Yale observation scale as a predictor of bacteremia and final outcome in 3-36 months old febrile children admitted in tertiary health centres: A hospital-based cross-sectional study

Objectives: The objective of the study was to assess predictability of bacteremia in febrile children in the age group of 3-36 months by application of Yale observation scale (YOS) and to predict clinical course during hospital stay and final outcome by YOS. Methods: A hospital-based cross-sectional study was carried out at Kasturba Medical College, Mangalore, Karnataka, for a period of 2 years (September 2013-September, 2015) in 100 febrile children in the age group of 3-36 months with probable infectious etiology admitted in ward/PICU. Children with any non-infectious causes of fever (vaccination, autoimmune, and immunodeficiency disorder) were excluded from the study. Cases were selected by simple random sampling. The primary study outcome was bacteremia based on positivity on blood culture and sensitivity sample drawn at admission. Secondary outcomes are clinical course in the hospital, use of antibiotics, need for mechanical ventilation, hospital stay, and mortality. Results: 100 cases were included in the study out of which 18 cases were bacteremic with a mean YOS of 26 (non-bacteremic - 11), mean hospital stay 19.5 days (non-bacteremic - 12 days). All 18 bacteremic children had YOS ≥20, but YOS ≥20 had 8 false positives cases. There was no significant interobserver variability in YOS assessment (Cronbach’s alpha - 0.993 showing good correlation with intraclass correlation coefficient - 0.986). Higher YOS scores had good sensitivity, specificity, positive and negative likelihood ratios, and area under curve for prediction of bacteremia at YOS >20 (100%, 90.2%, 10.2, 0.00, and 0.970), need for mechanical ventilation at YOS >21 (100%, 91.7%, 12.04, 0.00, and 0.969), need for scaling up antibiotics at YOS >21 (70.4%, 94.4%, 12.5, 0.31, and 0.822), and mortality at YOS >21 (90.9%, 85.4%, 6.2, 0.106, 0.878). Conclusion: YOS is a good tool to rule out bacteremia and to prognosticate the clinical course at the first visit. This simple scale can be of value in monitoring admitted patients for deteriorating clinical state and for assessing the need for referral to higher centers for further management.