Abstract Rationale Pulmonary infections are an important cause of morbidity and mortality in immunocompromised pediatric patients. Bronchoscopy with bronchoalveolar lavage (BAL) is a tool to characterize these infections. Although there have been studies that investigated the diagnostic yield of BAL, there are limited studies in pediatric population. The objective of this study was to investigate the BAL diagnostic yield and change in antimicrobial management in this population. Methods Single center retrospective study was conducted on immunocompromised patients aged 0-21 years with BAL between 2019-2024. Categorical variables presented as counts and percentages; continuous variables presented as medians and interquartile ranges. Associations between categorical variables were examined with the Pearson chi-squared test. Results Median age was 12 years (IQR 6-16), 53% male, and 66% Hispanic. Fifty-six percent were immunocompromised due to cancer chemotherapy. Pre-BAL, 19% had increased respiratory support, most commonly supplemental oxygen (76%). Common symptoms included cough (52%) and fever (47%). Most post-BAL studies were negative (only 14% positive bacterial, 7% positive fungal, and 2% positive acid-fast bacilli cultures). BAL was well-tolerated, with only 9% requiring increased respiratory support and 7.8% needing intensive care unit (ICU) escalation. Transfer to ICU was associated with increased respiratory support pre-BAL (58.2%, p = 0.00) and post-BAL (58.3%, p = 0.00). Positive microbiological results were associated with cough pre-BAL (76%, p = 0.015). Majority of those with positive cultures were not on antimicrobials pre-BAL (only 33% on antibiotics, 28% on antifungals). In patients with negative microbiological results, 60% were on antibiotics, and 40% were on antifungals pre-BAL. Post-BAL there was no association between change in antimicrobials and positive bacterial culture (19%, p = 0.082) or positive fungal culture (7.4%, p = 0.977). If addition of antibiotics was done, most had negative bacterial cultures (71%, p = 0.006). Few cases had antibiotics discontinued (15.9%) but if discontinued, all those cases had negative bacterial cultures. Conclusion BAL is a low-risk diagnostic procedure in immunocompromised pediatric patients with suspected pulmonary infections, demonstrated by minimal need for increased respiratory support or ICU escalation post-procedure. If required ICU, there was association with having increased respiratory support before BAL. The diagnostic yield of BAL was limited, particularly in patients who were receiving antimicrobial therapy prior to BAL. Our findings suggest a better microbiological yield when BAL is performed before the initiation of antimicrobials. Change in antimicrobials post-BAL was not frequent and appeared not to be dictated by BAL results. This abstract is funded by: None
Ahmed et al. (2026) studied this question.