Classifying the severity of asthma may be a conundrum for physicians and patients alike. Most classification systems incorporate many variables, including poor symptom control, medication type and dosage, and objective measurements of lung function. In 1991, the National Asthma Education Program (NAEP) published criteria for mild, moderate and severe asthma that included the frequency of symptoms, objectively measured pulmonary function and the frequency of corticosteroid use. However, given the highly variable nature of asthma, this system was considered imperfect. Gauging the severity of asthma according to self-reports of the frequency and severity of symptoms is appealing, but these reports typically reflect current levels of control or compliance with treatment. Osborne and associates compared the NAEP criteria with physician-assessed severity measures to determine which more accurately identified asthma severity.

All patients in a regional HMO population who had either been hospitalized in the previous two years or had taken medication in the previous year for asthma were eligible for the study. Patients were evaluated clinically by spirometry and completed a questionnaire about their respiratory symptoms and medications. Patients were asked whether they had experienced wheezing, coughing or shortness of breath at any time within the previous four weeks and whether they had experienced these symptoms at night. Three scaled indexes, based on the NAEP criteria, were developed to classify asthma severity. In addition, the medical records of all patients were reviewed by two pulmonologists who collected data on the frequency of asthma exacerbations and assessed the asthma severity.

Of the 914 patients eligible for the study, 193 patients met the criteria for enrollment. Overall, 107 patients were classified as having moderate or severe asthma when the physician-assessed severity rating scale was used. When the NAEP criteria were applied to this group, only 53 percent of these patients were classified as having either moderate or severe asthma. Of particular concern as a result of this comparison were the patients who were classified as having mild asthma but actually had severe asthma and vice versa. The frequency of use of corticosteroids and spirometry results correlated with the physician assessment of asthma severity. However, there was no correlation between level of current symptoms and the physician assessment, even when nighttime symptoms were included in the assessment.

The authors conclude that while symptoms are a key patient concern and an important focus of asthma management, they do not correlate with asthma severity. Rather, symptoms better measure control and compliance with medication. As protocols and practice guidelines for asthma management evolve, they must not rely heavily on asthma symptoms. Physician judgment continues to be a more reliable measure of asthma severity.