Free companion page

Companion page: How to Read an FDA 510(k) Clearance

Free to print: Appendix A: The Forty-Question Reading Checklist, from How to Read an FDA 510(k) Clearance.

This page is free and needs no sign-up. It reproduces material from How to Read an FDA 510(k) Clearance by Julian R. Sterling so you can print it or keep it open while you work.

To print it, use your browser's print button. Figures are as stated in the book; check the official source before relying on any of them.

Appendix A: The Forty-Question Reading Checklist

Work through these questions in order with the decision letter, the public database record and the 510(k) summary open side by side. After each question, one sentence says why it matters and which chapter explains it. The list is a reading aid, not regulatory, legal or investment advice, and where a question touches a change or a claim, the answer belongs to regulatory counsel.

The record and the pathway (questions 1 to 6)

  1. What kind of record is this: a 510(k) clearance, a De Novo grant or a premarket approval? The pathway decides which documents exist, what they can tell you and what the product is allowed to claim (Chapter 1).
  2. Which class does the record give the device, and does that fit the pathway? A class mismatch is the first sign that you are reading the wrong document or misreading the record (Chapter 1).
  3. Is the K number the same on the letter, the database record and the summary? A single wrong digit sends you to a different device and a different history (Chapter 2).
  4. Which classification regulation and product code are listed, and do they describe the device in front of you? These two entries place the device among its peers and tell you where to look for comparable clearances (Chapter 2).
  5. Which review team handled the file, as the record shows? The team points to the technical discipline that judged the device, which helps you judge how the testing section should read (Chapter 2).
  6. Was this a traditional 510(k), or a special or abbreviated one? The type tells you how much new evidence to expect and which parts of the file may lean on earlier work (Chapter 10).

The letter and the database (questions 7 to 12)

  1. What are the received date and the decision date, and do the letter and the database agree? Every later calculation about timing starts from these two dates (Chapter 2).
  2. Does the letter say the device is substantially equivalent, and does it treat the clearance as covering the stated indications? The letter is FDA's own voice, so it is the anchor against which every other document is read (Chapter 2).
  3. What does the letter say about requirements beyond the clearance? A clearance is permission to market under one pathway, not a finding that every other legal duty is met (Chapter 2).
  4. What decision code and decision description does the database record show? The record can differ in detail from the letter, and the difference is information (Chapter 2).
  5. Does the record point to a 510(k) summary or to a 510(k) statement? A summary is public and can be read in full, while a statement leaves you with far less detail (Chapter 2).
  6. When did the summary become visible, and how long after the decision? A late or missing summary changes what you can check, and the gap is worth noting (Chapter 9).

The predicate and equivalence (questions 13 to 18)

  1. Which device is the primary predicate, and which devices are only reference devices? The primary predicate carries the equivalence argument, while a reference device supports one feature at most (Chapter 3).
  2. Can you rebuild the whole chain of predicates back to the oldest clearance? Each step in a long chain may have added differences, and you should know where the original technology came from (Chapter 3).
  3. How old was the predicate when the new submission was received? Age alone proves nothing, but it prompts the question of how far technology moved between the two devices (Chapter 3).
  4. What does the public record show for the predicate: recalls, adverse event reports, withdrawn or changed clearances? A buyer inherits questions about the predicate's history as part of understanding the new device (Chapter 11).
  5. Which technological characteristics differ from the predicate, and how does the summary say each difference was handled? This is the point where equivalence is actually argued rather than asserted (Chapter 3).
  6. What exactly does each reference device support, and does the summary stay within that role? A reference device used beyond its role can hide a gap in the primary argument (Chapter 3).

Indications and labeling (questions 19 to 23)

  1. What do the cleared indications for use say, word by word? Every later question about claims, markets and growth runs back to this wording (Chapter 4).
  2. How do the cleared indications compare with the predicate's? A broader indication than the predicate's deserves a clear explanation in the file (Chapter 4).
  3. Did the indications change between submission and clearance? A narrowed indication tells you that the cleared scope is smaller than the one the maker asked for (Chapter 4).
  4. Is the device for prescription use, and who is the intended user? The user and the setting shape what the testing needs to show and what a seller may say (Chapter 4).
  5. Does the labeling or marketing of the product stay inside the cleared indications? Whether a claim fits is a question for regulatory counsel, and this question tells you when to ask it (Chapter 4).

Description, comparison and testing (questions 24 to 30)

  1. Does the device description tell you what the device is, what its parts are and how it is used? Without a clear description you cannot judge whether the comparison table is complete (Chapter 5).
  2. Is every row of the comparison table marked same or different, and are the rows complete? A table that leaves out an obvious feature should prompt a closer reading of the description (Chapter 5).
  3. For each difference, which test or data set does the summary point to? A difference with no matching evidence is the gap a careful reader looks for (Chapter 5).
  4. What are the bench test acceptance criteria, and how close does each result come to its limit? Results that sit near a limit say more than results with wide margins (Chapter 6).
  5. Which biocompatibility, cleaning or sterility topics apply, given how the device touches the patient? The type and length of contact decide which topics should appear at all (Chapter 6).
  6. Which electrical safety and electromagnetic standards are cited, and are they recognized consensus standards? A named standard is a claim you can check, and an unnamed one is a question (Chapter 6).
  7. Which tests are listed as not applicable, and does the stated reason make sense? A sound reason for leaving out a test is as informative as a test that was run (Chapter 6).

Software, cybersecurity and clinical data (questions 31 to 35)

  1. What does the file say about software verification and validation, and what does the software actually do? Software can be the main risk of a device, so the depth of this section matters (Chapter 7).
  2. Does the device connect to anything, and does the file treat it as a cyber device with a plan and a software bill of materials? Connectivity creates postmarket duties that continue long after clearance (Chapter 7).
  3. Does the device use a learning model, and if so, is a predetermined change control plan authorized? The plan sets what the maker may change later without a new submission (Chapter 7).
  4. Was clinical data included, and what reason does the file give? FDA asks for clinical data only in certain situations, so its presence or absence is a signal (Chapter 8).
  5. Does the clinical population match the cleared indications, subgroup by subgroup? Data gathered in one group cannot be stretched to another without a reason (Chapter 8).

Clock and after clearance (questions 36 to 40)

  1. How long did the review take in calendar days, and how many of those were FDA days after subtracting the time on hold? The two counts answer different questions about delay and about who held the file (Chapter 9).
  2. Did FDA send an additional information request, and what changed after the response? A request often shows the point on which the evidence was thin (Chapter 9).
  3. Which planned changes to the device, software or supply chain might need a new 510(k)? Whether a change does is a question for regulatory counsel, and the cleared record frames it (Chapter 11).
  4. What do the recall, adverse event, registration, listing and device identifier records show for the device and its predecessors? These public entries are the buyer's first check on how the company and the product behave in the market (Chapter 11).
  5. Do any contract terms, such as a milestone payment, depend on the date or the scope of the clearance? A date or an indication that lands differently from the plan can change the money, as the worked case shows (Chapter 12).