Validation of the modified Tal score for assessing bronchiolitis severity

Abstract

Introduction: different scores are available to assess the severity of bronchiolitis. In Uruguay the modified Tal score is used. Despite its widespread use, there are no data on its validity as a diagnostic test.

Objective: to validate the clinical modified Tal score in children less than 24 months hospitalized with bronchiolitis.

Material and methods: a cross-sectional study of children under 24 months hospitalized in two providers with 1st episode of bronchiolitis. In 2 stages: (1) Pilot, June-July 2016, to homogenize criteria and analyze concordance (N = 28). (2) Validation, June-July 2017 (N = 110). Reproducibility was measured with reliability using Cronbach’s alpha coefficient. When p <0.05 it was considered relevant. Validity was assessed through the logical appearance of the scale.

Results: inter-observer correlation for the global scale was 0.97 (p <0.01). Inter-observer correlations for each item were: respiratory rate 0.88, wheezing 0.79, oxygen saturation 0.97 and retractions 0.93 (p <0.01). Cronbach’s alpha coefficient for the full scale was 0.35. When removing one item at a time, the coefficient was: no respiratory rate 0.35; without wheezing 0.36; without oxygen saturation 0.33 and without retractions 0,07.

Discussion-conclusions: high inter-observer reliability was observed. This reinforces the need for prior training and the importance of expertise in the application of instruments with various clinical components. The reliability of the modified Tal score is low and the degree of correlation between its different items is inadequate. These results need to be critically analyzed to improve decision making.