How Accurate is the Self-Reported Diagnosis of Chronic Bronchitis?

How Accurate is the Self-Reported Diagnosis of Chronic Bronchitis?

How Accurate Is the Self-Reported Diagnosis of Chronic Bronchitis?* Arnel Bobadilla, MD; Stefano Guerra, MD, MPH; Duane Sherrill, PhD; and Robert Barb...

148KB Sizes 0 Downloads 66 Views

How Accurate Is the Self-Reported Diagnosis of Chronic Bronchitis?* Arnel Bobadilla, MD; Stefano Guerra, MD, MPH; Duane Sherrill, PhD; and Robert Barbee, MD, FCCP

Background: Although it is defined as cough and sputum production for at least 3 months per year for at least 2 consecutive years, difficulties exist in the use of the term chronic bronchitis for clinical diagnosis. In particular, the relationship between diagnosis and symptoms has been difficult to ascertain. Study objectives: To determine, in a large epidemiologic study, the degree to which a new self-reported diagnosis of chronic bronchitis (NCBR) satisfies the symptom criteria for that diagnosis, and to determine the relationship between self-reported physician-confirmed diagnoses and symptom criteria. Methods: We analyzed data obtained from the Tucson Epidemiologic Study of Obstructive Lung Diseases. Using responses to standardized respiratory questionnaires administered to 4,034 subjects, those with NCBRs were selected and assessed as to whether they met symptom criteria for that diagnosis. Descriptive statistics pertaining to gender, age, and smoking status were obtained. Furthermore, we determined how often symptom criteria were met among a subset of subjects with physician-confirmed self-reported diagnoses. Results: Of 481 subjects with NCBRs, only 56 subjects (11.6%) met the required symptom criteria. Men compared with women and current smokers compared with ex-smokers or neversmokers were more likely to meet symptom criteria. Four hundred fifteen of 481 subjects with NCBRs had physician-confirmed self-reported diagnoses. Of these, only 52 subjects (12.5%) met symptom criteria. Within the subgroup of subjects who met symptom criteria, higher percentages were observed in the older age groups, but this was not statistically significant. Conclusion: Only a minority of subjects with NCBRs satisfy the symptom criteria of cough and sputum production for at least 3 months per year for at least 2 consecutive years. This relationship holds true even among those with physician-confirmed self-reported diagnoses. (CHEST 2002; 122:1234 –1239) Key words: asthma; bronchiectasis; chronic bronchitis; COPD; cough; emphysema; rhinitis; sinusitis; sputum Abbreviations: BMRC ⫽ British Medical Research Council Questionnaire on Respiratory Symptoms; NCBR ⫽ new self-reported diagnosis of chronic bronchitis; TESOLD ⫽ Tucson Epidemiologic Study of Obstructive Lung Diseases

adham first introduced the term bronchitis in B the medical literature in 1808. Although bron1,2

chitis was generally recognized as being associated with tobacco smoking, pollution, and dusty occupations, for a long time no clear definition for chronic bronchitis existed. It was not until the development of the British Medical Research Council Questionnaire on Respiratory Symptoms (BMRC) that symptom criteria for disease diagnosis emerged. On the basis of responses to the BMRC and variants of it,2,3 *From the Arizona Respiratory Center, University of Arizona, Tucson, AZ. Manuscript received December 14, 2001; revision accepted March 22, 2002. Correspondence to: Robert A. Barbee, MD, FCCP, Arizona Respiratory Center, University of Arizona, 1501 N Campbell Ave, PO Box 245030, Tucson, AZ 85724-5030; e-mail: [email protected] 1234

the term chronic bronchitis evolved to mean a disease entity primarily involving the bronchi, manifested by cough and sputum, and for which no other pulmonary or cardiac causes could be found.2 The term chronic was defined as “occurring on most days for at least 3 months of the year for at least 2 consecutive years.”3–5 The Ciba Foundation Guest Symposium, an international conference of experts held in 1958, proposed a definition for chronic bronchitis as “the condition of subjects with chronic or recurrent excessive mucous secretion in the bronchial tree.”5,6 Because earlier studies showed that “neither chronic expectoration nor recurrent bronchial infections necessarily caused airflow obstruction,”7–10 demonstrable airway obstruction was not a requirement in these earlier definitions of chronic bronchitis. In 1986, the American Thoracic Society Clinical Investigations

defined COPD as a disorder characterized by abnormal test findings of expiratory flow that do not change markedly over several months of observation.11 At this point, along with emphysema, chronic bronchitis was recognized as one of the two main diseases included under COPD. The BMRC questionnaire inquired, among other things, about symptoms of cough and sputum production, with emphasis on both the time of day and the season of year that these symptoms were noted most. Similarly, the duration of symptoms (number of months and number of years) were noted as well. It is on the basis of affirmative responses to these symptom questions that chronic bronchitis prevalence rates have been reported. Over the years, variations of the BMRC questionnaire were developed. Some of these came to include questions pertaining directly to the diagnosis of chronic bronchitis (ie, whether subjects self-reported the diagnosis of chronic bronchitis and whether or not these diagnoses were physician confirmed). Little is known about how well self-reported diagnoses conform to the symptom criteria of cough and sputum production. The primary purpose of this study was to determine the degree to which new self-reported diagnoses of chronic bronchitis (NCBRs) and physicianconfirmed NCBRs satisfy the symptom criteria of cough and sputum production for at least 3 months per year for at least 2 consecutive years. Materials and Methods Data for our analyses were drawn from results of the Tucson Epidemiologic Study of Obstructive Lung Diseases (TESOLD), a longitudinal population study that enrolled a stratified cluster sample of 1,655 white Tucson, AZ households. These randomly selected non-Hispanic, white subjects were monitored for up to 20 years. Participants were administered standardized respiratory questionnaires during 12 different surveys 1 to 1.5 years apart. Details of population selection, enrollment process, and general methodology have been reported previously.12 From this large cohort, subjects ⬎ 15 years old with an NCBR in any survey other than the enrollment survey were selected for our study. A new diagnosis was based on either a self-report (ie, answering yes to the question, “Since the last questionnaire have you had chronic bronchitis?”) or a self-reported physician confirmation of the disease. Subjects with NCBRs were then classified into two groups based on whether or not they met symptom criteria for chronic bronchitis. To meet symptom criteria, the subjects had to answer “yes” to both of the following questions: (1) “Do you cough on most days as much as 3 months of the year?” and (2) “Do you bring up phlegm, sputum, or mucus from your chest on most days for as much as 3 months of the year?”; and the subjects had to answer “at least 2 years” to both of the following questions: (1) “For how many years have you had this cough?” and (2) “For how many years have you raised phlegm, sputum, or mucus from your chest?” We also looked into whether NCBRs were physicianconfirmed. For a diagnosis to be considered physician confirmed, the subjects must have answered “yes” to one or more of the www.chestjournal.org

following questions: (1) “During the past year, have you seen a doctor for chronic bronchitis?” (2) “Did a doctor ever tell you that you had chronic bronchitis?” or (3) “Since the last questionnaire, have you been treated for chronic bronchitis?” A “yes” answer to question 3 was further validated by stating that they saw a doctor for their chronic bronchitis. Descriptive statistics based on the following were obtained: gender distribution, age at the time of the NCBR, smoking status at the time of the NCBR (current smoker, ex-smoker, or neversmoker), and physician-confirmation status. To further characterize subjects with NCBRs, they were then classified into four symptom groups: (1) subjects who met both symptom criteria, (2) subjects who met only one symptom criterion, (3) subjects who had one or both symptoms but did not meet time duration for symptom(s), and (4) subjects who reported neither cough nor sputum. Among subjects with NCBRs who did not meet symptom criteria for chronic bronchitis, information on other reported symptoms such as shortness of breath with wheezing, chest congestion/“whistly” chest, and rhinitis were also obtained. Similarly, other reported comorbid illnesses such as asthma, emphysema, bronchiectasis, and sinusitis were analyzed. Statistics Descriptive statistics as outlined in “Materials and Methods” were obtained. Cross-tabulations and ␹2 analyses were performed to determine statistical significance.

Results The study population of subjects who met inclusion criteria was 4,034. Of these, 1,899 subjects (47.1%) were male. The age of the subjects ranged from 15.4 to 91.5 years (mean, 50 years). Table 1 shows the demographics and characteristics of the 481 subjects with NCBRs. There was no

Table 1—Demographics and Characteristics of Subjects With NCBRs*

Characteristics Patients with NCBRs Age, yr Gender Male Female Smoking status Current smoker Ex-smoker Neversmoker Physician-confirmation status Yes No

Met Symptom Criteria

Did Not Meet Symptom Criteria p Value

56 (11.6) 53.57 ⫾ 19.1

425 (88.4) 49.55 ⫾ 20.3

33 (17.7)† 23 (7.8)

153 (82.3) 272 (92.2)

27 (17)‡ 19 (12.4) 10 (6)

132 (83) 134 (87.6) 157 (94)

⬍ 0.01‡

52 (12.5) 4 (6.1)

363 (87.5) 62 (93.9)

NS

NS

0.001†

*Data are presented as No. (% within row) or mean ⫾ SD. NS ⫽ not significant. †p value for gender difference. ‡p value for smoking status. CHEST / 122 / 4 / OCTOBER, 2002

1235

significant difference between the mean ages of subjects who met and did not meet symptom criteria. Two hundred ninety-five female subjects (61.3%) had NCBRs compared with 186 male subjects (38.7%). Male subjects with NCBRs were more likely than female subjects to have met symptom criteria for chronic bronchitis (p ⫽ 0.001). Current smokers were more likely to have met symptom criteria than ex-smokers or neversmokers (p ⬍ 0.01). Of 481 cases of NCBRs, 415 cases (86.2%) were physician confirmed. Of these 415 cases, only 52 diagnoses (12.5%) met symptom criteria for the diagnosis. Although the percentage of subjects meeting the symptom criteria was higher among those with physician-confirmed diagnoses than for diagnoses that were not physician confirmed (12.5% vs 6.1%), this was not found to be statistically significant. Table 2 shows the breakdown of subjects with NCBRs by 15-year age subgroups. Within the age subgroups, higher percentages of subjects with NCBRs who met criteria were observed in the older age groups (14.3% in the ⬎ 45-year group and ⱕ 60-year group and 13.5% in the ⬎ 60-year group) compared with the younger age groups (8.5% in the ⬎ 15-year group and ⱕ 30-year group and 9.8% in the ⬎ 30-year group and ⱕ 45-year group). These differences, however, were not statistically significant. Figure 1 illustrates the breakdown of all 481 subjects with NCBRs by symptom groups. Fiftyseven of 481 subjects (11.9%) met only one symptom criterion (either cough or sputum production only), whereas 151 subjects (31.4%) had both cough and sputum production but did not have symptoms long enough to meet criteria. It is most striking that 208 of 481 subjects (43.2%) had neither cough or sputum production but had NCBRs. To better define the subset of subjects with NCBRs who did not meet symptom criteria for chronic bronchitis, we show in Table 3 other symptoms reported by these subjects. These may partly explain their perceptions regarding their NCBRs. Of the subjects, 50.4% reported concurrent symptom of whistly chest/congestion within 1 year of the date of

Table 2—Subjects by Age Group* Met Symptom Did Not Meet Total Within Age Age Groups, yr Criteria Symptom Criteria Group ⬎ ⬎ ⬎ ⬎

15 and ⱕ 30 30 and ⱕ 45 45 and ⱕ 60 60

8 (8.5) 13 (9.8) 12 (14.3) 23 (13.5)

86 (91.5) 119 (90.2) 72 (85.7) 148 (86.5)

*Data are presented as No. (% within row). 1236

94 (100) 132 (100) 84 (100) 171 (100)

Figure 1. Breakdown of 481 subjects by symptom groups.

the NCBR. One hundred ninety-five of 424 subjects (46%) had shortness of breath and wheezing within a year of the NCBR. A still greater percentage (70.1%) of subjects reported rhinitis within the preceding year. Interestingly, among those 208 subjects who did not meet symptom criteria and had neither cough or sputum, 81 subjects (38.9%) reported whistly chest, 129 subjects (65.1%) had rhinitis, and 82 subjects (39.4%) reported shortness of breath and wheezing (data not shown). It is common for patients with chronic bronchitis to report other comorbid illnesses. Table 4 shows every asthma, emphysema, bronchiectasis, and sinusitis diagnosis reported by the subjects with NCBRs. Thirteen of 56 subjects (23.2%), 11 of 56 subjects (19.6%), 13 of 52 subjects (25%), and 52 of 56 subjects (92.9%) who met symptom criteria reported ever having emphysema, asthma, bronchiectasis, and sinusitis, respectively. Subjects who met symptom criteria were more likely to report comorbid emphysema and sinusitis compared with those who did not (p ⫽ 0.001 and p ⫽ 0.046, respectively). This, however, did not hold true for those who also reported asthma or bronchiectasis.

Table 3—Other Symptoms Reported by Subjects Who Did Not Meet Symptom Criteria*

Symptoms Chest whistly/congested No Yes Rhinitis No Yes Short of breath wheezing No Yes

Subjects Who Did Not Meet Symptom Criteria Total Patients, No. 425 211 (49.6) 214 (50.4) 411 (14 missing) 123 (29.9) 288 (70.1) 424 (1 missing) 229 (54) 195 (46)

*Data are presented as No. (% within row) unless otherwise indicated. Clinical Investigations

Table 4 —Reported Comorbid Illnesses Among Subjects* Illnesses Ever emphysema Yes No Total Ever asthma Yes No Total Ever bronchiectasis Yes No Total Ever sinusitis Yes No Total

Met Symptom Criteria

Did Not Meet Symptom Criteria

13 (23.2)† 43 (76.8) 56

38 (8.9) 387 (91.1) 425

11 (19.6)‡ 45 (80.4) 56

118 (27.8) 307 (72.2) 425

13 (25)‡ 39 (75) 52 52 (92.9)§ 4 (7.1) 56

59 (15.2) 328 (84.8) 387 (42 missing) 350 (82.4) 75 (17.6) 425

*Data are presented as No. (%) unless otherwise indicated. †p ⫽ 0.001. ‡p ⫽ NS. §p ⫽ 0.046.

Discussion Four hundred eighty-one subjects (11.9%) of a total population at risk of 4,034 had NCBRs. However, only a small percentage of the subjects with NCBRs met the symptom criteria of cough and sputum production for at least 3 months per year for at least 2 consecutive years (56 of 481 subjects; 11.6%). To our knowledge, this is the first study that has attempted to determine the degree to which NCBRs in epidemiologic surveys satisfy the symptom criteria for that diagnosis. Given the popularity of the term chronic bronchitis and the fact that it is commonly used by laypersons, it should come as no surprise that the overwhelming majority (425 of 481 subjects; 88.4%) with NCBRs did not meet symptom criteria. Such mislabeling or overdiagnosis may be inherent in survey-derived diagnoses especially when dealing with a “popular” disease like chronic bronchitis. These overdiagnosis rates are perhaps greater in epidemiologic surveys that base their chronic bronchitis diagnoses on the presence of symptoms only and not on self-reported diagnoses as we have done. Although there were more women than men in the at-risk population (2,135 women vs 1,899 men) and more women had NCBRs, more men met symptom criteria for the disease. There is substantial evidence that shows a link between symptoms of cough and sputum production and smoking habit. This finding therefore can be explained in part by the fact that proportionally more men were current smokers and www.chestjournal.org

ex-smokers compared with women. This increased prevalence of cough and sputum among male smokers was similarly reported by Fletcher et al3 in their study of London postal workers in the 1950s. Furthermore, Palmer13 showed that the incidence of clinical bronchitis increased proportionally to the amount of tobacco smoked. By the same reasoning, it should come as no surprise that in our study that current smokers were more likely to have met symptom criteria compared with ex-smokers or neversmokers (p ⬍ 0.01). Although as expected, there were more current smokers and ex-smokers compared with neversmokers in the subgroup who met symptom criteria (17% current smokers and 12.4% ex-smokers compared with 6% neversmokers), the consistently higher percentages of subjects who did not meet symptom criteria (compared with those who met criteria) across all smoking status groups imply that factors other than smoking affect the perception of having the disease. Four hundred fifteen of 481 NCBRs were physician confirmed. Surprisingly, only 52 of the 415 subjects met symptom criteria for chronic bronchitis. This accounts for only 12.5% of all physicianconfirmed NCBRs. More striking is the fact that the majority (363 of 415 subjects; 87.5%) with NCBRs and whose diagnoses were physician confirmed did not meet symptom criteria. It is not clear whether this discrepancy can be explained by difference in patterns of diagnosis by Tucson-area physicians, or by difficulties in obtaining data from epidemiologic surveys to ascertain clinical criteria for diagnoses. It certainly is possible that given how widely the term chronic bronchitis is used that the diagnosis is often made too loosely, as is the case, for instance, of patients with acute viral or bacterial respiratory tract infections who present with cough and sputum. Similarly, because the term bronchitis was also used in other questions in the TESOLD questionnaires, it is possible that recall bias may have played a significant role in the overdiagnosis rate seen in our study. For instance, having answered “yes” to any of these other questions, (1) “Have you had an illness such as a chest cold, bronchitis, or pneumonia?” (2) “Have you ever had acute bronchitis?” and (3) “Have you ever been treated for an episode of acute bronchitis?” may have biased the subjects in reporting they had and/or were treated for new chronic bronchitis. Undoubtedly, a general consensus exists that determining accuracy of a clinical diagnosis through responses to respiratory questionnaires cannot, even under the best circumstances, be more than just an approximation. We must point out an important limitation in our methodology pertaining to the determination of physician-confirmation status. Positive physicianCHEST / 122 / 4 / OCTOBER, 2002

1237

confirmation status of an NCBR was based solely on affirmative answer(s) to one or more of the three questions outlined in “Materials and Methods.” In doing so, we relied entirely on patients’ recollections that they were either told and/or treated by a physician for chronic bronchitis. No actual review of patients’ medical records or conversations with their physicians was conducted. This is an inherent limitation because the TESOLD was, by design, an epidemiologic study. Across the different age groups, the percentages of subjects who met symptom criteria for chronic bronchitis were higher in the older age categories (14.3% between the ages of 45 years and 60 years, and 13.5% in patients ⬎ 60 years old, compared to 8.5% in patients between the ages of 15 years and 30 years, and 9.8% in young adults). These differences, although not statistically significant, are consistent with the classic description of chronic bronchitis as being more common among middle-aged subjects and older subjects. As Figure 1 shows, 151 of the 481 subjects (31.4%) with NCBRs had symptoms of cough and sputum production but did not have them long enough to satisfy the criteria for diagnosis. Furthermore, 208 subjects, accounting for 43.2% of all subjects, reported neither cough or sputum production and yet had NCBRs. These findings suggest that symptoms other than cough and sputum production may have affected the perception of having the disease. Interestingly, within the different symptom groups, persistently higher percentages of NCBRs were physician confirmed, compared with those that were not: 82.2% vs 17.8% for the group that reported no cough or sputum, 88.1% vs 11.9% for the group that had symptoms but not time duration, 89.5% vs 10.5% for the group that met only one symptom criterion, and 92.9% vs 7.1% for the group that met both symptom criteria (data not shown). Once again, the limitations regarding determination of physician-confirmation status may have contributed to these apparent physician overdiagnosis rates. As shown in Table 4, of the 425 subjects with NCBRs who did not meet symptom criteria, 46% reported shortness of breath/wheezing, 50.4% said they had whistly chest/congestion, and 70.1% reported rhinitis. It is possible that these subjects were falsely equating the presence of these symptoms with having the disease chronic bronchitis. Consistent with other reports that, especially among older subjects, several obstructive diseases commonly coexist,14 we found that 13 of 56 subjects (23.2%) with NCBRs who met symptom criteria also reported having emphysema. Compared with subjects who did not meet symptom criteria, this was found to be statistically significant. A slightly smaller 1238

number (11 of 56 subjects; 19.6%) reported having asthma. We point out as well that more than one fourth of the 425 subjects with NCBRs who did not meet criteria reported asthma as well (118 subjects; 27.8%). This raises the possibility that symptoms that are normally attributable to asthma (ie, shortness of breath and wheezing) were perceived as being characteristic of chronic bronchitis. Thirteen of 52 subjects (25%) with NCBRs who met symptom criteria and 59 of 387 subjects (15.2%) who did not meet symptom criteria also reported having bronchiectasis. Without radiologic confirmation, we could not ascertain how many merited this diagnosis; but because bronchiectasis is suggested clinically by persistent cough productive of sputum and recurrent infections, it is easy to imagine how this illness could have been confused with chronic bronchitis by subjects and physicians alike. An overwhelming majority of subjects with NCBRs (52 of 56 subjects [92.9%] who met criteria, and 350 of 425 subjects [82.4%] who did not meet criteria) also reported having sinusitis. Conceivably, the symptoms of postnasal drip and coughing that may accompany sinusitis were falsely equated with the disease chronic bronchitis. Burrows and Lebowitz14 reported characteristics of chronic bronchitis among residents of Tucson, AZ. Whereas their study pointed out differences in symptoms among subjects with chronic bronchitis living in the warm and dry region of Tucson, our study focuses on how well subjects with NCBRs meet symptom criteria for the diagnosis. Furthermore, while they studied all subjects, regardless of age, who were enrolled up to the time of the second survey (including subjects with preexisting chronic bronchitis at the first survey), our current study involved subjects ⬎ 15 years of age with NCBRs in any of the 12 surveys other than the enrollment survey (ie, excluding subjects with preexisting chronic bronchitis in the first survey). Just as they found that other symptoms such as wheezing may have affected the clinical diagnosis of chronic bronchitis, our study raises the possibility that other reported symptoms may have affected the perception of having the disease. Lastly, as in the report by Burrows and Lebowitz,14 we find that concurrent asthma and emphysema diagnoses are fairly common among adult subjects with chronic bronchitis. In summary, NCBRs rarely satisfy the symptom criteria of cough and sputum production for at least 3 months per year for at least 2 consecutive years. This holds true even among NCBRs that were physician-confirmed. No significant relationship exists between self-reported diagnoses and the required symptom criteria. Clinical Investigations

References 1 Badham C. Observations on the inflammatory affections of the mucous membrane of the bronchiae. London, England: Callow, 1808 2 Oswald NC, Harold JT, Martin WJ. Clinical pattern of chronic bronchitis. Lancet 1953; September; 265:639 – 643 3 Fletcher CM, Elmes PC, Fairbairn AS, et al. The significance of respiratory symptoms and the diagnosis of chronic bronchitis in a working population. BMJ 1959; August; 2:257–266 4 Definitions and classification of chronic bronchitis, asthma, and pulmonary emphysema: American Thoracic Society. Am Rev Respir Dis 1962; 85:762–768 5 Terminology, definitions, and classification of chronic pulmonary emphysema and related conditions: Ciba Guest Symposium Report. Thorax 1959; 14:286 –299 6 Fletcher CM. Difficulties in the diagnosis of asthma. Ciba Foundation Study Group 1971; 38:5–12 7 Fletcher CM, Peto R, Tinker CM, et al. The natural history of chronic bronchitis and emphysema: an 8-year study of working men in London. London, UK: Oxford University Press, 1976

www.chestjournal.org

8 Kauffmann F, Drouet D, Lellouch J, et al. Twelve years’ spirometric changes among Paris area workers. Int J Epidemiol 1979; 8:201–212 9 Peto R, Speizer FE, Cochrane AL, et al. The relevance in adults of air-flow obstruction, but not of mucus hypersecretion, to mortality from chronic lung disease. Am Rev Respir Dis 1983; 128:491–500 10 Fletcher CM, Pride NB. Definitions of emphysema, chronic bronchitis, asthma, and airflow obstruction: 25 years on from the Ciba Symposium. Thorax 1984; 39:81– 85 11 Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease (COPD) and asthma: This official statement of the American Thoracic Society was adopted by the ATS Board of Directors, November 1986. Am Rev Respir Dis 1987; 136:225–244 12 Lebowitz MD, Knudson RJ, Burrows B. Tucson epidemiologic study of obstructive lung diseases: I. Methodology and prevalence of disease. Am J Epidemiol 1975; 102:137–152 13 Palmer KNV. The role of smoking in bronchitis [abstract]. BMJ 1954; 1:1473 14 Burrows B, Lebowitz M. Characteristics of chronic bronchitis in a warm, dry region. Am Rev Respir Dis 1975; 112:365–370

CHEST / 122 / 4 / OCTOBER, 2002

1239