Early predictors of mortality in pneumococcal bacteraemia.

OBJECTIVES Pneumococcal bacteraemia carries a mortality of about 20%. Approximately 50% of deaths from pneumococcal bacteraemia occur within the first 48 h of admission. In order to influence outcome, critically ill patients should be identified at the time of presentation. This study enables the clinician to rapidly make an evidence-based assessment of a patient's prognosis, allowing the identification of patients who should be placed in a high-risk category at an early stage, when appropriate management is most likely to be effective. METHODS Data were collected from the medical record of history, physical examination, radiological examination and laboratory investigations done on initial presentation using a standardized proforma. The data were first examined by Pearson's Chi-squared test, with Yates' correction if needed. Variables found to be significantly associated with case fatality ( P < 0.05) by these methods were examined by stepwise logistic regression analysis in order to identify those factors which were independent predictors of case fatality. RESULTS The overall case fatality was 21%. Older age, apyrexia, tachypnoea, bilateral consolidation, hypoalbuminaemia, elevated aminotransferases, renal impairment, acidosis and leucopaenia were significantly associated with higher case fatality. Older age, acidosis and elevated serum alanine aminotransferase (ALT) were independent predictors of case fatality. Fifty-five percent of isolates belonged to serotypes 4, 6B, 9V, 14, 19F and 23F, to which good antibody levels have been documented in both young and elderly patients post-vaccination. Serotype 14 was most common, and was significantly associated with higher case fatality. Colder weather was associated with a higher incidence of both infection and case fatality. The case fatality amongst patients receiving ITU management was 44%. Less than 50% of patients who died received ITU management. CONCLUSIONS Despite the increased availability of new antibiotics and vaccines, the mortality of patients with pneumococcal bacteraemia remains unchanged. The parameters above allow early identification of patients with a higher case fatality; these patients may benefit from being placed in a "high-risk" category early on in their management. Vaccination of the elderly may reduce the incidence and/or mortality from pneumococcal bacteraemia. Further studies are required to understand the reasons for referral for intensive therapy in acute pneumococcal bacteraemia and whether ITU management affects outcome.