Bronchoscopy in immunocompromised host with pulmonary infiltrates.

Pulmonary complications related to immunosuppression may be secondary to infection, neoplasia, toxic effects of chemotherapy and radiotherapy or the primary disease itself. The diagnostic yield from bronchoscopic studies on immunocompromised hosts (ICH) with pulmonary infiltrates varies widely and the indication and timing for bronchoscopic procedures remain uncertain. We prospectively studied 60 consecutive ICH with pulmonary lesions over a 12-month period. Bronchoscopic studies were performed as soon as pulmonary lesions were detected and within 72 hours of antimicrobial treatment. The patients were divided into two groups: E (40 patients) and L (20 patients) were bronchoscoped on average 1.6 days (SD 0.8) and 16.7 days (SD 10.8) respectively after clinical detection and antimicrobial treatment. A total of 131 bronchoscopic procedures were performed. These included bronchoalveolar lavage 60, bronchoscopic lung biopsy 47, bronchial biopsy 8, brushing 8 and washing 8. Diagnostic yields for bronchoscopically obtained fluid and tissue histology were 45% and 49% respectively. Both complement each other resulting in a higher diagnostic yield of about 70% of the patients in both groups. Procedural complications were minor (13% of cases) and mortality was zero. Infections accounted for approximately two-thirds of the pulmonary lesions. Patients bronchoscoped earlier received less antimicrobial empiric therapy and had shorter hospitalisation. Despite delayed bronchoscopy in ICH in the late group, bronchoscopic results influenced in 85% of patients. However, based on earlier use of appropriate therapy, shorter hospitalisation and decreased costs, we recommend early bronchoscopy when response to empiric treatment has been unsatisfactory.