Review a patient's medication list, inspect a source-linked alert, and record the clinician's decision in one workflow. The current PhiClaw podiatry preview uses fictional patients only.
Preview status, checked October 22, 2026: the limited medication-safety workflow is built and synthetically tested. Real-patient clinical activation and validated predictive risk scores are not available.
Try the fictional medication-safety sandbox
The public sandbox uses page-memory history only and resets on reload. It has no connection to patient records. The authenticated chart implementation stores durable review evidence separately.
What the preview does
- Medication reconciliation: record ingredients, dose, route, active/stopped status and list source. An unknown list stays unknown.
- Interaction alerts: evaluate three specific TMP-SMX pairs against a documented drug-label snapshot. Every run carries the label version, publication date, check date and rule version.
- Allergy review: flag exact matches within the label's trimethoprim, sulfamethoxazole and sulfonamide-antibiotic hypersensitivity scope. Ambiguous allergy text needs independent review.
- Clinician decisions: record acknowledgment or override with a reason and clinician attestation. The authenticated demo stores the actor, time, saved input and source version; alerts remain visible.
- Unavailable-data handling: missing histories, unsupported drugs and an expired source never become a safety clearance.
Coverage is deliberately limited
The pilot covers TMP-SMX paired with warfarin, methotrexate or dofetilide, plus its specified hypersensitivity contraindication. It is not a complete drug database. No match means only that this small rule set found no match; it does not mean the medicines are safe together.
Dose adjustments, kidney function, age, pregnancy, disease-specific contraindications and other medication pairs require independent clinical review. Severity labels are local review priorities derived from the label, not a validated vendor ranking.
The source is DailyMed's sulfamethoxazole/trimethoprim manufacturer labeling, version 4, published January 29, 2026 and checked October 22, 2026. The pilot source expires November 21, 2026 unless reviewed and refreshed.
Separate from AI chat and prescribing
These checks run explicit rules over structured lists. General AI chat does not determine the alert result. The preview does not prescribe, place orders, change chart medications or contact patients.
Predictive analytics: a defined target, no active model
The proposed outcome is a first new diabetic foot ulcer within 12 months in adults with diabetes and no active ulcer at baseline. The prediction endpoint currently returns unavailable and no probability. A named model, deployment rights, external validation in the intended population, calibration evidence, a missing-data policy and clinician approval are required before a risk score can be offered.
What a clinical rollout still needs
A sufficiently broad and maintained clinical knowledge source with appropriate use rights; independent clinician validation of the rules and workflow; a verified connection to the clinic's exact EHR; and the practice's privacy, access and activation approvals. Connecting to OpenEMR or another EHR is a separate integration step. This preview makes no regulatory-approval or certification claim.
Read the drug-interaction and prediction evidence guide, or review the current plans. This feature preview does not change existing plan prices or usage allowances.
Walk through the podiatry workflow with fictional data.
Book a podiatry demo