pulm.cap.peds.v1

Pediatric community-acquired pneumonia

pediatricssubacuteacutepediatricneonataloutpatientacuteinpatient

Start encounter →Print handoutPRODUCTION

Pediatric CAP = age-stratified empiric-coverage engine. §5.5.1 effect sizes wired (≥6, all PMIDs PubMed-verified 2026-05-16): CAP-IT (Bielicki JAMA 2021 PMID 34726708) lower-dose re-treatment 12.6% vs higher 12.4% (diff 0.2%, 1-sided 95% CI →4.0%, NI margin 8%) and 3-d 12.5% vs 7-d 12.5% (diff 0.1%) — both non-inferior; severe-CAP subgroup signal lower-dose 17.3% vs higher 13.5% → caveat encoded; SAFER (Pernica JAMA Peds 2021 PMID 33683325) 5-d vs 10-d clinical cure 88.6% vs 90.8% per-protocol (RD −0.016, 97.5% CL −0.087), ITT 85.7% vs 84.1% — non-inferior; EPIC pediatric (Jain NEJM 2015 PMID 25714161) pathogen detected 81%, viral 73% > bacterial 15%, RSV <5 yr 37% vs ≥5 yr 8% (the age-conditioned prior); Griffin (NEJM 2013 PMID 23841730) post-PCV7 pneumonia hospitalisation fell 551.1/100,000 in children <2 yr (≈47,172 fewer/yr); APPIS/Addo-Yobo (Lancet 2004 PMID 15451221) WHO-severe pneumonia oral amoxicillin vs injectable penicillin treatment failure 19% in each (RD −0.4%, 95% CI −4.2 to 3.3). PRIOR PMID BUG FIXED: previous evidence.pmids listed 34499792 as "CAP-IT" — that PMID is an unrelated German occupational-health survey (verified via PubMed MCP). Correct CAP-IT = 34726708 (Bielicki JAMA 2021;326(17):1713-1724; DOI 10.1001/jama.2021.17843). Also corrected SAFER → 33683325 (was implied 34499792). Other prior PMIDs (DELIVER/ProMISe/POINT/CAPE-COD/REDUCE) were adult-sepsis/steroid trials not specific to pediatric CAP — replaced with the pediatric-CAP-specific set. Two conditional dependencies explicitly modeled (do-not-read-in-isolation): (1) PATHOGEN PRIOR | AGE-BAND — neonate GBS/E.coli/Listeria → infant viral-predominant (RSV) → preschool S. pneumoniae → school-age Mycoplasma; encoded in DIFFERENTIAL phase, age required_input rationale, regimen step age-gating, and sibling features (EPIC Jain PMID 25714161); (2) SEVERITY-MORTALITY | MALNUTRITION/HIV in the WHO/LMIC pathway — same WHO grade carries higher mortality in the malnourished/HIV child (severity_trigger who_severity_conditional_on_malnutrition_hiv; RISK_STRATIFICATION phase purpose). Immunisation status is a third prior-modifier (Griffin PMID 23841730) wired into the ceftriaxone-if-under-immunised regimen logic. §5.5.2 differential-as-data: 3 sibling_differentiation blocks (adult CAP pulm.cap.core.v1; bronchiolitis peds.bronchiolitis.v1; asthma-peds pulm.asthma.peds.v1) with per-feature discriminators + overlap handling. Additional cross-dossier routing via workups[].branches_to: id.sepsis.core.v1 (neonatal/septic), pulm.cap.core.v1 (adult transition), peds.bronchiolitis.v1, pulm.asthma.peds.v1, pulm.tuberculosis.v1. Sickle-cell acute chest syndrome encoded as a DISTINCT entity routing to heme.sickle-cell.core.v1 (severity_trigger sickle_cell_acute_chest_syndrome_overlap). All engine_ids grep-verified to resolve in src/lib/dossiers/ 2026-05-16. Special populations: neonate/<3 mo (sepsis-overlap broad coverage — ampicillin + gentamicin/cefotaxime, route id.sepsis.core.v1); under-immunised (ceftriaxone for non-vaccine serotypes/Hib — immunisation-conditioned prior); immunocompromised/asplenia (broaden + atypical); sickle cell (acute chest syndrome — distinct, route heme.sickle-cell.core.v1); malnutrition/HIV/LMIC (WHO pathway + conditional severity-mortality); recurrent/non-resolving pneumonia (FOLLOWUP phase — immunodeficiency/aspiration/structural workup + 4-6 wk CXR); penicillin allergy (clarify severity; cephalosporin or macrolide alternative — contraindication_rules). Regimen axis peds_cap_empirics_by_setting_age: 6 weight-based stepwise tiers (neonate sepsis-overlap / outpatient 3 mo-5 yr / outpatient ≥5 yr / inpatient immunised / inpatient under-immunised-or-severe / PICU complicated) + flat reference list; mg/kg/day with age-banded max caps, IV→PO switch, short-course duration (SAFER/CAP-IT), penicillin-allergy alternatives, oseltamivir antiviral. Matches pulm.cap.core.v1 regimen depth but PEDIATRIC weight-based throughout. RxCUI verification (RxNav REST /rxcui/{cui}/properties.json 2026-05-16): FIX amoxicillin-clavulanate 723→19711 (723 was amoxicillin ingredient IN; 19711 confirmed amoxicillin/clavulanate MIN — same correction as adult sibling). Verified-correct unchanged: amoxicillin 723 (IN), ampicillin 733 (IN), ceftriaxone 2193 (IN), azithromycin 18631 (IN), doxycycline 3640 (IN), vancomycin 11124 (IN), clindamycin 2582 (IN), oseltamivir 260101 (IN), gentamicin 4128 (IN), oxygen 7806 (IN). NO hand-authored CUIs. Run npm run research:rxnav:validate before dosing automation — combination MIN vs SCD selection may differ from ingredient IN; oxygen flagged non_pharm. SCHEMA-GAP NOTES: (1) _types.ts has no first-class field for age-conditioned pathogen-prior probabilities / WHO severity bands — encoded narratively in DIFFERENTIAL/RISK_STRATIFICATION phase purpose, required_input rationale, severity_triggers, regimen step applies_when, and the .depth.md age-stratified table; (2) no conditional-dependency graph type — the 2 dependencies modeled in severity_triggers + phase logic + the .depth.md schema-gap log; (3) RequiredCalculator.drives enum has no "diagnostic_gate" — qSOFA/CURB-65 reuse screening/risk_stratification (CURB-65 kept ONLY to document the deliberate adult-divergence); (4) RequiredInput.kind has no "panel"/"score" — viral panel encoded as kind:"lab" id:respiratory_viral_panel; (5) acuity widened to [subacute, acute] and population to [pediatric, neonatal] to reflect the neonatal sepsis-overlap tier. PRODUCTION caveats: (1) RxCUIs RxNav-verified/corrected 2026-05-16 but run npm run research:rxnav:validate before relying on dosing automation; (2) calc.qsofa/curb65/bsa + panels + protocol.septic_shock + workup.pediatric_fever/bronchiolitis/tb all confirmed in clinical-tools-registry.ts; pediatric age-band RR/PEWS severity checker + pediatric organ-dysfunction score not yet calculator-wired (registry edit out of scope); (3) manifest/design_brief pointers unchanged per dispatch scope (manifest authoring deferred); (4) sibling pediatrics-domain dossier peds.cap.v1 encodes the same disease from the pediatrics entry-point — these two are intentional companions, not duplicates.

Entry points (6)

symptom
Fever + cough + age-band tachypnea in a child (IDSA/PIDS 2011 PMID 21880587)
fever_cough_tachypnea_child
symptom
Retractions, nasal flaring, grunting, accessory-muscle use, head-bobbing (WHO IMCI severity)
work_of_breathing_child
imaging
New infiltrate on CXR or consolidation on lung ultrasound (IDSA/PIDS 2011 PMID 21880587)
new_infiltrate_peds
vital_abnormality
SpO2 <90-92% in a child with respiratory illness — admission threshold (IDSA/PIDS 2011 PMID 21880587)
hypoxia_child
symptom
WHO severe — chest indrawing, cyanosis, inability to feed/drink, lethargy/AMS, convulsions (WHO IMCI)
severe_pediatric_pneumonia
history
Neonate (<28 d) / young infant (<3 mo) with fever or respiratory distress → sepsis-overlap pathway (route id.sepsis.core.v1)
neonate_fever_respiratory

Required inputs (22)

agerequired
demographic • used at CONTEXT
Age tier is the PRIMARY pathogen prior: neonate GBS/E.coli/Listeria; <5 yr viral-predominant + S. pneumoniae; ≥5 yr Mycoplasma rises (EPIC Jain NEJM 2015 PMID 25714161 — RSV <5 yr 37% vs ≥5 yr 8%; IDSA/PIDS 2011 PMID 21880587)
weightrequired
demographic • used at TREATMENT
ALL pediatric antibiotic dosing is weight-based (mg/kg/day) with age-banded max caps (IDSA/PIDS 2011 PMID 21880587; CAP-IT PMID 34726708)
temperaturerequired
vital • used at CONTEXT
Fever pattern + height; neonate/<3 mo with T ≥38°C mandates full sepsis workup (route id.sepsis.core.v1)
rrrequired
vital • used at CONTEXT
WHO age-band tachypnea thresholds: ≥60 (<2 mo), ≥50 (2-12 mo), ≥40 (1-5 yr), ≥30 (≥5 yr) — replaces CURB-65 RR criterion (WHO IMCI; IDSA/PIDS 2011 PMID 21880587)
spo2required
vital • used at CONTEXT
SpO2 <90-92% on room air → admit; hypoxia is the strongest single severity marker in pediatric CAP (IDSA/PIDS 2011 PMID 21880587)
hrrequired
vital • used at CONTEXT
Tachycardia + perfusion assessment for sepsis/shock overlap (Surviving Sepsis Pediatrics 2020)
work_of_breathingrequired
symptom • used at CONTEXT
Retractions/grunting/nasal flaring/head-bobbing — WHO severity classification (WHO IMCI)
feeding_statusrequired
symptom • used at CONTEXT
Inability to feed/drink in a young child = WHO SEVERE pneumonia → admit (WHO IMCI)
mental_status_childrequired
symptom • used at RED_FLAGS
Lethargy / irritability / convulsions = WHO severe / impending failure (WHO IMCI)
immunisation_status_pedsrequired
history • used at CONTEXT
PCV13/PCV15/PCV20 + Hib + influenza coverage shifts the pathogen prior — under-immunised → empiric ceftriaxone for non-vaccine serotypes/Hib (Griffin NEJM 2013 PMID 23841730 — PCV cut <2 yr pneumonia hospitalisation 551.1/100,000; IDSA/PIDS 2011 PMID 21880587)
recent_abx_or_hospitalization
history • used at CONTEXT
Resistant-pathogen risk; HAP differential if hospitalised in last 90 d (route pulm.cap.core.v1 / HAP) (IDSA/PIDS 2011 PMID 21880587)
underlying_lung_disease
history • used at CONTEXT
CF/bronchiectasis (Pseudomonas), neuromuscular/aspiration risk, BPD, asthma — changes empiric coverage and the differential (IDSA/PIDS 2011 PMID 21880587)
sickle_cell_immunocompromise
history • used at CONTEXT
Sickle cell (acute chest syndrome — distinct entity, route heme.sickle-cell.core.v1), HIV, transplant, malignancy, asplenia → broaden coverage + atypical (IDSA/PIDS 2011 PMID 21880587)
malnutrition_hiv_lmic
history • used at CONTEXT
Malnutrition / HIV / LMIC setting raises WHO-pathway severity-mortality CONDITIONAL on the same WHO severity grade — do not read severity in isolation in the malnourished child (WHO IMCI; APPIS Addo-Yobo Lancet 2004 PMID 15451221)
wbc
lab • used at INITIAL_WORKUP
Bandemia + leukocytosis pattern; severity adjunct, NOT pathogen-specific (IDSA/PIDS 2011 PMID 21880587)
blood_culture
lab • used at INITIAL_WORKUP
Recommended in severe / hospitalised CAP and complicated pneumonia per IDSA/PIDS 2011 (PMID 21880587)
crp
lab • used at INITIAL_WORKUP
Severity adjunct; not for routine outpatient mild CAP (IDSA/PIDS 2011 PMID 21880587)
procalcitonin
lab • used at INITIAL_WORKUP
Bacterial-vs-viral probability + duration-shortening adjunct; selective use, does NOT override clinical judgment in radiographic CAP (IDSA/PIDS 2011 PMID 21880587)
respiratory_viral_panel
lab • used at INITIAL_WORKUP
RSV/influenza/SARS-CoV-2/parainfluenza/hMPV — viral predominance under 5 yr; positive viral panel + no toxicity may support narrowing/withholding antibiotics (EPIC Jain NEJM 2015 PMID 25714161 — viruses 73% > bacteria 15%)
cxr
imaging • used at INITIAL_WORKUP
Confirms pneumonia in moderate-severe/hospitalised; NOT required for outpatient mild CAP (IDSA/PIDS 2011 PMID 21880587)
pleural_ultrasound
imaging • used at BRANCHING_WORKUP
Lung/pleural ultrasound — sensitive for consolidation + parapneumonic effusion/empyema; preferred to define drainable collection (IDSA/PIDS 2011 PMID 21880587)
current_medsrequired
medication • used at CONTEXT
β-lactam allergy (clarify severity before avoiding), baseline meds, recent antibiotics (IDSA/PIDS 2011 PMID 21880587)

12-phase flow (12)

1FRAME
Confirm pediatric CAP scope — community onset (not HAP/VAP, not bronchiolitis, not aspiration/foreign-body, not active TB); set age tier (neonate <28 d / 1-3 mo / 3 mo-<5 yr / ≥5 yr). Neonate/young infant defaults to the sepsis pathway (route id.sepsis.core.v1) (IDSA/PIDS 2011 PMID 21880587)
inputs: age
advance: Community acquisition + age tier confirmed
2ENTRY
Trigger captured (fever + age-band tachypnea, hypoxia, infiltrate, WHO severe distress, or neonate fever) (IDSA/PIDS 2011 PMID 21880587; WHO IMCI)
inputs: age, weight
advance: Entry trigger captured
3CONTEXT
Capture age-band vitals + work-of-breathing + feeding (sets the WHO/PEWS severity prior), immunisation status (shifts pathogen prior — Griffin PMID 23841730), comorbidity (CF/sickle-cell/immunocompromise), recent antibiotics, allergy, household exposure (TB, pertussis), malnutrition/HIV/LMIC context (IDSA/PIDS 2011 PMID 21880587; WHO IMCI)
inputs: temperature, rr, spo2, hr, work_of_breathing, feeding_status, immunisation_status_peds, recent_abx_or_hospitalization, underlying_lung_disease, sickle_cell_immunocompromise, malnutrition_hiv_lmic, current_meds
advance: Risk + severity inputs captured
4RED_FLAGS
WHO/PALS emergent: septic shock, severe hypoxia (SpO2 <90%), apnoea/grunting in infant, cyanosis, inability to feed, lethargy/convulsions, complicated pneumonia (large effusion/empyema/necrotising). Neonate fever → full sepsis workup. Initiate Hour-1 bundle if shock (Surviving Sepsis Pediatrics 2020; WHO IMCI)
inputs: spo2, mental_status_child
actions: protocol.septic_shock, workup.pediatric_fever
advance: Emergent escalation initiated or excluded
5INITIAL_WORKUP
CXR (PA + lateral) ONLY if moderate-severe/hospitalised (not outpatient mild); CBC + CRP/procalcitonin + blood culture in severe; respiratory viral panel; pulse oximetry; lung/pleural ultrasound if effusion suspected. EPIC: viruses 73% > bacteria 15% under 5 yr — a positive viral panel reshapes the empiric decision (Jain NEJM 2015 PMID 25714161; IDSA/PIDS 2011 PMID 21880587)
inputs: cxr, wbc, crp, procalcitonin, respiratory_viral_panel, blood_culture
actions: panel.cbc, panel.inflammation, workup.pediatric_fever
advance: Imaging (if indicated) + labs sent
6BRANCHING_WORKUP
CT chest if complicated/non-resolving; pleural ultrasound + drainage if moderate-large effusion/empyema; bronchoscopy if foreign-body suspected or non-resolving + immunocompromised; TB workup if endemic exposure/chronic cough/weight loss/contact (route pulm.tuberculosis.v1); bronchiolitis branch if <2 yr first-wheeze + viral prodrome (route peds.bronchiolitis.v1); asthma branch if recurrent wheeze + atopy (route pulm.asthma.peds.v1) (IDSA/PIDS 2011 PMID 21880587)
inputs: pleural_ultrasound
actions: workup.bronchiolitis, workup.tb, panel.pleural
advance: Branch tests obtained when triggered
7DIFFERENTIAL
Age-conditioned: neonate GBS/E.coli/Listeria; infant <5 yr RSV/influenza/hMPV/parainfluenza + S. pneumoniae (viral-predominant — EPIC RSV <5 yr 37%); school-age ≥5 yr S. pneumoniae + Mycoplasma/Chlamydophila rising. Mimics: bronchiolitis (<2 yr first wheeze + URI prodrome), viral URI/wheeze, asthma exacerbation, aspiration/foreign-body (focal/recurrent same-lobe), TB (chronic + contact + weight loss), empyema/necrotising (persistent fever + effusion) (Jain NEJM 2015 PMID 25714161; IDSA/PIDS 2011 PMID 21880587)
inputs: age
advance: Age-conditioned working pathogen/diagnosis category assigned
8RISK_STRATIFICATION
Severity tier by AGE-BAND physiology (NOT CURB-65/PSI — adult-only): outpatient (mild, SpO2 ≥92%, feeding, no retractions); inpatient (moderate — SpO2 <90-92%, dehydration/poor feeding, age <3-6 mo with significant illness, moderate-severe retractions, failed outpatient therapy); PICU (apnoea, shock, mechanical ventilation, complicated pneumonia with respiratory compromise). Severity-mortality is CONDITIONAL on malnutrition/HIV in the WHO/LMIC pathway (WHO IMCI; APPIS Addo-Yobo PMID 15451221; IDSA/PIDS 2011 PMID 21880587)
inputs: spo2, rr, work_of_breathing, feeding_status
actions: calc.qsofa
advance: Age-band severity tier documented and disposition implied
9TREATMENT
WEIGHT-BASED empirics by age × setting × severity × atypical-risk: outpatient 3 mo-5 yr → high-dose amoxicillin 90 mg/kg/d (CAP-IT — lower 35-50 vs higher 70-90 mg/kg/d both NI, re-treatment 12.6% vs 12.4%; PMID 34726708); outpatient ≥5 yr → amoxicillin + macrolide if atypical; inpatient fully-immunised → ampicillin 200 mg/kg/d; under-immunised/severe → ceftriaxone 50-100 mg/kg/d; +azithromycin if atypical/school-age; vancomycin/clindamycin if MRSA/complicated; oseltamivir if influenza; O2 target ≥92%; isotonic IVF; empyema → drainage ± fibrinolysis/VATS. Duration: 5 d non-inferior to 10 d (SAFER PMID 33683325), 3 d NI to 7 d in non-severe (CAP-IT PMID 34726708) (IDSA/PIDS 2011 PMID 21880587)
inputs: weight, age, respiratory_viral_panel
actions: calc.bsa
advance: Weight-based empiric regimen + duration + complication plan documented
10DISPOSITION
Outpatient if mild + tolerates PO + reliable caregiver/follow-up; ward if SpO2 <90-92%, dehydration/poor feeding, age <3-6 mo with significant illness, failed outpatient therapy, complicated pneumonia not in shock; PICU if shock, mechanical ventilation, apnoea, complicated pneumonia with respiratory compromise (IDSA/PIDS 2011 PMID 21880587; WHO IMCI)
inputs: spo2, feeding_status
advance: Disposition set
11MONITORING
Clinical response at 48-72 h; if not improving → reassess for resistance, atypical, complication (effusion/empyema/necrotising), or alternative diagnosis (bronchiolitis/asthma/foreign-body/TB). IV→PO switch once afebrile ≥24 h + tolerating PO + improving WOB/oxygen weaned; complete short course (5 d uncomplicated per SAFER; 10-21 d if complicated) (SAFER PMID 33683325; IDSA/PIDS 2011 PMID 21880587)
advance: Improvement achieved or therapy escalated
12FOLLOWUP
Pneumococcal (PCV) + Hib + influenza + COVID catch-up immunisation review (Griffin PMID 23841730 — PCV materially cut hospitalisation); household smoke-exposure cessation; immunodeficiency / aspiration / structural workup if recurrent or non-resolving pneumonia; CXR follow-up at 4-6 wk ONLY for round/lobar/necrotising or non-resolving pneumonia — not routine if clinically resolved (IDSA/PIDS 2011 PMID 21880587)
advance: Follow-up scheduled and prevention/recurrence addressed