The use of infection probability score and sequential organ failure assessment scoring systems in predicting mechanical ventilation requirement and duration.

BACKGROUND

This study examines the efficacy of two different scoring systems in predicting mechanical ventilation need (MVN) and duration (DMV) in a surgical intensive care unit (ICU).

METHODS

This prospective observational study included 144 consecutive patients admitted to the ICU for more than 24 hours for whom the Sequential Organ Failure Assessment (SOFA) score and the Infection Probability Score (IPS) were calculated on admission and every 48 hours until discharge or death. Sensitivity and specificity of the two scoring systems were measured by the chi-square method. The Youden index and area under the Receiver Operating Characteristic (ROC) curve were also obtained.

RESULTS

For prediction of MVN, the best cut-off points were 2.5, 3.5, 2.5 for SOFA and 10.5, 10.5, 9.5 for IPS on the day of admission (0) and days 2 and 4, respectively. For the prediction of MVN, SOFA on days 0, 2 and 4 yielded significantly better results in the area under the ROC curve and Youden index than those of IPS (p<0.05). Neither of the two scoring systems provided good discrimination in prediction of more than 3 days respiratory support under MV.

CONCLUSION

For prediction of MVN, the SOFA scoring system on days 0, 2 and 4 has better accuracy than IPS.