1. Thiadens HA, de Bock GH, Dekker FW, Huysman JAN, van Houwelingen JC, Springer MP, et al. Identifying asthma and chronic obstructive pulmonary disease in patients with persistent cough presenting to general practitioners: descriptive study. BMJ 1998; 316: 1286-1290. (25 April.)
2. McFadden ER. Exertional dyspnea and cough as preludes to acute attacks of bronchial asthma. N Engl J Med 1975; 292: 565-568.
3. McKenzie S. Coughbut is it asthma? Arch Dis Child 1994; 70: 1-2[Medline].
4. Chang AB, Phelan PD, Carlin JB, Sawyer SM, Robertson CF. Randomised controlled trial of inhaled salbutamol and beclomethasone for recurrent cough. Arch Dis Child 1998; 79: 6-11[Abstract/Free Full Text].
5. Chang AB, Phelan PD, Sawyer SM, Robertson CF. Cough receptor sensitivity and airway hyper-responsiveness in children with non-specific recurrent cough. Am J Respir Crit Care Med 1997; 155: 1935-1939[Abstract].

Authors' reply

EDITORRothenberg wonders why we did not use a control group. We chose the cut off values in the pulmonary tests in such a way that only 5% of the general population might show airway obstruction by chance. Epidemiological studies have shown that bronchial hyperresponsiveness is present in the general population and that bronchial challenge testing cannot precisely separate asthmatic from non-asthmatic people in the community.¹ A paper on the prevalence of hyperresponsiveness and symptoms in the general population in the Netherlands showed that hyperresponsiveness was present in 16% of the adult population²; by contrast, 42% of our population had a low PD₂₀ (the provocative dose causing a 20% fall in forced expiratory volume in one second).

Our population was not a general population, since they attended a general practitioner with a troublesome cough, which is not the case in population studies. In clinical medicine virtually all people with asthma show airway hyperresponsiveness (PD₂₀ 5.6 µmol methacholine) when they have symptoms. For these reasons, a control group is not necessary. We agree with Rothenberg that follow up is important to determine prognosis for patients who cough who are and are not given a diagnosis of asthma or chronic obstructive pulmonary disease; we hope to report the results later.

We agree with Chang and Masters that in children with persistent cough overtreatment with inhaled corticosteroids is not uncommon in tertiary clinical practice. Misdiagnosis of asthma may also occur in adults in tertiary centres,³ but in general practice underdiagnosis is probably more frequent than overdiagnosis in coughing patients.⁴ It is dangerous to generalise findings from tertiary centres to primary care, since the prevalence of diseases in these settings differs. To avoid misdiagnosis, objective pulmonary function testing is necessary so that results of (methacholine) provocation tests are available to improve the diagnostic possibilities in general practice.

It was not our purpose even to suggest that most adults with asthma or chronic obstructive pulmonary disease have cough as the predominant symptom. We tried to find the key features to identify those with these disorders in a sample of patients with a persistent cough. Thus we agree with Chang and Masters that persistent cough as an isolated symptom has low predictive value for a diagnosis of asthma. These authors state that nocturnal cough did not significantly contribute in our study because of the definitions used. This might be an explanation, but nocturnal cough is a bad predictor of asthma not only in adults but in children as well.⁵