Predictors of outcome in the initial assessment of patients with infective endocarditis
Walton BI., Wallace SM., Kharbanda RK., Hardy R., Wilson AP., Swanton RH.
Infective Endocarditis (IE) carries a high morbidity and mortality. Prompt identification of high risk patients may improve prognosis by allowing changes in management strategies. The aim of this study was to define early markers of high risk. Methods: Consecutive patients with infective endocarditis presenting between 1981-99 to a tertiary centre were retrospectively studied. Clinical, echocardiographic and haematological data within 48 hours of admission were obtained. Outcome measures were mortality at discharge and six months. Data was analysed using univariate and multivariate logistic regression. Results: We obtained complete data on 201 of 215 cases (93.5%). 93% were positive for the Clinical Duke Criteria. 74 cases were from referring hospitals. Mean age was 52 years and 133 were male. 174 were culture positive - 45 were S. aureus. Valves infected were Aortic (83 cases). Mitral (71), Tricuspid (18) and multiple valves (29). 31% were prosthetic valve endocarditis. 53% of patients underwent surgery. Mortality at discharge was 19.4% and at 6 months 28.2%. Days ill prior to admission, age, sex, body mass, valve infected (type and position), visible vegetation, infecting organism, left ventricular function or renal function were not predictors of adverse mortality. However, both abnormal white cell count (WCC) < 3 or > 11 x 10 9 /L and albumin (SA), < 30g/l were significant predictors of mortality at discharge and 6 months as assessed by Odds Ratio (OR), (see table). Mortality Discharge O.R. Six months O.R. P value WCC 9.1 * 4.03 + *0.001, + 0.01 SA 3.17* 2.76 + *0.04, + 0.05 Conclusions: Factors that have previously been shown to influence prognosis in patients with endocarditis do not appear to do so early in hospital admission. Simple haematological indices which are readily available in routine clinical practice allow reliable, cheap and powerful prediction of high risk patients.