Relationship of cold frequency and virus detection to exacerbations in COPD patients

Background: Cold frequency is positively associated with risks of acute exacerbations (AE) in COPD. We measured AE risk and symptom severity during AE in COPD patients with and without self-reported cold-like symptoms (CLS). Methods: Forty-six community-based GOLD 1-4 COPD patients transmitted daily health diaries by fax from December 2006 to April 2007, 50 transmitted diaries using BlackBerrys from December 2007 to November 2008 and 71 from December 2008 to April 2009. Sputum and nasal fluid collection for virology occurred if AE or colds were suspected. The relative increase in the AE rate in the presence of CLS was examined using a Poisson regression model. Generalised estimating equation models were used to examine differences in symptom severity scores. Results: Compliance with diary transmission exceeded 95%. Among 263 AE meeting Anthonisen criteria 79 (44%) of 180 with coincident CLS were virus positive, as were 26 (31%) of 83 without CLS. The population mean AE risk increased 1.3 times (95% CI 1.3,1.4 p<.001) for each CLS episode. Within subject risk of AE was 30 times (95% CI 20,47 p<.001) greater when CLS were present than when not. During AE with CLS present the median symptom score was 0.86 (range -2.7 to 6.0), with no CLS 0.29 (range -1.1 to 2.7). The estimated mean difference in severity score between AE with/without CLS was 0.51 (p<0.001), between virus positive/negative AE 0.52 (p<0.001). Conclusions: Positive virus testing correlates poorly with the presence of CLS. CLS are associated with a higher risk of COPD AE and increased respiratory symptom severity during these. COPD patients should be a priority for treatment with emerging drugs that target causes of CLS.