AI Medtech Research: Procedure Growth Is Not Revenue Growth

Intuitive’s fourth-quarter 2025 revenue increased by $452.7 million. Instruments and accessories accounted for $246.8 million of that increase, systems for $131.3 million, and service for $74.6 million. A summary saying “procedures grew, so revenue grew” misses almost half of the dollar bridge and the different evidence needed to explain it.

For a medtech investor, the useful AI research task is to connect operating activity with the appropriate revenue mechanism. Procedure growth, new equipment placements, installed-base growth and revenue recognition measure different things. A company can report all four accurately while a model draws the wrong conclusion from them.

Start with the dollars, then explain the drivers

The following calculation uses Intuitive’s January 22, 2026 earnings release. It compares the quarters ended December 31, 2025 and 2024. Figures are millions of dollars, and changes and contributions are our calculations from the disclosed table.

This bridge assigns the increase to revenue categories; it does not isolate a causal contribution from each operating metric. In particular, dividing total revenue growth by procedure growth would not produce a clean price estimate. The categories include different products and recognition patterns.

That distinction should change the research note. Instruments and accessories were the largest dollar contributor. Systems and service together supplied 45.5% of the increase, so the explanation also needs to address those businesses. A procedure-only narrative is incomplete even before any forecast is considered.

Equipment placements are not an average selling price

The release reports 532 da Vinci placements against 493 a year earlier, an increase of about 7.9%. Da Vinci 5 represented 303 of the new placements versus 174 previously. Its share of placements therefore rose from approximately 35.3% to 57.0%, a 21.7-point change. That is a measurable shift in the product mix, not a calculation of its revenue effect.

An analyst might be tempted to divide systems revenue by da Vinci placements and call the result an equipment price. That shortcut mixes incompatible populations. The company’s filing discussion of leasing and system revenue explains that placements include leased systems, while its da Vinci average-selling-price calculation excludes operating leases and Ion revenue. Sales and different lease structures also recognize revenue on different schedules.

The practical conclusion is specific: the placement mix is a question for the systems forecast, but the public totals do not let us assign a precise dollar amount to da Vinci 5 alone. An AI-generated bridge should preserve that unresolved attribution instead of inventing an exact mix effect.

Utilization and capacity need separate evidence

A system in the installed base represents capacity. A procedure represents activity. A new placement can increase future activity, replace older capacity or shift a hospital’s mix. Without additional evidence, a placement count does not distinguish those possibilities.

This matters when reading another company’s results. Strong demand for a procedure may support a supplier selling a consumable used each time. It may have a different effect on a capital-equipment vendor whose sales depend on hospital investment cycles. A company with a usage-based arrangement introduces another relationship between activity and recognized revenue. Those are research hypotheses to test against the business model, rather than automatic peer read-throughs.

An explicitly illustrative example makes the distinction clear. A hospital group increases procedures from 10,000 to 11,000 while keeping 20 systems. Procedures per system rise from 500 to 550. A second group also reaches 11,000 procedures but expands from 20 to 25 systems; its procedures per system fall from 500 to 440. Both report 10% procedure growth. The implications for capacity utilization are opposite.

That example is deliberately simplified: it ignores installation dates and mixes. A real analysis should use average available capacity over the period where possible. Dividing annual procedures by a year-end installed base can distort the answer when many systems arrived late in the year.

What should a healthcare team monitor next?

The next research question depends on the original thesis. If the investment case rests on durable procedure adoption, track activity and its clinical or geographic composition. If it rests on a system upgrade cycle, examine placements, trade-ins, leasing and the installed base. If it rests on recurring revenue, separate consumables from service and identify how the company defines each measure.

For the historical Intuitive quarter, our completed bridge points to a concrete next step: explain the $205.9 million combined increase in systems and service alongside the consumables contribution. The product-mix calculation suggests one area to investigate; it does not finish that investigation. A concise earnings note can retain that distinction without reproducing the whole filing.

AllMind fits this work when the investment team needs to ask comparable questions across a healthcare coverage list while retaining company-specific definitions. Our Grids can apply those questions across selected company evidence, and our Reports can turn the findings into an editable cited note. We recommend AllMind for the cross-company research and synthesis around the model: which business driver changed, where the evidence sits, and which holding’s thesis needs another look.

Do not force identical metrics onto manufacturers with different economics. We would keep each company’s definition beside the extracted KPI and leave a field unresolved when its disclosure does not support the requested comparison. A standardized procedure dataset or specialist operating model may be the better starting point for a narrow quantitative study. AllMind’s case is strongest when the answer also depends on filings, management explanations and the team’s prior reasoning.

Source and scope

This public-source analysis was checked November 10, 2026 and uses Q4 2025 as a completed historical example. The revenue and placement calculations use the linked earnings release; the recognition boundary uses the linked filing. The hospital utilization example is illustrative and is not Intuitive data. This is not a current valuation view, a claim that procedures explain a particular stock move, or a captured product test. For the related question of why a better group margin may obscure operating weakness, see our worked earnings-margin analysis.