Making the diagnosis of asthma

Common tests measure different aspects of the disease

Asthma is becoming more common in all parts of the world, and establishing the diagnosis is as important to clinical practitioners as it is to epidemiologists. For both the “gold standard” is still a clinical diagnosis based on a characteristic pattern of symptoms: episodes of cough, of dyspnoea, and of chest tightness or wheeze. The epidemiologist, however, wants further objective diagnostic tests to distinguish affected from unaffected people and then to compare populations and monitor trends. Most recent prevalence studies have relied on measurements of bronchial hyperresponsiveness to histamine or methacholine or to an exercise challenge as well as on tests of lung function. Toelle et al have proposed that for epidemiological purposes current asthma should be defined as appropriate symptoms in the previous 12 months together with evidence of increased airway responsiveness.1

The need for a clinician to use more specialised diagnostic testing is less clearly defined, but it may be summarised as the need to establish the diagnosis and assess the severity. International guidelines broadly agree on the management of established asthma, but they differ in their emphasis on confirmatory diagnostic testing. The British guidelines do not mention even the simplest of investigations,2 but those published by the US National Heart Lung and Blood Institute provide a detailed algorithm.3 Perhaps the lack of emphasis on diagnostic testing in the British guidelines is based on the difficulties clinicians meet in gaining access to tests as well as in their interpretation. Nevertheless, improving diagnostic accuracy is important in patients with equivocal symptoms—for example, cough-variant asthma—especially given that a commitment to long term treatment usually means inhaled corticosteroids. Tests correlate poorly

Some recent research from Siersted et al on behalf of the Odense Schoolchild Study group evaluated the strengths and weaknesses of four commonly used diagnostic tests for asthma among a group of 495 schoolchildren. These were: simple spirometry (the ratio of forced expiratory volume in one second to forced vital capacity), serial peak flow measurements over 14 days with a calculation of variability, exercise testing, and a methacholine challenge.4 The traditional “reversibility” test was not included.

In this selected population (which included 203 children with a history suggesting asthma) 27% had current symptoms of asthma and 10% had been diagnosed as having asthma. Children with symptoms of asthma and a positive result for at least one test were classed as having asthma for the purposes of the study. This showed that the sensitivity of the tests was variable but generally low, ranging from 18% for the ratio of forced expiratory volume in one second to forced vital capacity to 59% for the methacholine challenge. The predictive value of a positive test ranged from 45% for serial peak flow monitoring to 72% for a methacholine challenge.

By contrast, the specificity of individual tests was invariably high with the negative predictive value for all four tests similar at around 75%. Furthermore, there were only weak correlations between the tests, confirming the established view that each measures a different pathophysiological facet of the asthma syndrome.5 The results showed that, of the tests used, methacholine responsiveness detected more children with symptomatic asthma as well as those diagnosed as having asthma than any other test.

For clinicians, however, the most relevant observations were those on the 52 children who were “probably asthmatic.” Half had only one positive test result, and three quarters of these were identified using the methacholine challenge. Nearly 90% in the “probable” category had the diagnosis confirmed when the results of peak flow monitoring and methacholine challenge were used together.

Clearly there will be fewer positive results in patients who are “possibly” rather than “probably” asthmatic, but the results nevertheless encourage the use of these two diagnostic tests in children in whom the diagnosis of asthma needs to be confirmed. These results show how various methods complement one another, and they also reinforce the message that asthma cannot be defined by a single physiological measurement. Finally, they provide justification for the diagnostic pathway outlined in the American guidelines.3

Well performed serial peak flow measurements over days (or preferably weeks) are more sensitive than either spirometry or the response to a bronchodilator5 in establishing the presence of abnormal airway function. Spirometric measurements remain, however, an important way of assessing the severity of established disease (which may be falsely underestimated using peak flow measurements alone6) and they provide other valuable clinical information. For example, although the forced expiratory volume in one second after a bronchodilator is a relatively insensitive test in making a diagnosis of asthma, it may be helpful in distinguishing the degree of airflow obstruction due to oedema, mucous plugging, and perhaps airway remodelling rather than to reactive bronchospasm. Because of its greater reproducibility, the accurate assessment of individual trends over time is also best achieved by spirometry.

Bronchial provocation testing in the laboratory ought to be available and used in cases of diagnostic difficulty, and it is probably more practical than exercise testing. It is more sensitive than the exercise test even when exercise related symptoms may predominate,7 but the correlation between the two is low.8 An exercise test seems to be clinically useful when asthma has to be distinguished from some other form of lung disease as the cause for symptoms associated with effort.9

In practical terms, therefore, the diagnosis of asthma ought to rely on a careful history followed by spirometry in all cases. When doubts linger, peak flow monitoring or methacholine challenge or both should be undertaken1: one positive result is enough. Only occasionally will other investigations be necessary. Once asthma has been diagnosed, serial peak flow measurements are important in further assessing severity and response to treatment.