The video questionnaire: one approach to the identification of the asthmatic phenotype
Professor Beasley Department of Medicine, Wellington School of Medicine, PO Box 7343, Wellington South, New Zealand.
Provide feedback or get help You are viewing our new enhanced HTML article.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Since clinicians, physiologists and pathologists cannot agree on how to define asthma, it is not surprising that genetic epidemiologists also have difficulty. Furthermore, the definition of asthma has become more complex as our understanding of its pathophysiology has increased. However, despite this increased complexity, the characteristic features of symptomatic reversible airflow obstruction and bronchial hyperresponsiveness (BHR) by which one recognizes or diagnoses the disease, remains the basis of the current WHO definition of asthma [ 1]:
‘Asthma is a chronic inflammatory disorder of the airways in which many cells play a role, in particular mast cells, eosinophils, and T lymphocytes. In susceptible individuals this inflammation causes recurrent episodes of wheezing, breathlessness, chest tightness, and cough particularly at night and/or in the early morning. These symptoms are usually associated with widespread but variable airflow limitation that is at least partly reversible either spontaneously or with treatment. The inflammation also causes an associated increase in airway responsiveness to a variety of stimuli.’
A fundamental problem is that, even when a single definition such as that proposed by the WHO is accepted, the diagnosis of asthma involves an overall assessment of the patient's medical history, physical examination, and lung function results, and there are no universally accepted rules for combining the information from these various sources. A related problem is that the term ‘asthma’ unifies what may be a disparate group of disorders which produce similar clinical features. Furthermore, the phenotypic expression of asthma may vary by age [ 2].
By necessity most genetic epidemiological studies have defined the asthmatic phenotype on clinical information which can either be collected or measured at a single point in time. The main options in this regard are the administration of questionnaires based on self-reporting of symptoms or a previous diagnosis of asthma, tests of lung function including the response to bronchodilator therapy, bronchial responsiveness testing, physician assessment, or markers of related conditions such as atopy. In this manuscript issues relevant to each of these approaches are briefly discussed, with the exception of atopy which is the topic of other reviews in this series. Findings from preliminary studies of the use of a video questionnaire, a novel approach to the identification of the asthma phenotype, are also presented.
A number of symptoms including wheezing, chest tightness, breathlessness, and coughing with or without sputum are recognized by physicians as indicative of asthma. Of these the most important symptom for the identification of the asthmatic subject is wheezing [ 3, 4], and most written questionnaires therefore have focused on this symptom. Whereas reversible airflow obstruction usually (but not always) gives rise to wheeze, the converse is not true [ 5]. For example, a study in New Mexico [ 6] found that wheezing (apart from colds) occurred in 74% of current doctor-diagnosed asthmatics, but also occurred in 18% of smokers and 10% of non-smokers without diagnosed asthma. Thus, wheezing is not synonymous with diagnosed asthma; in some instances, wheezing may indicate asthma which has not been diagnosed, but it may also indicate other diseases (particularly chronic bronchitis and emphysema in persons aged 45 years or more) or may be unrelated to other symptoms or disease; conversely, diagnosed asthmatics may not experience or recognize wheeze as a symptom. Another difficulty with the use of self-reported wheezing is that of recall. For example, even when identical questions on the occurrence of wheezing in the last year are asked 2 weeks apart, concordance rates of less than 90% may be achieved [ 7]. Within a previously identified symptomatic group, about a third of subjects may not report asthmatic symptoms when administered the same questionnaire several years later [ 8]. This implies that a significant number of additional asthmatic subjects will be identified through the repeat administration of a symptom-based questionnaire, and illustrates the difficulties associated with identifying subjects with a disease which may relapse or remit over time.
Nevertheless, wheezing remains the symptom which is most characteristic of asthma. Self-reported wheezing during the last 12 months has reasonable sensitivity and specificity for BHR, although this may vary depending on the language in which the questionnaire is administered [ 7, 9, 10]. In addition to self-reported wheezing, most questionnaires include additional questions on the frequency of wheezing and the circumstances in which wheezing occurs (wheezing after exercise, waking with wheezing). Such questions have been included in the standardized ECRHS and ISAAC questionnaires for use in adults [ 9] and children [ 11], respectively, although it is uncertain how they should be used to identify the asthmatic individual.
The use of ‘diagnosed asthma’ in written questionnaires has good specificity for identifying children who have asthma, in that children with diagnosed asthma are likely to have the disease, at least in western countries [ 9]. However, a major problem with its use is that it has poor sensitivity, in that a substantial number of both children [ 12] and adults [ 13] with asthma have never had the diagnosis made. In addition, no consistent standardization exists in the use of ‘diagnosed asthma’ as it is dependent on numerous factors including the individual's self-recognition of key symptoms, the diagnostic skills or practice of the doctor, and the individual's recollection of the doctor's opinion [ 14]. These factors are likely to account for the findings from international asthma prevalence comparisons that differences in diagnosed asthma may be much greater than differences in reported asthma symptoms and BHR [ 15].
Measurements of lung function before and after inhaled bronchodilator can be employed to identify reversible airflow obstruction. However, if taken on a single occasion, such measures may have limited application in studies of the asthmatic phenotype, since by definition the airflow obstruction in asthma is reversible and may not be present on the day of assessment. Indeed, one of the characteristic features of reversible airflow obstruction in asthma is its diurnal variation, in that the degree of airflow obstruction is greatest at night or on wakening, and least during the day when lung function testing is likely to be undertaken in genetic studies. As a result, a significant proportion of asthmatics are likely to have lung function within normal limits on any one day, and not exhibit the recognized 15% improvement in FEV1 or PEF values that are required for a physiological diagnosis of asthma to be made. A better alternative is to undertake repeat measurements of lung function before and after bronchodilator, at different times of day over a period of time (e.g. 2 weeks). It is this variation and reversibility of changes in lung function over time which typifies asthma, and this cannot be captured in a one-off survey on a particular day.
Another approach is to measure the degree of bronchial responsiveness, to provide a marker of one of the physiological characteristics of asthma. However, although bronchial hyperresponsiveness (BHR) has been considered to represent the ‘gold standard’ marker for asthma, it is neither wholly sensitive nor specific for asthma ( Table 1) [ 16]. Although the level of BHR correlates with asthma severity and the need for asthma therapy [ 17], asthma may also occur without BHR [ 18], with one study reporting this phenomenon in only a third of children with recurrent wheezing [ 3]. BHR may also occur in subjects without a history of asthma symptoms [ 3], particularly in those with allergic rhinitis [ 19], or in relatives of asthmatic subjects [ 20]. Bronchial responsiveness may vary considerably over time in subjects with current asthma and the level may correlate poorly with asthma symptoms [ 21]. Furthermore, BHR may be significantly modified by drug therapy [ 22], and smoking habits [ 23].
Table 1. .
Relationship between symptoms before the age of 9 and response to bronchial challenge in 791 9-year-old children. From Sears et al
A further problem is that asthma is assumed to be a qualitative concept (someone either has or does not have asthma during a particular time period) whereas bronchial responsiveness is a continuous measure and it may be necessary to select an arbitrary cut-off point. Nevertheless, BHR testing does at least provide a marker of one physiological characteristic associated with asthma and remains one method to assist in the identification of the asthmatic phenotype.
Combination of symptoms and BHR
A combination of recent symptoms and abnormal BHR has also been suggested as a method for identifying the asthmatic phenotype [ 24]. This has obvious advantages since neither wheezing or BHR alone appear to allow the diagnosis of asthma to be made with certainty. In support of this approach, it has been shown that children with both current wheeze (in the previous 12 months) and BHR showed a greater degree of BHR, more symptoms, diagnosed asthma, and asthma medication use than either marker alone ( Table 2). In this study, the interpretation of the findings for the ‘wheeze only’ and ‘BHR only’ groups is of interest. For asthma symptoms, diagnosed asthma, and asthma medications the ‘wheeze only’ group showed greater severity than the ‘BHR only’ group suggesting that wheeze may be a more sensitive marker of asthma than BHR.
Table 2. .
BHR, lung function, self-reported symptom frequency, diagnosed asthma and medication use in 352 children aged 7–12 years old. From Toelle et al
Another option is to use the same approach as that employed in clinical practice, in which a diagnosis is made following assessment by a physician. This approach would be likely to involve a standardized history and examination, with the interpretation of the results of other methods of assessment, such as serial lung function testing or response to bronchodilator therapy. The major problem with this approach is that internationally accepted criteria for a clinical diagnosis of asthma by a physician have not yet been established. It should however, be possible for such criteria to be established and agreed internationally. Although this approach would be relatively time-consuming and costly when employed in genetic epidemiology studies, it is likely to be more informative than other methods in terms of identifying the asthmatic individual.
In response to the limitations associated with the established methods for determining the asthmatic phenotype, a video questionnaire involving the audiovisual presentation of clinical signs and symptoms of asthma has been developed [ 7, 10, 25]. The philosophy on which the video questionnaire is based is that by showing rather than describing clinical asthma, a more valid and repeatable recognition of asthma could be achieved. The questionnaire involves five sequences of clinical asthma symptoms in young persons: wheezing at rest; wheezing after exercise; waking with wheezing; waking with cough; a severe asthma attack. After each sequence, participants are asked to specify whether their breathing had ever been like that of the person in the video; if so they are asked further nested questions as to whether this had occurred in the last year, and whether this had occurred one or more times a month. The term asthma or ‘wheezing’ is not mentioned at any stage in the video questionnaire. In this way, individuals are identified who have experienced episodes of what doctors would recognize as clinical asthma.
In a validation exercise in English-speaking children the video questionnaire (AVQ2.0) was found to be as sensitive and specific for predicting BHR, and substantially more repeatable than the IUATLD written questionnaire [ 7]. More recently an international version of the video has been produced, featuring asthmatics from various continents and ethnic groups (AVQ3.0) [ 9].
This international version of the video questionnaire has been validated in a recent study from Hong Kong. A total of 189 children completed the video questionnaire as well as the ISAAC written quesitonnaire, and underwent bronchial responsiveness testing with methacholine [ 10]. Only 14 children (44%) with diagnosed asthma were hyperresponsive but they comprised the majority (88%) of children with BHR. Correlations between responses to corresponding questions in the written and video questionnaires were fair to poor with Kappa indices of 0.06–0.44. The question on ‘asthma ever’ (i.e. diagnosed asthma) in the written questionnaire had the highest validity for BHR when compared with individual questions on asthma symptoms in the two questionnaires ( Table 3). Similar figures for sensitivity, specificity and Youden's index for BHR were seen between the corresponding questions in the video and written questionnaires except for the question on ‘severe wheeze’ where the Youden's index for the video questionnaire was significantly higher than that for the written questionnaire. The ability to predict BHR was not enhanced by increasing the number of positive responses to the five questions, because this led to a reduction in sensitivity of a similar magnitude to the increase in specificity.
Table 3. .
Sensitivity, specificity and Youden's index for BHR of each of the five questions relating to asthma symptoms within the previous 12 months with the international video (AVQ3.0) and ISAAC written questionnaires (WQ) * Comparison between the Youden's indices of AVQ3.0 and WQ by paired t
In summary, there are a number of different methods available for use in genetic epidemiology studies for determining individuals with asthma. One novel method which has recently been developed is the video questionnaire, which attempts to identify individuals who have experienced episodes of clinical asthma. Initial validation studies of the video questionnaire have been encouraging suggesting that it may have some advantages over traditional methods of assessment, particularly with regard to the comparability of information across populations. Nevertheless, further validation studies are required, ideally involving a full clinical assessment by a physician as the ‘gold’ standard, rather than bronchial responsiveness testing or other methods of assessment. In the meantime, the most practical and valid method of identifying asthmatic individuals for genetic studies may well involve using both video and written questionnaires, together with bronchial responsiveness testing, repeated measurements of lung function over time, including the response to bronchodilator treatment, as part of a standardized assessment by a physician. However, the use of this approach will clearly require further evaluation before it can be applied with confidence in genetic epidemiology studies.
The Wellington Asthma Research Group is supported by a Programme Grant from the Health Research Council of New Zealand and by a major grant from the Guardian Trust (Trustee of the David and Cassie Anderson Memorial Trust). CKW Lai is supported by Research Grant Council Earmarked Grant 96/97 No. CUHK 232/96 M.