Predictive value of acid-base status in critically ill neonates
Background: Neonates have a high affinity for acid-base disorders. Although reports about the correlation between acid-base abnormalities and outcome are contradictory, metabolic acidosis remains a strong indicator of poor prognosis in critically ill patients.
Aim: is to analyze the predictive value of acid-base parameters in predicting outcome in critically ill neonates.
Patients and methods: Data were prospectively collected from 101 consecutive neonates, admitted to the Department of intensive therapy and care, Pediatrics clinic of University Clinical Center of Tuzla during 2011. We analyzed acid-base parameters by the traditional and Stewart's method and tested their predictive value in predicting outcome.
Resaults: Subjects who died had much worse all acid-base parameters. Differences in the derived parameters of acid-base analysis were also significant. Iin univariate logistic regression analysis the base excess (BE) and corrected anion gap (AGcorr) were statistically significant predictors of mortality, in contrast to the strong ion gap (SIG). In multivariate logistic regression analysis, AGcor was no longer a significant predictor (OR = 1.052, 95% CI = 0.94 to 1.18, p = 0.37), while the BE, according to this analysis, was an independent prognostic factor for mortality (OR = 0.725;% 95 CI = 0.62 to 0.85, p <0.001), which practically means, that for every decrease in BE by one unit, risk of death increased by 37.9%.
Conclusion: Acid-base status is a reliable assessment method of current health condition, may indicate signs of deterioration or crisis, and predict mortality risk. Metabolic acidosis remains strong predictor of mortality among critically ill neonates.
Key words: predictive value, acid-base status, neonate