Poor Sensitivity of Symptoms in Early Detection of COPD

Abstract

The prevalence of chronic obstructive pulmonary disease (COPD) has been increasing. However, COPD is often underdiagnosed. The objective of this study was to determine how many outpatients had persistent airflow limitation and could be diagnosed as COPD by post-bronchodilator spirometry. We also evaluated whether the newly diagnosed patients had any symptoms. All outpatients with liver or general diseases over 40 years old who regularly visited to our hospital were tested for pulmonary function by spirometry. Patients with airflow limitation by the first screening spirometry had further examinations including post-bronchodilator spirometry and chest radiograph by pulmonary specialists. A total of 288 patients accepted a first spirometry. The most common chronic diseases of these patients were chronic hepatitis (33.7%), fatty liver (26.4%), liver cirrhosis (8.3%), diabetes (3.5%) and hypertension (3.1%). Approximately half of the patients had a smoking history. 44 of 288 patients (15.3%) showed airflow limitation by pre-bronchodilator spirometry. Of these, 8 patients did not show airflow limitation by a repeat pre-bronchodilator spirometry nor did 5 patients by post-bronchodilator spirometry. The rest were diagnosed as COPD (80.6%), asthma (16.1%) and bronchiectasis (3.2%). The prevalence of COPD was 8.7%. Approximately half of the patients (13/25, 52.0%) diagnosed as COPD had never complained of any respiratory symptoms. Because symptoms such as dyspnea on exertion, cough and sputum are less sensitive for the diagnosis of COPD, the propagation of spirometry in a general practice/setting should be recommended for establishing the diagnosis rate of COPD rather than relying on the presence of respiratory symptoms.

Key words: :

• Chronic obstructive pulmonary disease
• Spirometry
• Airflow limitation
• General practice

INTRODUCTION

Chronic obstructive pulmonary disease (COPD) is an important disease, and the prevalence and mortality are increasing worldwide ([1]). A 1996 survey conducted by the Japanese Ministry of Health and Welfare reported approximately 220,000 diagnosed cases of COPD, which equates to a prevalence of 0.2% ([2]). In contrast, according to the Nippon COPD Epidemiology (NICE) study that was performed by using spirometry to measure the prevalence of COPD in Japanese adults, there were 5.3 million patients with COPD in Japan and the overall prevalence was 8.6% in persons over 40 years old ([3]). These findings suggest that patients with COPD were underdiagnosed and subsequently undertreated in Japan.

However, the NICE epidemiological survey used pre-bronchodilator spirometry and thus not based on the current definition of COPD (the Global Initiative for Chronic Obstructive Lung Disease; GOLD) ([1]). The GOLD guidelines recommend spirometry for the diagnosis of COPD in persons who have symptoms such as dyspnea, cough and sputum. In the GOLD guidelines, the diagnosis of COPD is confirmed when, after bronchodilator, the ratio of forced expiratory volume in 1 second (FEV₁) to forced vital capacity (FVC) is < 0.7 ([1]). Therefore, the prevalence of COPD shown in the NICE study might be overestimated since spirometry was not performed post-bronchodilator.

The American Thoracic Society (ATS)/European Respiratory Society (ERS) recommends that spirometry should be performed in all persons with a history of exposure to cigarette smoke and/or environmental pollutions, a family history of COPD, sputum production, or shortness of breath ([4]). On the other hand, several studies have suggested that early COPD can be detected by questionnaires ([5], [6], [7], [8]). These questionnaires are mainly based on patients' symptoms and/or a history of exposure to cigarette smoke or other risk factors. Another study has indicated the sensitivity of patients' symptoms for the diagnosis of COPD is lower than that by spirometric diagnosis ([9]). Therefore, not all patients with COPD without any symptoms or history of exposure can be detected by questionnaires alone.

The aim of this study was to determine how many outpatients could be diagnosed with COPD by post-bronchodilator spirometry. Further, we evaluated whether symptoms were useful in suggesting a diagnosis of COPD.

METHODS

Subjects and study design

A convenience sample of outpatients over 40 years old who regularly visited the non-respiratory section of the Third Department of Internal Medicine in Wakayama Medical University Hospital from April 2005 to October 2006 were recruited. This section treats the patients with liver diseases and other general diseases such as diabetes and hypertension. This study was approved by the ethics committee of our institution. Informed written consent was obtained from all subjects. Patients already diagnosed as COPD, asthma, or any other chronic respiratory diseases were excluded. After giving informed consent, spirometry was performed. During the study period, a total of 288 subjects were recruited and underwent spirometry.

Patients who had airflow limitation by the first spirometry were further examined by pulmonary specialists on separate day (2nd day). If patients continued to show airflow limitation on the 2nd day, post-bronchodilator (400 μ g, salbutamol)-spirometry was also performed. If patients still showed airflow limitation (FEV₁/FVC < 0.7) after bronchodilator, they were carefully queried concerning their respiratory symptoms. Furthermore, if needed, they had a chest radiograph and chest computed tomography for differential diagnosis. The diagnosis of COPD, asthma, or other disease was made independently by 2 different pulmonary specialists.

Spirometry

Pulmonary function testing was performed by using a dry rolling-seal spirometer (System 7; Minato Medical Science, Osaka, Japan). FEV₁ and FVC were measured on at least three acceptable FVC maneuvers. The greatest values of FEV₁ and FVC were adopted for the subsequent analysis based on the American Thoracic Society's recommendation ([10]). The value of FEV₁ % predicted was calculated and used to grade the severity of airflow limitation based on the GOLD definition ([1]). We used the predicted values of FEV₁ of the Japanese Respiratory Society ([11]).

RESULTS

288 patients consented to participate in the study and had a spirometry. The characteristics of the analyzed subjects are shown in Table 1 and Table 2. The most common chronic diseases of these patients were chronic hepatitis (33.7%), fatty liver (26.4%), liver cirrhosis (8.3%), diabetes (3.5%) and hypertension (3.1%), with mean age of 62.5 + 9.9 (mean + SD). 44.4% were female and 55.6% were male. The percentages of current smokers, ex smokers and non-smokers were 23.6%, 24.0% and 52.4%, respectively.

Table 1 Baseline disease of subjects

	Number (%)
Total	288
Chronic hepatitis	97 (33.7)
Fatty liver/Hyperlipidemia	79 (26.4)
Liver cirrhosis	24 (8.3)
Hepatocellular carcinoma	24 (8.3)
Autoimmune hepatitis	13 (4.5)
Diabetes	10 (3.5)
Hypertension	9 (3.1)
Alcoholic hepatitis	8 (2.8)
Others	27 (9.4)

Table 2 Characteristics of subjects

	All subjects	FEV1/FVC < 0.70
Total	288	44 (15.3)
Age (years)
40–49	31	0 (0.0)
50–59	74	6 (8.1)
60–69	105	16 (15.2)
70–	78	22 (28.2)
Gender
Female	128	18 (14.1)
Male	160	26 (16.3)
Smoking status
Never	151	15 (9.9)
Ex	69	14 (20.3)
Current	68	15 (22.1)
Smoking amount (pack-years)
0–24	43	6 (14.0)
25–49	56	10 (17.9)
50–	38	13 (34.2)

*% of all subjects.

44 of 288 patients (15.3%) showed airflow limitation by screening spirometry. Patients over 70 years old had the highest rate of airflow limitation (28.2%) of all generations. The percentages of female and male patients who had airflow limitation were similar. The percentage of ever smokers who had airflow limitation was higher than that of never smokers. Patients who had more than 50 pack-years smoking histories showed the highest rate (34.2%) of airflow limitation.

Of the 44 patients with airflow limitation on a first spirometry, 8 patients did not show airflow limitation on the separate day. In addition, 5 patients did not show airflow limitation after bronchodilator. The remaining 31 patients with airflow limitation were diagnosed as COPD (80.6%), asthma (16.1%) and bronchiectasis (3.2%) (Figure 1). Diagnoses were not different between the pulmonary specialists in the present study.

Figure 1 Study flow diagram. The number in each panel indicate the number of patients.

The prevalence of COPD in our study was 8.7% (25/288). In the 25 COPD subjects, 21 patients were GOLD stage I and the remaining were stage II. 4 of 25 COPD patients (16.0%) never had a smoking history (Figure 2A). Approximately half of the COPD patients (13/25, 52.0%) did not complain of any symptoms (Figure 2B). All 13 COPD patients without symptoms were stage I COPD.

Figure 2 Smoking status and symptoms of subjects with COPD. Left panel indicates smoking status of 25 patients who had diagnosed COPD (A). Right panel indicates symptoms of patients who had diagnosed COPD (B). The number in each panel indicate the patient number.

DISCUSSION

In the present study, we investigated the prevalence of COPD based on the GOLD guidelines ([1]). At the first spirometry, 15.3% of the patients showed airflow limitation. However, of these, 29.5% of the patients had no airflow limitation on a second pre-bronchodilator spirometry or post-bronchodilator spirometry. Subsequently, 25 of 288 (8.7%) patients were diagnosed as COPD. Half of these patients never complained of any symptoms. In addition, 16% of these patients had no smoking history.

The major aim of this study was to determine how many outpatients in our practice setting have persistent airflow limitation by spirometry and could be diagnosed as COPD by post-bronchodilator spirometry based on the GOLD guidelines. 44 of 288 patients (15.3%) showed airflow limitation in the screening pre-bronchodilator spirometry.

Of these, 8 of 44 (18.2%) patients showed no airflow limitation in the spirometry performed on the 2nd day. In addition, 5 of 44 patients did not show airflow limitation in the post-bronchodilator spirometry. The reason for the transient airflow limitation is uncertain. It has been demonstrated that there is a circadian rhythm in the human pulmonary function, which may cause a day-to-day variability in the airflow limitation ([12]).

The day-to-day variation of the airway caliber or pre-disease stage may be the reason for the transient airflow limitation. Subsequently, 25 of 44 patients (8.7%) in their cohort were diagnosed as COPD as defined by the GOLD guidelines. In smokers and ex-smokers, the prevalence of COPD were 17.6% and 13.0%, respectively.

Recently, the prevalence of COPD has been investigated worldwide ([3], [13], [14], [15], [16], [17], [18], [19], [20]). A recent systematic review of the global prevalence of physiologically defined COPD in adults over 40 years old was estimated about 9–10% ([21]). In Japan, the NICE study was performed by using spirometry to clarify the prevalence of COPD in Japanese adults. In the NICE study, the prevalence of COPD was estimated as 5.3 million, which was 8.6% in people over 40 years old. However, because the results of the NICE study were based on only pre-bronchodilator testing, there is a possibility that the prevalence of COPD might be overestimated. Indeed, one study recently reported that the prevalence of GOLD-defined COPD was 7.0%, which was 27% lower than that defined without bronchodilators ([22]). In another recent publication, it was demonstrated that the overall prevalence of airflow limitation was reduced from 21.7% to 14% by bronchodilator testing ([23]). In the present study, we investigated the prevalence of COPD with spirometry performed three times including a post-bronchodilator procedure. Thus, it can be considered that the results of the present study more precisely show the prevalence of COPD.

It has been reported that the prevalence of COPD in a Japanese general practice setting was 27%, and that 61% of these subjects were moderate to severe COPD according to the GOLD guidelines ([24]). This prevalence is much higher and the severity of the COPD was relatively greater than those of our present study. However, in Takahashi's study, the subjects examined were limited to those over 40 years old who had a smoking history and/or had respiratory symptoms including chronic cough and sputum. In contrast, in our present study, all outpatients over 40 years old who regularly visited our general practice were investigated without reference to symptoms or smoking history. Thus, the prevalence of COPD in the present study was less than that in Takahashi's study.

Several studies have been performed for the early detection of COPD by using questionnaires ([5], [6], [7], [8]). Because these questionnaires were mainly based on symptoms and smoking history, subjects with COPD who had no symptoms and/or smoking history could not be detected. Buffels et al. have shown in the DIDASCO study that 42% of newly diagnosed cases of obstructive lung disease would not have been detected without spirometry ([25]). In the present study, 52% of COPD patients did not have symptoms. Furthermore, 16% of COPD patients had no smoking history. Although it is difficult to compare the results of the present study with those of the DIDASCO study, which did not perform post-bronchodilator spirometry, the prevalence of COPD in our present study (8.7%) was higher than in the DIDASCO study (7.4%). Use of questionnaires only may underestimate the prevalence of COPD.

It has been recently reported that spirometry is underused in spite of the recommendation of GOLD 2006 ([26]). Lee et al. have suggested that spirometry is inconsistently used in the diagnosis of COPD or the management of COPD patients, and that much of the current COPD diagnosis and management is based solely on symptoms, rather than a combination of symptoms and objective lung function assessment ([27]). In Japan, it has been shown that only one half of physicians recommend the use of spirometry for the diagnosis of COPD ([28]). However, as shown in our present study, half of the patients diagnosed by spirometry never complained of any symptoms. In addition, 16% of these patients had no smoking history. It should be noted that all of the patients with COPD were 50 years of age or older, possibly due to the small number of younger patients.

In the present study, many of the patients had chronic liver diseases due to the prevalence of these conditions in outpatients visiting our hospital for non-respiratory diseases. It is known that a pulmonary vascular disorder, called hepatopulmonary syndrome, is seen in severe liver disease ([29]). Therefore, there is a possibility that the presence of many patients with liver disease may have affected the present results. However, the clinical feature of hepatopulmonary syndrome is hypoxemia accompanied by an impairment of gas exchange, not obstructive lung disturbance. Therefore, we did not expect an influence on our present results due to the existence of many patients with liver disease. It has been shown in GOLD 2006 that various comorbidities coexist in COPD, which affect the pathophysiology and the management of COPD ([1]). However, it is still unknown whether the existence of various chronic diseases may affect the prevalence and symptoms of COPD. Further investigation will be needed to clarify the prevalence and symptoms of COPD in general populations and the influence of the existence of different diseases on the prevalence of COPD.

In conclusion, our present results suggest an active enforcement of spirometry in a general setting should be recommended for improving the diagnosis rate of COPD, because of the lack of sensitivity of symptoms such as dyspnea on exertion, cough and sputum is less.

The authors are grateful to Mr. Brent Bell for reading the manuscript. We also thank Drs. T. Ichikawa, S. Yanagisawa, T. Hirano, H. Sugiura, K. Matsunaga, and Y. Minakata for their useful discussion for preparing the manuscript.