Introduction Respiratory viruses have been associated with increased symptoms and a decline in lung function in patients with cystic fibrosis (CF). The optimal means of diagnosis of respiratory viruses in CF is unclear. We compared the suitability of sputum, nose- and throat-swabs for the diagnosis of respiratory viruses in adults with CF.

Methods 100 adults with CF provided sputum, nose- and throat-swabs every 2 months between December 2010 and June 2011 within a prospective observational study. Samples were repeated if new respiratory symptoms developed between regular visits. Specimens were analysed using PCR assays for adenovirus, influenza, metapneumovirus, parainfluenza, respiratory syncytial virus and rhinovirus. Positive viral identification rates for each specimen type were compared. FEV₁, inflammatory markers and symptom scores for upper respiratory tract infection (URTI) and pulmonary exacerbation (PEx) were recorded at each visit.

Results 210 sets of samples were collected. A respiratory virus was identified at 52 (24.8%) visits of which rhinovirus accounted for 63%, influenza A 10.5%, metapneumovirus 9% and influenza B 7%. Among virus-positive cases, sputum was positive in 34 (65%), nose swab in 25 (48%) and throat swab in 21 (40%). A single specimen type was positive in 32 (62%) cases; two specimens in 12 (23%) and all three specimens in only 8 (15%). Sputum alone was positive in 17 (33%) cases, nose-swab alone in 8 (15%) and throat-swab alone in 7 (13%). An increasing number of positive specimens was associated with higher mean (SD) URTI scores (4.9 (5.0) vs 6.7 (6.6) vs 10.3 (5.5) for 0, 1 and =2 positive specimens respectively; p=0.046 for =2 vs 1 specimens) and higher PEx scores (2.8 (2.8) vs 3.2 (2.2) vs 5.1 (1.5); p=0.002). FEV₁, CRP and WCC were similar between these groups. There were no significant differences in lung function, symptoms or inflammatory markers when viruses were detected in sputum compared with the upper airways.

Conclusions Sputum is superior to nose- and throat-swabs for the diagnosis of respiratory viruses in adults with CF but all three are required for optimal identification rates. Viral positivity in =2 specimens is associated with higher upper and lower respiratory symptom scores.