Core Recommendations and the AI Clinical Assistant

Clinical software needs two things that pull in different directions: consistency and context. Treat every problem as a question for generative AI and you get inconsistency. Treat every problem as a static rule and you miss the nuance of the individual resident. Pharmacore separates these jobs on purpose. Core Recommendations give the governed baseline. The AI Clinical Assistant supports resident-specific inquiry. The pharmacist decides.

Core Recommendations: the governed baseline

Core Recommendations are the things that should not be overlooked in a medication regimen review. They reflect 15 years of hands-on long-term care pharmacy practice and are maintained by Pharmacore's pharmacist team. They help identify the recurring medication, monitoring, documentation and regulatory considerations that deserve a pharmacist's attention: the kind of item a consultant checks on every chart, and the kind that is easiest to miss on the last chart of the day.

Because they are governed, they behave the same way for every pharmacist and every resident. A new consultant and a long-serving one start from the same baseline. That is the consistency half.

The AI Clinical Assistant: resident-specific inquiry

The AI Clinical Assistant supports a different kind of work. It helps the pharmacist explore a specific resident's information, ask questions and draft content for review. It is most useful when the clinical question depends on the relationship among medications, diagnoses, laboratory values, orders and history, which is exactly where a static rule runs out. That is the context half.

It is an assistant in the plain sense of the word. It gathers, surfaces and drafts. It does not decide.

Why the two layers are kept separate

A single engine that tries to do both jobs tends to do neither well. If the baseline is generated on the fly, it drifts, and two pharmacists end up with two different checklists. If a resident-specific question is forced through fixed rules, the rules fire on everything or on nothing useful. Keeping the layers apart gives a practice governed consistency for what must never be missed, contextual assistance for what deserves deeper inquiry, and one clear line of accountability for the final decision. We wrote about the principle behind this in clinical decision support should strengthen a pharmacist's judgment.

The pharmacist decides, and the product is built around that

Neither layer replaces professional judgment. Pharmacore is designed so that the pharmacist can inspect the context behind a suggestion, understand its basis, change or reject it, and publish only the conclusion the pharmacist approves. That control is part of the product, not a disclaimer added after the fact. For a practical way to test any vendor on this point, see six practical tests for AI recommendation review in pharmacy.

What this looks like in a review

1. Open the resident. Core Recommendations have already identified the baseline considerations for this chart.
2. Work the baseline. The pharmacist accepts, edits or sets aside each consideration with the context in view.
3. Ask the harder question. The AI Clinical Assistant helps explore the relationship among this resident's medications, diagnoses, labs and history, and drafts text for the pharmacist to review.
4. Decide and document. The pharmacist's conclusion, in the pharmacist's words, goes into the note and the report.

The combination is a practical operating model: governed consistency for what should not be overlooked, contextual assistance for what deserves deeper inquiry, and pharmacist accountability for the final decision. It is also the answer to a question we hear often from practices weighing a change, covered in choosing software around the consultant pharmacist's working day.

Frequently asked questions

What are Core Recommendations in Pharmacore?

A governed baseline of recurring medication, monitoring, documentation and regulatory considerations for medication regimen review, maintained by Pharmacore's pharmacist team and applied consistently to every chart.

What does the AI Clinical Assistant do?

It supports resident-specific inquiry. The pharmacist can explore a resident's medications, diagnoses, laboratory values, orders and history, ask questions and have content drafted for review.

Does Pharmacore's AI make clinical decisions?

No. The pharmacist inspects the context, understands the basis of a suggestion, changes or rejects it, and publishes only the conclusion the pharmacist approves.

Can a pharmacist override a Core Recommendation?

Yes. Core Recommendations identify what deserves review. The pharmacist decides what applies to the resident in front of them.

The best way to see the two layers working together is a clinical walkthrough using a synthetic resident case. Book a walkthrough.

About the author

Tim Merritt, PharmD, is a co-founder of Pharmacore and a practicing consultant pharmacist. He founded and built the company with Adam Brown and Dave Robbins.