Medication Control Command Center
Signals → Context → Score → Action. RXBeacon Intelligence converts dispensing, inventory, timing, user, patient, and exception signals into explainable review priorities for pharmacy, compliance, and audit teams.
Last refresh 05:12
Medication Accountability Index (MAI)
Executive measure of controlled-medication events that can be traced from documented need and authorization through dispensing, administration or disposition, and final inventory accountability. The unresolved portion is prioritized for structured human review by RXBeacon Intelligence.
1,253Fully reconciled medication events
18Awaiting documentation
8Authorization exceptions
5Unexplained inventory / orphan events
Property Profile
Highland Hospital
Rochester, NY • Community teaching hospital sample environment
Dispensing locations7Cabinets / medication access points
Care units monitored5ED, ICU, OR/PACU, med-surg
Active authorized users18Current shift and recent activity
Operating profileOversightPyxis • MAR • Orders • Inventory
Shift Focus Areas
Executive roll-up of what matters right now by shift.
Highest activity
Emergency Department and ICU accounted for most controlled removals this shift.
Stable
Top medication
Hydromorphone had the highest controlled-dispense volume with expected matching administration activity.
Tracked
Exception watch
One incomplete waste-witness event and two documentation-lag events remain open for morning review.
Watch
Escalation risk
No P1 events. One P2 inventory discrepancy remains inside the current response target.
Monitor
Controlled dispenses today
42
▲ 4.8% vs 30-day weekday avg
Open discrepancies
3
1 exceeds 60-minute resolution target
Inventory variance
−2
2 dosage units pending reconciliation
High-review transactions
5
Beacon Index ≥ 70
Authorized users active
18
Across 7 dispensing locations
Today — Dispensed by Controlled Medication
Dosage units removed from dispensing cabinets. Waste and returns are tracked separately.
Dispensed unitsWaste documented
Priority Review Queue
Ranked by RXBeacon Intelligence transaction risk with explicit contributing factors.
Risk Distribution
All controlled-substance transactions today.
Unit Activity
Dispensing volume by care location.
Control Health
Key indicators for reconciliation and documentation.
Transactions with linked order98.4%
Waste with timely witness94.1%
Overrides today6
Late returns > 60 min2
Unresolved cabinet variance1
Data Intake, AI Mapping & Immutable Record Control
Bring Pyxis/cabinet reports, MAR extracts, order data, pharmacy inventory, staffing, waste/return files, or other approved reports into one review layer. AI-assisted mapping proposes where each field belongs; a reviewer approves the mapping before established records are appended.
Files staged
0
Awaiting reviewer approval
AI mapped fields
0
Proposed, never silently committed
Low-confidence fields
0
Require manual confirmation
Manual entries
0
Source = authorized user entry
Deleted records
0
Deletion disabled by design
Demo Report Center — Synthetic Pyxis-Style Activity
Use a demonstration-only controlled-medication activity report with synthetic patient tokens and staff IDs. It contains normal dispensing plus selected exceptions so an end user can see RXBeacon mapping, review prioritization, and append-only ingestion. This is not an official BD Pyxis export and contains no real PII/PHI.
18 synthetic eventsDispense • Waste • ReturnOverride scenarioNo-order exceptionDelayed witness
Drag & Drop Approved Reports
Drop CSV, TXT, JSON, or report exports here. The browser sandbox reads the file locally and simulates AI field mapping; production would use an approved hospital integration/OCR/AI service with access controls and PHI safeguards.
Drop report files here
or click to select files
Pyxis transaction • inventory • MAR • orders • staffing • waste/return • reconciliation
No files staged.
🔒
Source preservation: Once approved, the original source record becomes immutable. Corrections are stored as a new amendment/reversal linked to the original record with user, timestamp, reason, and approval. Established source data is never overwritten or deleted.
AI Mapping Preview
AI identifies likely fields and the RXBeacon areas affected. Confidence below the threshold is automatically held for review.
Stage a report to view proposed mapping.
Five-Way Medication Accountability Match
The objective is not merely to count pills. RXBeacon tests whether every controlled-medication event has a defensible clinical/operational path from authorization through final inventory accountability.
1 • Ordered / Authorized
98.4%
Provider order, PRN parameters, emergency/override authority, timing and unit access.2 • Dispensed
100%
Cabinet removal linked to user, patient token, medication, quantity, cabinet and time.3 • Administered / MAR
96.8%
Dispense matched to administration or documented non-administration within expected window.4 • Waste / Return
94.1%
Unused medication accounted for with witness, return, reason and timing.5 • Physical Inventory
97.9%
Expected closing count reconciles to cabinet/physical count and approved adjustments.RXBeacon Intelligence review trigger: An orphan event—dispensed without valid authorization, dispensed but not administered/wasted/returned, administration without a matching dispense, or physical count variance—moves into the exception queue with source evidence attached.
Manual Entry — Append Only
Use when a source report is unavailable, incomplete, or requires an authorized correction. Manual entry cannot edit or delete an established record.
No destructive editing: After Save, this entry is locked. If something is wrong, create an amendment linked to its Record ID. The original remains visible in the audit trail.
Provenance & Change Ledger
Every source file, AI proposal, reviewer approval, manual entry, reveal event, amendment and reconciliation should be attributable.
| Record | Time | Action | Source | User | Status |
|---|
Append an Amendment / Correction
Corrections never overwrite the established record. The amendment becomes a new immutable record linked to the original.
Medication Accountability & Authorization
Core objective: establish that every controlled medication was clinically ordered or otherwise authorized, removed by an authorized user, tied to the intended patient, administered or dispositioned, and reconciled to inventory.
Medication Accountability Index
97.6%
1,253 / 1,284 fully traced
Orphan events
5
No complete five-way match
Authorization exceptions
8
Order, role, unit, timing or access issue
Shift attestations
92%
2 units awaiting sign-off
Median order → dispense
11m
Contextualized by care setting
Five-Way Match — Need to Final Accountability
A mismatch creates a review signal; it does not determine clinical appropriateness or deny patient care.
1 • Ordered / Authorized
98.4%
Valid order/PRN parameters or documented emergency authority; user role, unit, patient and timing fit.2 • Dispensed
100%
Cabinet event identifies medication, quantity, user, patient token, cabinet and time.3 • Administered / MAR
96.8%
Removal has a corresponding administration or documented non-administration.4 • Waste / Return
94.1%
Unused amount has timely witness/return and reason with chain of custody.5 • Physical Inventory
97.9%
Expected closing count reconciles to physical/cabinet count and approved adjustments.Orphan Transaction Engine
Finds events that cannot be fully matched across orders, cabinet activity, MAR, waste/return and inventory.
| ID | Time | Exception | User / Patient | Medication | Gap | Risk |
|---|
Why This Score?
Select an orphan or authorization event. RXBeacon Intelligence exposes the contributing facts, mitigation and source evidence.
Select a review event.
Scores are prioritization signals only.
Scores are prioritization signals only.
Authorization Validation
Tests whether the user, medication, patient, location, timing and order/override authority align.
| Time | User | Patient | Order | Role / Unit | Timing | Result |
|---|
Order → Verification → Dispense → MAR → Disposition
Chronology matters. Reversed sequencing or excessive gaps can be surfaced without assuming misconduct.
Shift-to-Shift Controlled Medication Accountability
Opening + receipts/restock + returns − dispenses − waste ± approved adjustments = expected closing, followed by physical count and accountable sign-off.
| Unit / Shift | Opening | Receipts | Returns | Dispensed | Waste | Expected | Physical | Variance | Attestation |
|---|
Live Dispensing Ledger
A unified chronology of medication removals, returns, waste, overrides, orders, staff, patients, and cabinet location.
| Time | Medication | Qty | Nurse / User | Patient | Unit / Cabinet | Why / Order Context | Event | Beacon Index |
|---|
Inventory Reconciliation
Each defined period begins with a cabinet count. Dispenses decrement inventory; documented returns add inventory; waste is removed but remains accountable.
Control equation: Opening + Restock + Returns − Dispensed − Documented Waste ± Approved Adjustment = Expected Closing. Expected is compared with physical/cabinet count to identify variance.
Reconciliation Worksheet
Synthetic shift: 23:00–07:00. Variance is never automatically treated as diversion; it creates a review event.
| Medication | Opening | Restock | Returns | Dispensed | Waste | Expected | Physical | Variance |
|---|
Variance Review
RXBeacon Intelligence uses context before elevating a discrepancy.
Nurse / Authorized User Risk
Staff Risk Index (SRI) is an audit-prioritization signal based on patterns, peer context, documentation, overrides, discrepancies, timing, and witnessed waste.
SRI Leaderboard
Higher score = higher review priority, not a finding of wrongdoing.
| User | Unit | Dispenses | Overrides | Open Disc. | After-hours | SRI |
|---|
Selected User — Explainable Factors
Select a nurse in the leaderboard.
Same-Unit / Same-Shift Peer Baseline
Volume is normalized for worked hours, unit, shift, medication mix and comparable patient-care environment before peer deviation contributes to SRI.
| User | Comparable hours | Dispenses / 12h | Peer median | Variance | Override rate | Witness lag | Context |
|---|
Patient Controlled-Medication Utilization
Masked identifiers preserve minimum-necessary access. Authorized staff can reveal a fictional patient name in this sandbox; every reveal is treated as an auditable event.
Privacy design note: Last four SSN digits are shown here only because they were requested for the concept. In production, prefer a hospital-assigned token/MRN surrogate and keep SSN outside the analytics layer whenever operationally possible.
Patient Utilization Review Index
Measures unusual controlled-medication utilization patterns for pharmacist/compliance review; it is not an addiction, diagnosis, or medical-necessity score.
| Patient | Encounters | Controlled doses | Distinct drugs | Units | Review Index |
|---|
Selected Patient Timeline
Select a patient to view a synthetic medication timeline.
Cross-Unit Encounter Path
Follows the medication-accountability chain as a patient moves among care areas. It is a reconciliation view, not a clinical-need determination.
Visual Analytics
See what was dispensed, where it was dispensed, who dispensed it, and how utilization changes across time.
Dispensing Trend — Today
Controlled medication dosage units by time interval.
Medication Mix
Share of controlled-medication dispensing for the selected period.
Dispensed By User
Top authorized users by volume, with risk context kept separate from raw volume.
Patient Need vs Dispensed — Review View
Compares ordered/eligible administrations in the synthetic data to cabinet dispenses. A mismatch is a review signal, not a clinical judgment.
Waste Witness Intelligence
Examines waste attestations as a relationship network: who witnesses whom, how quickly, whether both users were plausibly present, and whether reciprocal pairs repeat unusually often.
Waste events today
17
Across 5 controlled medications
Timely witness
94.1%
Within configured workflow window
Unusual reciprocal pairs
2
Above same-unit baseline
Cross-unit witness events
1
Requires presence validation
Median witness lag
4m
90th percentile: 19m
Witness Pair Network
Higher pair score means the relationship deserves contextual review first.
Witness Detail
Select a pair to see the RXBeacon Intelligence explanation.
Select a witness pair.
Waste / Witness Review Table
| Primary user | Witness | Events | Reciprocal | Median lag | Same unit / shift | Presence evidence | Risk |
|---|
Prescriber / Ordering Provider Layer
Adds ordering context to the RXBeacon relationship model so exceptions can be reviewed across provider → patient → authorized user → medication → cabinet → time.
Provider Pattern Review
Raw controlled-medication volume is separated from risk. Scores emphasize authorization timing, unusual overrides, repeated order-after-dispense sequences and cross-unit relationships.
| Provider | Controlled orders | Patients | Overrides linked | Order after dispense | Units | Review Index |
|---|
Selected Provider — Explainable Context
Select a provider.
Relationship Signals
This view highlights repeated combinations without treating correlation as wrongdoing.
| Provider | User | Medication | Patient cohort | Pattern | Mitigating context | Priority |
|---|
Data Quality & AI Trust Dashboard
RXBeacon Intelligence should show whether it trusts the evidence before it trusts the risk score. Missing, conflicting, duplicate or stale inputs are surfaced independently from behavioral risk.
Overall evidence quality
97.1%
Across currently approved sources
Missing required fields
7
Primarily witness and staffing context
Source conflicts
3
Held from auto-resolution
Possible duplicates
2
Require source comparison
AI low-confidence mappings
0
Never silently committed
Source Reliability
| Source | Completeness | Freshness | Conflicts | Quality | Status |
|---|
Current Data Exceptions
Data Quality Rules
A risk score is tagged with evidence quality so reviewers can distinguish “high-risk pattern with strong evidence” from “uncertain pattern caused by incomplete data.”
Complete
✓
Required identifiers, timestamps, medication, quantity and source available.Consistent
↔
Order, cabinet, MAR and inventory sources agree within configured tolerances.Unique
1×
Potential duplicate records are held rather than automatically merged.Fresh
⏱
Source timestamp and ingestion time remain inside approved latency thresholds.Traceable
🔒
Original source, AI mapping, reviewer approval and amendments remain attributable.Audit & Exception Management
Turns cabinet discrepancies and risk flags into accountable review workflows with ownership, due dates, evidence, and closure rationale.
Open cases
7
3 inventory • 2 overrides • 2 documentation
Past due
1
Oldest: 74 minutes
Closed today
9
Median closure 22 min
Evidence completeness
96%
Witness, order, cabinet log, MAR
Repeat patterns
2
Same user / same med within 30 days
| Case | Opened | Type | User | Patient | Medication | Risk | Owner | Status | Next Step |
|---|
Immutable Evidence Package
One review package ties the exception to original source identifiers, orders, MAR, cabinet activity, inventory, waste/witness information, score rationale, reviewer notes, amendments and disposition.
Choose “Evidence” on a review case.
Help, Manuals & Definitions
Operational guidance for every RXBeacon function tab, platform definitions, ownership, privacy, and use limitations.
Platform Ownership, Use Limits & Privacy Disclaimer
Evidentify LLC is the sole owner of the proprietary HEP360 and RXBeacon platforms and associated intellectual property. This includes, without limitation, the HEP360 and RXBeacon names and branding; model architecture; scoring and weighting concepts; the Medication Accountability Index (MAI); Beacon Index; risk, reconciliation, authorization, exception, audit, evidence, data-mapping, workflow, dashboard, reporting, and decision-support structures; user-interface designs; documentation; methods; and derivative components, except where a separate written agreement expressly states otherwise.
Third-party product, system, company, hospital, or service names referenced in this prototype are included only to illustrate potential interoperability, source data, or use environments. Those third-party names and marks remain the property of their respective owners, and their appearance does not imply ownership by Evidentify LLC, endorsement, affiliation, integration, or production deployment.
- RXBeacon is a decision-support and oversight tool intended to track risk, reconcile dispensing activity, and assess whether controlled-medication dispensing appears documented, authorized, and accountable. It organizes and highlights source evidence for human review; it does not create a factual inference of diversion, misconduct, medical necessity, impairment, addiction, or wrongdoing.
- The platform does not provide clinical advice, medical necessity determinations, diagnosis, prescribing decisions, employment conclusions, disciplinary findings, or legal conclusions.
- Scores, indicators, and exceptions are review signals only. They support human follow-up and should never be used as the sole basis for patient-care, staffing, or punitive action.
- Production use should follow minimum-necessary principles for PII/PHI. Masked patient identifiers, role-based access, audit logging, approval workflows, and source preservation should remain enabled at all times.
- Where possible, use a hospital-generated token or MRN surrogate rather than Social Security Number data inside analytics layers. Any reveal of identity should be authorized, logged, and limited to a legitimate operational need.
Key Definitions
Medication Accountability Index (MAI)Executive measure showing the percentage of controlled-medication events that can be traced from authorization through final accountability.
Beacon IndexExplainable review-priority score applied to a transaction or event to highlight where human review should focus first.
Orphan EventA dispense, administration, waste, return, or inventory movement that cannot be fully matched across the expected source records.
Authorized RevealControlled workflow allowing privileged staff to view masked patient identity when operationally justified and auditable.
Append-only RecordkeepingEstablished source data is preserved. Corrections occur by amendment, reversal, or linked follow-up entry rather than deletion.
Source-System ReconciliationComparison of source-system counts and ingested records to confirm completeness and identify unmatched records.
Witness LagElapsed time between waste event initiation and witness confirmation.
Data Quality / TrustAssessment of whether underlying evidence is complete, timely, non-duplicate, and consistent enough for reliable review.
How to Use the Platform
- Select the hospital/property and data window from the top header.
- Review the Command Center and Shift Executive Snapshot for current posture.
- Use Data Intake + AI to drag and drop approved reports and review mapping before appending.
- Validate control integrity in Medication Accountability, Inventory Reconciliation, and Audit & Exceptions.
- Use Nurse Risk, Patient Utilization, Waste Witness Intelligence, and Prescriber Layer to understand context—not to make standalone decisions.
- Use Implementation Controls for retention, reconciliation, escalation, access-review, and audit-ready evidence generation.
All sample records in this sandbox are fictional and illustrative.
Command Center
Purpose: Gives leaders a single operational view of medication-control posture, unresolved issues, and current trends.
- Review MAI, top KPIs, risk distribution, and review queue.
- Use the shift snapshot to brief incoming leadership or pharmacy operations.
- Escalate from here into detailed tabs as needed.
Data Intake + AI
Purpose: Ingest approved reports and propose where fields belong.
- Drag and drop source files.
- Review AI mapping and confidence.
- Approve only after confirming the file type and affected modules.
- Remember: approved records append; they do not overwrite or delete prior source data.
Medication Accountability
Purpose: Verifies that dispensing equals documented need, valid authorization, and downstream accountability.
- Review five-way matching.
- Open orphan-event detail to understand why a transaction was flagged.
- Use the order timeline to see chronology gaps or reversals.
Live Dispensing
Purpose: Displays the event ledger for who dispensed what, where, why, and for which masked patient.
- Filter by unit, risk, or search terms.
- Use as the primary chronology of controlled removals and related event context.
Inventory Reconciliation
Purpose: Confirms that cabinet and physical inventory align with documented activity.
- Run reconciliation.
- Review expected versus physical counts.
- Investigate variances before sign-off.
Nurse Risk
Purpose: Prioritizes staff activity for review using pattern-based controls.
- Select a user to view explainable factors.
- Compare activity against same-unit/same-shift peers.
- Do not use the score alone for employment action.
Patient Utilization
Purpose: Reviews controlled-medication utilization in a masked and auditable manner.
- Select a patient to view their timeline.
- Use authorized reveal only when operationally justified.
- Not a diagnosis, addiction, or medical-necessity determination tool.
Visual Analytics
Purpose: Summarizes dispensing patterns across time.
- Toggle today, week, month, and year views.
- Use charts to spot concentration, spikes, and directional changes.
Waste Witness Intelligence
Purpose: Examines waste witness relationships and timing.
- Look for repeated reciprocal witnessing, delayed sign-off, or staffing mismatches.
- Select a witness pair for added context.
Prescriber Layer
Purpose: Adds order-origin context by linking provider, patient, nurse, medication, and dispensing location.
- Use this view to evaluate patterns that are relational rather than isolated to one user.
Data Quality
Purpose: Tells reviewers whether the evidence is reliable enough to trust the resulting analytics.
- Review completeness, duplication, conflicts, freshness, and confidence levels.
- Resolve evidence issues before relying on scores.
Audit & Exceptions
Purpose: Tracks cases, exceptions, linked evidence, workflow notes, and resolution status.
- Create or open a case.
- Document review action.
- Export or preserve evidence packages as needed.
Implementation Controls
Purpose: Demonstrates governance and control operations behind the analytics.
- Monitor retention, reconciliation, alert escalation, access review, and audit-ready reporting.
- Use implementation evidence to show the control environment is operating.
Intelligence Model
Purpose: Explains the model logic, factors, and rationale supporting review prioritization.
- Use for transparency and governance.
- Factor weights are illustrative in the sandbox.
Help, Manuals & Definitions
Purpose: Central reference library for all functions.
- Contains ownership language, privacy notes, operating limitations, and end-user guidance.
Implementation Controls
Operational governance for RXBeacon: preserve the record, reconcile every source, escalate unresolved risk, recertify access, and produce defensible audit evidence.
RET-01 • Data RetentionPolicy-based preservation, legal holds, immutable archive, approved disposition workflow.
REC-01 • Source ReconciliationPyxis, MAR, orders, identity/staffing and inventory must balance to ingestion totals.
ALT-01 • Alert EscalationSeverity, ownership, SLA, acknowledgment and escalation path remain attributable.
IAM-01 • Access ReviewRole-based access, privileged access, break-glass use and periodic recertification.
RPT-01 • Audit ReportingReproducible packages with scope, source counts, exceptions, approvals and provenance.
Implementation Control Register
Control owners, frequency, evidence and testing are explicit so the operating model can be audited independently of the Beacon score.
| Control | Objective | Owner | Frequency | Required evidence | Last test | Status |
|---|
Data Retention & Legal Hold
Retention periods are configurable and must be approved by hospital records management, legal, privacy, pharmacy and compliance. RXBeacon users cannot delete established data.
| Record class | Configured retention | Storage state | Legal hold | Disposition |
|---|
Design rule: routine users have no delete capability. Expiration does not equal immediate deletion; eligible records move to a governed disposition queue only after retention, legal-hold and approval checks pass. The sandbox does not destroy records.
Retention Control Evidence
Evidence expected for each retention cycle.
Immutable Provenance RequiredPolicy version • record class • retention rule • hold status • reviewer • approval • timestamp • disposition reference
Source-System Reconciliation
Confirms that source record counts and key monetary/medication quantities agree with RXBeacon ingestion and downstream matching. Unmatched records create exceptions rather than disappearing.
| Source | Source records | Ingested | Matched | Unmatched | Last received | Reconciliation | Status |
|---|
Latest reconciliation is based on the current sandbox snapshot.
Alert Escalation Queue
RXBeacon Intelligence alerts are assigned a severity, acknowledgment target, resolution target, owner and next escalation point. Escalation actions append to the audit trail.
| Alert | Severity | Age | Owner | SLA | Status | Action |
|---|
Escalation Matrix
Illustrative workflow targets; production thresholds are configurable by Highland leadership.
Safety rule: the platform escalates review; it does not accuse an employee, diagnose a patient, or autonomously restrict clinically indicated medication.
User Access Review & Privileged Access
Periodic certification of role, minimum-necessary access, privileged permissions and break-glass identity reveal. Changes are requested and logged; historical access evidence is retained.
| Role / access group | Users | Privileged | Last review | Next due | Exceptions | Status | Review |
|---|
Access review should include inactive/stale accounts, role changes, terminated users, segregation-of-duties conflicts, break-glass use, failed login patterns and privileged-role membership.
Audit-Ready Reporting
Generate a reproducible evidence package containing scope, control posture, source reconciliations, exceptions, access review, retention/hold status, alert history and provenance.
No implementation-control report has been generated in this session.
Audit Package Contents
Every package is point-in-time and identifies exactly what evidence was included.
Implementation Evidence Trail
Control actions are added to the same append-only provenance model used for medication records. This table shows implementation-control activity from the current session.
| Time | Control | Action | Actor | Evidence / reference | Status |
|---|
RXBeacon Intelligence Model
Explainable, configurable risk scoring for controlled-medication oversight. V1.3 uses synthetic weights and fictional records for demonstration.
1. IngestCabinet transactions, inventory, order context, MAR, staffing
2. NormalizeMedication, dose, time, unit, user, patient token
3. ScoreWeighted factors + multipliers + mitigation
4. ComparePeer, patient history, cabinet baseline, expected inventory
5. ActReview queue, reconciliation, audit case, closure evidence
Illustrative RXBeacon Transaction Risk Formula
Weights are intentionally transparent and should be validated with hospital pharmacy, nursing, compliance, legal, privacy, and clinical leadership before production.
TxRisk = clamp(0,100,
Base
+ (Override × 14)
+ (AfterHours × 8)
+ (UnresolvedDiscrepancy × 28)
+ (WasteWitnessLag × 12)
+ (OrderMismatch × 24)
+ (PeerDeviation × 10)
+ (RepeatPattern × 10)
+ (HighPotencyMultiplier × 1.20)
− (VerifiedReturn × 8)
− (TimelyWitness × 6)
− (DocumentedClinicalContext × 4)
)
Risk Domains
RXBeacon Intelligence keeps independent signals available for review instead of collapsing the context into a single opaque number.
Inventory integrity25%
Transaction integrity20%
User / peer deviation20%
Patient utilization pattern15%
Time / location context10%
Documentation / mitigation10%
Staff Risk Index (SRI)
Longitudinal staff review score. Factors include overrides, discrepancies, after-hours pattern, peer deviation, witness lag, returns, waste, and repeat exceptions.
A high SRI means “review this pattern first.” It does not mean diversion occurred.
Patient Utilization Review Index
Longitudinal pattern of controlled medications received across encounters, units, time, medication classes, dose frequency, and linked orders.
Designed for stewardship/compliance review—not diagnosis, addiction labeling, or denial of clinically indicated care.
NSD — Narcotic Surveillance Deviation
Measures deviation between expected and observed activity for a cabinet, medication, shift, user, or patient context.
NSD helps identify “different than expected.” Reviewers determine whether the reason is normal workflow, documentation error, waste, return, count issue, or a true control concern.
Medication Accountability Index (MAI)
Percentage of controlled-medication events fully traced from authorization/need evidence through final administration, disposition and inventory accountability.
MAI measures accountability completeness—not medical necessity.
Evidence Quality Index (EQI)
Independent confidence measure for completeness, consistency, uniqueness, freshness and provenance of source evidence.
Low EQI can reduce confidence or hold a score for manual review.
Relationship Risk
Looks for unusual repeated relationships across provider, patient, nurse/user, witness, medication, cabinet, unit and time.
Relationship patterns prompt review; they are never treated as proof by themselves.