HCA Healthcare
Company-reportedContinuous decisioningCoordination compressedHuman quality loopSepsis surveillance should follow patient data continuously rather than wait for nursing rounds or shift changes.
Clinical operations · United States and United Kingdom
Collections: Queue eliminated · Regulated autonomy
Executive brief
HCA replaced shift-timed sepsis screening with continuous surveillance and a staffed telemetry bridge. The model signals; bedside clinicians still diagnose and treat.
AI value · Earlier sepsis identification
Company-reportedApproximately six hours earlier, according to HCA Healthcare
This is an operator-reported comparison with no published method, denominator, or statistical detail in the retrieved sources. A larger figure circulates in trade press but could not be verified.
Before
Screens manually for sepsis during rounds or shift changes using the current chart and bedside condition. → Assesses suspected sepsis and decides whether to initiate diagnosis and treatment.
After
Continuously evaluates vital signs, laboratory results, nursing reports, and other clinical data for patterns. → Receives the signal, checks the available information, and accelerates notification to the bedside team. → Evaluates the patient and decides whether to diagnose and treat sepsis.
Human boundary
SPOT signals; telemetry technicians escalate; bedside clinicians diagnose and treat. The system does not hold clinical decision rights.
Why it matters
Replacing periodic screening with continuous surveillance only works when a named human role owns the signal and the consequential decision stays clinical.
This case is company-reported. Use it for the operating-model shift; do not treat the numbers as independently measured.
Before
How the work ran before the change.
Nurse
Screens manually for sepsis during rounds or shift changes using the current chart and bedside condition.
ControlClinical screening cadence follows rounds and staffing handoffs.
Bedside clinician
Assesses suspected sepsis and decides whether to initiate diagnosis and treatment.
ControlClinician owns diagnosis and treatment.
What changed
Sepsis surveillance should follow patient data continuously rather than wait for nursing rounds or shift changes.
Decision rightSignal becomes continuous; clinical decision stays human
After
How the same work runs now.
SPOT
Continuously evaluates vital signs, laboratory results, nursing reports, and other clinical data for patterns.
ControlThe operator describes SPOT as AI or machine learning in 2019–2020 and as rules-based algorithms in 2025; this case treats it as algorithmic surveillance rather than autonomous clinical AI.
Telemetry technician
Receives the signal, checks the available information, and accelerates notification to the bedside team.
ControlA trained human intermediary owns escalation.
Bedside clinician
Evaluates the patient and decides whether to diagnose and treat sepsis.
ControlClinical authority remains with the bedside team.
Exception path
Signals route through trained telemetry technicians to bedside clinicians. Public sources do not document timeout, duplicate-alert, disagreement, or downtime handling, so those exception controls remain an explicit evidence gap.
Work removed
Decision authority
SPOT signals; telemetry technicians escalate; bedside clinicians diagnose and treat. The system does not hold clinical decision rights.
Before
Nurse
Screens manually for sepsis during rounds or shift changes using the current chart and bedside condition.
Control: Clinical screening cadence follows rounds and staffing handoffs.
Bedside clinician
Assesses suspected sepsis and decides whether to initiate diagnosis and treatment.
Control: Clinician owns diagnosis and treatment.
After
SPOT
Continuously evaluates vital signs, laboratory results, nursing reports, and other clinical data for patterns consistent with sepsis.
Control: The operator describes SPOT as AI or machine learning in 2019–2020 and as rules-based algorithms in 2025; this case treats it as algorithmic surveillance rather than autonomous clinical AI.
Telemetry technician
Receives the signal, checks the available information, and accelerates notification to the bedside team.
Control: A trained human intermediary owns escalation.
Bedside clinician
Evaluates the patient and decides whether to diagnose and treat sepsis.
Control: Clinical authority remains with the bedside team.
Work that left the path
Human role before
Nurses screened periodically during rounds and shift changes, then bedside clinicians assessed suspected cases and owned treatment.
Human role after
Telemetry technicians receive continuous signals and bridge them to bedside clinicians, who retain diagnosis and treatment decisions.
AI roleContinuously detects sepsis-consistent patterns in clinical data and sends a signal into a staffed escalation path. Public HCA descriptions differ on whether the method is machine learning or rules-based.
Traditional manual screening during nursing rounds or shift changes→Approximately six hours earlier, according to HCA Healthcare
Operator descriptions published from 2020 through 2025; SPOT in use from 2018 · HCA Healthcare hospital deployment
This is an operator-reported comparison with no published method, denominator, or statistical detail in the retrieved sources. A larger figure circulates in trade press but could not be verified.
No published baseline rate or expected-mortality model→HCA estimated that SPOT, together with evidence-based clinical interventions, helped save 8,000 lives
The five years preceding HCA's May 2019 announcement · 2.5 million patients across HCA Healthcare hospitals
This is an operator estimate explicitly attributed jointly to SPOT and concurrent evidence-based clinical interventions. The retrieved source provides no denominator or estimation method, so the result cannot be attributed to the algorithm alone.
Replacing periodic screening with continuous surveillance only works when a named human role owns the signal and the consequential decision stays clinical.
Anti-pattern
Do not attribute enterprise mortality improvement to an alerting system when the operator reports a joint clinical-intervention program and does not publish the estimation method.
Questions
Portability conditions
Reputation risk
medium
Evidence and authority
Watch · updated
Watch status: verify the cited source and deployment condition before reusing this case.
3 primary; publication outcomes are reported.
Bundle 1.0.0 · reviewed 2026-08-23 · stable ID e6027e1950637490
Princess Margaret Cancer Centre / University Health Network · Continuous decisioning
A machine can generate a candidate plan, but physicians and peer review select the plan delivered.
Google DeepMind and EMBL-EBI · Continuous decisioning
Researchers can use precomputed structure predictions, then decide what experimental validation is required.
Google Research Flood Hub · Continuous decisioning
A global model can transfer hydrology into ungauged basins.
Mayo Clinic primary-care practices · Continuous decisioning
A routine ECG can trigger evaluation for asymptomatic low ejection fraction before specialist suspicion.