12 Supervised reading and final sign-off

The eleven modules before this one teach vocabulary and rules. None of them can teach the thing that makes a reader: the moment when a page looks one way to you and another way to the attending, and you find out why. This module is a method for producing that moment on purpose, many times, from the first week of training.

The routine

The log is the product. A resident who has read sixty studies without one has sixty impressions; a resident with a log has a map of their own errors.

What to do with a disagreement

When the draft and the attending read differ, the first question is which of the two is right, and the honest answer is sometimes neither, or that the record does not decide it. Go back to the page together and apply the rule the disagreement turns on: the six IFCN criteria from Module 5 if the question is whether a transient is epileptiform, the field and video from Module 3 if it is whether a waveform is cerebral, the ACNS thresholds from Module 10 if it is whether a run is a seizure. A disagreement resolved by rule is a lesson; a disagreement resolved by rank is not.

The monthly review

Each month, the resident and a faculty reader go through the log looking for recurring errors, not single ones. Two patterns are common enough to expect: overcalling sharply contoured normal activity as epileptiform, which is the subject of Module 5 and a documented route to misdiagnosing epilepsy, and missed artifact, which Module 3 addresses. When the review finds a pattern, the response is to reread that module and to bias the next month's studies toward the records that expose it.

The curriculum is structured around explicit sign-offs and a reading log because the published evidence says that most programs have neither.

In a 2024 survey of US and Canadian child neurology and neurodevelopmental disabilities program directors, 62% used no objective EEG competency measure and 93% set no minimum number of reads. Programs required a mean of 7.3 weeks of EEG training (Katyal 2024). In an online routine-EEG examination given to adult and child neurology residents, residents scored 72% against experts' 82%, and the authors concluded that residents need more than 8 weeks, ideally more than 12, of EEG training for minimal routine-EEG competency (Nascimento 2023, as summarized in Katyal 2024). Set beside each other, the two papers say that the typical program provides less training time than the examination data suggest is needed, and rarely checks the result.

What competency looks like on paper

The ACGME Child Neurology Milestones describe EEG competence in five levels, and the outline of this curriculum maps its modules onto them (ACGME 2020):

The ILAE competency roadmap places the same skills at its Level 2, proficiency, including background and sleep at all ages, artifacts and normal variants, and interictal and ictal patterns (Blümcke 2019). The milestones describe what a competent reader can do; they do not say how many studies it takes to get there, and no published minimum has been validated.

What this means for you

Two things follow. First, the number of weeks on a rotation is a poor proxy for competence, so the sign-offs in this curriculum test what you can do on real records rather than how long you have been in the reading room. Second, because the median program does not measure EEG competence objectively, you may need to build the measurement yourself: the log, the sign-offs and the objective tests in the last section are that measurement.

There is no validated minimum number of studies. In the 2024 survey, 93% of child neurology programs set none (Katyal 2024). That is not a reason to have no number; it is a reason to set a local one, state it, and track it, so that the program can look back and see whether it was enough.

A starting mix, offered as a suggestion rather than a standard

The mix is shaped by where residents go wrong. The under-2 requirement exists because the normal record changes fastest in the first two years, as Module 4 shows, and because the artifacts of infancy in Module 3 are not learned from adult records. The neonatal studies must span postmenstrual ages, because a reader who has only seen term records cannot judge a 32-week background against Module 6. The continuous EEG days are there because ICU patterns and seizure burden, the subject of Module 11, cannot be learned from 20-minute studies. The three named syndromes are the ones whose EEG features Module 9 treats as syndrome-defining, and a reader should have seen each at least once before signing off.

Tracking

The log from the first section is the tracking tool. Add a column for study type and one for postmenstrual or chronological age, and the mix counts itself. Review the totals at the monthly meeting alongside the error patterns, because the two interact: a resident who keeps overcalling temporal transients needs more adolescent records, not more of the same, and a resident who has never logged a disagreement on a neonatal study has probably not read enough of them.

Treat the numbers as a floor to revise, not a ceiling to reach. The point of setting a local target is to be able to say, a year later, whether the residents who met it read competently, and to move the target if they did not.

Sign-off in this curriculum has two layers: objective tests that can be scored by a machine, and sign-offs on real records that only a faculty reader can give. Neither replaces the other.

Objective testing

The tests measure recognition and rule application against a fixed key. They cannot measure whether you write a usable report or localize on a real bipolar chain, which is why the record-based sign-offs exist.

Final sign-off checklist

Each item is the faculty-verified competency from its module. Tick it only when a faculty reader has verified it; the site records your tick as self-reported.

Reading the checklist as a curriculum

The items are not interchangeable. Modules 1 to 5 are signed off on skills that every later read depends on, and a resident who has not been signed off on localization (Module 1) or on the IFCN criteria (Module 5) is not ready for the abnormal-EEG sequence. The neonatal pair (6 and 7) and the abnormal sequence (8 to 10) can be completed in either order, and both lead into Module 11. Module 12 is signed off last, because its target counts the reading done during all the others. When every box is ticked by a faculty reader, the resident has met the curriculum's definition of a competent pediatric EEG reader: not a number of weeks, but a set of things they can do on real records.