Abstract

FERNANDEZ, Alicia et al. PIM 2 score yield in a Pediatric Intensive Care Unit. Arch. Pediatr. Urug. [online]. 2016, vol.87, n.3, pp.234-239. ISSN 0004-0584.

Objective: to assess the PIM 2 score yield for the prediction of mortality in a population of critical pediatric patients. Methods: a prospective study was conducted, from January 1 to December 31, 2013. All patients who were between 29 days and 14 years old and were admitted to the Pediatric Intensive Care Unit at Asociación Española were included in the study. Those patients who died before 12 hours, those who were admitted for monitorization or observation after minor endoscopic or surgical procedures, or for the placement of catheters and neonates were excluded from the study. To assess yield of the PIM 2 score both its calibration and its discrimination, by means of a Hosmer-Lemeshow test for goodness of fit and the construction of the ROC curve and the area under the curve were analysed. Results: 184 patients were included for the analysis. The mortality rate of this populaiton was 5.4% (CI 98% 1.88-8.98). Average hospitalization was 8.3 days (range 12 hours to 27 days) and in the case of the patients who died, it was 11.3 days (range 1-20). Application of the Hosmer-Lemeshow test resulted in a Chi square test value of 5.37 (p=0.71). The area under the ROC curve was 0.90. For a 0.5 cut off point the yield of the test evidenced a global validity of 96.7 (CI 95% between 93.77- 99.57), a positive predictive value of 100% (87.5-100) and a negative prediciton value of 96.67% (93.77-99.57). Sensitivity for this cut off point was 40% (4.64-75.36) and a specificity of 100% (99.71-100). Conclusions: the PIM 2 score has evidenced, in the studied population and adequate global calibration and discrimination. Therefore, its usefulness as a tool for measuring and assessing assistance quality is still valid. However, its practical applicability was limited in patients allocated to “low”risk deciles, where the score presented discrimination problems and a high rate of false negatives.

Keywords : INFANT MORTALITY; RISK ASSESSMENT; PEDIATRIC INTENSIVE CARE UNITS.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )