract manufacturer controls, and incoming and final acceptance procedures. FDA may conduct a pre-approval inspection depending on device risk profile, inspection history and recall history. A 30-day notice, at a FY2027 fee of $10,188, is available only where the change does not require a supplement. The fee schedule for the alternatives is the clearest statement of relative burden: 180-day supplement $95,510; real-time supplement $44,571; panel-track supplement $509,386; a new PMA $636,732. Annual establishment registration is $13,785.

Two implications follow for tariff strategy.

The first is that the cost of qualifying an alternative source is not the user fee. It is the fee plus process validation, biocompatibility and sterilisation revalidation where materials change, stability work where relevant, the internal quality and regulatory labour to assemble a thirteen-category submission, the risk of a pre-approval inspection, and the elapsed calendar time during which the original source must continue to be paid. For a Class III implant, a duty of 10 to 25 percent on a component may simply be cheaper than qualification. That is the honest reason complex devices do not re-source: not that it is impossible, but that the arithmetic frequently does not clear.

The second is that this asymmetry should shape which tariff mitigation levers a manufacturer reaches for. Where regulatory change control makes physical re-sourcing uneconomic, the available levers are customs and commercial rather than operational — classification review, origin and country-of-origin planning, first-sale valuation, transfer pricing, bonded warehousing and entry-date control, foreign trade zone status elections, and drawback where available. Zimmer Biomet’s disclosed playbook is instructive precisely because it is a customs playbook: optimise country of origin, adjust transfer pricing, dual and redundant sourcing, pre-position inventory ahead of tariff dates, and evaluate European sourcing for China-destined product.

7. Margin capacity: who can actually absorb a tariff

The single most useful way to rank exposure is not by product category but by gross margin, because gross margin is the measure of how much duty a business can swallow before it must either raise price or shrink.

At the top, Edwards Lifesciences held gross margin guidance at 78 to 79 percent through the tariff period and disclosed exposure of around $0.025 per share — a rounding error. Zimmer Biomet at roughly 71 percent, Intuitive Surgical at 68 to 69 percent, Dexcom at 64.1 percent and Abbott at 57.1 percent all have room. These businesses absorbed, offset and in several cases raised guidance.

At the bottom, the picture inverts. ICU Medical disclosed a first-quarter 2026 tariff expense equal to approximately 2 percent of adjusted revenue and a full-year range of $40 to $50 million against an adjusted earnings guide of $7.75 to $8.45 per share — a burden that consumes a material share of profit. Teleflex reported a 470 basis point year-on-year gross margin decline in the first quarter of 2026, driven primarily by tariffs and quality remediation. Steris reported $46 million and 80 basis points for fiscal 2026 and guided to $60 to $65 million for fiscal 2027 assuming no refunds. Becton Dickinson took 160 basis points of adjusted gross margin in the second quarter of fiscal 2026 and has said pricing is “a very important topic for us heading into FY ’27.” Commodity glove manufacturers operate in the low teens on gross margin and have no capacity at all.

The distributors sit in a separate and more precarious position again, running low-single-digit gross margins on a pass-through model; a duty they cannot immediately pass through is not a margin compression event but a loss event. Danaher is one of the few names in the broader sector to have explicitly cited surcharges as a mitigation, alongside footprint changes and cost cutting; Cardinal Health, facing up to $300 million, announced layoffs.

The practical conclusion for a manufacturer assessing its own position is to stop asking which product category it sits in and start asking what proportion of gross profit a given duty rate consumes, product line by product line. A 15 percent duty on a 78 percent gross margin implant is an inconvenience. The same duty on a 12 percent gross margin consumable is an exit decision.

8. The demand side: why pass-through is blocked

The reason absorption rather than pass-through has been the sector’s revealed strategy is not restraint. It is structural, and it operates on two levers.

The first is contracting. Hospital supply is bought under multi-year agreements, frequently through group purchasing organisations, with pricing fixed for the term. AdvaMed’s central argument in the docket was that hospitals “operating under multi-year purchasing contracts and fixed reimbursement structures cannot easily absorb sudden cost spikes.” The corollary, less often stated, is that those same contracts prevent the manufacturer from repricing mid-term. Whether a tariff can be passed on depends on whether a specific contract contains a material-change, price-adjustment or force-majeure mechanism that a duty increase triggers — and that is a contract-by-contract question, not an industry-level one. Premier asked Commerce for clear definitions of which components are dutiable precisely because the allocation of tariff cost within existing contracts is ambiguous.

The second is reimbursement. Device prices are embedded in bundled procedure payments — diagnosis-related groups on the inpatient side, ambulatory payment classifications on the outpatient side — set administratively and updated annually. A hospital that pays more for an implant does not receive more for the procedure. Needham’s Mike Matson identified this as the reason medtech is worse positioned than pharmaceuticals under Section 232: pharmaceutical manufacturers could trade drug price concessions for tariff relief, but a device maker cannot cut a knee replacement price by 20 percent in a way that saves the government money, because the government is not paying for the knee separately. The pharmaceutical 232’s 0 percent tier for companies accepting most-favoured-nation pricing plus onshoring has no medtech analogue, and could not easily be constructed.

The measured outcome bears out the constraint. Vizient’s July 2025 projection for 2026 put overall healthcare supply chain inflation at 2.41 percent, with surgical supplies at 3.28 percent and pharmacy separately at 3.35 percent. A Premier survey in January 2025 had found 82 percent of healthcare industry experts expecting tariff-related import expenses to raise hospital costs by 15 percent within six months, a figure the American Hospital Association continued to cite in its July 2026 filing alongside a finding that 90 percent of supply chain professionals anticipated procurement disruption. The realised number is roughly a fifth of the feared one. That delta went onto manufacturers’ income statements.

For context on the base, US hospital medical and surgical supply costs rose from roughly $40 billion in 2020 to approximately $57 billion in 2025, an average of about 8.2 percent a year, averaging $16.5 million per hospital in 2024 across more than 3,600 institutions. Supplies represented 13 percent of hospital expenses in 2024, or approximately $202 billion, against labour at 56 percent and $890 billion. AHA reporting in March 2026 put 2025 increases at 9.9 percent for medical supplies, 7 to 15 percent for equipment and 13.6 percent for drugs — and notably attributed these to more expensive global supply chains, transportation and raw material costs rather than naming tariffs as the driver. Medicare covered 83 cents of every dollar of hospital cost in 2023. There is no headroom on the buy side, which is the reason the sell side has had to eat it.

9. Mitigation: what works, what does not, and the stacking map

9.1 The stacking architecture

As of 5 August 2026 an imported medical device can face up to six independent duty layers: the most-favoured-nation base rate; the Section 301 forced-labour layer at 10 or 12.5 percent; country-specific Section 301 duties including the China Lists and the strategic medical rates; Section 232 duties on metals, and where applicable pharmaceuticals; Section 338 on Canadian goods from 19 August 2026; and antidumping or countervailing duties. The order of analysis is classification, MFN, country-specific 301, forced-labour 301, Section 232, Section 338, then AD/CVD.

The interaction rules are not intuitive, and getting them wrong in either direction is expensive.

Goods already subject to Section 232 duties are exempt from the Section 301 forced-labour layer. This is a genuine and counterintuitive planning lever: a metal-containing device captured by the metals 232 derivative lists escapes the 10 to 12.5 percent forced-labour duty. Section 338 likewise does not apply to goods under Section 232, but it does stack with Section 301 and is not waived by USMCA origin.

The forced-labour layer stacks additively with MFN and with country-specific Section 301 duties. For the EU, Japan, Korea, Switzerland and Taiwan it is a cap net of MFN rather than an adder — which, as noted, delivers nothing on the many device lines where MFN is near zero.

The pharmaceutical Section 232 operates as a ceiling rather than an adder, applying the higher of the 232 rate or Column 1, and the lower where two 232 rates apply. It is unique among the 232 programmes in this respect.

Two stale assumptions to purge from internal memos: IEEPA is gone entirely as of 20 February 2026, and the Section 122 mutual-exclusivity carve-out died with Section 122 on 24 July 2026. Any stacking analysis dated in the first half of 2026 needs rewriting.

9.2 Tools that work

Entry-date and withdrawal-date control is the highest-leverage near-term lever, because the applicable rate is fixed at entry or withdrawal for consumption. Bonded warehousing assesses duty at withdrawal rather than arrival, which makes it the natural instrument for timing around known cliffs — 19 August 2026 for Section 338 on Canadian goods, 29 September 2026 for the pharmaceutical 232’s general effective date, and any medtech 232 effective date in the September window.

First-sale valuation reduces the dutiable base to the manufacturer-to-middleman price and works simultaneously against every ad valorem layer, which makes it disproportionately valuable in a six-layer stack. Its critical limitation, given the shape of the likely remedy, is that it is useless against per-unit specific duties of the kind the Coalition for a Prosperous America proposed for gloves and syringes.

Origin and transfer-pricing optimisation is the lever the industry has actually used. Zimmer Biomet disclosed country-of-origin optimisation and transfer pricing adjustments; GE HealthCare enhanced USMCA duty-free qualification and reconfigured bonded logistics routes.

Foreign trade zones offer duty deferral and, in principle, inverted-tariff relief. In practice Section 232 goods have historically been required to enter in privileged foreign status, which defeats the inverted-tariff benefit; Section 338 goods must enter in privileged foreign status on or after 19 August 2026. The status election has real consequences and should be modelled, not defaulted.

Duty drawback recovers up to 99 percent of duties on subsequent export within a five-year window, and matters unusually much to this sector because medtech is a net exporter. The critical caveat is that drawback has been statutorily unavailable for Section 232 duties and for most Section 301 duties. That is precisely why both the US Chamber and MDMA asked Commerce to make drawback available under any medtech 232 action. Do not assume eligibility; confirm it programme by programme.

Nairobi Protocol treatment under HTSUS 9817.00.96 is the most underused medtech-specific provision. It provides duty-free entry for articles specially designed or adapted for the use or benefit of persons with permanent or chronic physical or mental impairments, with chronic conditions such as diabetes and cardiovascular disease qualifying and transient conditions not. Qualifying examples include blood glucose monitors, oxygen devices, pacemakers, insulin pumps and prosthetic limbs. Zimmer Biomet cited the Nairobi Protocol on its first-quarter 2026 call in the context of steel-related Section 232 exposure, and the American Hospital Association asked in the docket that the treatment be preserved. The important caution is that whether Chapter 98 provisions override Section 232 or Section 301 duties is unresolved and contrary to general practice under the current regime; a binding ruling request is the only safe path.

Protective litigation filing is, for the refund question, decisive rather than optional, since the largest tranche of IEEPA recovery is available on the government’s position only to CIT plaintiffs.

9.3 Tools that no longer exist

The public inclusions process for Section 232 metals derivatives was terminated on 6 April 2026, and there has never been a product-exclusion process under the current metals programme. Advocacy on derivative classification must now run informally through Commerce and USTR. For the pharmaceutical 232, the analogous relief is company-specific negotiated onshoring agreements, for which the application deadline was 12 June 2026. Manufacturers should assume that under any medtech 232 there may be no formal exclusion channel at all — which is exactly why the Coalition for a Prosperous America asked for one to be created and why the industry associations asked for scope to be narrowed at the design stage rather than litigated afterwards.

10. Risks and options the standard analysis omits

Shortage risk is a distinct failure mode from cost risk. The 2024 disruption to Baxter’s North Cove facility after Hurricane Helene demonstrated how a single site can produce a national shortage of intravenous solutions. FDA’s section 506J authority requires manufacturers to notify the agency of discontinuances and interruptions in the supply of devices critical to public health, and FDA maintains a device shortage list — but the authority is a reporting mechanism, not a supply guarantee. A duty high enough to make a low-margin consumable uneconomic to import does not produce domestic substitution at speed; it produces exit. Premier’s request that Commerce conduct impact studies to surface shortages before acting is the correct instinct, and manufacturers of low-margin single-source consumables should expect to be asked by customers and regulators alike whether a tariff would cause them to exit a line.

Small and mid-size manufacturers face a different problem from large caps. The mitigation playbook described in Section 9 presumes customs and trade expertise, treasury capacity to carry duty on inventory, litigation budget to file protectively at the CIT, and negotiating leverage with suppliers. Large caps have all four. A sub-$100 million device company, a contract manufacturer working to a customer’s fixed price, or an importer of a single consumable line has none. MDMA’s argument in the docket that tariffs would push firms to stop serving other markets from US facilities is a small-company argument, and the mechanism is real: when a US-based manufacturer’s imported components become dutiable and its exports cannot recover drawback, offshoring the whole operation becomes the rational response. That is the “tariffs meant to reshore medtech would send it offshore” case, and it applies with most force to the companies least able to be heard in the docket.

Reprocessing is an underrated cost lever with tariff logic behind it. The US reprocessed medical devices market was approximately $0.81 billion in 2025 and is forecast at $0.94 billion for 2026 and $2.30 billion by 2031, a compound growth rate of roughly 16.6 percent. Reprocessor member companies reported generating $495.5 million in hospital savings across nearly 11,500 facilities in 2025. Cardiovascular devices account for around 56.5 percent of the market, with electrophysiology procedure costs reducible by roughly 30 percent; acute care hospitals represent about 62.9 percent of volume and ambulatory surgery centres are the fastest-growing segment at around 18.9 percent. The regulatory pathway is widening — Innovative Health reached its fiftieth FDA clearance in January 2026 and Restore Robotics secured approvals for robotic instrument reprocessing. For original equipment manufacturers, the tariff-driven cost pressure on single-use devices is a structural argument in favour of the reprocessing channel, which competes directly with new-unit sales. Manufacturers should treat reprocessing growth as a demand-side consequence of their own pricing decisions.

Compounding cost pressures make 2026 a bad year to be measuring tariffs in isolation. European Union Medical Device Regulation and In Vitro Diagnostic Regulation compliance continues to absorb regulatory and quality capacity that might otherwise be available for tariff-driven change control — a real constraint on how many second-source qualifications a company can run at once. Memory and semiconductor cost inflation, freight, and oil were all named by GE HealthCare and Intuitive Surgical as 2026 input pressures, with GE HealthCare’s total input-cost inflation bucket at roughly $250 million and tariffs the minor component. Hospital capital budgets remain constrained by the reimbursement gap described in Section 8. A manufacturer that solves its tariff problem in an environment of 8 percent annual supply cost inflation and constrained hospital capital has solved the smaller half of its cost problem.

The import-dependence statistics used in this debate do not agree with each other, and manufacturers should know which one they are being quoted. FDA data cited by the American Hospital Association in July 2026 puts approximately 49 percent of medical devices used in the United States as imported, while nearly 70 percent of devices marketed in the United States are manufactured exclusively overseas. The University of North Carolina’s Center for the Business of Health, writing on 1 November 2025, used 62 percent. AdvaMed’s framing is that roughly 70 percent of medtech used domestically is made domestically. These are not all measuring the same thing — units versus value, devices used versus device models marketed, and finished goods versus components diverge sharply — and the difference between “49 percent imported” and “70 percent made exclusively overseas” is doing a great deal of rhetorical work in the docket. Any submission or customer communication that cites one of these figures should say which measure it is.

11. Scenarios

Scenario A: narrow remedy on gloves and syringes with allied exclusions. This is the shape best supported by the comment record and by the administration’s existing behaviour, having already taken Chinese gloves and syringes to 100 percent under Section 301. Design features to expect include specific per-unit duties as well as or instead of ad valorem rates, trusted-partner exclusions for USMCA, EU, Japan and Korea, anti-circumvention provisions aimed at transshipment, and possibly the stockpiling and federal offtake elements that both MDMA and the Coalition for a Prosperous America requested. Impact: severe for PPE and basic consumables importers, negligible for capital equipment and implants. Mitigation note: per-unit duties defeat valuation-based mitigation, so the levers are origin, exclusion advocacy and contract pass-through language.

Scenario B: no action, docket allowed to lapse. Given the commercial aircraft precedent and analyst readings that the administration is not targeting the sector adversarially, this is a live possibility rather than a tail case. Impact: the industry’s tariff cost in 2027 is then determined almost entirely by the metals 232 derivative treatment, the Section 301 forced-labour layer, and the outcome of the refund litigation — none of which would be affected. Manufacturers should note that Scenario B is not a benign outcome; it is the status quo, and the status quo already costs the sector billions annually.

Scenario C: broad phased ad valorem tariff on devices. Needham’s expectation of gradual escalation on the pharmaceutical model. Impact: material margin compression across all tiers, absorbed by large caps over two to three years, existential for low-margin consumables and distributors. Watch for a tiered structure with reduced rates conditional on US manufacturing commitments, mirroring the pharmaceutical 20 percent onshoring tier — and note that the pharmaceutical 0 percent most-favoured-nation-pricing tier has no device analogue because device prices are not separately reimbursed.

Scenario D: scope expansion into diagnostics and reagents. Underweighted in most analyses because diagnostics is usually omitted from the category framework. Reagents are explicitly in the medtech 232 scope and simultaneously adjacent to the pharmaceutical 232’s chapter 29 and 30 universe. A remedy that captures reagents and laboratory consumables would hit clinical laboratories, IVD manufacturers and research institutions with almost no substitution options and almost no public preparation.

12. Watch list: dates that matter

13. Recommended actions for manufacturers

Immediate, before the September window. Run a full HTS classification screen of the product portfolio against the Section 232 metals Annexes I-A and I-B, modelling the difference between metal-content and full-customs-value assessment on every metal-containing line — this is likely to surface more cost than any medtech 232 scenario. Screen the three Section 338 Canadian annexes for medical lines before 19 August. Read Annex I and Annex IV of the pharmaceutical proclamation against any combination product, prefilled device, kit or reagent in the portfolio. Confirm whether protective filings have been made at the Court of International Trade for IEEPA refund recovery, and file protests on Section 122 liquidated entries within the 180-day window.

Contract remediation. Identify every customer contract under which the company bears tariff cost with no pass-through mechanism, and every supplier contract under which it bears a supplier’s duty. Ahead of the 29 September pharmaceutical effective date and any medtech action, this is the highest-value commercial work available, because contract terms fixed now will govern through the entire remedy period.

Segment the internal exposure model by gross margin, not product category. Express duty exposure as a percentage of gross profit by product line. The lines where a plausible duty rate consumes more than a quarter of gross profit are the ones requiring a decision — reprice, re-source, or exit — and the decision needs to be made before the rate lands, not after.

Sequence regulatory change control against the tariff calendar. Because second-source qualification for Class II devices runs on a 510(k) or letter-to-file timeline and Class III devices on a 180-day supplement timeline with possible pre-approval inspection, the qualifications that could plausibly complete within twelve months should be started now for the highest-exposure lines. Recognise that EU MDR and IVDR workload competes for the same regulatory capacity, and prioritise explicitly rather than implicitly.

Do not model second-quarter 2026 margins forward. Strip IEEPA refunds out of the internal baseline and out of any competitor benchmark. Several competitors have taken refunds into reported margin and several have not, and the resulting apparent divergence is an artefact.

Engage on remedy design, not remedy existence. The docket evidence suggests the industry’s realistic influence is over scope definitions, allied-country exclusions, USMCA treatment, drawback availability, phasing, and whether a formal exclusion process exists — not over whether action occurs. The US Chamber’s ten asks are a template for that kind of engagement. Where a company’s exposure is concentrated in a category that a narrow remedy would capture, the argument to make is about domestic capacity reality — the $850 million glove programme’s outcome is the strongest available evidence that a tariff without offtake commitments produces cost without capacity.

Test the Nairobi Protocol properly. For any product plausibly within HTSUS 9817.00.96 — glucose monitors, insulin pumps, pacemakers, prosthetics, oxygen devices — seek a binding ruling rather than relying on secondary commentary about whether Chapter 98 treatment survives Section 232 and Section 301. The provision is the only medtech-specific duty-free channel in the tariff schedule and it is under-used.

14. Corrections to commonly cited claims

This section documents where widely repeated figures in the medtech tariff discussion are inaccurate, outdated or incomplete, with the better-sourced position.

“Roughly 99 percent of medical gloves used in the US are imported.” Supported. Domestic production is approximately 1 percent of roughly 120 billion nitrile gloves consumed annually (reported 7 July 2026).

“Malaysia now supplies about 60 percent of US medical gloves.” Understated and incomplete. Malaysia supplied 67.5 percent of US medical nitrile and latex glove imports in January–July 2025 (data published 12 September 2025), with Thailand at 16.9 percent and Vietnam at 9.8 percent. More importantly, the figure conceals a category split: China supplied only 2.8 percent of medical nitrile and latex but 90.8 percent of vinyl gloves in the same period. Glove exposure must be assessed by glove type.

“Malaysian glove production still depends on Chinese nitrile butadiene rubber feedstock.” Directionally supported but not precisely quantified in public sources. China supplies raw materials to Malaysian, Thai and Vietnamese glove producers, and Blue Star NBR’s Virginia project was to be the first domestic producer of medical-grade NBR in more than thirty years — which corroborates the absence of US feedstock capacity. A specific Chinese share of global NBR feedstock supply to glove producers could not be verified and should not be asserted as a number.

“The US relies on Chinese producers for more than half of critical medical consumables.” Provenance not established, and the aggregate obscures a highly uneven product-level picture. Better-sourced product-level figures, cited by the American Hospital Association from 2023 data: China supplied the majority of N95 and other respirators, one-third of disposable face masks, two-thirds of non-disposable face masks and 94 percent of plastic gloves used in US healthcare. Note also that Chinese direct share of medical nitrile gloves has since collapsed to under 3 percent. Use product-level figures rather than the aggregate.

“Mexico currently carries a 25 percent tariff rate, and Europe around 20 percent.” Outdated and incorrect as of August 2026. USMCA-originating Mexican goods are exempt from the 10 percent Section 301 forced-labour layer entirely; non-originating goods bear duty on non-US content subject to a 15 percent minimum effective rate. EU goods carry 10 percent capped net of MFN, within a 15 percent general framework ceiling. Neither figure should be used.

“A 10 to 25 percent cost increase is the relevant range for consumables.” Understates current reality. Chinese medical and surgical gloves and Chinese syringes and needles have carried 100 percent Section 301 duty since 1 January 2026 and September 2024 respectively; respirators and disposable masks are at 50 percent.

“The US sources roughly half its tungsten from China.” Incorrect. USGS Mineral Commodity Summaries 2026 reports China at 26 percent of US tungsten imports over 2021–2024. US net import reliance exceeds 50 percent of apparent consumption overall — a different statistic. China’s actual leverage is its roughly 79 percent share of world mine production and its February 2025 export controls on selected tungsten items. The vulnerability is real; the stated mechanism is not.

“Complex capital equipment is the most insulated segment.” Wrong on exposure magnitude, right on absorptive capacity. Capital equipment makers reported the largest absolute tariff bills in the sector — GE HealthCare approximately $500 million gross for 2025, Siemens Healthineers €400 million for fiscal 2026, Stryker approximately $400 million for 2026, Intuitive Surgical 100 basis points of gross margin guidance. What insulates them is 68 to 79 percent gross margins and pricing power, not low exposure.

“Complex capital equipment is least likely to see rapid tariff-driven change.” Contradicted by events. The clearest tariff-attributed manufacturing relocation in the sector was Siemens Healthineers moving Varian radiotherapy production from Baja California to Palo Alto, announced 16 May 2025, one week after its €300 million tariff warning.

“Section 232 is the pending question that will determine medtech tariff exposure.” Incomplete. As of August 2026 no medtech 232 report or proclamation has been confirmed, while the industry has already absorbed roughly eighteen months of cost under IEEPA, Section 122 and Section 301, and the largest live cost change is the 6 April 2026 shift to full-customs-value assessment on Section 232 metals derivatives.

Diagnostics and laboratory reagents are omitted from most category analyses despite being expressly named in the Section 232 scope, sitting at the classification boundary with the pharmaceutical 232, and having organised separately in the docket through the College of American Pathologists.

Sources 

Dates given are publication or last-update dates. Figures should be re-verified before use in filings or contracts.

Primary and government

• • • Federal Register, “Notice of Request for Public Comments on Section 232 National Security Investigation of Imports of Personal Protective Equipment, Medical Consumables, and Medical Equipment, Including Devices,” 90 FR 46383, 26 September 2025. [Link]
    • Bureau of Industry and Security, Section 232 Investigations status page. [Link}
    • Supreme Court of the United States, Learning Resources, Inc. v. Trump, No. 24-1287, decided 20 February 2026. [Link}
    • White House, “Adjusting Imports of Pharmaceuticals and Pharmaceutical Ingredients into the United States,” proclamation, 2 April 2026. [Link}
    • US Geological Survey, Mineral Commodity Summaries 2026, tungsten chapter.[Link}
    • FDA, “Medical Device User Fee Amendments (MDUFA): Fees,” FY2027 schedule. [Link}
    • FDA, “Manufacturing Site Change Supplements: Content and Submission,” guidance, December 2018. [Link}
    • FDA, “Medical Device Supply Chain and Shortages,” including section 506J reporting. [Link}
    • Office of the United States Trade Representative, Section 301 forced-labour findings and final action, June–July 2026. [Link}
    • USTR, “Ambassador Greer Issues Statement on USMCA Joint Review,” July 2026. [Link}
    • European Commission, “Commission restricts Chinese participation in medical devices procurement,” 19 June 2025. [Link}
    • International Trade Administration, Canada Country Commercial Guide, Medical Devices, updated 28 April 2026. [Link}

Legal analysis

• • • White & Case, “Trump administration initiates Section 232 investigation on medical supplies.” [Link]
    • White & Case, “United States Modifies Steel, Aluminum and Copper Section 232 Tariffs,” 7 April 2026. [Link}
    • White & Case, “USMCA 2026 Joint Review: United States Declines to Extend Agreement,” 2 July 2026. [Link}
    • White & Case, “Trump Administration Imposes 50% Tariffs on Certain Canadian Products: First Use of Section 338.” [Link}
    • White & Case, “United States Finalizes Section 301 Tariff Increases on Imports from China.” [Link}
    • Covington & Burling, “Current and Forthcoming Section 232 Actions by the Trump Administration,” 23 April 2026. [Link}
    • Holland & Knight, “Section 232 Watch: Industry Warily Eyes a Slate of Summer Deadlines,” 8 June 2026. [Link}
    • Holland & Knight, “IEEPA Tariff Refund Update: Government Appeals,” 15 June 2026. [Link}
    • Holland & Knight, “And the Tariff Beat Goes On,” 30 July 2026. [Link}
    • Holland & Knight, “’50 Percent Opening Bid’: Canadian Imports Subject to Section 338 Tariffs,” July 2026. [Link}
    • Skadden, “The Supreme Court Ends IEEPA Tariffs,” February 2026. [Link}
    • Crowell & Moring, “Trump Administration Imposes Section 232 Tariffs on Patented Pharmaceutical Imports,” April 2026. [Link}
    • Perkins Coie, “Restructured and Additional Section 232 Tariffs: Aluminum, Steel and Copper.” [Link}
    • Morgan Lewis, “Tariff Refund Battle Continues: Government Appeals Order,” June 2026. [Link}
    • Hogan Lovells, “China announces new restrictions on the participation of certain EU-origin medical devices in government procurement activities,” 2025. [Link}
    • Mondaq, “Exclusion Of Chinese Medical Devices From Public Procurements,” 12 August 2025. [Link}
      
    • RSM, “Nairobi Protocol tariff relief for medtech importers.” [Link}

Industry associations and docket filings

• • • AdvaMed, “AdvaMed Responds to Commerce Department Section 232 National Security Investigation,” 17 October 2025. [Link]
    • US Chamber of Commerce, “Comments on Section 232 Medical Supplies and Devices Investigation,” 20 October 2025. [Link}
    • American Hospital Association, “AHA Comments on Proposed Tariffs on Medications, Medical Devices,” 6 July 2026. [Link}
    • American Hospital Association, “AHA Responds to Department of Commerce’s RFI on PPE and Medical Equipment,” 17 October 2025. [Link}
    • American Hospital Association, 2025 Cost of Caring Report, published 9 March 2026. [Link}
    • Premier Inc., “Premier weighs in on Section 232 investigation into medical devices and tariffs,” 16 October 2025. [Link}
    • Premier Inc., supply chain special report on tariffs, January 2025. [Link}
    • Vizient, “Vizient projects continued cost pressures across the healthcare supply chain in 2026,” 29 July 2025. [Link}
    • Coalition for a Prosperous America, Section 232 comments on PPE, medical consumables and medical equipment, October 2025. [Link}
    • Business Roundtable, Section 232 comments. [Link}

Trade and financial press

• • • MedTech Dive, “Section 232 probe reignites tariff uncertainty for medtech firms,” 26 September 2025. [Link]
    • MedTech Dive, “Medtech industry closely watching Section 232 investigation.” [Link}
    • MedTech Dive, “One year in: How medtech companies are coping with tariff challenges,” 9 April 2026. [Link}
    • MedTech Dive, “AdvaMed urges tailored tariff approach for medtech,” 24 February 2026. {Link}
    • MedTech Dive, “GE HealthCare expects smaller tariff impact in 2026,” 5 February 2026. [Link}
    • MedTech Dive, “Siemens Healthineers predicts tariff impact will double,” 6 November 2025. {Link}
    • MedTech Dive, “Siemens Healthineers to move production to US from Mexico,” 2025. {Link}
    • MDDI, “Philips posts strong Q2 growth, relies on tariff refunds to boost margins and cash flow,” 28 July 2026. [Link}
    • MDDI, “Inside Zimmer Biomet’s playbook for navigating global trade tensions.” [Link}
    • MDDI, “Trump Finds New Way to Tariff Medical Devices via Section 232.” [Link}
    • Medical Economics, “What the Supreme Court’s tariff ruling means for medtech and device costs,” 23 February 2026. [Link}
    • Becker’s Hospital Review, “$850M later, US medical glove production still lags,” 7 July 2026, reporting Bloomberg. [Link}
    • Bloomberg, “Why It’s So Difficult to Produce 100% American-Made Medical Gloves,” 7 July 2026. {Link}
    • The Edge Malaysia, “Glove majors take opportunity to reclaim margins, market share amid US-China trade war,” February 2025. [Link}
    • 360Dx, “Siemens Healthineers ups FY26 EPS guidance, lowers revenue outlook,” 31 July 2026. [Link}
    • GE HealthCare, second quarter 2026 results, 29 July 2026. [Link}
    • Philips, second quarter 2026 report, 28 July 2026. [Link}

Data and research

• • • INTCO Medical, “U.S. Disposable Gloves Import Data Brief (Jan–Jul 2025),” 12 September 2025. [Link]
    • UNC Kenan Institute, Center for the Business of Health, “Tariffs on Medical Devices and Supplies: Healthcare Cost Implications,” 1 November 2025. {Link}
    • Definitive Healthcare, “Annual Hospital Medical Supply Cost Changes,” 29 December 2025. [Link}
    • IQVIA, “Tariffs, Trade and Transformation,” 4 September 2025. [Link}
    • Tetakawi, “Medical Device Manufacturing in Mexico: The Complete Industry Guide,” 10 March 2026. [Link}
    • Mordor Intelligence, “U.S. Reprocessed Medical Devices Market,” updated 22 May 2026. [Link}
    • Journal of Political Risk, Scott Maier, “How a $0.03 Nitrile Glove Could Shut Down America’s Reindustrialization,” 9 February 2026. [Link}

16. Method and limitations

This report was compiled on 5 August 2026 from public sources: Federal Register notices and presidential proclamations, Supreme Court and Court of International Trade materials, US Geological Survey and FDA publications, law firm client alerts from firms active in the Section 232 dockets, industry association filings in Commerce docket 250924-0160 and regulations.gov docket BIS-2025-0258, company earnings releases and call transcripts for the period from the first quarter of 2025 through the second quarter of 2026, and trade press.

Four limitations should be understood before the report is relied upon.

First, whether Commerce has transmitted its Section 232 report to the President cannot be determined from public sources, because transmission is not required to be public. The report’s central status finding — that no report or proclamation has been publicly confirmed as of 5 August 2026 — is an absence-of-evidence finding, not a confirmation of inaction.

Second, three verification items were identified and not closed, and each is material enough that a manufacturer should close it directly rather than relying on this document: whether specific medical device HTSUS lines appear in Annexes I-A and I-B of the 6 April 2026 metals proclamation; whether medical lines appear in the three Section 338 Canadian annexes taking effect 19 August 2026; and whether chapter 30 headings 3002, 3005 and 3006 and HS 3822 diagnostic reagents appear in Annex I or Annex IV of the pharmaceutical proclamation. All three require line-by-line annex review against a specific product portfolio.

Third, several country-level rate cells in Section 3.5 rest on trade-service aggregators rather than primary CBP guidance. They are internally consistent across independent sources and are reliable for planning, but specific rates should be verified against CBP CSMS messages before being used for entry filings.

Fourth, one important claim could not be substantiated and is flagged rather than repaired: the specific Chinese share of nitrile butadiene rubber feedstock supply to Southeast Asian glove producers. The dependency is corroborated qualitatively from several directions but no reliable quantitative figure was located.