Importance of pre-existing co-morbidities for prognosis of septicemia in critically ill patients.

OBJECTIVE: To determine admission characteristics associated with the outcome of septicemia in critically ill patients and more specifically assess the prognostic value of pre-existing co-morbidities. DESIGN: 5 year-retrospective cohort study. SETTING: Surgical Intensive Care Unit (ICU-20 beds) in a 1600 bed-tertiary care center. PATIENTS: Among 5457 patients admitted to the ICU between 1984 and 1988, 176 (3.2%) met prospectively-defined criteria for blood culture-proven septicemia (8.77 per 1000 patient-days). Overall septicemic patients had a 5-fold increased risk of death compared to non-septicemic patients (relative risk 5.03, 95% confidence intervals 4.17-6.07, p < 0.0001), and this estimate persisted after stratification according to age, sex, primary diagnosis and conditions of admission to the ICU (emergency/elective). RESULTS: Prognostic factors recorded on admission to ICU that were associated with mortality from septicemia among 173 patients were older age, higher admission Apache II score, gastrointestinal surgery, ultimately and rapidly fatal diseases and the number of co-morbidities in addition to the principal diagnosis (active smoking, alcohol abuse, non-cured malignancy, diabetes mellitus, splenectomy, recent antibiotic therapy, major surgery, or major cardiac event). In the multivariate analysis with logistic regression procedures, Apache II and co-morbidities were identified as the two independent predictors of mortality. CONCLUSIONS: Pre-existing co-morbidities assessed at the admission to the ICU significantly improved the prediction of mortality from septicemia compared to Apache II score alone.

Late adult onset of Langerhans cell histiocytosis mimicking glioblastoma multiforme.

Langerhans cell histiocytosis (LCH) with multiple organ involvement is a rare disorder in adults. Extrapituitary involvement of the central nervous system (CNS) is uncommon. We report the unusual case of a 55-year-old woman presenting with a left-sided hemiataxia-hemiparesis, left hemisensory loss and short-lasting episodes of an alien left hand due to lesions of the internal capsule and the right thalamus, extending into the mesencephalon associated with extensive surrounding edema, without pituitary involvement. The neuroradiological image suggested glioblastoma multiforme. Brain biopsy revealed inflammatory tissue and "pseudotumoral" multiple sclerosis was suspected. Biopsy of concomitant lung and bone lesions disclosed Langerhans cell histiocytosis. The treatment with pulsed steroids in association with mycophenolate mofetil led to a sustained, clinical neurological remission.

Bedside prediction of mortality from bacteremic sepsis. A dynamic analysis of ICU patients.

The prognosis in patients with sepsis depends on severity of acute illness, underlying chronic diseases, and complications associated with infection. Adjusting for these factors is essential for evaluation of new therapies. The purpose of the present study was to determine variables readily identifiable at the bedside that predict mortality in intensive care unit (ICU) patients with sepsis and positive blood cultures. For a 5-yr period, all patients of a surgical ICU presenting with positive blood cultures and sepsis were systematically analyzed for clinical variables and organ dysfunctions at the day of onset of sepsis and bacteremia and during the subsequent clinical course. The prognostic value of these variables was determined using logistic regression procedures. Of the 5,457 admissions to the ICU, 176 patients developed sepsis with positive blood cultures (3.2 per 100 admissions). The fatality rate was 35% at 28 days after the onset of sepsis; in-hospital mortality was 43%. Independent predictors of mortality at onset of sepsis were previous antibiotic therapy (odds ratio [OR], 2.40; 95% confidence interval [CI95], 1.59 to 3.62; p = 0.034), hypothermia (OR, 1.43; CI95, 1.04 to 2.44; p = 0.030), requirement for mechanical ventilation (OR, 2.97; CI95, 1.96 to 4.51; p = 0.009), and onset-of-sepsis APACHE II score (OR, 1.21; CI95, 1.13 to 1.29; p < 0.001). Vital organ dysfunctions developing after the onset of sepsis influenced outcome markedly. The best two independent prognostic factors were the APACHE II score at the onset of sepsis (OR, 1.13 per unit; CI95, 1.08 to 1.17; p = 0.0016) and the number of organ dysfunctions developing thereafter (OR, 2.39; CI95, 2.02 to 2.82; p < 0.001). In ICU patients with sepsis and positive blood cultures, outcome can be predicted by the severity of illness at onset of sepsis and the number of vital organ dysfunctions developing subsequently. These variables are easily assessed at the bedside and should be included in the evaluation of new therapeutic strategies.

Predicting survival of patients with sepsis by use of regression and neural network models.

OBJECTIVES: (1) To predict at the time of diagnosis of sepsis the subsequent occurrence of multiple organ failure and patient death; and (2) to compare the prediction accuracies of standard multiple logistic regression (MLR) and neural network (NN) models. METHODS: The data were collected during a 5-year period for all patients (n=173) who met prospectively determined criteria for sepsis and had positive blood culture results while admitted in the surgical intensive care unit at the University Hospital of Geneva, Switzerland. These data formed the basis for a retrospective cohort study described elsewhere. The MLR model was adapted from existing data. An NN model of the feed-forward, back-propagation type was constructed for predicting the outcome of sepsis with bloodstream infection. Both models were constructed from randomly chosen subsets of patients and subsequently were evaluated on the remaining (independent) patients. RESULTS: Survival after sepsis was predicted with an accuracy of 80% by the NN model, which used only information collected at the time of the diagnosis of sepsis. The development of multiple organ failure after the diagnosis of sepsis was predicted accurately (81.5%) with either the MLR or the NN model. Both the MLR and the NN methods depended on the interpretation of a likelihood quantity, requiring the choice of a threshold to make a survival prediction. The accuracy of the MLR models was very sensitive to the threshold value. The accuracy of the NN models was not sensitive to the choice of threshold, because they generated likelihood predictions that were distributed far from the middle range where the threshold was placed. CONCLUSION: Compared with MLR models, the NN models were slightly more accurate and much less sensitive to the arbitrary threshold parameter.