ABSTRACT
Care providers for children with bronchiolitis use various tools to evaluate respiratory status. The use of a single tool by different types of care provider requires a high level of inter-observer agreement, an aspect rarely studied. This study, involving 82 physicians, nurses, and respiratory therapists aimed to evaluate inter-observer agreement for clinical evaluations in children hospitalized for a first episode of bronchiolitis. Respiratory evaluation included three frequently reported parameters of respiratory status: respiratory rate, retraction signs, and wheezing. The frequency of concordance for observers from the same and from different care provider groups was assessed using a weighted kappa statistic and considering all possible combinations of care providers. We also calculated inter-provider agreement as a function of patient age, regardless of care provider type. Overall inter-observer agreement for all provider pairs was 93.1%, with a weighted kappa statistic of 0.72 (95% CI, 0.66-0.78), indicating substantial agreement, with no difference as a function of pair composition. Inter-observer agreements for the various age groups ranged from 87% to 93%, with kappa scores ranging from 0.62 to 0.78. We conclude that a simple clinical evaluation for respiratory status assessment has a high level of inter-observer agreement within and between physicians, nurses and respiratory therapists. Thus, once the validity of this test has been confirmed in a large population sample, it should be possible to use this test to monitor children hospitalized with bronchiolitis and as an endpoint in clinical trials.
