Predictive models for mortality and length of hospital stay in an Egyptian burns centre

Our aim was to obtain a statistical profile of survivors and deaths among burn victims and to develop predictive models for mortality and length of hospital stay. All patients admitted to the Burns Unit of Alexandria Main University Hospital over a 1-year period were included. Of 533 cases, mean length of hospital stay was 15.5 +/- 21.6 days and the mortality rate was 33%. Total surface area burnt, inhalation burns, age, sex, depth and degree of burn wounds were the significant independent predictors of mortality in multiple logistic regression analysis. The significant independent predictors of the length of hospital stay were clothing ignition, total surface area burnt, sex, degree and depth of burn and inhalation burns

Maternal haemoglobin and premature child delivery

The effect of anaemia [haemoglobin <11 g/dl] during pregnancy on preterm delivery was prospectively studied in 921 low-risk pregnancies. The association between anaemia and prematurity was measured at three different haemoglobin levels, early in pregnancy and during the third trimester. The risk of prematurity was higher with haemoglobin <9 g/dl in early pregnancy, followed by haemoglobin <10 g/dl during the same period; the least risk was recorded during the third trimester when haemoglobin was <10 g/dl. In multivariate logistic regression analysis, haemoglobin was the most important predictor of prematurity after controlling for other confounders. When applying the logistic predictive equation, a woman with all the significant risk factors included in the multivariate analysis had a 72% probability of giving birth to a preterm baby