Critical Analysis of PIM2 Score Applicability in a Tertiary Care PICU in Western India

Abstract

Objective. Children have limited physiological reserve that deteriorates rapidly. Present study profiled patients admitted to PICU and determined PIM2 score applicability in Indian setting. Patients and Methods. Prospective observational study. Results. In 742 consecutive admissions, male : female ratio was 1.5 : 1, 35.6% patients were ventilated, observed mortality was 7%, and 26.4% were <1 year. The profile included septicemia and septic shock (29.6%), anemia (27.1%), pneumonia (19.6%), and meningitis and encephalitis (17.2%). For the first year, sensitivity of PIM2 was 65.8% and specificity was 71% for cutoff value at 1.9 by ROC curve analysis. The area under the curve was 0.724 (95% CI: 0.69, 0.76). This cutoff was validated for second year data yielding similar sensitivity (70.6%) and specificity (65%). Logistic regression analysis (LRA) over entire data revealed various variables independently associated with mortality along with PIM2 score. Another logistic model with same input variables except PIM2 yielded the same significant variables with Nagelkerke R square of 0.388 and correct classification of 78.5 revealing contribution of PIM2 in predicting mortality is meager. Conclusion. Infectious diseases were the commonest cause of PICU admission and mortality. PIM2 scoring did not explain the outcome adequately, suggesting need for recalibration. Following PALS/GEM guidelines was associated with better outcome.