Clinical Commander

All dossiers
ent.otitis-externa.core.v1

Acute otitis externa (incl. necrotizing/malignant OE)

general_internal_medicineacutesubacuteadultpediatricacuteoutpatientinpatient

Single ENT engine spanning the FULL severity arc of external-ear-canal infection: topical-first uncomplicated diffuse AOE (AAO-HNS strong rec AGAINST initial systemic antimicrobials) AND the necrotizing (malignant) OE emergency (Pseudomonas skull-base osteomyelitis in the diabetic/immunocompromised/elderly host) — two distinct regimen axes (topical ladder + systemic anti-pseudomonal). Skull-base-osteomyelitis source control + prolonged systemic therapy, sepsis bundle, glycemic-emergency management, and the middle-ear sibling are recognised then routed OUT by engine_id (id.osteomyelitis.core.v1, id.sepsis.core.v1, endo.dm2.core.v1, ent.otitis-media.core.v1, neuro.bells-palsy.core.v1) — not re-authored here. RxCUIs are guideline-confident anchors flagged for next-session live RxNav re-confirmation (per shard convention — RxNav reached this session for the core anti-pseudomonal set): ofloxacin 7623, ciprofloxacin 2551, ceftazidime 2191, piperacillin-tazobactam 33533, clotrimazole 2356. Acetic-acid otic and aural toilet/wick/analgesia/glycemic-control/systemic-antifungal are correctly modelled as non_pharm (no single stable MIN / owned by routed engines) rather than fabricating a code. Bayesian linkage (AOE vs AOM vs otomycosis vs necrotizing OE pre-test priors by host; LR+/LR− for granulation at the bony–cartilaginous junction, pain out of proportion, cranial neuropathy, canal edema/discharge; imaging + systemic-therapy decision thresholds; cross-engine routing edges by engine_id) is documented in the co-located _design-brief.md + _research-bundle.md; first-class TS LR fields remain schema-blocked (same constraint as derm.cellulitis.core.v1 / id.cellulitis.core.v1). Effect sizes (≥5): topical antimicrobial+steroid vs placebo OR 11 (95% CI 2.00–60.57) and acetic-acid inferiority OR 0.25–0.29 at weeks 2–3 (Kaushik Cochrane 2010, PMID 20091565); ciprofloxacin-otic clinical cure 86.6% vs PNH 81.1%, treatment difference 5.6% (95% CI −0.9 to 12.1) (Drehobl 2008, PMID 19032135); ciprofloxacin+fluocinolone faster otalgia resolution median 5.0 vs 5.9 d, p=0.002 (Chu JAMA Netw Open 2022, PMID 35834251); Pseudomonas aeruginosa 38% of AOE isolates, S. aureus 7.8% (Roland & Stroman 2002, PMID 12169893); necrotizing-OE mortality historically ~50% → 21% modern (Loh 2013, PMID 23558287) and 4% in conservatively-managed fungal MOE with 78.7% immunosuppressed (Sideris 2024, PMID 39534824); necrotizing-OE resolution 83% with algorithm-driven care and Tc-99m positive in 56% of CT-negative cases (Haleem 2025, PMID 40344787).

Entry points (5)

  • symptom
    Acute ear pain + otorrhea + tragal/pinna tenderness with diffuse ear-canal edema/erythema (classic diffuse AOE — Rosenfeld AAO-HNS CPG 2014; Roland & Stroman Laryngoscope 2002 — Pseudomonas-predominant)
    acute_otalgia_otorrhea_canal_edema
  • symptom
    Severe otalgia out of proportion to exam ± granulation tissue at the bony–cartilaginous junction of the canal floor — necrotizing OE pivot finding (Long Am J Emerg Med 2020; Loh Otolaryngol HNS 2013)
    ear_pain_out_of_proportion_with_granulation
  • history
    Previously diagnosed otitis externa NOT responding to topical therapy at 48–72 h — mandatory reassessment trigger (Rosenfeld AAO-HNS CPG 2014 key-action statement)
    otitis_externa_not_responding_at_48_72h
  • history
    Otalgia/otorrhea in a diabetic, immunocompromised, or elderly host — high necrotizing-OE prior, low imaging threshold (Long Am J Emerg Med 2020; Sideris Cureus 2024 — 78.7% immunosuppressed)
    diabetic_or_immunocompromised_with_otalgia
  • symptom
    New facial weakness or lower-cranial-nerve palsy with ear pain/discharge — necrotizing OE skull-base spread until proven otherwise (Loh Otolaryngol HNS 2013; route-in from neuro.bells-palsy.core.v1)
    cranial_neuropathy_with_ear_disease

Required inputs (15)

  • otalgia_severity_and_proportionalityrequired
    symptom • used at ENTRY
    Pain assessment is the single AAO-HNS strong-recommendation action; pain OUT of proportion to canal findings is the strongest bedside necrotizing-OE pointer (Rosenfeld AAO-HNS CPG 2014; Long Am J Emerg Med 2020)
  • ear_canal_exam_edema_discharge_granulationrequired
    symptom • used at CONTEXT
    Diffuse canal edema/discharge = AOE; localized pustule = furunculosis; granulation tissue at the bony–cartilaginous junction = necrotizing-OE pivot (Rosenfeld AAO-HNS CPG 2014; Loh Otolaryngol HNS 2013)
  • tympanic_membrane_statusrequired
    symptom • used at CONTEXT
    Intact vs perforated/tube-bearing TM gates ototoxic-drop avoidance AND separates AOE from acute otitis media with perforation (Rosenfeld AAO-HNS CPG 2014 — non-ototoxic preparation if TM non-intact)
  • diabetesrequired
    history • used at CONTEXT
    Diabetes is the dominant necrotizing-OE risk factor and the dominant modifiable substrate (microangiopathy, impaired neutrophil function, alkaline cerumen); flips management toward skull-base osteomyelitis (Long Am J Emerg Med 2020; Loh Otolaryngol HNS 2013)
  • immunocompromiserequired
    history • used at CONTEXT
    HIV, transplant, chemotherapy, chronic steroids → necrotizing-OE prior elevated; also raises fungal MOE prior (Sideris Cureus 2024 — 78.7% of fungal MOE immunosuppressed)
  • agerequired
    demographic • used at CONTEXT
    Elderly age is an independent necrotizing-OE risk; pediatric AOE is the AAO-HNS target population (age ≥2) and changes the otomycosis/foreign-body differential weighting (Rosenfeld AAO-HNS CPG 2014; Long Am J Emerg Med 2020)
  • water_exposure_or_canal_traumarequired
    history • used at CONTEXT
    Swimming, humid climate, hearing aids, cotton-bud trauma, dermatitis macerate the canal and seed Pseudomonas/Staph — the modifiable AOE substrate (Rosenfeld AAO-HNS CPG 2014; Roland & Stroman Laryngoscope 2002)
  • cranial_nerve_examinationrequired
    symptom • used at RED_FLAGS
    CN VII (then IX/X/XI/XII jugular foramen, VI petrous apex) palsy signals skull-base spread and the worst-prognosis necrotizing-OE subset — mandatory neuro screen in any at-risk host (Loh Otolaryngol HNS 2013; Long Am J Emerg Med 2020)
  • temperaturerequired
    vital • used at RED_FLAGS
    Uncomplicated AOE is usually afebrile/non-systemic; fever + systemic features raise the necrotizing/septic prior and the disposition tier (Rosenfeld AAO-HNS CPG 2014; Long Am J Emerg Med 2020)
  • esr
    lab • used at INITIAL_WORKUP
    ESR correlates with necrotizing-OE disease activity and is the primary treatment-response/duration marker (serial down-trend gates antibiotic cessation) (Loh Otolaryngol HNS 2013; Haleem Am J Otolaryngol 2025)
  • crp
    lab • used at INITIAL_WORKUP
    CRP tracks necrotizing-OE activity alongside ESR and trends faster; serial CRP gates IV→PO step-down and end-of-therapy decisions (Loh Otolaryngol HNS 2013)
  • glucose_or_a1c
    lab • used at INITIAL_WORKUP
    Hyperglycemia/HbA1c quantifies the dominant necrotizing-OE driver; poor glycemic control worsens prognosis and triggers cross-engine glycemic optimization (Long Am J Emerg Med 2020; Loh Otolaryngol HNS 2013)
  • ear_canal_culture
    lab • used at BRANCHING_WORKUP
    Canal swab/culture (and biopsy of granulation tissue) identifies Pseudomonas vs MRSA vs fungal MOE and guides systemic anti-pseudomonal therapy + excludes EAC malignancy (Roland & Stroman Laryngoscope 2002; Sideris Cureus 2024)
  • temporal_bone_ct_and_nuclear_or_mri
    imaging • used at BRANCHING_WORKUP
    CT temporal bone for bony erosion; Tc-99m bone scintigraphy when CT equivocal; MRI/gallium SPECT for soft-tissue extent and treatment-response — confirms necrotizing OE and stages skull-base involvement (Haleem Am J Otolaryngol 2025; Long Am J Emerg Med 2020)
  • creatinine
    lab • used at TREATMENT
    Renal function for systemic ciprofloxacin / ceftazidime / piperacillin-tazobactam dose adjustment in necrotizing OE (Long Am J Emerg Med 2020; Loh Otolaryngol HNS 2013)

12-phase flow (12)

  1. 1FRAME
    Frame the external-ear-canal infection across its full severity arc: uncomplicated diffuse AOE is a TOPICAL-ONLY disease (AAO-HNS strong rec AGAINST initial systemic antimicrobials — Rosenfeld 2014); the engine's parallel job is to detect the host/exam signal of NECROTIZING (malignant) OE — Pseudomonas skull-base osteomyelitis in the diabetic/immunocompromised/elderly — which flips management entirely. Distinguish FROM acute otitis media (middle ear) and FROM referred otalgia / EAC malignancy. Skull-base-osteomyelitis source control, sepsis, and glycemic emergencies are routed OUT by engine_id, not authored here.
    advance: OE scope confirmed; AOE-vs-necrotizing-OE dual frame established; not-this-engine concerns routed by engine_id
  2. 2ENTRY
    Recognise the entry trigger — acute otalgia/otorrhea/canal-edema (likely diffuse AOE) vs pain-out-of-proportion + granulation (necrotizing-OE pivot) vs not-responding-at-48-72h (mandatory reassessment) vs at-risk host (diabetic/immunocompromised/elderly) vs cranial neuropathy (skull-base spread); capture pain severity and proportionality up front (the AAO-HNS strong-rec pain action).
    inputs: otalgia_severity_and_proportionality
    advance: entry trigger identified; pain assessed and analgesia recommended by severity
  3. 3CONTEXT
    Build the host + substrate prior: canal exam (diffuse edema/discharge = AOE; localized pustule = furunculosis; granulation at bony–cartilaginous junction = necrotizing pivot); TM status (gates ototoxic-drop avoidance + separates from AOM); diabetes; immunocompromise; age; water/trauma/dermatitis substrate. This phase assigns the AOE-vs-otomycosis-vs-necrotizing-OE pretest probability.
    inputs: ear_canal_exam_edema_discharge_granulation, tympanic_membrane_status, diabetes, immunocompromise, age, water_exposure_or_canal_trauma
    actions: panel.glucose_a1c
    advance: host substrate + AOE/necrotizing pretest prior assigned
  4. 4RED_FLAGS
    Necrotizing-OE + systemic red-flag screen: cranial-nerve exam (CN VII first → IX/X/XI/XII jugular foramen → VI petrous apex — skull-base spread, worst prognosis); fever/systemic toxicity; granulation tissue + pain out of proportion in a diabetic/immunocompromised host. Positive necrotizing screen → high-suspicion pathway + route to id.osteomyelitis.core.v1 for prolonged systemic therapy + ID co-management; qSOFA≥2/sepsis → route to id.sepsis.core.v1; isolated CN VII without ear disease → consider neuro.bells-palsy.core.v1 (but necrotizing OE must be excluded first in any at-risk host).
    inputs: cranial_nerve_examination, temperature
    actions: calc.qsofa, workup.bells_palsy
    advance: necrotizing-OE + sepsis + cranial-neuropathy red flags screened and routed by engine_id if positive
  5. 5INITIAL_WORKUP
    Uncomplicated AOE needs NO labs/imaging — clinical diagnosis, topical therapy (AAO-HNS CPG 2014). When any necrotizing modifier is present: ESR + CRP baseline (the disease-activity/treatment-response markers — Loh 2013), glucose/HbA1c (quantify the dominant driver), CBC + inflammation panel, creatinine for systemic dosing. Aural toilet + microscopy at the same visit to clear debris and enable drop delivery.
    inputs: esr, crp, glucose_or_a1c
    actions: panel.cbc, panel.inflammation, panel.glucose_a1c
    advance: uncomplicated AOE → straight to TREATMENT; necrotizing modifier → baseline ESR/CRP/glucose sent + aural toilet done
  6. 6BRANCHING_WORKUP
    Necrotizing-OE confirmation tree: canal/granulation-tissue culture (Pseudomonas vs MRSA vs fungal MOE — Roland & Stroman 2002; Sideris 2024) AND mandatory biopsy of granulation to exclude EAC squamous-cell carcinoma; CT temporal bone for bony erosion; Tc-99m bone scintigraphy when CT equivocal (Haleem 2025 — 56% positive in CT-negative); MRI/gallium SPECT for soft-tissue extent and serial treatment-response. Otomycosis branch: fungal elements on microscopy → topical antifungal + aural toilet (NOT antibacterial drops). Refractory/atypical "OE" → biopsy mandatory before extending therapy.
    inputs: ear_canal_culture, temporal_bone_ct_and_nuclear_or_mri
    actions: workup.cellulitis_necfasc, workup.lymphadenopathy
    advance: necrotizing OE confirmed/excluded + skull-base extent staged, OR otomycosis/EAC-malignancy alternative assigned and routed
  7. 7DIFFERENTIAL
    Terminal differential with named pivots: uncomplicated diffuse AOE (diffuse canal edema + tragal tenderness + topical-responsive pivot) vs necrotizing/malignant OE (diabetic/immunocompromised host + granulation at bony–cartilaginous junction + pain out of proportion + ESR↑ pivot) vs otomycosis (itch-dominant + visible fungal hyphae/spores + topical-antibacterial failure pivot) vs furunculosis (localized pustule on a hair-bearing canal pivot) vs acute otitis media ± perforation (middle-ear pathology + bulging/perforated TM + NO canal-edema pivot — route ent.otitis-media.core.v1) vs EAC squamous-cell carcinoma / referred otalgia (refractory granulation + no inflammatory response + biopsy pivot) vs eczematous/contact dermatitis OE (chronic itch + dermatitis elsewhere pivot)
    advance: single best diagnosis selected; necrotizing OE explicitly excluded or confirmed; co-existence (AOE precipitating necrotizing OE in a diabetic) flagged
  8. 8RISK_STRATIFICATION
    Stratify: uncomplicated AOE = low-risk outpatient. Necrotizing OE is high-risk by definition — layer prognostic modifiers: cranial-nerve involvement (CN VII and especially lower CN palsy → worse outcome), clival/contralateral/intracranial extension, poor glycemic control, multidrug-resistant Pseudomonas, persistently elevated ESR/CRP. qSOFA/NEWS2 for systemic-toxicity tier. Severe pain, immunocompromise, or any necrotizing modifier lowers the admission threshold.
    inputs: temperature, cranial_nerve_examination
    actions: calc.news2, calc.qsofa
    advance: AOE risk tier OR necrotizing-OE prognostic modifier overlay assigned
  9. 9TREATMENT
    TWO-track treatment. (A) Uncomplicated diffuse AOE: topical ladder — fluoroquinolone-otic (ofloxacin/ciprofloxacin) ± dexamethasone for 7 d (AAO-HNS CPG 2014; Drehobl 2008 — ciprofloxacin ≥86% cure noninferior to neomycin/polymyxin; Chu JAMA 2022 — cipro+steroid faster pain resolution), acetic-acid drops as a low-cost option (Cochrane 2010 — inferior beyond week 1), aural toilet + wick when canal occluded, NON-OTOTOXIC preparation if TM non-intact/tube, analgesia by pain severity (strong rec), NO systemic antimicrobial (strong rec against). Otomycosis → topical antifungal + aural toilet. (B) Necrotizing OE: PROLONGED systemic anti-pseudomonal therapy — IV ceftazidime or piperacillin-tazobactam ± oral/IV ciprofloxacin (Loh 2013; Long 2020), aggressive glycemic control (route endo.dm2.core.v1), ENT + ID co-management, ESR/CRP-guided duration (often 6+ weeks), surgical debridement only for sequestra/abscess — source-control + prolonged-therapy ownership routed to id.osteomyelitis.core.v1.
    inputs: creatinine, tympanic_membrane_status
    advance: AOE → topical regimen + analgesia + aural-toilet/wick started, no systemic antibiotic; necrotizing OE → systemic anti-pseudomonal + glycemic control + ENT/ID co-management initiated and routed
  10. 10DISPOSITION
    Uncomplicated AOE → outpatient with 48–72 h return-if-not-better safety net (Rosenfeld AAO-HNS CPG 2014). Necrotizing OE (suspected or confirmed) → admit for IV anti-pseudomonal therapy, imaging, ENT/ID; route to id.osteomyelitis.core.v1 for prolonged-therapy ownership. Systemic toxicity / qSOFA≥2 → admit + route id.sepsis.core.v1. EAC malignancy on biopsy → head-and-neck oncology referral. Document the necrotizing-OE exclusion explicitly in every at-risk host.
    inputs: temperature, cranial_nerve_examination
    advance: disposition documented; necrotizing OE either excluded with rationale or admitted+routed
  11. 11MONITORING
    Uncomplicated AOE: clinical improvement expected by 48–72 h; persistence at 48–72 h mandates reassessment (confirm diagnosis, exclude necrotizing OE / otomycosis / EAC malignancy / non-ototoxic-drop need / non-adherence) NOT silent therapy extension (AAO-HNS CPG 2014 key-action statement). Necrotizing OE: serial ESR + CRP as the objective treatment-response markers gating antibiotic duration and step-down (Loh 2013 — correlate with disease activity); repeat MRI/nuclear imaging for persistent symptoms; surveillance for new cranial neuropathy or contralateral spread.
    inputs: esr, crp
    actions: panel.inflammation
    advance: AOE improving by 48–72 h OR reassessment triggered; necrotizing OE ESR/CRP trending down toward cessation criteria
  12. 12FOLLOWUP
    AOE recurrence prevention: dry-ear precautions, treat the eczematous/contact-dermatitis substrate, acidifying/drying drops for recurrent swimmer's ear, avoid cotton-bud trauma, hearing-aid hygiene. Necrotizing OE: long-term ENT/ID surveillance for recurrence (relapse can occur months later — re-image and re-check ESR/CRP on any symptom return), sustained glycemic optimization (route endo.dm2.core.v1 — the dominant modifiable recurrence driver), and cranial-nerve rehabilitation follow-up where deficits persist.
    inputs: diabetes, water_exposure_or_canal_trauma
    advance: AOE prevention plan documented; necrotizing OE long-term ENT/ID + glycemic-optimization surveillance plan established and routed