Consideration should be given to the possibility of plague, particularly if there is a sudden increase in rodent mortality in an endemic region. The diagnosis should be considered in healthy subjects who suddenly become very severely ill with fever, extremely enlarged painful lymph nodes, brutal pneumonia or if a rapid succession of deaths occurs within one family.

*

Extensive leukocytosis is present. Microscopic examination of aspirated fluid from a bubo, sputum, cerebrospinal fluid and/or peripheral blood shows bipolar Gram-negative bacilli. The buboes do not contain liquid pus. Some sterile saline (1 ml) is injected into a bubo in order to obtain an aspirate. In the words of Yersin, the fluid contains "une véritable purée de microbes". Sometimes the bacteria can be detected in a thick or thin blood smear. They then have the appearance of a "safety pin" (bipolar granules). A staining method that reveals this clearly is the Wayson stain (based on basic fuchsin mixed with methylene blue in 95% ethanol and phenol). The organism is then light blue with darker terminal granules.

*

Culture is desirable for formal proof in view of the implications of a potentially threatening epidemic. Serology is possible in specialised laboratories (e.g. ELISA for detecting antibodies to the F1 antigen). Approximately 5% of survivors do not seroconvert. Serology permits a retrospective diagnosis, but is not useful for the acute, individual patient. There is also a technique available involving a dipstick coated with antibodies which can be used to detect the F1 antigen. This test can use sputum or serum, as early as the second day of the disease. The result is known in 15 minutes and is thus clinically very useful for the individual patient and any contacts. F1-deficient mutants occur very rarely and cannot be detected with this dipstick method.