FDA Medical Device Oversight Guide: How Devices Are Checked

Maybe you saw a phone ear camera in an ad that mentioned FDA oversight. Or a clinic mentioned a device, and you wondered who checks it before it reaches you.

The rules are real, but they are not always easy to read on a box or a product page. Many people search for this exact question and find legal text instead of plain answers.

The short answer: the FDA oversees medical devices in the United States, sorts them by risk, and reviews many of them before they are sold. This guide walks you through that check, one step at a time.

Step 1: Start with what the device says it does

You have seen a device and want to know whether it is official. Start with its intended use, the job it is cleared or designed to do. The FDA’s rules depend on that job.

A medical device intended use and FDA classification guide explains those two terms for connected exam tools. For how phone-based ear imaging has been studied, read the smartphone-enabled otoscopy evidence limits guide and the digital otoscopy evidence limits guide.

A Federal Register final order dated March 25, 2026 reclassified optical diagnostic devices for melanoma detection from class III to class II. The order also renamed them software-aided adjunctive diagnostic devices and says they are subject to premarket notification. That device type is for skin lesions, not ears, but it shows how the FDA gives a device type its class and its name.

If your question began with an ear symptom, the device is only one part of it. A guide on what can be mistaken for an ear infection and another on how to tell if you have an ear infection cover the symptom side.

Next, see how the FDA watches devices before and after they reach you.

Step 2: Learn how the FDA watches a device before and after sale

You want to know whether anyone checks a device once it is on the market. The answer has two parts: review before sale, and monitoring after.

A 2012 systematic review in PLoS Medicine found quality problems in premarket submissions in the US. The same review said assessment of regulatory performance is limited by a lack of data on post-approval safety outcomes. In plain terms, the record of how devices perform in daily use is thinner than the record of how they were reviewed.

That is why postmarket monitoring matters. A 2016 review says clinicians should participate in premarket and postmarket activities, particularly adverse-event and device-failure reporting. Start with postmarket surveillance for medical devices to see what that monitoring is designed to learn. A device used at home or in a community setting also depends on clear instructions, covered in the home and community training and instructions guide. Software that changes over time raises its own questions, covered in software changes in connected diagnostic devices.

Once you know how oversight works, you can check a specific product against the official records.

Step 3: Check a specific product in FDA records

You have a product name and a claim, and you want to know whether the FDA’s records match.

Start with the guide to using FDA device databases to verify a connected diagnostic claim. Then check the device’s intended use against the clinical decision support or diagnostic device software guide, because software that analyzes images differs from software that supports a clinician’s decision. AI features raise their own questions, covered in the AI-enabled diagnostic devices transparency guide.

Labels and workflow matter too. The guide on what labeling should explain when a device gives an error shows what to look for on the product’s own paperwork.

A matching record still leaves a question about the accuracy numbers. That is the next step.

Step 4: Read an accuracy claim without being fooled by it

You have seen a number such as accuracy or sensitivity, and you want to know what it means for you.

The plain-language guide to reading a clinical study from the research library is a good companion to the steps below. Start with the difference between analytical and clinical performance.

Then look at the design. A guide on reading a diagnostic-accuracy study explains the questions to ask. Indeterminate results, such as unreadable images, should appear in the study counts. And a diagnostic cutoff can change the result when it moves.

Next, look at who was in the study and who left it.

Step 5: Look at who was in the study and who left it

You want to know whether a study reflects people like you. Gaps in the sample, the images, and the follow-up change the picture.

Check the image quality guide for a usable ear image to see why a picture can fail to be read. Bench tests are different from clinical studies, as the bench testing and simulation guide explains. Look also for blinded, independent review of results.

Missing data matter. Representation matters too, covered in the bias and representation guide.

A 2012 systematic review made a similar point about regulation: few studies had measured it quantitatively. Treat any single study as one piece of the record.

Supplements follow a different path, and that path has its own limits.

Step 6: Supplements and ear products follow different rules

You may be considering a supplement or a product with ingredients, not a device. The rules are not the same, and the reader needs to know that first.

A 2023 narrative review in Innovations in Pharmacy says the FDA’s dietary supplement oversight rests on the 1994 DSHEA law. It reports 79,071 adverse event reports to the FDA’s CFSAN system from 2004 to 2021. The review’s authors argue that supplement review should look more like prescription medicine review.

A systematic review of QT studies looked at 26 published supplement studies. More than 30% of them showed some ECG changes, and the authors say continued cardiovascular safety assessment is needed. For what a marketing claim does and does not prove, the research library’s guide to ingredient evidence versus marketing claims is useful.

If you are ready to look at a specific ear product, the Audiquiet ingredients, amounts and evidence guide is the place to read the ingredient details.

Whatever you use, you also need to stay safe with it.

Step 7: Stay safe with devices at home and in clinics

You want to know what could go wrong and how to spot it. Recalls, usability, and data security are the three areas to check.

A systematic review found conflicting views of device safety based largely on recall data. Published research has evaluated device safety using recall data, although its findings were conflicting. The guide on reading a safety alert or recall shows how to match your device.

Usability is a safety question. The human factors guide covers why. Data also needs protection, covered in the cybersecurity questions guide and the medical image storage and communication guide. Devices that exchange data raise questions covered in the interoperability questions guide. For a camera-based tool specifically, read whether ear cameras are safe and the guide on image capture and clinical context.

Finally, decide what your next step is.

Step 8: Decide your next step

You now know how oversight works and how to check a claim. The last step is to decide what to do with a symptom or a product.

For a health concern, the next step is a visit with a clinician. A device check does not replace that visit, and only a clinician can tell you what is causing a symptom. Ear timelines are covered in the guides on how long inner ear infections last, whether swimmer’s ear goes away on its own, and why swimmer’s ear pain worsens at night. For a laboratory tool, the mobile clinical microscopy guide covers the real-world setting.

Your next step

1. Write down the exact brand, model, and claim of the device or product.
2. Check the brand and model in FDA databases and recall notices.
3. Read the study design behind any accuracy number, including who was left out.
4. For supplements, read the ingredient amounts and the safety record.
5. Take any health symptom to a clinician.

Every guide on this topic

If you are checking a device claim:

When a Connected Diagnostic Device Gives an Error, Clinical Decision Support or Diagnostic Device Software, How to Use FDA Device Databases, AI-Enabled Diagnostic Devices: Transparency Questions, Software Changes in Connected Diagnostic Devices, Medical Device Intended Use and FDA Classification, Training and Instructions in Home and Community Device Evidence.

If you want to know what happens after sale:

Postmarket Surveillance for Medical Devices, How to Read a Medical Device Safety Alert or Recall, Medical Device Interoperability: Questions to Ask.

If you are reading a study or accuracy number:

Analytical Performance vs. Clinical Performance, How to Read a Diagnostic-Accuracy Study, Verification Bias in Diagnostic Research, Indeterminate Diagnostic Results, Diagnostic Cutoffs in Connected Devices, Bench Testing and Simulation Clinical Evidence, Blinding and Independent Review of Device Study Results, Missing Data and Withdrawn Participants, Bias and Representation in Connected Diagnostic Evidence, Positive Predictive Value and Prevalence, Digital Otoscopy Image Quality.

If you want device safety at home or in a clinic:

Human Factors in Connected Diagnostic Devices, Cybersecurity Questions for Connected Diagnostic Devices, Medical Image Storage and Communication, From Image Capture to Clinical Context, Are Ear Cameras Safe?.

If you have an ear concern or are considering an ear product:

What Can Be Mistaken for an Ear Infection, How to Tell if You Have an Ear Infection, Digital Otoscopy Connected Exam Devices: Evidence Limits, Smartphone-Enabled Otoscopy: Clinical Evidence Limits, How Long Do Inner Ear Infections Last, Does Swimmer’s Ear Go Away on Its Own, Swimmer’s Ear Pain at Night, Audiquiet Ingredients, Amounts and Evidence, Mobile Clinical Microscopy Laboratory Performance.

Disclosure: This page is education from Connected Diagnostics Evidence. It recommends no product and does not diagnose or treat any condition.

By Connected Diagnostics Evidence Editorial Team