Management and outcome of spontaneous pneumothoraces at three urban EDs

Objective:  Spontaneous pneumothoraces predominantly affect young people. Substantial morbidity arises from the disease, invasive procedures and hospitalization. The literature is inconsistent regarding optimal management. This retrospective study aimed to define factors affecting the outcome of these patients. Methods:  Patients were identified from databases at three EDs for explicit retrospective medical record review. Iatrogenic and traumatic pneumothoraces were excluded. Data collected included demographic details, treatment and outcome. The primary outcome was failure of initial treatment, defined as the need for a second treatment modality (including inpatient pleurodesis for persistent air leak) or re‐presentation within 5 days of treatment cessation. Associations with the primary outcome in primary spontaneous pneumothorax were assessed using logistic regression. Results:  One hundred and twenty‐one spontaneous pneumothoraces were identified. There was poor correlation between clinician estimates and objective measurement of pneumothorax size. Pneumothorax size, measured using the average interpleural distance method, was the only independent predictor of treatment failure. Initial treatment modality demonstrated a confounder relationship with outcome. Subgroup analysis for patients treated with continuous pleural drainage compared small and large calibre drainage tubes, with no significant difference found. Conclusions:  Objective measurement of pneumothorax size was the only independent predictor of treatment failure, with initial treatment modality having a confounding effect. Algorithms regarding initial treatment modalities are usually based on pneumothorax size and presence or absence of symptoms. The present study illustrates the importance of objective assessment of pneumothorax size in both clinical research and clinical practice.