Three main features must be considered in the diagnosis of asthma: symptoms, variable airflow obstruction and airway inflammation.[1–3] Airway inflammation is not yet readily tested in routine clinical practice and will not be considered further here. However, skin testing may be an adjunct to diagnosis and is discussed in this section.
Symptoms
Common symptoms of asthma include wheezing, chest tightness, dyspnea and cough. The characteristics of these symptoms, which are variable, often paroxysmal and provoked by allergic or nonallergic stimuli such as cold air and irritants, are useful in diagnosis. Nocturnal occurrence is common. Measuring the patient's response to a therapeutic trial may be helpful in diagnosis. Nonpulmonary symptoms that suggest a predisposition to allergy - rhinitis, conjunctivitis and eczema - are also common in, but not specific to, asthma patients. In patients with symptoms that are persistent or that do not respond to simple treatment, objective confirmation of variable airflow obstruction is required.4
Variable airflow obstruction
Objective measurements are needed to confirm the diagnosis of asthma in all patients and to assess its severity. Objective documentation of variable airflow obstruction can be obtained through measurement of FEV1, PEF or hyperresponsiveness to methacholine inhalation challenge.
Forced expiratory volume in 1 second
Variable airflow obstruction can be illustrated by improvement in FEV1 15 minutes after an inhaled β2-agonist or after a 7- to 14-day course of inhaled glucocorticosteroid or ingested prednisone. A 12% or greater improvement in FEV1 (i.e., at least 180 mL) from the baseline after administration of a β2-agonist is considered significant5 (i.e., outside the 95% confidence interval (CI) for repeatability in people without asthma). However, there are no data to confirm that a bronchodilator response outside this 95% CI is indicative of asthma, and some suggest basing diagnosis on a greater than 15% increase in FEV1.5
Because there is greater variability in FEV1 over a longer time interval (days or weeks v. minutes), longer-term changes in FEV1, either without any specific therapeutic intervention or after glucocorticosteroids, must be greater than 20% (at least 250 mL). A trial of glucocorticosteroid involves maximizing the patient's response to a bronchodilator and obtaining a baseline FEV1, then carrying out a follow-up measurement after a 2-week course of prednisone (taken at the rate of 30 to 40 mg/d) to determine significant response.5
Peak expiratory flow
Home measurement of PEF may also be used to document variable airflow obstruction.[6, 7] Variable airflow obstruction is confirmed when the 95% CI of the mean percentage difference between the highest and lowest of 4 PEF values (morning and afternoon, before and after using a bronchodilator) is > 12%.7 However, some recommend a 20% variability to confirm the diagnosis of asthma.8 The importance of appropriate technique and the limitations of PEF are discussed further under "Home monitoring."
Airway hyperresponsiveness
In patients with normal airflow while resting, excessive responsiveness to a bronchoconstrictor can be documented using a methacholine inhalation challenge.8 This test should be done when symptoms are present or have occurred within a few days. Usually the test is available only in specialized centres, which may limit its utility. This test should be made available to primary care physicians who see most patients with mild asthma and where the measurement of responsiveness is most useful.9 Tests for airway responsiveness may give normal results in patients with glucocorticosteroid-responsive cough due to eosinophilic bronchitis.10
Evaluation of asthma severity
There is no agreement about how best to assess overall asthma severity. Assessment of asthma severity before or without treatment usually takes into account 3 factors, including 2 considered in the diagnosis: symptoms, physiologic indicators of airway disease and asthma morbidity. Thus, some algorithm based on frequency and severity of symptoms (including the need for inhaled β2-agonist rescue therapy), degree of airflow obstruction and indices of morbidity (admissions to hospital, need for intubation, emergency room visits, time away from work or school, etc.) can be used to classify asthma severity (Table 1).
Because asthma is controllable, the factors that define its severity before treatment become markers of its control in the treated patient. The amount of anti-inflammatory medication required to control symptoms is often added to the severity algorithm. However, a case has been made that the primary measure of asthma severity in the treated patient should be the minimum anti-inflammatory medication required to achieve ideal control (See Fig. 1, page S4).11
Diagnosis in children
The consensus group believes that, in children able to perform reproducible spirometry, the diagnosis can be established by the same method used for adults. When spirometry is not reproducible, for example in a young child, the diagnosis rests on careful and sometimes repeated history taking and physical examination. Some factors that are particularly useful in establishing a diagnosis in young patients are severe episodes of wheezing, wheezing after 1 year of age, more than 3 episodes of wheezing in a given year, a family history of asthma or atopy, a personal history of asthma or atopy, maternal smoking, clinical benefits from acute bronchodilator therapy, clinical evidence of improvement after anti-inflammatory treatment, chronic cough (especially nocturnal or associated with exercise) and wheezing when viral etiology is unlikely. The likelihood of a diagnosis of asthma increases with the number of these factors present (level V).
Suggestions for future research
• What will be the role of noninvasive markers of airway inflammation (e.g., expired nitric oxide, induced sputum and the quantification of its constituents) on the diagnosis and evaluation of asthma?
• Can the diagnosis of asthma be confirmed through noninvasive means in patients unable to perform reproducible spirometry (e.g., young children)?
Recommendations
• Objective measurements are needed to confirm the diagnosis of asthma and to assess its severity in all symptomatic patients (level III) using:
Spirometry: A 12% (preferably 15%) or greater (at least 180 mL) improvement in FEV1 from the baseline 15 minutes after use of an inhaled short-acting β2-agonist, a 20% (250 mL) improvement after 10-14 days of inhaled glucocorticosteroid or ingested prednisone when symptoms are stable or a 20% (250 mL) or greater "spontaneous variability" is considered significant (level IV).
Peak expiratory flow (PEF): When spirometry and methacholine testing are unavailable, variable airflow obstruction (i.e., ideally 20% or greater diurnal variability) can be documented by home-measured PEF (level II), although this method is not as sensitive or reliable as FEV1.
Airway hyperresponsiveness: Measurement of airway responsiveness to methacholine in specialized pulmonary function laboratories may help to diagnose asthma (level III).
• Appropriate allergy assessment is warranted in patients with asthma (level III) and must be interpreted in light of the patient's history (level III).
• The primary measure of asthma severity in the treated patient should be the minimum therapy required to achieve acceptable control (level III).