Identifying asthma and chronic obstructive pulmonary disease in patients with persistent cough presenting to general practitioners: descriptive study

Measurements

Data on respiratory symptoms were collected by means of a modified Medical Research Council questionnaire at baseline and at 8 weeks.8 Additional questions concerned information about respiratory symptoms (wheeze, coughing, and breathlessness) after contact with non-specific stimuli such as smoke, dust, fog, exercise, and specific allergens (house dust, hay, cats, dogs, birds, guinea pigs, moulds). Subjects were asked about family history of asthma and personal and family history of atopy. Dyspnoea was defined as reported breathlessness—an attack of dyspnoea, woken up by dyspnoea, or persistent dyspnoea—during at least the past 2 weeks. Patients were asked about current smoking behaviour and cumulative smoking (pack years of smoking were calculated as the product of years of smoking and the mean number of cigarettes per day divided by 20).

All participants were given a standardised physical examination (inspection, percussion, and auscultation of the thorax) by one practitioner (JANH). Special attention was paid to the absence or presence of rhonchi and prolonged expiration. Spirometry was carried out with a Microlab 3300 (Sensormedics, Rochester, UK). Forced expiratory volume in one second (FEV₁) and forced vital capacity (FVC) were measured until three reproducible recordings (with a difference <5%) were obtained; the highest values were used in the analysis. Predicted values of FEV₁ and FVC were those of the European Respiratory Society.9 Bronchodilator response was measured 15 minutes after the patient inhaled 400 μg salbutamol by a spacer device (Volumatic, Glaxo, Netherlands).

Two weeks later patients filled in a short questionnaire about their clinical state and again underwent physical examination and spirometry. A methacholine provocation test was carried out according to the method of Yan.10 Twofold increments of methacholine chloride were administered from a starting dose of 0.06 μmol to a cumulative dose of 15.6 μmol. The test was discontinued if FEV₁ fell by 20% or more from the post-saline value or after a cumulative dose of 15.6 μmol methacholine had been administered (PD₂₀). Men with an FEV₁ <1.5 litres and women with an FEV₁ <1.2 litres were excluded from the challenge test.11

Eight weeks after the baseline measurements, the patients were examined, completed another questionnaire, and underwent spirometry.

Definitions

Bronchodilator response was positive if FEV₁ improved ≥9% of predicted value after 400 μg salbutamol was inhaled. 12 13 Hyperresponsiveness was defined as PD₂₀ ⩽15.6 μmol methacholine.11 Patients were considered to have asthma if in the previous year they had had one or more episodes of wheeze, cough, or dyspnoea during the day or at night lasting for more than 3 weeks and if PD₂₀ ⩽15.6 μmol methacholine or bronchodilator response was positive. Chronic obstructive pulmonary disease was defined as FEV₁ <70% of predicted at baseline and <75% after 8 weeks; reversibility at baseline and after 8 weeks <9% predicted; and improvement in FEV₁ <12% predicted, either spontaneously or after drugs were given, after 2 and 8 weeks.14

Statistical analysis

Data for this study were analysed with SPSS 4.0 (SPSS, Chicago, USA). Differences between subjects with and without asthma or chronic obstructive pulmonary disease were analysed with Student's t tests and non-parametric tests (Mann-Whitney) as appropriate. Differences in proportions were tested using χ² tests. As it is often difficult in general practice to distinguish between asthma and chronic obstructive pulmonary disease after one or two visits, most analyses combined asthma and chronic obstructive pulmonary disease.

To determine if single symptoms and signs differed significantly between subjects with and without asthma or chronic obstructive pulmonary disease, odds ratios were estimated by logistic regression, adjusted for age, sex, and pack years of smoking.

Combinations of symptoms, signs, and patients' characteristics that corresponded with high or low risks of asthma or chronic obstructive pulmonary disease were sought. For this purpose, all variables associated with asthma or chronic obstructive pulmonary disease were used in a logistic regression model with asthma or chronic obstructive pulmonary disease as the dependent variable. The model was built in a stepwise fashion by including variables for which P <0.05. The values of the β coefficients of these variables in the regression equation were used to generate a scoring formula. For this scoring formula the probabilities for each value were estimated by logistic regression.

We verified how many subjects were identified correctly by using different cut off values of the scoring formula. For this purpose, the observed values according to the scoring formula were divided into six groups. Probabilities for having asthma or chronic obstructive pulmonary disease were calculated for each score by using the six group scores as cut off values. All tests were two sided and P values of <0.05 were considered significant.

Findings

The main finding from this study is that asthma or chronic obstructive pulmonary disease is present in nearly half the patients with coughing for a period of at least 2 weeks attending a general practitioner, and most of these can be identified by history taking and physical examination only. The best scoring formula was generated from a few variables. Thus, symptoms such as current wheeze and dyspnoea, symptoms elicited by allergens, prolonged expiration, cumulative smoking, and female sex will help the general practitioner to predict which patients are at risk of having asthma or chronic obstructive pulmonary disease.

The comparability of our findings with those of two other studies suggest that our findings are applicable to other populations. Two previous studies have shown a similar prevalence of asthma and chronic obstructive pulmonary disease associated with coughing. 3 7 Irwin et al found asthma in 24% and chronic obstructive pulmonary disease in 5% of patients with persistent coughing.3 This is suprising because they investigated patients in a hospital setting. In Irwin's study the patients had coughed for at least 3 weeks; the mean duration of coughing was 3 months. In our study it was impossible to assess the mean duration of coughing, but all patients had been coughing for at least 2 weeks and 28% of the study group had been coughing over the previous 3 months. Hallet and Jacobs investigated 46 patients referred from primary care clinics with a current episode of clinically diagnosed acute bronchitis and a history of at least two similar episodes diagnosed as bronchitis within the preceding 5 years; whether any of these patients had asthma was not known. Thirty patients (65%) could be diagnosed clinically as having mild bronchial asthma.7 In this study, subjects were considered to be hyperresponsive if PD₂₀ ⩽25 mg/ml. The cut off value we used (PD₂₀ ⩽15.6 μmol methacholine) is comparable to a dose of 8 mg/ml histamine.

Possible bias

Since there is no agreed definition for asthma and chronic obstructive pulmonary disease, these disorders must be defined carefully.15 In the definitions used in this study, airway obstruction, airway hyperresponsiveness, or significant bronchodilator response, together with a history of at least one episode of protracted airway symptoms in the past year, were the determinants of a diagnosis. Some of the subjects may have been hyperresponsive only temporarily, but as we measured PD₂₀ at least 4 weeks after the start of their symptoms, a virus induced hyperresponsiveness is unlikely.11 Transient hyperresponsiveness after a viral infection might reflect asthma, perhaps at an early stage. As hyperresponsiveness (with or without symptoms) has been shown to have a high predictive value for later development of asthma and a rapid decline in FEV₁, 16 17 recognition of such patients seems important. The cut off value for PD₂₀ that we chose is common in clinical settings in Europe, but other studies have used a cut off value of PD₂₀ ⩽3.9 μmol methacholine. 10 18 When we used this value the prevalence of asthma or chronic obstructive pulmonary disease decreased to 35%, still a considerable proportion of the study population. The “symptom” previous respiratory episodes for at least 3 weeks over the past year was included in the clinical part of the definition; in all guidelines it is considered to predict asthma or chronic obstructive pulmonary disease.

Separate models for asthma and chronic obstructive pulmonary disease were not necessary, as pulmonary function testing is needed in both disorders to confirm or reject the diagnosis and to initiate a long term follow up. Also, in our population the number of patients with chronic obstructive pulmonary disease was too low to validate the differential diagnosis.

A limitation of this study is the lack of a validation phase in the development of the scoring formula. The size of the study group made analysis impossible. Another study is needed to confirm the generalisability of the model. The advantage of the evaluated criteria (the key variables) is that they are exclusively clinical and therefore easy to assess in routine practice. The results show that just one symptom (for example, reported wheeze or dyspnoea over the past 2 weeks) has a low positive predictive value, whereas a combination of the variables is highly predictive.

Surprisingly, a family history of asthma and a personal or family history of atopy did not significantly contribute to the diagnosis. This may be because patients with these symptoms were diagnosed as having asthma or chronic obstructive pulmonary disease much earlier and were therefore excluded from this study. Although asthma attacks over the past 2 weeks, reported wheeze or (nocturnal) dyspnoea over the past year, and symptoms provoked by non-specific stimuli had significant odds ratios, these symptoms were not included in the model. This could be because they are highly related to the key variables.

The prevalence of asthma was higher in women than in men. This has also been found in epidemiological studies.11 As the number of asthmatic patients (mostly women) was much higher than the number of people with chronic obstructive pulmonary disease (mostly men), it is understandable that female sex was an important risk factor in this population. The modest but significant value of pack years of smoking was due to the patients with chronic obstructive pulmonary disease. In this population cumulative smoking contributed more significantly than current smoking.