A Patient Had a Complication. What Should the Record Show?

Published September 25, 2026 · 6 minute read · General information, not legal advice. Your licensing regulators decide what applies to you.

Most people who search this have already had the phone call. A patient reports swelling, a burn, a reaction, pain that did not settle, a result nobody expected. The clinical response is under way or already finished. The question left over is what the file is supposed to look like afterwards, and it usually gets asked late at night by someone who has never had to build one.

This guide is about the record, not the clinical response. It does not say what to do for a patient, what to give them, when to send them somewhere else, or how to judge severity. Those are clinical decisions that belong to the clinicians involved and to the protocols their prescriber signed. What follows is the paperwork side: what the file should contain when the event is over, and who the different parts of it are written for.

Two records, two readers

The treatment record is the patient's. It continues as it always did: what was observed, what was done, what was communicated, what follow-up was arranged. Its reader is the next clinician to see that patient.

The incident record is the practice's. Its readers are people outside the treating relationship, and there may be several of them at different times: an insurer, a prescriber reviewing the practice's protocols, a licensing body, a prospective buyer during due diligence, a lawyer. None of them were in the room. All of them will be reading years later.

Practices that keep only one of these end up in trouble from both directions. Where there is no incident record, the internal account of the event survives only as hallway memory. Where the internal account has been written into the patient's chart, the chart now contains speculation, blame, and internal review notes that were never meant for a clinical reader.

What the incident record usually contains

• When. Date and time of the treatment, and separately, the date and time the problem was identified and by whom. These are rarely the same and the gap between them is one of the first things an outside reader looks at.
• Who was present. Everyone in the room, by name and role, not just the person treating.
• What was being done. The service performed, the device or product category involved, and the identifier of the treatment record it belongs to.
• What happened, in plain description. What was observed, in the order it was observed, without conclusions about cause.
• What was done in response, and by whom, with times. If anyone outside the practice was contacted, who and when.
• What the patient was told, and by whom, and what they said back.
• Product and device identifiers. Lot or serial numbers, and whether anything was retained rather than discarded.
• Follow-up arranged, and whether it happened.
• Who was notified inside the practice, including the prescriber, and when.
• Signature and date of the person writing it, who should be someone who was there.

The field most often missing is the identifier that ties this record to the treatment record and to the evaluation that authorized the service. Without it, tracing one document to another is a job for a person with an afternoon and a filing cabinet.

Two things do not belong in it. Speculation about cause, which will be read as a finding by everyone who sees it later. And any rewriting of what was already written; a correction is an addendum, dated and signed, that leaves the original entry legible underneath it.

Reporting channels that exist outside the practice

Beyond whatever the practice's own agreements and its insurer require, national channels exist for problems involving the products themselves.

The U.S. Food and Drug Administration runs MedWatch, which it describes as its medical product safety reporting program for health professionals, patients and consumers. It takes reports on prescription and over-the-counter medicines, biologics, medical devices, combination products and cosmetics. The agency's forms page distinguishes the routes: health professionals, consumers and patients "can submit voluntary reports using MedWatch Online or by downloading, completing and then submitting FDA Form 3500 (health professional) or 3500B (consumer/patient)," while Form 3500A is the route by which "industry, user facilities personnel, and Investigational New Drug (IND) researchers may use the FDA Form 3500A to submit mandatory reports of adverse events and device malfunctions."

That is cited here as a description of a federal program, not as a statement about what any particular practice has to do. Whether a mandatory reporting duty attaches to a given practice, and to whom else an event has to be reported, is decided by the bodies that license the professions involved and by the practice's own agreements. The regulator directory lists those bodies, and the guide on how to verify what your state actually requires covers getting the answer in writing from the body that issues it.

Where the incident record sits in the file

An incident record on its own is thin. What makes it readable is the set of documents it points to: the evaluation that authorized the service, the consent signed for it, the treatment record, the protocol the response followed, and the oversight record showing who authorized whom. A practice that can hand over all six as a set is in a different position from one that can hand over a narrative. The guide on the documents an insurer asks for first covers the order those requests usually arrive in.

Where our documents fit

The Emergency Documentation Set contains blank incident and correction records with these fields, and the Emergency Preparedness Module holds the surrounding records the incident record points back to. They are blank where the clinical and jurisdictional decisions belong. They do not tell you how to treat a complication or whom you are obliged to notify.

This week

1. Open the practice's current incident form, if there is one, and check it against the field list above. Note which fields are missing, starting with the identifier that ties it to the treatment record.
2. Write down, on one page, who inside the practice is told when something goes wrong, in what order, and who writes the record. Put a name against each step rather than a role nobody has been assigned.
3. Ask the prescriber whose protocols the practice runs on to confirm in writing how they want to be notified and what they want to receive.

Questions this guide answers

Should a complication be written in the treatment record or a separate incident record?

Both, and they do different jobs. The treatment record continues as clinical documentation of what happened to the patient and what was done. The incident record is a separate account of the event itself, written for readers outside the treating relationship. Keeping them separate stops the patient's chart from filling with internal analysis.

Should the original treatment note be corrected after a complication?

An entry that is already written is not replaced. Corrections are made as a dated, signed addendum that leaves the original entry readable, with the correction and the reason for it visible. A record that shows signs of having been rewritten after the fact is worth less than one that shows a plain correction.

Who has to be told when a patient has a bad outcome?

That depends on the professions involved, the products used, the practice's insurer, and its own agreements. Notification duties are not the same everywhere and are not the same for every practice. The practice's regulators, its carrier, and its attorney are the ones who can say which apply.

Related guides

• What Is a Good Faith Exam, and What Does the Record Have to Show?
• The Four Documents an Insurer Asks a Med Spa For First
• Aesthetic Injector Training Options Compared