Pilot readiness pack · protocol

A controlled medicines evaluation with explicit inclusion, ownership and stopping rules.

The purpose is not to prove Sitora right. It is to determine whether the three workflows create safe, measurable and operationally worthwhile benefit in a real NHS setting.

Minimum data specification Technical implementation

Duration20 weeks

Mobilise to decision

Includes baseline, shadow mode, controlled intervention and evaluation.

Products3

Separate workstreams

Each has its own denominator, intervention and outcome.

DecisionPre-set

Scale / modify / stop

Thresholds agreed before intervention begins.

SafetyHuman

Professional control

No autonomous prescribing or medication-state changes.

Three pilot use cases

Each question has a defined population and endpoint.

Medicine LoopCross-setting medication change not fully propagated

Cross-setting medication change not fully propagated

Include

Adults with a clinically authorised medication stop, start, dose or formulation change originating in hospital or another participating service and requiring downstream action.

Exclude

Changes with unclear provenance, emergency situations outside the agreed workflow, or records that cannot be safely linked.

Primary endpoint

Time from authorised change to all required downstream reconciliations complete.

Secondary endpoints

Open discrepancies, obsolete repeats/supply/admin states prevented, workload per case, safety events and escalation burden.

WasteGuardPotential repeat supply not required this cycle

Potential repeat supply not required this cycle

Include

Selected repeat medicines in agreed stable cohorts where dispensing history, timing, medication state and stock/use information are available.

Exclude

High-risk medicines or cohorts excluded by the clinical safety case, urgent supply, or cases where a delay could create unacceptable harm.

Primary endpoint

Verified supplies safely deferred or avoided after human review.

Secondary endpoints

False-positive rate, pharmacist/prescriber workload, patient experience, subsequent urgent supply, cost avoidance and safety.

Medicines Resource IntelligenceEvidence-backed pathway alternative

Evidence-backed pathway alternative

Include

One or two high-volume or resource-intensive medicine pathways with an established evidence base and a locally plausible alternative.

Exclude

Pathways without agreed clinical eligibility criteria, insufficient baseline data or unresolved safety/governance requirements.

Primary endpoint

Observed local change in the pre-agreed resource measure after implementation.

Secondary endpoints

Pharmacy/nursing minutes, chair/bed capacity, consumables, unused doses, acquisition cost, patient impact, safety and actual capacity reuse.

Operating model

Every signal has an owner and a clock.

Detect

Sitora creates a signal or open medication-change event from agreed source data.

Triage

Rules classify urgency, provenance, completeness and responsible service.

Review

Named professional reviews the case when clinical judgement is required.

Act

The responsible service reconciles, defers, implements or rejects the proposed action.

Close

Outcome, workload, safety and benefit classification are recorded before the case closes.

Service levels

Suggested pilot response rules.

Urgent discrepancy: immediate routing into existing local clinical escalation, not a new Sitora-only process.
Routine medication-change action: acknowledged within 1 working day, reconciled within a locally agreed window.
WasteGuard review: resolved before the relevant dispensing event where operationally feasible.
Resource opportunity: no live change until pathway owner and clinical governance approve the intervention.
Unresolved ownership: escalated to the pilot clinical lead.

Closure definition

A message sent is not a closed loop.

A case closes only when the responsible downstream workflow has been reconciled or a documented decision explains why no change is required.

Evaluation design

Measure the denominator, workload and balancing measures.

Measure	Minimum pilot requirement
Baseline	Normally 8-12 weeks retrospective or prospective baseline where data quality permits.
Denominator	All eligible events/items/pathway episodes, not only successful interventions.
Intervention rate	Signals reviewed, accepted, rejected and unresolved.
Workload	Staff minutes by role, escalation burden and additional contacts.
Safety	Medication incidents, urgent supply, readmission/attendance signals where relevant, complaints and clinical concerns.
Benefit	Cash release, cost avoidance and capacity release kept separate.
Comparison	Pre/post, matched comparison or stepped implementation chosen before go-live.

Stopping rules

Reasons to pause or stop early.

Unexpected medication safety signal or material patient harm concern
False-positive burden materially exceeds the agreed threshold
Required source data proves unreliable or incomplete
Workload transfer outweighs measured benefit
Clinical or information-governance conditions are no longer met
The intervention cannot be isolated enough to evaluate credibly

Week 20

The answer may legitimately be stop.

The pilot is successful if it produces a trustworthy commissioning decision, including evidence that the opportunity is too small or too burdensome to pursue.

Ready for technical discovery

Next confirm whether the minimum data can actually be supplied.

The data specification is deliberately minimal. It should be adapted to local systems rather than forcing an unnecessary replacement architecture.

Open data specification Return to pilot overview