DEMO · REAL MOVER CASE

Quality intelligence and patient safety
for intraoperative anesthesia care.

LINCR turns continuous intraop monitoring + intervention data into the three things every anesthesia group needs and almost none can produce at scale: automated quality-improvement reports for the medical director, safety alerts for events that need clinical review, and defensibility documentation the malpractice carrier can stand behind. Every output starts from the same foundation — a clinical differential per event, generated the way a CAA actually reasons at the bedside.

This demo runs against the MOVER dataset (UC Irvine, 55,483 anesthesia cases). We sample 300 cases, detect every IOH event (sustained MAP <65 mmHg for ≥5 min), generate a differential per event, and aggregate up into the QI · safety · defensibility outputs you'd hand to a customer.

01 · CASE TIMELINE

Case #0 minutes of intraoperative physiology.

Each line is the actual recorded value from the MOVER dataset. Pink shaded windows are detected IOH events (sustained MAP <65 mmHg for ≥5 minutes). Drug administrations are stem markers along the bottom.

Mean Arterial Pressure (mmHg)
Threshold line at MAP = 65 mmHg. IOH event windows shaded.
Heart rate & ETCO2
Heart rate (teal) and end-tidal CO2 (amber) during the case. The HR × ETCO2 joint signature is what the endotype attribution reasons on.
Heart rate (bpm) ETCO2 (mmHg)
Drug administrations & fluid record
Vasopressors (phenylephrine, ephedrine, norepi, vasopressin) above zero; sedatives (propofol) and crystalloid below zero.
Phenylephrine Ephedrine Norepi/Vasopressin Propofol Crystalloid
02 · CLINICAL DIFFERENTIAL PER EVENT

Every event gets a clinical differential — the same reasoning a CAA writes in the chart, generated automatically.

This is the foundational unit of LINCR's output. For each IOH event, the platform triangulates across all available vitals + intervention signals and produces a structured differential: the most-likely cause as primary, a competing hypothesis as secondary when evidence supports it, and explicit rule-outs for the rest. Each line carries the evidence FOR and AGAINST that diagnosis. This is what flows downstream into the QI report, the safety alert, and the defensibility record.

Generated by a Python orchestrator (differential_engine.py) that wraps the signal-extraction layer. Output: primary + optional secondary + ruled-out + the lab/EKG signals that would resolve any remaining ambiguity. M2+ adds EHR lab integration; M3+ adds EKG waveform analysis and CF2 counterfactual scoring. The differential output is structured (JSON), so downstream rendering is trivially adaptable to any reporting surface (PDF for the QI committee, dashboard for the medical director, expert-witness exhibit for the malpractice attorney).

Full decision logic, signal definitions, and priority order are documented in section 02B below.

02B · ATTRIBUTION LOGIC

The same decision tree a CAA reasons through at the bedside, made reproducible in code.

For each detected IOH event (sustained MAP <65 mmHg for ≥5 minutes), the heuristic computes eight signals from the surrounding minutes of vitals + intervention record, then applies a priority-ordered decision tree. Bradycardia is checked first as a deterministic override. Hypovolemia and myocardial depression both score multiple weak signals — the distinguishing feature is whether HR rose (compensation) or fell (failure) in response to the MAP drop. Vasodilation is the default when HR is preserved. Unattributable is surfaced honestly when key data is missing rather than defaulting against insufficient evidence — in the audit it doubles as a data-quality flag for the customer.

Signals consumed per event

The audit triangulates across 13 intraoperative vitals + intervention channels. Where a definitive lab or waveform signal would normally apply (lactate, base deficit, Hgb trend, troponin, BNP, EKG ST/rhythm, echo), we flag it explicitly — those require EHR + waveform integration shipping in M2–M3+. The rules below operate on what continuous monitoring alone provides.

DETECTION GATE
MAP
Median + minimum during event. Gates IOH detection itself: sustained < 65 mmHg for ≥ 5 minutes is what flags an event in the first place.
COMPENSATION MARKER
HR Δ
Percent change in median HR during the event vs the 10-min baseline before. Rising HR = compensatory tachycardia (volume or vasodilation). Falling HR = failed compensation (cardiogenic) or vagal/AV-block.
COMPENSATION MARKER
HR absolute
Median HR during event. Gates the bradycardia trigger at < 50 bpm regardless of baseline. Definitive rhythm classification (sinus brady vs AV block vs nodal vs vagal) requires EKG morphology — M3+ deliverable.
PERFUSION SIGNAL
ETCO2 Δ
Percent change in median end-tidal CO2 vs baseline. Pulmonary perfusion proxy — falls when cardiac output drops because less CO2 reaches the alveoli per breath. Sensitive marker for both hypovolemia and cardiogenic causes.
PERFUSION SIGNAL
SpO2 drift
Peripheral perfusion sign tracked two ways: signal-loss percentage during event (pulse oximetry fails when peripheral pulsation drops) AND baseline drop in median value. Both indicate the patient's extremities aren't getting adequate flow.
PERFUSION SIGNAL
Pulse pressure Δ
SBP − DBP change vs baseline. Narrowing = low stroke volume (suggests hypovolemia or cardiogenic depression). Widening = decreased SVR (suggests vasodilation, sepsis, anaphylaxis). Prefer arterial line; falls back to NIBP.
VOLUME STATUS
SVV
Stroke volume variation during event. Fluid responsiveness marker — SVV ≥ 13% strongly suggests the patient would respond to a volume bolus. Computed from arterial-line waveform; available in cases with art line monitoring.
VOLUME STATUS
CVP Δ
Central venous pressure change. Falling CVP + falling MAP = preload failure (hypovolemia). Rising CVP + falling MAP = forward-flow failure (heart cannot move volume forward, cardiogenic).
VOLUME STATUS
EBL recent
Cumulative recorded blood loss in 15 min before event onset. Active surgical bleeding is one of the most specific hypovolemia signals when documented — clinically, hard to mistake for any other endotype.
VOLUME STATUS
Urine output rate
Urine output in 30-min event window vs the prior 30 minutes. Sustained drop indicates low renal perfusion — either from hypovolemia or sustained low MAP independent of cause.
DRUG EXPOSURE
Propofol load
Cumulative propofol administered in the 10 minutes before event onset. IV anesthetic-vasodilator. Recent propofol bolus is the most common drug-driven cause of intraop hypotension, especially after induction or during deep maintenance.
DRUG EXPOSURE
Agent Et
Median end-tidal volatile agent (sevo / iso / des) during the event. Inhaled anesthetic-vasodilator burden. ≥ 0.7% Et indicates maintenance or above — sufficient to cause continuous vasodilation in most patients.
DRUG EXPOSURE
Fentanyl load
Cumulative fentanyl in 10 min before event. Opioid contribution to both vasodilation (mild) and bradycardia (vagal). Recent bolus narrows the differential when HR drop accompanies the MAP drop.
01 of 13 · auto-rotating every 5 seconds Click any dot to lock · double-click to resume
DEFINITIVE LAB / WAVEFORM SIGNALS NOT IN MOVER
These are the signals an anesthesiologist would also use clinically; they require EHR or waveform integration:
  • Lactate, base deficit, Hgb trend — definitive for hypovolemia / shock states (EHR labs, M2+)
  • Troponin, BNP, echo — definitive for cardiogenic cause (EHR labs + imaging, M2+)
  • EKG ST-segment, rhythm classification, T-wave morphology — definitive for myocardial ischemia and bradycardic rhythm type (waveform analysis, M3+)

Decision priority — first match wins

01 of 05 · auto-rotating every 9 seconds Click any tab to lock
01 Bradycardia HR THRESHOLD + RAPID-DROP DETECTION
Detection criteria
  • HR during event < 50 bpm absolute, OR
  • HR Δ ≤ −25% from baseline (rapid drop, even if absolute ≥ 50)
TriggerEither criterion alone is sufficient
ConfidenceHIGH if both criteria met, otherwise MEDIUM
Definitive signalEKG morphology distinguishes sinus brady vs AV block vs nodal vs vagal — M3+ deliverable. HR threshold alone calls the event but does not classify the rhythm.
TreatmentChronotrope (atropine, ephedrine, glycopyrrolate)
02 Hypovolemia MULTI-SIGNAL TRIANGULATION — perfusion + compensation + volume
Detection criteria
  • SVV ≥ 13% — fluid responsiveness
  • EBL ≥ 100 ml in last 15 min — recorded blood loss
  • Compensatory tachycardia (HR Δ ≥ +10%) AND ETCO2 drop (Δ ≤ −8%)
  • SpO2 signal loss ≥ 30% during event — peripheral perfusion drop
  • SpO2 baseline drop ≥ 2% — mild peripheral perfusion drop
  • Pulse pressure narrowing ≤ −20% — low stroke volume
  • CVP drop ≤ −3 mmHg — preload failure
  • Urine output dropped ≥ 30 ml in 30-min window — renal perfusion drop
TriggerAny one criterion is sufficient. Multiple criteria triangulate to higher confidence.
ConfidenceHIGH if ≥ 3 signals fire, otherwise MEDIUM
Definitive lab signalsLactate, base deficit, Hgb trend — require EHR lab integration (M2+).
TreatmentCrystalloid / colloid / blood per loss estimate
03 Myocardial depression FAILED-COMPENSATION + CARDIOGENIC-PATTERN PROXY
Detection criteria
  • Failed compensation: HR Δ < −10% AND ETCO2 Δ ≤ −10%
  • Pulse pressure narrowing ≤ −25% without HR compensation (HR Δ ≤ +5%) — low stroke volume + no inotropic reserve
  • CVP rising ≥ +3 mmHg while MAP falling — classic forward-flow failure (heart can't move volume forward)
TriggerAny one criterion is sufficient. Multiple criteria raise confidence.
ConfidenceHIGH if ≥ 2 signals fire, otherwise MEDIUM
Definitive signalsEKG ST-segment change, T-wave inversion, new arrhythmia — waveform morphology, M3+ deliverable. Troponin, BNP, intraop echo — lab + imaging via EHR integration, M2+.
TreatmentInotrope (epinephrine, dobutamine), reduce anesthetic depth, treat underlying ischemia
04 Vasodilation DRUG-EVIDENCE BASED — not a default attribution
Drug evidence (any one required)
  • Propofol load > 50 in 10 min before event — IV anesthetic-vasodilator
  • Volatile agent ≥ 0.7% Et during event — inhaled anesthetic at maintenance or above
  • Fentanyl load > 30 in 10 min before event — opioid vasodilation contribution
Supporting signal (any one required)
  • HR preserved (Δ between −5% and +15%) — rules out concurrent hypovolemia or cardiogenic
  • Pulse pressure widening ≥ +10% — classic low-SVR sign (DBP drops more than SBP)
TriggerDrug evidence AND at least one supporting signal — both required, no triggering on physiology alone
ConfidenceHIGH if ≥ 2 drug signals AND pulse-pressure widening present, otherwise MEDIUM
NoteVasodilation is the most common cause of intraop hypotension because patients are continuously exposed to anesthetic vasodilators. We require drug evidence rather than calling it as a default for HR-preserved hypotension — absence of drug exposure with low MAP and preserved HR may instead reflect early hypovolemia, sepsis, or anaphylaxis.
TreatmentVasopressor (phenylephrine, vasopressin, norepinephrine)
05 Unattributable DATA-QUALITY FLAG — honest output when signal is missing
TriggerNo rule above fired with sufficient evidence. Either critical channels (HR, perfusion, SpO2, pulse pressure, drug record) were not populated around the event window, or the signal pattern does not match any of the four endotype signatures.
Why this existsRefusing to attribute when evidence is insufficient is a feature, not a bug. In the aggregate report, the percent unattributable is itself a metric — it tells the customer how complete their EHR data capture is, and which sites need monitoring or documentation upgrades. The right answer to incomplete data is “don't know,” not “default to vasodilation.”
ConfidenceAlways LOW — surfaced for data-quality review
ActionNo clinical recommendation issued. Event flagged for data audit.

Why this priority order

Bradycardia first because the HR threshold is unambiguous and missing it has high consequence (different drug class entirely). Hypovolemia next because volume-loss signals (SVV, EBL, pulse-pressure narrowing, CVP drop, urine output drop) are highly specific when present — a positive SVV or active EBL is hard to mistake for anything else. Myocardial depression after hypovolemia because their signatures overlap (both can show ETCO2 drop and pulse-pressure narrowing); the distinguishing feature is whether HR rose (compensation) or fell (failure), and whether CVP rose with falling MAP (cardiogenic). Vasodilation last among the affirmative endotypes because it requires drug evidence to fire — we don't default to it just because HR is preserved. Unattributable when no rule fires with evidence, so the system never claims an endotype the underlying data cannot support.

03 · SITE-LEVEL PATTERNS

Aggregate the differentials across N cases → the protocol-change story falls out of the data.

The medical director and the QI committee don't need 419 individual differentials. They need to know what's actually driving IOH burden across the group, where the outliers are, and which one protocol change has the most leverage. This is the next layer up.

Distribution of primary causes

From the differentials of N events across N cases.

Outlier events flagged for clinical review

Events that fall outside expected severity or duration patterns. These are the cases the QI committee should review first — the ones most likely to involve a near-miss or a documentation gap that matters.

Top protocol-change recommendation

Highest-leverage change identified by the audit. Based on which endotype dominates the burden + what intervention pattern in the data correlates with the events.

04 · WHAT THE CUSTOMER ACTUALLY RECEIVES

Three downstream outputs from the same differential foundation: QI report, safety alert, defensibility record.

The differentials per event and the site-level patterns above are the raw material. The customer-facing deliverables are below — what actually shows up in the medical director's inbox, the chief of anesthesia's safety review queue, and the malpractice carrier's expert-witness file.

OUTPUT 01 Quality Improvement Report For: Medical Director, QI Committee
OUTPUT 02 Patient Safety Alert For: Chief of Anesthesia, On-Call Lead
OUTPUT 03 Defensibility Documentation For: Malpractice Carrier, Risk Management, Legal

All three outputs derive from the same structured differential JSON produced for each event. Adding a new output surface is a rendering layer change, not a model retrain — e.g., a CMS PSI reporting pack, a research dataset export for an academic partner, or a real-time SMS to the on-call attending all reuse the same primitive. The data layer (LINCR's differential record) is the durable asset; the surfaces are commodity.