About & methods

What this is

NeuroBlu Insights translates a clinical research audit into something patients and advocates can use. The audit asked a simple question across the whole field of psychiatry: when a guideline says a specific care step should happen, how often does it actually happen? We found 57 well-defined care gaps — recommended steps that often get missed — and turned each one into an advocacy card with plain-language questions you can bring to an appointment. These are system-level patterns, not a verdict on any individual clinician: most gaps trace back to caseload, fragmented records, and the simple absence of anyone tracking the checklist — not to anyone failing at their job.

Where the numbers come from

The analysis sits on de-identified, real-world records from NeuroBlu, a large behavioral-health database drawn from U.S. mental-health care. For each care gap we defined an eligible population (for example, everyone on a sustained antipsychotic) and measured how often the recommended step was documented in the expected window.

A few things worth knowing about how we read these numbers:

The measured rate is a floor, not a ceiling. If a test isn’t in the record, we count it as not done. Some of those tests may have happened elsewhere and gone unrecorded — so the true gap is at most what we report, and often the realistic story is “documented care,” which is what patients actually experience.
“Achievable floor” benchmarks. For the flagship gaps we compared every clinic with enough patients and found the top-performing 10%. When the best clinics miss a step only a little and the median misses it a lot, the gap is about the system, not the patient — and it’s fixable.
Harm, where we could measure it. Several flagship cards report how much more often harm followed when a step was missed (for example, higher rates of new diabetes among antipsychotic patients whose blood sugar drifted unmonitored). These are associations from observational data, stress-tested with sensitivity analyses, not proof of cause in any one person.
Disparities. Where the data allowed, we checked whether some groups are missed more than others. Several gaps are not evenly distributed.

The flagship cards carry the richest numbers. The remaining cards ship with the core “should” and how often it’s missed, and are progressively enriched over time.

A medical-safety statement, in plain terms

This site is information, not medical advice. It describes patterns across large populations — it cannot know your history, your reasons, or your clinician’s judgment. Please read the following as seriously as we mean it:

Never start, stop, or change a medication based on anything here. Some of the gaps on this site are about the danger of stopping a medicine suddenly — abruptly stopping a benzodiazepine, for instance, can be dangerous. Changes belong in a conversation with your prescriber.
The cards give you questions to ask, never instructions to follow.
A care step appearing for your situation does not mean your care was wrong. The right answer is sometimes “we considered it and decided against it” — the point is that it was considered, with you.

Your privacy

Everything you enter on the Build my checklist page stays in your browser (localStorage). Nothing you check is sent to us or anyone else — there is no account, no tracking of your selections, and no server that sees them.

If you need help now

988 — Suicide & Crisis Lifeline (call or text), free and 24/7.
Text HOME to 741741 — Crisis Text Line.
findtreatment.gov — find mental-health and substance-use treatment near you.