Accuracy of cough reporting by carers of Indigenous children

The study aims to determine the reliability of a 24-h history of reported cough presence and quality (wet/dry) compared with objectively recorded cough, and evaluate factors that may influence cough reporting.

A digital recorder (ICD-PX720, Sony, Tokyo, Japan) was attached to 47 Indigenous children for 24 h during admission at Royal Darwin Hospital. After recording, carers reported their child's cough details. Cough counts were objectively measured and details of cough reports by carer (cough present/absent, quality wet/dry, cough severity determined by visual analogue scale and verbal category descriptive score) were the main outcomes measured. Other factors examined were: carer's sex, age, education, smoking, carer's cough, parent versus non-parent, respiratory illness in child and mean parent-proxy cough-specific quality of life questionnaire. Data were entered into STATA (V.10, STATA Corp., College Station, TX, USA). Cohen's kappa (κ) coefficients and Spearman's rank correlation coefficient (rs ) were used for data analysis.

Reporting of cough by Indigenous carers (compared with cough monitoring) was slight when a low cough threshold (0.25 coughs/h) was used (κ = 0.17, 95% CI -0.15, 0.49) and moderate when a higher cough threshold was used (κ = 0.57, 95% CI 0.28, 0.88). Carers' evaluation of wet cough disagreed with clinician's evaluation (κ = -0.24, 95% CI -0.58, 0.09). Subjective reporting of cough severity moderately correlated with objective cough rates (rs = 0.41 to 0.44, 95% CI 0.11, 0.67). None of the factors examined was associated with reliability of cough reporting.

The unreliability of reporting of cough highlights the need for community education on the importance of cough to improve health-seeking behaviour for early detection and treatment of respiratory disease.