PT  - JOURNAL ARTICLE
AU  - Pieter Goeminne
AU  - Sandrien Abbeel
AU  - David Ruttens
AU  - Lieven Dupont
TI  - Evaluation of three categorical sputum colour charts: is less more?
DP  - 2013 Sep 01
TA  - European Respiratory Journal
PG  - P4915
VI  - 42
IP  - Suppl 57
4099  - http://erj.ersjournals.com/content/42/Suppl_57/P4915.short
4100  - http://erj.ersjournals.com/content/42/Suppl_57/P4915.full
SO  - Eur Respir J2013 Sep 01; 42
AB  - Introduction: Sputum colour charts (SCC) are used in the follow-up of patients with bronchiectasis. The best known are the chart by Murray et al. (mSCC) with 3 categories and the one by Stockley et al. (sSCC) with 9 categories. (1,2) We compared these SCC with a new chart containing 17 categories (nSCC)(abstract n° 852384), comparing intra-rater reliability (IntraRR) and inter-rater reliability (InterRR).Methods: Twenty cystic fibrosis patients were asked to collect their morning sputum. All samples were scored 5 times in a random order by 2 blinded raters, scoring for the 3 SCC. IntraRR and InterRR was analyzed by calculating the kappa.Results: All 3 charts had significant IntraRR (all p<0.0001). The highest IntraRR was seen for mSCC (κrater1=0.91 and κrater2=0.78). Similar IntraRR was seen for the other charts (sSCC κrater1=0.62 and κrater2=0.63; nSCC κrater1=0.72 and κrater2=0.65). The InterRR was high for the mSCC (p<0.0001; κ=0.91), lower but significant for the sSCC (p=0.037; κ=0.59) and not significant for the nSCC (p=0.15; κ=0.26). Plotting out number of categories and p-value, we estimate that the p=0.05 threshold lies at ten categories.Conclusion: Higher InterRR is seen when the SCC contains less categories. IntraRR reliability was independent of number of categories. SCC cannot contain more than 10 categories to remain reliable.1 Murray MP et al. Eur Respir J 2009;34:361-42 Stockley RA et al. Thorax 2001;56:366-72.