ABSTRACT
Chronic obstructive pulmonary disease (COPD) patients experiencing several episodes of acute clinical derangement suffer from increased morbidity, mortality, and accelerated decline in lung function. Nevertheless, the relationship between co-morbidity profile and exacerbation rates in the frequent exacerbator phenotype is poorly characterized, and evidence-based management guidelines are lacking. We sought to evaluate the co-morbidity profile and treatment patterns of “frequent exacerbators” with severe or very severe airflow limitation. We conducted a cross-sectional, multicenter study in 50 Italian hospitals. Pulmonologists abstracted clinical information from medical charts of 743 COPD frequent exacerbators. We evaluated the exacerbation risk and center-related variations in diagnostic testing. One-third of patients (n = 210) underwent a bronchodilator response test, and 163 (22%) received a computerized tomography (CT) scan; 35 had a partial response to bronchodilators, while 119 had a diagnosis of emphysema; 584 (79%) lacked sufficient diagnostic testing for classification. Only 17% of patients did not have any coexistent disease. Cardiovascular conditions were the most frequent co-morbidities. A history of heart failure [odds ratio (OR): 1.89; 95% confidence interval (CI) 1.48–2.3] and affective disorders (OR: 1.66; 95% CI 1.24–2.1) was associated with the frequency of exacerbations. Center membership was strongly associated with exacerbation risk, independent of casemix (variance partition coefficient = 29.6%). Examining the regional variation in health outcomes and health care behavior may help identify the best practices, especially when evidence-based recommendations are lacking and uncertainties surround clinical decision-making.
Introduction
Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality worldwide, especially in smokers over 40 years of age Citation(1). In contrast with other leading causes of death, like neoplasia or cardiovascular diseases, mortality secondary to COPD is continuously increasing Citation(2): it is estimated that COPD will be the fourth cause of death in 2030, due to the increase in tobacco smoking and the decrease in mortality from other causes Citation(3).
The clinical course of COPD alternates periods of clinical stability and acute exacerbations characterized by worsening of symptoms, increased risk of hospitalization, and high socioeconomic costs. Although most COPD patients may experience at least one exacerbation in their lifetime, a subgroup of this population appears to be particularly susceptible to several repeated episodes of acute clinical derangement, leading to the definition of the frequent exacerbator phenotype. In the ECLIPSE study, the frequent exacerbator phenotype represented 12% of the sample. These patients suffer from increased COPD-related morbidity, mortality, and accelerated decline in lung function Citation(4).
There is sparse evidence that selected co-morbidities may increase the likelihood of exacerbations in COPD patients and that the overall burden of coexistent diseases may affect patients' recovery from episodes of acute derangement. In almost all patients, COPD is associated with other chronic disorders sharing the same risk factors, such as smoking habit and aging (Citation5–7). These co-morbidities, either pulmonary or extrapulmonary, negatively affect patients' survival and clinical management and lead to enormous economic consequences, in terms of both increased medical expenses and reduced productivity Citation(2,5,8). Moreover, some clinical interventions recommended for COPD may have negative effects on concomitant pathologies Citation(9). However, the co-morbidity profile and its impact on the exacerbation rate among frequent exacerbators are poorly characterized.
In the present study we sought to characterize the epidemiology of phenotypic subgroups and to assess variations in clinical practice patterns among Italian frequent COPD exacerbators.
Patients and methods
Study design and patients
We conducted a cross-sectional, observational, multicenter study performed in 50 Italian centers for the diagnosis and management of lung diseases. All patients seen at the clinics from March 2012 to December 2013 were screened for eligibility. We enrolled patients defined as frequent exacerbator, according to the following inclusion criteria: age of 45–85 years, known COPD documented by past lung function tests showing fixed nonreversible respiratory obstruction, FEV1 < 50% in a previous test anytime, symptoms of chronic bronchitis (chronic productive cough for 3 months at least in each of the 2 years prior to the index visits), on treatment with any bronchodilators in the past 12 months, and a history of at least two exacerbations (moderate/severe) in the last 12 months prior to index visit. A pulmonologist at each clinic retrospectively abstracted clinical information of COPD patients from clinical charts and recorded clinical information obtained during the visit in an electronic data collection form.
Clinical and sociodemographic characteristics
We recorded the following characteristics: number of exacerbations in the past 12 months, age, sex, diagnostic workup including any imaging technique, parameters derived from basal spirometry and bronchodilator tests, body mass index (BMI), co-morbidities (i.e., hypertension, heart failure, coronary artery disease, arrhythmia, other cardiovascular disorders, diabetes, metabolic syndrome, psychiatric conditions, chronic kidney disease, and osteoporosis), and smoking status (current or past smoker).
Treatment regimens
COPD therapy was identified based on the Anatomical Therapeutic Chemical (ATC) codes as follows:
| 1. | Long-acting bronchodilators: any of R03AC12, R03AC13, R03AC18, R03AC19, R03BB04, R03BB05, R03BB07. | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| 2. | Short-acting bronchodilators: any of R03AC02–06, R03AC11, R03AC14–R03AC16, R03BB01, R03BB02. | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| 3. | Corticosteroids: any of R03BA01–R03BA09, H02A. | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| 4. | Other: any of C01CE, R03BC01–R03BC03, R03BX, R03DX, R03DC, R03DA, J01, J07, N07BA, N06AX12, R03C. | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| 5. | Other treatments (such as phosphodiesterase inhibitors, leukotriene modifiers, antibiotic therapy, cromoglicic acid, antitussive, xanthines, nicotine, and drugs used for nicotine dependence): any of C01CE, R03DX07, R03BC01, R03BC03, R03BX, R03DC, R03DA, J01, J07, N07BA, N06AX12, R05DB. | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| 6. | Combinations:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
We then evaluated the occurrence of monotherapies and combination therapies based on co-occurrence of different medication codes or codes suggestive of fixed combination treatments.
Definition of clinical phenotypes
Patients were classified into four different clinical phenotypes according to computerized tomography (CT) scan and bronchodilator response test results:
| 1) | Frequent exacerbators with partial response to bronchodilator test with either severe or very severe respiratory function impairment based on FEV1% (delta FEV1% pre-post greater than 12% or FEV1% pre-post greater than 200 ml; partial bronchodilator response phenotype); | ||||
| 2) | Frequent exacerbators with emphysema phenotype defined with CT scan; | ||||
| 3) | Frequent exacerbators phenotype with no emphysema or any response to bronchodilators (patients with negative CT scan and delta FEV1% pre-post less than 12% and FEV1% pre-post less than 200 ml; bronchitis phenotype); | ||||
| 4) | Frequent exacerbators unclassified (no adequate diagnostic testing for classification). | ||||
We further classified patients on the basis of FEV1%, according to Global initiative for chronic Obstructive Lung Disease (GOLD) airflow limitation stages Citation(3): 1) severe COPD for 30% ≤ FEV1% ≤ 50% and 2) very severe COPD for FEV1% ≤ 30%.
Statistical analysis
A descriptive analysis was performed for all considered variables, presenting the absolute frequencies in case of categorical variables, as well as the mean with standard deviation in the case of the continuous variables. Differences in variable distributions across phenotypes and COPD severity were tested with chi-square or analysis of variance (ANOVA) test when appropriate. In order to evaluate the proportion of patients without a thorough clinical characterization, we estimated the center-specific prevalence of the unclassified phenotype with random-intercept logistic regression models accounting for the clustered nature of the data (level 1: patients; level 2: center). The between-center variation was estimated with the variance partition coefficient (VPC). We used a hierarchical model building approach to progressively enter the variables in the model. To estimate the unadjusted between-center variation in the prevalence of the unclassified phenotype, we run an “empty” model, which did not consider any patients' covariate. In order to account for casemix differences across centers, we first included demographic variables (age, sex, BMI, and FEV1%), and finally, we added the co-morbidities and treatment regimens.
In order to evaluate potential correlates of exacerbation risk and the between-center variation in exacerbation rate, we ran a second multilevel hierarchical logistic regression model adopting the same analytic strategy previously described. The dependent variable for this analysis was an indicator variable, denoting the occurrence of three or more exacerbations in the 12 months prior to enrollment. We evaluated whether the case-mix adjusted rate of diagnostic testing (% of unclassified phenotype) was associated with the case-mix adjusted exacerbation risk, with Spearman's correlation coefficient.
Results
Sample characteristics and clinical phenotypes
Clinical and sociodemographic characteristics of 743 patients classified into clinical phenotypes and GOLD airflow limitation severity classes are reported in , respectively. About 18% (n = 134) of the sample had a very severe expiratory airflow limitation (FEV1% < 30%). Patients with very severe disease were younger, had lower BMI, and were more likely enrolled in the study during the course of a hospitalization ().
Table 1. Patients characteristics across clinical phenotypes.
Table 2. Patients characteristics across GOLD stages of airflow limitation.
One-third of patients (n = 210, 29%) underwent a bronchodilator response test. Among them, 35 (17%) had a partial response to bronchodilators. Among 163 (22%) patients undergoing a CT scan, 119 (73%) had a diagnosis of emphysema. There were only five patients with no evidence of emphysema at the CT scan and also had insufficient response to bronchodilators. The remaining patients were defined “unclassified” (584, 79%) because of the lack of CT scan and results of the bronchodilator test.
Patients with a partial response to bronchodilators were less likely to have very severe respiratory function impairment. All of them have been enrolled during a regular follow-up outpatient visit; subjects with emphysema were younger and had lower BMI and more severe airflow limitation.
The prevalence of the unclassified phenotype significantly varied across centers (). The VPC was 44.1% in the intercept-only model and 45.3% after casemix adjustment.
Figure 1. Unadjusted and casemix-adjusted prevalence of the unclassified phenotype across participating centers. Distribution of unclassified patients across centers: (A) Crude prevalence of patients without sufficient diagnostic testing for phenotypical characterization across centers. (B) Estimated casemix-adjusted prevalence of unclassified patients. We obtained predicted center-related probabilities from a random-intercept logistic regression model adjusted for age, sex, co-morbidities, FEV1%, body mass index, smoking habit, and therapy regimens.
Co-morbidities
Only a minority of patients (n = 124; 17%) had no co-morbidities, whereas 252 (34%) had one co-morbidity, 198 (27%) had two co-morbidities, and 169 (23%) more than two co-morbidities.
The most frequent coexistent disease was hypertension (65.8%), followed by heart failure (19.7%), diabetes (19.5%), coronary artery disease (17.5%), psychiatric conditions (14.9%), and other cardiovascular disease (arrhythmia 4.9%, other 7.3%); metabolic syndrome (i.e., central obesity plus any two of elevated triglycerides, reduced high density cholesterol, hypertension, and raised fasting glucose), renal failure, and osteoporosis were less frequently reported.
The distribution of co-morbidities varied across clinical phenotypes. Cardiovascular disorders including heart failure were less prevalent among patients with a partial response to bronchodilator and among the “unclassified” phenotype; patients with emphysema more likely had osteoporosis and cardiovascular co-morbidities (). Additionally, we found that hypertension (67.7% vs. 57.5%, respectively) and diabetes (21.4% vs. 11.2%) were more prevalent among patients with severe airflow limitation compared to those with very severe obstruction according to the GOLD classification system ().
Therapy
Patients enrolled in the study took on average 2.8 (SD = 1.4) medications for COPD. Long-acting bronchodilators alone or in combination with other drugs were the most prescribed class of medications (69.0%, n = 521), followed by corticosteroids (20.1%, n = 152) and short-acting bronchodilators (15.1%, n = 114). Among patients taking corticosteroid therapy, 79.3% also used either short-acting or long-acting bronchodilators in combination. There were 139 patients taking other COPD-related treatments such as leukotriene modifiers or phosphodiesterase inhibitors (18.7%).
Treatment regimen significantly varied across different clinical phenotypes (). Patients with emphysema were more likely to receive complex therapeutic regimens entailing a larger number of medications compared to other subgroups (p < 0.05). The majority of patients with partial response to bronchodilator took corticosteroids as a stand-alone therapy (57.1%), or in association with one or more bronchodilators (37.1%). Patients with emphysema were more likely prescribed combinations of bronchodilators and corticosteroids (68.9%) or double long-acting bronchodilators (11.8%). Among unclassified patients, 14.5% and 8.4% received either single or double long-acting bronchodilators. Additionally, over 60% of patients in this subgroup received bronchodilators in combination with corticosteroids. No differences emerged among treatment regimens between the two COPD severity groups ().
Risk of exacerbations
The mean number of exacerbations was 2.6/year (SD = ±0.94). The prevalence of patients with three or more exacerbations across clinical phenotypes is reported in . We found no significant differences in the risk of ≥3 exacerbations across clinical phenotypes and COPD severity classes.
In the multivariable model, only a clinical history of heart failure [odds ratio (OR): 1.89; 95% confidence interval (CI) 1.48–2.3) and mood/affective disorders (OR: 1.66; 95% CI 1.24–2.1) was associated with the likelihood of frequent exacerbations. The casemix-adjusted center-specific prevalence of patients experiencing ≥3 exacerbations is reported in . The casemix-adjusted VPC was 29.6%. The correlation between the casemix-adjusted center-specific prevalence of patients with ≥3 exacerbations and the casemix-adjusted center-specific prevalence of the unclassified phenotype was rho = −0.26, p = 0.07.
Figure 2. Casemix-adjusted prevalence of patients with three or more exacerbations in the previous year across participating centers. Points represent the estimated casemix-adjusted prevalence of patients with three or more exacerbations in the previous year at each center. We obtained predicted center-related probabilities from a random-intercept logistic regression model adjusted for age, sex, co-morbidities, FEV1%, body mass index, smoking habit, and therapy regimens.
Discussion
After the definition of the frequent exacerbator COPD group Citation(4), this is the first Italian study aimed at characterizing the phenotypical profile of this population. Geographical variations in the prevalence and clinical behavior of COPD have been documented underscoring the need to conduct national epidemiological studies to inform health care and policy decision-making.
Among the 743 COPD frequent exacerbators with severe airflow limitation enrolled in this study, 83% had at least one co-morbidity, hypertension and cardiovascular diseases being the most represented. Despite being consistent with previous reports concerning the frequent exacerbators phenotype Citation(10), the prevalence of coexistent diseases found in our study was much higher compared to that generally observed among the general COPD population. Co-morbidity burden is an important aspect of care for patients with COPD. It has been shown that mortality risk associated with co-morbidities extends beyond the effect of common risk factor among patients with COPD Citation(11). The majority of frequent exacerbators took complex combination regimens. Additionally patients enrolled in our study reported multiple comorbidities. As a consequence, frequent exacerbators may experience non-adherence, treatment side effects and quality of life impairment due to a severe medication burden.
We observed a differential co-morbidity profile across subgroups defined by clinical characteristics. Patients with severe expiratory airflow obstruction showed higher BMI, hypertension, diabetes, and metabolic syndrome rates compared to patients with very severe obstruction; this was partially explained by the lower prevalence of patients with emphysema–hyperinflation among subjects with severe airflow limitation. Although emphysema assessed with CT was more prevalent among GOLD stage 4, there was no complete overlap between GOLD stage 4 and the emphysema phenotype Citation(3). Additional explanations include clustering of lifestyle factors such as smoking, occurring along with alcohol consumption and high-fat diets Citation(11), and selection bias due to the increased mortality of very severe patients with such co-morbidities.
Conversely, there was a higher prevalence of affective disorders among patients with very severe expiratory airflow obstruction, a result consistent with that of previous studies, showing a high prevalence of mood and anxiety conditions. The causal pathway linking the degree of FEV1% impairment and mood disorders is not well understood and merits further investigation since the co-occurrence of depression in patients with COPD is associated with impoverished quality of life and reduced survival.
Similar to previous studies, patients with emphysema represented a large share of frequent exacerbators and presented with worse clinical conditions. Despite younger, patients with emphysema had lower basal FEV1%, lower BMI, a slightly higher heart failure rate, and an increased prevalence of osteoporosis, a condition possibly secondary to malnutrition, severe chronic inflammation, physical inactivity, and corticosteroid therapy Citation(12,13). The loss of elastic properties of the lung makes alveolar emptying difficult and determines hyperinflation, a condition that is closely related to carbon monoxide diffusion capacity, functional impairment, and mortality, independent of the severity of expiratory airflow obstruction Citation(14). There is no agreement on the best treatment for patients with emphysema and frequent exacerbations. There is substantial evidence that the use of single or double long-acting bronchodilators is beneficial for patients with emphysema–hyperinflation phenotype. The use of corticosteroids is more controversial even in patients with frequent exacerbations, and there is evidence that addition of corticosteroids does not provide additional benefit to long-acting bronchodilator therapy in this patients' group. Nevertheless, the results of our analysis showed that patients with emphysema were often treated with complex regimens, entailing the combinations of bronchodilators with corticosteroids, and only a minority of cases used double long-acting bronchodilators without steroid supplementation.
A potentially important result of our study was that most patients remained unclassified due to the lack of diagnostic testing. Our analysis showed that the single, most important factor affecting the likelihood of receiving such diagnostic testing was center membership, whereas both clinical and sociodemographic variables played a very little role. Since appropriate risk stratification may affect treatment choices and patient's prognosis, our results raise the question whether such variation in diagnostic testing is justified by sensible decision-making or imply inappropriate clinical practice in some centers. Currently, it is not clear yet whether clinical profiling solely based on bronchodilator test response and the presence of emphysema at CT scan provides prognostic data, informing clinical decision-making independent of clinical signs and symptoms Citation(15–20).
Bronchodilator test response is usually considered a cornerstone for the diagnosis of COPD, asthma, and ACOS. Additionally GOLD guidelines suggest that spirometry should be performed every year. However, this position has been questioned by evidence suggesting that short-acting bronchodilator responsiveness testing is scarcely reliable (21), may be inaccurate for differential diagnosis (15),and may not be predictive of clinical response to long-acting bronchodilators (22) and may not be predictive of clinical response to long-acting bronchodilators Citation(22).
Additionally, although patients with emphysema represent a clearly distinguished phenotype with specific clinical trajectory and health care needs, there is no specific guidance as to whom and when to perform a CT scan in this population, a circumstance that may partially explain the wide variability in diagnostic testing pattern across centers. According to previous studies, CT scan seems to have a better prognostic value than spirometry alone. In fact, CT scan findings of emphysema are associated with reduced functioning Citation(23) and increased mortality Citation(19) among COPD patients, independent of spirometric evidence of airflow obstruction and hyperinflation. However, emphysema solely defined on the basis of a CT scan may not correspond to a separate clinical entity and therefore may not be effectively treated with a specific therapeutic regimen Citation(24).
Finally, consistent with previous research Citation(4,25,26), we have found evidence that a higher exacerbation rate was associated with worse airflow obstruction. However the strength of association was weak in our study and lost significance in the multivariable model, possibly due to restricted range in FEV1% imposed by the inclusion criteria of the study. Additionally, we found that affective disorders were associated with higher exacerbation risk, a finding previously observed in several studies. Although it is thought that the relationship between affective disorders and exacerbation risk is mediated by reduced adherence, dysregulation of inflammatory response, and lifestyle disruption Citation(27,28), the mechanism underlying such association has not been characterized yet. Additionally, we have found that patients with three or more exacerbations had higher rates of cardiovascular disease. The relationship between exacerbation history and subsequent myocardial injury is well documented, and it is mediated by increased arterial stiffness and pro-inflammatory stimuli during exacerbation episodes Citation(29). Whether cardiovascular disease is a consequence of the pathophysiological changes induced during exacerbations or both are epiphenomena of the same underlying mechanism in COPD patients is a matter of further research. The relationship between exacerbation rate and sociodemographic and other clinical factors, including phenotypes based on CT scan lung densitometry and bronchodilator test response, was not significant. Although we found little evidence of the influence of patient-related factors on exacerbation risk, we observed a strong between-center variation in the occurrence of this outcome. The casemix-adjusted VPC showed that unmeasured center-related factors account for approximately 30% of variability in exacerbation prevalence. Additionally, center-specific prevalence of patients with three or more exacerbations in the past year was inversely associated with the casemix-adjusted center-specific prevalence of the unclassified phenotype. This observation may be indicative of increased diagnostic testing among patients with worse clinical outcomes, yet the reason underlying the differential distribution of exacerbation frequency across centers remains unexplained by our data. Previous studies have documented regional and temporal variations in exacerbation rates Citation(30,31) that may be attributed to heterogeneous practice patterns Citation(30), variations in the distribution of risk factors Citation(30,31), and unmeasured confounding. A study conducted on the US Veteran Affairs Network showed that between-center differences in exacerbation rates were robust to adjustment for climate, age, co-morbidities, and baseline exacerbation rates (a proxy for severity) in a historical cohort study, providing strong evidence that practice patterns have a strong influence on patients' outcomes Citation(30).
Limitations
Although we adjusted for several patient-related factors, residual confounding cannot be ruled out as an alternative explanation of our findings. Previous studies have shown that the risk of exacerbations is increased in older patients Citation(25), subjects with chronic mucus secretion Citation(25), those with a history of hospitalization or exacerbations in the previous year Citation(4,26), a higher St. George's Respiratory Questionnaire (SGRQ) score Citation(4), history of gastroesophageal reflux Citation(4), and use of corticosteroids as a maintenance therapy Citation(26). Most of these variables were not available for our analysis. Additionally, the enrollment period of the study lasted for almost 2 years and was not homogeneous across centers, making it difficult to estimate the confounding factors due to the impact of seasonal variations in transient risk factors such as infectious disease temporal dynamics and weather conditions. Additionally, there are several potentially appropriate classifications of clinical phenotypes among COPD patients, which may diverge from the definition adopted in this study Citation(32). For this reason, differences across different frequent exacerbators subgroups reported in this study should be interpreted cautiously. A further important limitation preventing us to characterize patients' characteristics and risk profiles across clinical phenotypes was the large prevalence of patients with insufficient testing for classification. As a consequence, the bronchitis subgroup and the partial response to bronchodilators subgroup as well, represented a small share of patients, leading to potentially unreliable statistical inferences.
Conclusions
In this cross-sectional survey of frequent COPD exacerbators, co-morbidities, in particular cardiovascular diseases, were very common. Although individual coexistent diseases and the overall burden of co-morbidities did not contribute much to exacerbation risk, it is well known that patients with multiple diseases besides COPD have poorer long-term outcomes and are often difficult to manage due to polypharmacy and poor prescription adherence. The distribution of co-morbidities observed in our study significantly varied across clinical phenotypes and severity classes. Our results suggest that the GOLD staging system, although simple and intuitively sound, does not capture the extreme variability of patients' clinical presentation, and it did not contribute significantly to treatment decision-making. Our study further highlighted the importance and the difficulties of patients' profiling for COPD health care. In line with the uncertainties related to COPD phenotyping, we observed a wide between-center variability in both diagnostic testing patterns and exacerbation rates beyond casemix differences across centers. Examining regional variation in health outcomes and health care behavior may help identify best practices, especially when there are no well-established evidence-based recommendations and uncertainties surround clinical decision-making for selected patient subgroups. Given that numerous patients' profiling algorithms and procedures have been proposed, it is not surprising that such an uncertainty reflects in inconsistent practices across centers. However, the wide casemix-adjusted between-center variation in exacerbation rate suggests that practice patterns may significantly contribute to health outcomes of patients. In-depth examination of practice patterns in centers with the lowest exacerbation rates could provide useful guidance on medical decision-making algorithms or interventions that could be used across the health care system.
Acknowledgments
F. Blasi, S. Centanni, F. Falcone, and G. D. Maria developed study concept, participated in study design, contributed to the interpretation of results, and approved the final version of the manuscript. L. Neri conducted data analysis and drafted the first version of the manuscript. Members of the PREFER study group participated in study design, contributed to the interpretation of results, and approved the final version of the manuscript.
Declaration of interest
F. Blasi, S. Centanni, F. Falcone, and G. D. Maria received consultancy fees as members of the scientific advisory board. L. Neri received consultancy fees for data analysis.
Funding
This study was partially funded by Takeda Italia.
References
- Hillas G, Perlikos F, Tsiligianni I, Tzana N. Managing comorbidities in COPD. Int J Chron Obstruct Pulmon Dis 2015; 10(1):95–109.
- Negewo N, Gibson P, McDonald V. COPD and its comorbidities: impact, measurement and mechanisms. Respirology 2015; 20(8):1160–1171.
- Global Strategy for the Diagnosis, Management and Prevention of COPD. Global Initiative for Chronic Obstructive Lung Disease (GOLD). 2016. Available from: http://www.goldcopd.org/
- Hurst JR, Vestbo J, Anzueto A, Locantore N, Müllerova H, Tal-Singer R, et al. Susceptibility to exacerbation in chronic obstructive pulmonary disease. N Engl J Med 2010; 363(12):1128–1138.
- Divo M, Cote C, de Torres JP, Casanova C, Marin JM, Pinto-Plata V, et al. Comorbidities and risk of mortality in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2012; 186(2):155–161.
- Decramer M, Janssens W. Chronic obstructive pulmonary disease and comorbidities. Lancet Respir Med 2013; 1(1):73–83.
- Faner R, Cruz T, López-Giraldo A, Agustí A. Network medicine, multimorbidity and the lung in the elderly. Eur Respir J 2014; 44(3):775–788.
- Guarascio AJ, Ray SM, Finch CK, Self TH. The clinical and economic burden of chronic obstructive pulmonary disease in the USA. Clinicoecon Outcomes Res 2013; 17(5):235–245.
- Valderas JM, Starfield B, Sibbald B, Salisbury C, Roland M. Defining comorbidity: implications for understanding health and health services. Ann Fam Med 2009; 7(4):357–363.
- Fumagalli G, Fabiani F, Forte S, Napolitano M, Balzano G, Bonini M, et al. INDACO project: COPD and link between comorbidities, lung function and inhalation therapy. Multidiscip Respir Med 2015; 10(1):4.
- Müllerová H, Shukla A, Hawkins A, Quint J. Risk factors for acute exacerbations of COPD in a primary care population: a retrospective observational cohort study. BMJ Open 2014; 4(12):e006171.
- Gu S, Li R, Leader JK, Zheng B, Bon J, Gur D, et al. Obesity and extent of emphysema depicted at CT. Clin Radiol 2015; 70(5):e14–e19.
- Wan ES, Cho MH, Boutaoui N, Klanderman BJ, Sylvia JS, Ziniti JP, et al. Genome-wide association analysis of body mass in chronic obstructive pulmonary disease. Am J Respir Cell Mol Biol 2011; 45(2):304–310.
- Jose J, Escobar A, Bashir M, Claes D. Abnormal rv/tlc ratio is a better criterion to diagnose obstruction in patients with asthma. Chest 2011; 140( 4_MeetingAbstracts):226A.
- Richter DC, Joubert JR, Nell H, Schuurmans MM, Irusen EM. Diagnostic value of post-bronchodilator pulmonary function testing to distinguish between stable, moderate to severe COPD and asthma. Int J Chron Obstruct Pulmon Dis 2008; 3(4):693–699.
- Calverley PMA, Burge PS, Spencer S, Anderson JA, Jones PW. Bronchodilator reversibility testing in chronic obstructive pulmonary disease. Thorax 2003; 58(8):659–664.
- Tashkin DP, Celli B, Decramer M, Liu D, Burkhart D, Cassino C, et al. Bronchodilator responsiveness in patients with COPD. Eur Respir J 2008; 31(4).
- Sorino C, D'Amato M, Steinhilber G, Patella V, Corsico AG. Spirometric criteria to diagnose airway obstruction in the elderly: fixed ratio vs. lower limit of normal. Minerva Med 2015; 14(1).
- Haruna A, Muro S, Nakano Y, Ohara T, Hoshino Y, Ogawa E, et al. CT scan findings of emphysema predict mortality in COPD. Chest 2010; 138(3):635–640.
- Tsushima K, Sone S, Fujimoto K, Kubo K, Morita S, Takegami M, et al. Identification of occult parechymal disease such as emphysema or airway disease using screening computed tomography. COPD 2010; 7(2):117–125.
- Paul A, Agusti A, Edwards L, Tal-Singer R, Yates J, Bakke P, et al. Bronchodilator responsiveness as a phenotypic characteristic of established chronic obstructive pulmonary disease. Thorax 2012; 67(8):701–708.
- Burgel PR, El Azzi G, Le Gros V, Deslée G. Short-acting bronchodilator reversibility does not predict the response to long-term treatment by indacaterol 150 µg in COPD: REVERBREZ study. Eur Respir J 2013; 42(57):693–699.
- Diaz AA, Bartholmai B, San José Estépar R, Ross J, Matsuoka S, Yamashiro T, et al. Relationship of emphysema and airway disease assessed by CT to exercise capacity in COPD. Respir Med 2010; 104(8):1145–1151.
- Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson LJ, Loscalzo J, et al. Harrison's Principles of Internal Medicine, 17th ed. New York, NY: McGraw Hill, 2009.
- Miravitlles M, Guerrero T, Mayordomo C, Sánchez-Agudo L, Nicolau F, Segú JL. Factors associated with increased risk of exacerbation and hospital admission in a cohort of ambulatory COPD patients: a multiple logistic regression analysis. Respiration 2000; 67(5):495–501.
- Niewoehner DE, Lokhnygina Y, Rice K, Kuschner WG, Sharafkhaneh A, Sarosi GA, et al. Risk indexes for exacerbations and hospitalizations due to COPD. Chest 2007; 131(1):20–28.
- Laurin C, Moullec G, Bacon SL, Lavoie KL. The impact of psychological distress on exacerbation rates in COPD. Ther Adv Respir Dis 2011; 5(1):3–18.
- Xu W, Collet JP, Shapiro S, Lin Y, Yang T, Platt RW, et al. Independent effect of depression and anxiety on chronic obstructive pulmonary disease exacerbations and hospitalizations. Am J Respir Crit Care Med 2008; 178(9):913–920.
- Patel AR, Kowlessar BS, Donaldson GC, Mackay AJ, Singh R, George SN, et al. Cardiovascular risk, myocardial injury, and exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2013; 188(9):1091–1099.
- Joo MJ, Lee TA, Weiss KB. Geographic variation in chronic obstructive pulmonary disease exacerbation rates. J Gen Intern Med 2007; 22(11):1560–1565.
- Donaldson GC, Wedzicha JA. The causes and consequences of seasonal variation in COPD exacerbations. Int J Chron Obstruct Pulmon Dis 2014; 9(1):1101–1110.
- Miravitlles M, Soler-Cataluña JJ, Calle M, Molina J, Almagro P, Quintano JA, et al. A new approach to grading and treating COPD based on clinical phenotypes: summary of the Spanish COPD guidelines (GesEPOC). Prim Care Respir J 2013; 22(1):117–121.