Cryptosporidiosis: multiattribute evaluation of six diagnostic methods.

Six diagnostic methods (Giemsa staining, Ziehl-Neelsen staining, auramine-rhodamine staining, Sheather's sugar flotation, an indirect immunofluorescence procedure, and a modified concentration-sugar flotation method) for the detection of Cryptosporidium spp. in stool specimens were compared on the following attributes: diagnostic yield, cost to perform each test, ease of handling, and ability to process large numbers of specimens for screening purposes by batching. A rank ordering from least desirable to most desirable was then established for each method by using the study attributes. The process of decision analysis with respect to the laboratory diagnosis of cryptosporidiosis is discussed through the application of multiattribute utility theory to the rank ordering of the study criteria. Within a specific health care setting, a diagnostic facility will be able to calculate its own utility scores for our study attributes. Multiattribute evaluation and analysis are potentially powerful tools in the allocation of resources in the laboratory.