Whither the WBC differential?--some alternatives.

Most efforts to automate the WBC differential have involved the use of devices to identify and enumerate WBC types, the underlying assumption being that the cell type percentages constitute the important end results. A different assumption exists. It is that the clinical usefulness of this procedure derives primarily from the fact that a trained morphologist has conducted a detailed examination of the slide. Based on this second assumption, a different approach to automation was investigated. This involved the use of a computer system to augment the morphologist in the slide examination process. This augmentation was accomplished by: 1. Determination of linked abnormalities on 117,000 WBC counts and presentation of linkages to the morphologist when a particular abnormality was identified on the hemogram or the slide. 2. Information feedback which compared each technologist with others in the laboratory performing similar tasks. 3. Varying the number of cells examined per slide, based on the apparent abnormality of the slide. The linkage patterns found, their clinical usefulness, and the impact on technologist performance and test quality are discussed.