Objective: To determine the value of the platelet count at admission for the assessment of the severity of disease in acute meningococcal infections.
Design: Retrospective and prospective, descriptive patient study.
Setting: University Hospital Intensive Care Unit (ICU).
Patients: All patients (n = 92) with acute meningococcal disease from 1985 to 1997, who arrived at the ICU within 12 h after hospital admission and had more than one platelet count during the first 12 h.
Measurements and results: After admission, platelets dropped in 95% of the patients. At admission, 2/41 (5%) of the non-hypotensive patients and 13/51 (25%) of the hypotensive patients had platelets fewer than 100 x 10(9)/l. During the following 12 h, these percentages increased to 15% and 71%, respectively. Fatalities had, at admission, a median platelet count of 111 x 10(9)/l (range, 19-302 x 10(9)/l), whereas the nadir, occurring at median 7.0 h (range, 1.3-12 h), was 31 x 10(9)/l (range, 12-67 x 10(9)/l). Plasma TNF, measured shortly after admission, correlated better with the platelet nadir (r = -0.65, p < 0.0001) than with the platelet count at admission. Similarly, serum lactate correlated better with the platelet nadir.
Conclusions: As platelets drop after admission, the use of the platelet count at admission for the assessment of the prognosis in acute meningococcal disease may be misleading. Frequently repeated platelet counts are a better tool for evaluating the severity of disease.