Clear answers without making people read the entire proposal.

Sitora is a proposed intelligence layer designed to sit above existing healthcare systems. It does not replace prescribing, dispensing, EPR or pharmacy platforms. It connects agreed operational signals, identifies candidate resource-loss events, supports governed interventions and records whether any resource was genuinely recovered.

Medicine Loop asks whether another repeat supply is actually required before it is dispensed. Medicines Resource Intelligence asks whether a clinically necessary medicine is being delivered through the most appropriate resource pathway, including route, formulation, preparation location, aseptic production, nursing time, chair time, bed use and unused preparations.

No. The proposed model does not autonomously stop clinically required medicines, alter a prescription or make a clinically significant medication decision. It identifies anomalies or missing information and routes clinically significant cases to an authorised pharmacist, prescriber or other appropriate professional.

No. Campaigns rely mainly on patient behaviour and awareness. Medicine Loop is intended to make the system itself more intelligent by using prescribing, dispensing, expected consumption, medication changes, reconciliation and patient-stock signals to intervene before avoidable supply occurs.

Prescribing systems record intent and dispensing systems record supply, but these records do not consistently prove that the next supply is needed. Remaining stock, changed doses, discontinuations, hospital discharge changes and actual consumption can sit outside the next repeat decision unless someone actively reconciles them.

It looks for whole-pathway opportunities such as ready-to-administer products, clinically appropriate route changes, dose banding, batching, preparation-location redesign, avoidable local preparation, unnecessary chair or bed time, repeated cancellations and prepared-but-unused doses.

Potentially, but AI is not required for every function. Many early use cases are better handled by deterministic rules, thresholds and data matching. AI may help with prioritisation, pattern detection, explanation and complex combinations of signals. Safety-critical decisions remain governed by clinical rules and authorised professional review.

No. The proposed architecture is intentionally additive. Sitora should consume the minimum permitted data from existing systems, generate intelligence and feed governed actions back into existing workflows rather than forcing organisations to replace core clinical infrastructure.

The minimum dataset depends on scope, but likely includes repeat-prescribing data, issue and dispensing history, medication changes, reconciliation events, expected days supply, selected patient-stock confirmation, injectable or SACT pathway data, preparation records, administration timing, cancellations, unused preparations, staffing inputs and relevant finance data.

Only where necessary for the agreed use case and under the appropriate lawful basis, governance and access controls. The pilot design should minimise personal data, pseudonymise where feasible, restrict role-based access, maintain audit logs and document data flows through a DPIA.

Clinical safety must be built in from the start. The intended use would need to be assessed to determine the exact regulatory and clinical-safety obligations. The current pilot model assumes formal clinical-risk management, named Clinical Safety Officers, hazard logs, escalation rules and human review of clinically significant interventions.

It could if designed badly, which is why workload is a core pilot outcome. The system should aim to let routine cases pass with minimal friction and only escalate cases where there is a meaningful anomaly or potential safety/resource issue. A solution that saves medicine cost but creates more professional workload than it releases has failed economically.

The pilot should use conservative thresholds, exception-based review, suppression of repeated low-value alerts and continuous measurement of alert acceptance and workload. The objective is not maximum detection. It is high-value, actionable detection with a manageable signal-to-noise ratio.

They will not be rolled into one headline number. Each benefit must be classified as cash-releasing, cost avoidance or capacity/productivity release. Finance validation is required for monetary claims, and capacity only counts as recovered where the freed resource is demonstrably available for another patient or activity.

It is the proposed audit unit for each verified intervention. It records the original signal, intervention, clinical approval where relevant, avoided or changed activity, resource effect, safety outcome, benefit type and verification status. This creates a traceable evidence base instead of relying on top-down estimates.

It records selected prepared medicines that are not administered and classifies why they became unused, for example cancellation, deterioration, dose change, scheduling issue, expiry or readiness failure. Repeated causes can then be targeted with specific operational interventions.

These pathways can consume several scarce resources at once, including pharmacy production, nursing preparation, infusion-chair time and bed capacity. National work has already identified substantial potential from transforming aseptic and injectable-medicine pathways, so this is a credible area for a measurable pilot.

The current proposal is a 20-week programme: 4 weeks mobilisation, 4 weeks baseline measurement, 8 weeks controlled intervention and 4 weeks evaluation. It would involve one acute provider plus a small primary-care and community-pharmacy network.

A: Medicine Loop for selected repeat-medicine cohorts. B: Medicines Resource Intelligence for selected high-volume or resource-intensive injectable/SACT pathways. C: Prepared-but-unused medicines with structured root-cause capture and intervention.

Success means clinically safe, measurable and reproducible net benefit. That includes evidence of unnecessary supply prevented, resource use reduced, unused preparations lowered, workload remaining manageable, data quality being sufficient for scale and the value of recovered resources exceeding the intervention cost.

That is a valid result. The programme has an explicit Scale / Modify / Stop decision. If the opportunity is small, the operational burden is too high or safety concerns outweigh benefit, the correct decision is to stop rather than manufacture a business case.

A dashboard shows what happened. Sitora is intended to connect detection to action and verification: observe the signal, explain the likely cause, trigger a governed intervention and then prove whether anything changed afterwards.

The ideal structure includes an executive sponsor, senior pharmacy or clinical lead, medicines-optimisation representation, digital and information-governance leads, a Clinical Safety Officer, finance, analytics/evaluation and operational owners from the participating acute and primary-care pathways.

Approval for a governed 20-week medicines discovery-and-pilot programme, nomination of participating sites and accountable leads, access to the minimum agreed datasets, an approved intervention protocol, and pre-agreed definitions for safety, workload, capacity release and financial benefit.

Yes. The wider Sitora research applies the same Observe, Detect, Explain, Intervene and Verify model to discharge delays, missed appointments, theatres, workforce and other resource-loss areas. Medicines is proposed as the first focused vertical because it has clear transactions, identifiable owners and measurable resource pathways.