1. Floviken
  2. Keel
  3. Emergency Room
  4. Karenin, Sergei
Karenin, Sergei [TEST]
MRN5410477
47 M
DOB22-Jan-1979
CodeFull
AllergiesNKDA
ER attendingDr. Sjöberg
Arrived17:47
Dwell71h 33m
LocER

AI activity for this patient.

Lowfire

Neutropenic-fever empiric coverage agent

Deterministic IDSA / ASCO empiric coverage. When an ANC below 500 pairs with a temperature at or above 38.3 °C, the agent proposes a class-level antibiotic bundle for the clinician to sign or override. The algorithm is the source of truth; Claude only voices the recommendation.

Triggered. Bundle proposed for sign / override.

Lowfire · Neutropenic Fever · 2026-06-12

ANC 150, T 39.0°C — neutropenic fever criteria met.

ANC
150 cells/µL
Temperature
39.0 °C
Systolic BP
122 mmHg
Hemodynamically stable
yes
MDR history
MRSA
Indwelling catheter / line
yes

Antibiotics

  • β-lactam-1 · IV · X g every Y hours
  • glycopeptide-1 · IV · X mg/kg loading dose, then trough-guidedselected for MRSA coverage

Cultures

  • Blood cultures × 2one peripheral, one from each lumen of indwelling line
  • Urinalysis with culture

Ancillary

  • CBC with differentialrepeat in 6 hours
  • Comprehensive metabolic panel
  • Lactate

Nursing

  • Vital signs every 1 hour × 4 hours, then every 4 hours
  • Reverse isolation precautionsneutropenic patient

Pharmacy

  • Verify renal dosing on empiric antibiotics
  • Cross-check empiric coverage against patient's MDR historyDocumented colonization: MRSA

Decisions

  • Trigger met: ANC 150 cells/mm³ (<500) and temperature 39.0°C (single ≥38.3°C).
  • Risk stratification: high-risk inpatient IV pathway (v1 default for the simplified algorithm).
  • Empiric antipseudomonal β-lactam-1 monotherapy with β-lactam-1.
  • Recent MRSA colonization — added glycopeptide for gram-positive coverage. Additional indications: indwelling catheter.
  • Catheter present — paired blood cultures (line and peripheral) to evaluate for line infection.

Out of scope

  • Local antibiogram and unit-specific resistance patterns are not in the algorithm; consider when finalizing coverage.
  • Pediatric dosing is not covered by this algorithm (adults only).
  • MASCC scoring and full risk-stratification beyond hemodynamic stability are not implemented in v1.

Second opinion

Simulated colleague voice; class-level antibiotic names only. No real drug names are sent to the model.

Sign writes class-level antibiotics + "Neutropenic fever" problem. Override writes nothing.