Bronchoalveolar lavage fluid differential cell count. How many cells should be counted?

Abstract

OBJECTIVE: To investigate the number of cells to be counted in cytocentrifuged bronchoalveolar lavage (BAL) fluid preparations in order to reach a reliable enumeration of each cell type. STUDY DESIGN: A total of 136 BAL fluid samples for patients with suspected pneumonia or interstitial lung disease were investigated. Differential cell counts were performed on May-Grünwald-Giemsa-stained cytocentrifuged preparations by 2 observers, each differentiating 500 cells. Reliability for the enumeration of each cell type was expressed as phi value, as calculated in generalizability theory. RESULTS: For polymorphonuclear neutrophils (PMNs), alveolar macrophages, lymphocytes and eosinophils, an acceptable phi value of > or = .95 was reached at a count of 300 cells by 1 observer. Mast cells reached a phi value of only .674 at a count of 500 cells by 1 observer, precluding a reliable count. At a count of 500 cells by 1 observer, squamous epithelial cells, bronchial epithelial cells and plasma cells displayed phi values of .868, .903 and .816, respectively. CONCLUSION: At a count of 300 cells, PMNs, alveolar macrophages, lymphocytes and eosinophils are reliably enumerated in cytocentrifuged BAL fluid samples.