Comparison of two respiratory scores in children with acute bronchiolitis / Tan Wee Nee

• Main Author: Tan, Wee Nee
• Format: Thesis
• Published: 2017
• Subjects: R Medicine (General)
• Online Access: http://studentsrepo.um.edu.my/8913/4/tan_wee_nee.pdf; http://studentsrepo.um.edu.my/8913/
• Institution: Universiti Malaya

Background : Respiratory scores are an objective method of detecting severity of disease and response to treatment. However, choosing the most accurate system can be confusing with the availability of many different scores of varying complexities. Objectives: The objectives of this study were to a) compare 2 respiratory scores i.e.Kristjansson Respiratory Score (KRS) and Children's Hospital of Wisconsin Respiratory Scores (CHWRS), at predicting admission, b) correlate oxygen saturation at with length of stay (LOS), need for admission andnon-invasive ventilator, c) correlate change in respiratory score within 24 hours admission and LOS, d) determine the interrater reliability of the 2 scores, e) determine the association between aetiology and length of stay and e) identify components of the respiratory score which correlated with admission, in children admitted with bronchiolitis. Design: This is a single center, cross-sectional study at Pediatric Trauma Unit University Malaya Medical Centre, from March 20 l 7 to August 2017. Sample Population: Children aged l till 18 months old presenting with acute bronchiolitis were included. Patients with: chronic respiratory di case, congenital heart disease, more than 2 episodes of wheezing, diagnosis of asthma, symptoms more than 14 days and parents who refused consent were excluded. Methods: Each patient was assessed by 2 doctors using the 2 scoring systems(KRS and CHWRS) in the emergency department and within 24 hours after admission. Admission was decided by the managing doctor. , Demographic date, length of stay, nasopharyngeal cultures and virology, treatment received during admission and clinicalprogress were collected. Results: One hundred and twenty-nine children met the inclusion criteria but only 122 patients were recruited. Median (IQR) age of patients was 9 (6,12) months old and majority were discharged (69.7%). The are under receiver operating characteristic curve (aROC) for predicting admission was 0.832 for Children's Hospital of Winconsin Respiratory Score and 0.760 for Kristjansson Respiratory Score. Low saturation (Spo2<95%) was associated with need for admission (p=0.008) and for non-invasive ventilation (p= 0.027). The inter-rater reliability between the first and second assessors for CHWRS (Intraclass Classification [ICC] 0.918) was higher than for KRS (ICC: 0.829). The highest interrater reliability for CHWRS were chest x-ray (ICC 0.918) and heart rate (ICC O.892); whereas for the KRS, general condition (ICC 0.749) and respiratory rat (ICC00.742) were the highest. There was no association between length of stay and change in respiratory score from admission to 24 hrs after admission ( CHWRS r= 0.137, p 0.418; KRS r= 0.157, p= 0.352). Co-infection (bacterial and virus) was associated with a longer hospital tay (p=0.032). Breath sounds and surgical status in the CHWRS and breath sound and skin colour in the KRS were poor at predicting admission. Conclusion CHWRS had a better discriminative power in predicting admission and higher interrater reliability compared to KRS. Low saturation was significantly a ssociated with the need for admission and non-invasive ventilation. Co-infection (bacterial and virus) was associated with a longer hospital stay. Breath sound and skin colour were poor at predicting outcome. (Word count 488)