What a doctor's certification file should prove

I read certification files with one question in mind. If this surgeon has a bad case in eighteen months, and someone asks how they came to be certified, does this file answer them?

Most files do not. They prove attendance. Dates, a course title, a signature, sometimes a photograph of a group holding certificates. That is a record of a person having been present, which is not the same as a record of a person being ready.

The file is an argument, not an archive

A certification file makes a claim: this doctor can be trusted with an unsupervised case at our standard. Everything in the file is either evidence for that claim or it is filler.

That framing changes what goes in. It also changes who the file is written for. Not the trainer, who was there and remembers. Not the trainee. It is written for a stranger reading it later, possibly under pressure, possibly with a lawyer present.

Six things that belong in it

Documented cases with measured performance. Each case with the numbers that matter for the procedure, taken the same way each time so cases can be read against each other. For surgical work in my field that means extraction and implantation rates against the clock, transection rate, density achieved against density planned, and out-of-body time. Numbers without a defined method of capture are decoration.

Case mix, stated deliberately. Six cases that were all straightforward prove that the surgeon can do straightforward cases. The file should show the range: different grades, different donor situations, at least one case that did not go to plan. If the training programme never produced a difficult case, say so in the file rather than letting the absence pass as competence.

A competency tracker with honest stages. Observed, assisted, performed under supervision, performed solo. Each competency, each stage, dated. The value of this is that it makes regression visible. A trainee who is solo on extraction and still assisted on implantation at week four has a specific development area, and the file should say it plainly.

Photographic evidence to a fixed protocol. Same angles, same distance, same lighting, every case. The protocol matters more than the camera. Photographs taken at whatever angle looked good cannot be compared to anything, and a certification file built on uncomparable images is a scrapbook.

The assessor's judgement, in words. The numbers do not speak for themselves. Somewhere in the file a named person has to write what they think, including the reservation. "Extraction is well above standard. Implantation speed is below target and should be reviewed at ninety days." A file with no reservation in it is either a genuinely exceptional trainee or an assessor who was not paying attention, and the reader cannot tell which.

Scope, stated explicitly. What the certificate covers and what it does not. Which procedures, which case complexity, under what supervision arrangement. Most disputes about certification are actually disputes about scope, and they are entirely preventable with one paragraph.

Two things that should not be in it

Comparison with other trainees. Each doctor is assessed against the standard, never against a colleague. Ranking language creeps in easily, and once it does, the file stops being evidence about this surgeon and becomes an opinion about a cohort. It also makes the file unusable if it is ever read outside the organisation.

Praise that is doing emotional work. Feedback should be honest and it should be kind, and those are compatible. What is not useful is softening a real finding so that a good week is not spoiled. If implantation speed is below target, the file says so. The trainee already knows; writing it down is what allows it to be fixed.

The speed and quality problem

Every training programme runs into this. A trainee is safe but slow, or fast and starting to be careless. The temptation is to certify on the number that is easiest to measure.

My rule is that safety numbers are gates and speed numbers are development areas. A transection rate outside the band stops certification. A slow implantation rate does not stop it, provided the file names it, sets a review point, and the supervising clinic knows to watch it. Speed improves with volume. The habits that produce a high transection rate do not improve on their own.

Sign-off with a name on it

One person signs. That person has read the evidence and can be asked, later, to explain the decision. Committees produce certificates that nobody defends, because responsibility spread across five people is responsibility held by none.

Give the signer the right to say not yet. If the programme has a fixed end date and the doctor is flying home on Friday, the temptation to certify anyway is enormous. The answer is a partial certification with a named scope and a follow-up requirement, which is honest, rather than a full one, which is not.

What the file is really for

I used to think of these files as compliance. They are not. They are the mechanism by which a clinical standard survives being copied into a country you cannot visit, taught by someone you did not train, and applied on a day you were not there.

That is the whole business of running quality across a network. The file is where it either holds or quietly stops.

Questions people ask

What belongs in a surgical certification file?

Documented cases with measured performance, a competency tracker showing progression from observed to solo, photographic evidence to a fixed protocol, the named assessor's judgement, and an explicit statement of what the certificate does and does not cover.

How many cases are enough for certification?

Case count alone proves nothing. What matters is case mix, whether performance held across them, and whether the last cases were run without help. A surgeon with six varied documented cases and a clean solo session is better evidenced than one with thirty undocumented ones.

Who should sign a certification?

One named person with clinical authority, who has seen the evidence and can be asked to defend the decision later. Committee sign-off without a named individual produces certificates nobody stands behind.