What Are Diagnostic and Therapeutic Devices?

Diagnostic and therapeutic devices split cleanly by purpose: a diagnostic device finds out what is wrong, and a therapeutic device does something about it. That distinction is not academic, because it decides nearly everything a buyer has to check next — the class the device sits in, the premarket record that proves it may be sold, and the maintenance cycle that keeps it accurate.

Last updated: October 3, 2026

How diagnostic devices work

A diagnostic device detects, measures or images a biological signal and turns it into something a clinician can read. Imaging systems reconstruct anatomy, monitors track a vital sign continuously, and laboratory analysers identify a marker in a sample.

Two features separate the current generation from the previous one. The first is portability, which moves the test to the bedside and the home. The second is integration, which writes the result into the patient record instead of onto a printout.

Which devices count as therapeutic

Therapeutic devices act on the patient rather than describe them. Pacemakers regulate rhythm, infusion pumps deliver a controlled dose, dialysis machines replace a filtration function, defibrillators correct a lethal rhythm, and surgical lasers such as laser therapy systems remove or coagulate tissue.

Their common thread is that the therapy is delivered by hardware, so device accuracy is the dose. A pump that drifts delivers the wrong amount; a laser that drifts burns the wrong depth. That is why the maintenance interval on a therapeutic device is a clinical parameter rather than a facilities line item.

How hard each device works is what sets that interval. A pacemaker runs continuously inside one patient and is managed through a device clinic. An infusion pump sits in near-constant use across a ward and therefore cycles through preventive maintenance most often. A dialysis machine is confined to a renal unit and is serviced on a schedule driven by treatment volume.

The FDA class decides what you may buy, and from whom

The counter-intuitive part of medical device procurement is that the marketing description of a device carries no regulatory weight at all. What carries weight is the class the FDA assigned, and the class determines which submission the manufacturer had to make before selling it.

Classification is risk based. Class I covers the lowest-risk devices and applies general controls; Class II applies general controls plus special controls; Class III covers the highest-risk devices and requires premarket approval. If a device is Class I or II and is not exempt, a 510(k) is required for marketing.

In practice, most diagnostic devices a clinic buys sit in Class II, while the higher-risk therapeutic devices reach Class III. Reading the class before the brochure is what tells a buyer which documents the seller should be able to produce. It is also the reason advanced diagnostic tools list differently from therapeutic equipment on a sourcing platform.

Three public records that settle provenance

Once the class is known, the paperwork reduces to records anyone can search. The FDA’s UDI System identifies devices from manufacturing through distribution to patient use, and the label of most devices carries a unique device identifier in human- and machine-readable form; the identifier can be looked up in AccessGUDID. Devices@FDA searches premarket notifications and approvals together, so a 510(k) or PMA can be confirmed against the model. Quarterly Registration & Listing confirms the establishment that made the device was registered at the time.

Four checks follow from those records, and each one is requestable before payment: the UDI or model and serial number, the premarket reference for the model, a dated statement of any open recall or early alert, and the establishment registration behind the manufacturer claim. A sourcing platform can shorten the search, but it does not replace the answers.

Maintenance: the 6–12 month cycle, and what voids it

Regular calibration, cleaning and software updates are what keep a diagnostic device accurate, and the usual preventive maintenance interval is every 6 to 12 months. Certified technicians should perform the repairs, and the records should be kept, because an audit trail is what proves the device was maintained to the manufacturer’s protocol rather than merely cleaned.

Two conditions quietly void that protection. A calibration performed without the record, and a software update installed outside the manufacturer’s channel. Either one can take a compliant device out of compliance without changing how it looks.

What changes when a device crosses into Great Britain

A device with an impeccable FDA record can still be the wrong purchase for a UK hospital. The MHRA’s guidance on regulating medical devices in the UK sets out the pieces a buyer should expect to see referenced: registrations in Great Britain, a UK responsible person for devices from outside the UK, UKCA marking and conformity assessment bodies, CE marking and notified bodies, labelling requirements, and post-market surveillance and vigilance.

The two files answer different questions, and a cross-border procurement has to assemble both.

Innovations worth weighing, not assuming

Therapeutic devices are moving toward wearable therapies, minimally invasive implants, and closed-loop insulin pumps that adjust doses automatically. Nanotechnology allows targeted drug delivery with fewer systemic effects, and wireless charging and robotics improve efficiency and outpatient outcomes. None of those claims should be accepted from a brochure; each maps to a premarket record that can be checked in the FDA databases before a purchase is committed.

References

• FDA: Classify Your Medical Device — the three classes and the controls attached to each.
• FDA: Unique Device Identification System (UDI System) — labelling and the AccessGUDID search.
• FDA: Medical Device Databases — Devices@FDA (510(k) and PMA) and quarterly registration and listing.
• FDA: Medical Device Recalls and Early Alerts.
• GOV.UK / MHRA: Regulating medical devices in the UK.

FAQ

Doesn’t a CE mark or a 510(k) number mean a device is safe to use?

No. A clearance or approval means the manufacturer demonstrated safety and effectiveness for an intended use at the time of submission, and the FDA records it in Devices@FDA. It does not certify the individual unit in front of you, its service history or its software revision, all of which still have to be checked.

Which record tells me whether a model was actually cleared?

Devices@FDA, which searches premarket notifications and premarket approvals together. Match the model against the submission rather than trusting the listing text, and use the quarterly registration and listing database to confirm the establishment that made it was registered.

Is a Class I device always the simplest to buy?

Not automatically. Class I is the lowest risk and is usually exempt from premarket notification, but exemptions come with limitations, and a Class II device that is not exempt needs a 510(k). The class tells you the pathway, not the amount of paperwork a particular sale requires.

Why does a device with a clean FDA record still need a UK file?

Because Great Britain runs its own registration, UK responsible person, UKCA marking and post-market surveillance requirements. A device cleared in the US can be perfectly lawful there and still lack the arrangements a UK hospital needs, so both files have to be assembled.

How often should diagnostic equipment be serviced?

The general interval is every 6 to 12 months, following the manufacturer’s protocol, with certified technicians performing the work. Keep the calibration and software records, because an undocumented adjustment can take a compliant device out of compliance without any visible change.

Part of the Medical Imaging & Diagnostics Equipment guide.