Life-Threatening Hemoptysis in a Pediatric Referral Center

Abstract

Objectives: Hemoptysis is uncommon in children, even among the critically ill, with a paucity of epidemiological data to inform clinical decision-making. We describe hemoptysis-associated ICU admissions, including those who were critically ill at hemoptysis onset or who became critically ill as a result of hemoptysis, and identify predictors of mortality.

Design: Retrospective cohort study. Demographics, hemoptysis location, and management were collected. Pediatric Logistic Organ Dysfunction-2 score within 24 hours of hemoptysis described illness severity. Primary outcome was inhospital mortality.

Setting: Quaternary pediatric referral center between July 1, 2010, and June 30, 2017.

Patients: Medical/surgical (PICU), cardiac ICU, and term neonatal ICU admissions with hemoptysis during or within 24 hours of ICU admission.

Interventions: No intervention.

Measurements and main results: There were 326 hemoptysis-associated ICU admissions in 300 patients. Most common diagnoses were cardiac (46%), infection (15%), bronchiectasis (10%), and neoplasm (7%). Demographics, interventions, and outcomes differed by diagnostic category. Overall, 79 patients (26%) died inhospital and 109 (36%) had died during follow-up (survivor mean 2.8 ± 1.9 yr). Neoplasm, bronchiectasis, renal dysfunction, inhospital hemoptysis onset, and higher Pediatric Logistic Organ Dysfunction-2 score were independent risk factors for inhospital mortality (p < 0.02). Pharmacotherapy (32%), blood products (29%), computerized tomography angiography (26%), bronchoscopy (44%), and cardiac catheterization (36%) were common. Targeted surgical interventions were rare. Of survivors, 15% were discharged with new respiratory support. Of the deaths, 93 (85%) occurred within 12 months of admission. For patients surviving 12 months, 5-year survival was 87% (95% CI, 78-92) and mortality risk remained only for those with neoplasm (log-rank p = 0.001).

Conclusions: We observed high inhospital mortality from hemoptysis-associated ICU admissions. Mortality was independently associated with hemoptysis onset location, underlying diagnosis, and severity of critical illness at event. Additional mortality was observed in the 12-month posthospital discharge. Future directions include further characterization of this vulnerable population and management recommendations for life-threatening pediatric hemoptysis incorporating underlying disease pathophysiology.