Sputum induction (SI) by inhalation of nebulised hypertonic saline is an efficient and non-invasive way of obtaining samples to assess airway inflammation and diagnose pulmonary infections such as tuberculosis (TB).1 2 Data published on the safety of SI are largely derived from subjects with obstructive lung disease. Yet on this basis a Working Group of the European Respiratory Society (ERS) recommended that all patients undergoing SI should receive pre-procedure bronchodilation with monitoring throughout by either serial forced expiratory volume in 1 s or peak expiratory flow rate (PEFR) measurements.3 Subjects with other respiratory conditions may not be at the same risk of bronchoconstriction as the original reference group. If so, then the recommended steps could be removed from routine SI, leading to a simplified procedure with no compromise in patient safety. Here we report our findings from routine clinical practice.

We prospectively observed 100 consecutive and unselected adult subjects undergoing SI to investigate possible TB. SI was performed using 3% hypertonic saline delivered via an ultrasonic nebuliser (Sunrise Medical, Wollaston, UK). To ensure good infection control, bronchospasm was assessed with disposable Mini-Wright peak flow meters (Clement Clarke International, Harlow, UK). PEFR was measured at baseline and then at 5, 10, 15 and 20 min. The preselected criteria for discontinuing SI were: (1) a fall in PEFR of >15% at any time point, (2) patient choice, (3) if the patient were deemed by the clinician performing the test to be unable to continue. Bronchodilators and resuscitation equipment were available if required. Equipment costs for generic salbutamol metered dose inhaler devices and the PEFR meter were calculated from the British National Formulary (March 2007) and manufacturer’s data.

Median age of the 100 subjects was 43 years (range 18–68) and 60 were male. Six were ex-smokers and 26 were current smokers. Median baseline PEFR was 450 l/min (range 220–710). No subject had to stop the procedure because of respiratory symptoms or a fall in PEFR of >15% from baseline. The median change in PEFR was −6.6% (range −14% to +29%) with 76 subjects having a fall in PEFR, eight no change and 16 an increase in PEFR at 20 min. None of the patients required a bronchodilator post-procedure or developed symptoms thereafter. Three of 100 patients stopped at 15 min: one with headache, one because of nausea and one as they had collected an adequate sample.

Our data argue against the current ERS guidelines. We found a low rate of bronchospasm or complications following SI in our patients being investigated for TB (including current and ex-smokers). This suggests that it is safe to omit routine bronchodilation and PEFR measurements during SI in these subjects. We estimate that removing these steps produces a saving of approximately £10 in cost and 10 min in time per procedure (which usually takes 30 min without this step). We would be interested to know whether other centres have had a similar experience in their clinical practice.