The clinical course of forty autopsy cases with leukemia was investigated to assess factors associated with the development of visceral mycosis. From the results of multivariate analysis, the number of total leukocytes (XI), neutrophils (X 2), lymphocytes (X 3), and monocytes (X 4) in peripheral blood, the number of consecutive days indicating leukocyte count below 1,000/μ-l and the body temperature were selected as effective factors to discriminate cases with visceral mycosis from others. The following formula derived from the linear discriminant analysis of the above factors gave the best result to classify autopsy cases. Z=-100.5427-0.00401· (X2)-0.01057· (X 4) +0.05622· (X 6) +2.61331· (Wl) If Z is more than 0, the case is classified in the mycosis group. Using this formula, all of 21 autopsy cases with mycosis were classified correctly to the mycosis group and the only 2 of 19 autopsy cases without mycosis were incorrectly identified as being in the mycosis group. Applying the formula to clinical cases with leukemia, 7 out of 8 cases with mycosis and 12 out of 15 without mycosis were predicted correctly. These results suggest that the newly established risk index will be useful in making an early prediction and starting the empiric therapy for mycosis in leukemic patients.
ASJC Scopus subject areas
- Infectious Diseases