Initiation of Triple Therapy Maintenance Treatment Among Patients With COPD

Claims data analysis showed that 60% of patients with chronic obstructive pulmonary disease (COPD) receiving triple therapy had no evidence of exacerbation or only 1 exacerbation not resulting in hospitalization.

ABSTRACTObjectives: Triple therapy is indicated for patients with very severe chronic obstructive pulmonary disease (COPD). Use of this treatment in the appropriate patient population is important to ensure optimal outcomes. This study quantified the use of triple therapy and assessed concordance with 2013-2016 Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommendations within a national health plan.

Study Design: Retrospective cohort study using data from a large national health plan.

Methods: To estimate the prevalence of triple therapy using claims data, patients in the first of 2 cohorts were indexed on their first diagnosis of COPD between January 1, 2012, and December 31, 2014, and required to have 24 months postindex continuous enrollment. To assess concordance with GOLD recommendations, a second cohort was created and indexed on the date of triple therapy initiation between January 1, 2013, and November 30, 2016, and required to have 12 months preindex and 1 month postindex continuous enrollment. For both cohorts, patients were aged 40 years or older, with no International Classification of Diseases code for asthma, cystic fibrosis, or lung cancer during the study period.

Results: In the first cohort of 92,248 patients with COPD receiving any COPD maintenance medication, 17% were prescribed triple therapy. In the second cohort (n = 19,645), the majority (60%) of patients on triple therapy were classified as GOLD group A or B (ie, no evidence of any exacerbation or only 1 exacerbation not resulting in hospitalization at baseline).

Conclusions: Results showed that triple therapy was often prescribed among patients classified as GOLD group A or B. Additional research is required, however, to further assess whether these patients may have had an exacerbation that was not evident in claims data. Treatment of COPD should be individualized to optimize outcomes and reduce adverse events.

Am J Manag Care. 2020;26(4):e106-e112. https://doi.org/10.37765/ajmc.2020.42837

Takeaway Points

This study provided updated estimates of triple therapy use among patients with chronic obstructive pulmonary disease (COPD) and assessed the concordance of triple therapy use with Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommendations.

  • Of patients with COPD who received any COPD maintenance therapy, 17% were prescribed triple therapy.
  • Among patients prescribed triple therapy, 60% had no evidence of exacerbation or only 1 exacerbation that did not lead to hospitalization prior to initiation of initial therapy.
  • Additional research is required, however, to further assess whether these patients may have had an exacerbation that was not evident in claims data.
  • The extent of potential overtreatment warrants better provider education around GOLD recommendations to ensure appropriateness of care.

Chronic obstructive pulmonary disease (COPD) is a preventable and treatable disorder characterized by persistent respiratory symptoms and airflow limitations.1 Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines classify patients with COPD into 1 of 4 groups (A, B, C, and D) based on exacerbation history and current symptoms.1 Triple therapy, defined as treatment with a long-acting β-agonist (LABA), a long-acting muscarinic antagonist (LAMA), and an inhaled corticosteroid (ICS), is indicated only for patients with very severe COPD (GOLD group D).1 Furthermore, ICS-containing treatment is only recommended for patients with 2 or more COPD exacerbations or 1 or more exacerbations requiring hospitalization (GOLD groups C and D).1 Despite these treatment recommendations, there is evidence that triple therapy is broadly prescribed to patients with COPD in GOLD groups A and B.2-4 Concern over this prescribing practice is based on the lack of evidence demonstrating a favorable benefit-risk profile with the use of triple therapy in patients with less severe COPD. Even when clinically indicated in very severe COPD, ICS exposure has been shown to increase risk of certain adverse events, namely pneumonia. Similarly, nonadherence to GOLD guidelines is associated with increased healthcare resource utilization (HRU) and medical costs.5-7 Collectively, evidence suggests that adherence to GOLD guidelines is recommended to attain optimal clinical and economic outcomes in the COPD population.

Data on real-world triple therapy use are limited in the literature. Make et al8 reported that the proportion of patients with COPD receiving triple therapy was 12.5% among commercially insured patients and 9.7% among patients with Medicare, based on health plan data from 2004 to 2005. More recently, Simeone and colleagues5 estimated that 7.5% of patients with COPD received triple therapy, based on 2009-2013 claims data (primarily from commercial health plans) from the IMS PharMetrics Plus database. The present study sought to provide an updated look at the prevalence of triple therapy in a predominantly Medicare population, using data from a large health plan, to understand the demographic and clinical characteristics of triple therapy users, and to assess the extent of concordance with the GOLD guidelines in the treatment of COPD.

Specifically, the objectives of the study were (1) to estimate the proportion of patients with COPD who received a triple therapy regimen overall and stratified by none/mild cases versus moderate to severe cases based on exacerbation history and (2) to describe patient characteristics, exacerbation history, and HRU prior to initiation of a triple therapy regimen to assess the concordance of triple therapy use with GOLD recommendations.

METHODS

Data Source

This was a retrospective cohort study based on the Humana Inc administrative claims database from 2012 to 2016. The database contains integrated medical, pharmacy, behavioral, and laboratory-related claims, representing more than 12 million current and former Humana patients. The deidentified patient-level data include detailed cost, use, and outcomes data for healthcare services performed in inpatient and outpatient settings; prescription drug claims; and information on patient enrollment. The claims are linked through a patient identifier that remains constant regardless of benefit design, gap in coverage, or change of employer. These data include information on patients enrolled in commercial, Medicare Advantage, and prescription drug plans; have national coverage with a high proportion of patients from Texas, Florida, and Ohio; and form one of the largest Medicare Advantage claims databases.

Sample Selection

Two separate patient cohorts were created to achieve the objectives of this study. Cohort 1 included all patients with diagnosed COPD. Use of triple therapy was examined as an outcome measure among this cohort. Cohort 1 allowed us to estimate the proportion of patients with COPD receiving a triple therapy regimen. Cohort 2 included all patients who initiated triple therapy. The characteristics of these patients prior to the initiation of triple therapy were examined. With cohort 2, we examined patients’ demographic and clinical characteristics during the 12 months before the initiation of triple therapy and assessed concordance with GOLD recommendations based on these characteristics. Specific inclusion and exclusion criteria for these 2 cohorts are described as follows. (For the inclusion criteria, COPD diagnosis was defined as a claim with International Classification of Diseases, Ninth Revision [ICD-9] diagnosis code 491.xx [chronic bronchitis], 492.xx [emphysema], or 496.xx [COPD, unspecified]; or International Classification of Diseases, Tenth Revision [ICD-10] diagnosis codes J41.x, J42, J43.x, or J44.x in primary or secondary position.)

Cohort 1 (for objective 1). Patients were included if they met the following criteria during the identification period (January 1, 2012, to December 31, 2014): (1) 1 or more inpatient medical claims with a COPD diagnosis or 2 or more outpatient medical claims (physician office, emergency department [ED], or other outpatient) with a COPD diagnosis (the 2 COPD medical claims were required to have occurred on separate dates); (2) continuous enrollment in a health plan for a period of 24 consecutive months following the index date (the date of the first observed COPD diagnosis during the identification period for patients identified from inpatient medical claims, or the second medical claim for patients identified through outpatient medical claims for COPD) during the study period; and (3) treatment with a LAMA, LABA, ICS, LABA/ICS, LABA/LAMA, or phosphodiesterase-4 (PDE-4) inhibitor within 24 months post index.

Patients were excluded if they met any of the following criteria: (1) younger than 40 years at COPD diagnosis date or (2) presence of at least 1 medical claim with a diagnosis of asthma (ICD-9 code 493.xx; ICD-10 code J45.x), cystic fibrosis (ICD-9 code 277.0; ICD-10 code E84.x), or lung cancer (ICD-9 code 162.xx, except 162.0x; ICD-10 code C34.x) during the study period.

Study measures for cohort 1. Cohort 1 was primarily used to estimate the proportion of patients with COPD who received a triple therapy regimen. Similar to the definition used by Simeone et al,5 patients were deemed to be on triple therapy if the days they were supplied all 3 therapy components (LABA, LAMA, and ICS in free-dose or fixed-dose combination) overlapped for 30 or more days (eg, LAMA + LABA/ICS, LAMA + LABA + ICS, LAMA/LABA + ICS) cumulatively over the study period.

Cohort 2 (for objective 2). Patients were included if they met the following criteria during the identification period (January 1, 2013, to November 30, 2016): (1) evidence of triple therapy (defined the same as explained in the previous paragraph, with the index date for cohort 2 defined as the date of initiation of the third component of triple therapy); (2) 1 or more inpatient medical claims with a COPD diagnosis or 2 or more outpatient medical claims (physician office, ED, or other outpatient) with a COPD diagnosis (the 2 COPD medical claims were to have occurred on separate dates); (3) continuous enrollment in a health plan for a period of 12 consecutive months prior to the index date and 1 month after the index date; and (4) 40 years or older at the index date.

Patients in cohort 2 were excluded if they had at least 1 medical claim with a diagnosis of asthma, cystic fibrosis, or lung cancer during the study period.

Study measures for cohort 2. Demographic characteristics assessed for the study included age, gender, and geographic region. Clinical characteristics included Deyo-Charlson Comorbidity Index (DCI) score,9 common comorbid conditions (eAppendix A [eAppendices available at ajmc.com]), COPD etiology, and COPD severity. COPD etiology was derived by assessing all medical claims with a diagnosis code indicating COPD and classified as chronic bronchitis (ICD-9 code 491.xx; ICD-10 codes J41.x, J42, J44.0, or J44.1), emphysema (ICD-9 code 492.xx; ICD-10 code J43.x), unspecified (ICD-9 code 496.xx; ICD-10 code J44.9), or mixed (any combination of these). COPD severity was measured based on the frequency and severity of exacerbations (mild, moderate, or severe) prior to initiation of triple therapy. For the purpose of this study, exacerbations not leading to hospitalization were classified as mild, and exacerbations leading to hospitalization were classified as moderate if there was no concurrent diagnosis of a respiratory failure and severe if there was concurrent diagnosis of a respiratory failure. These definitions are consistent with those used by Simeone et al5 to allow for comparison of results, and the full algorithm for classifying severity of exacerbation can be found in eAppendix B. Note that these exacerbation definitions deviate from those introduced in the 2017 GOLD report, which were not available at the time of the Simeone et al study. We then used exacerbation severity to approximate GOLD ABCD groups. Patients who experienced no exacerbations or 1 mild exacerbation were approximated as being in GOLD groups A and B, whereas those who experienced more than 1 mild exacerbation or experienced a moderate or severe exacerbation were approximated as being in GOLD groups C and D.

Use of maintenance and rescue COPD medications (including fixed-dose combinations) was assessed based on pharmacy claims and included LABA, LAMA, ICS, short-acting β-agonists (SABA), short-acting muscarinic antagonists (SAMA), PDE-4 inhibitors, methylxanthines, and oral and injectable systemic corticosteroids (eAppendix C). Other baseline medications of interest were use of antibiotics, antidiabetic drugs, and cardiovascular drugs. These data were collected to help provide a more complete clinical profile for patients with COPD.

All-cause and COPD-related HRU at baseline were examined. All-cause visits were defined as any inpatient, ED, or outpatient visit related to any diagnosis; COPD-related visits were identified as inpatient, ED, or outpatient visits in which a diagnosis code for COPD was the primary diagnosis code associated with that visit (or discharge claim for an inpatient visit). Additionally, spirometry testing and oxygen therapy were reported based on Healthcare Common Procedure Coding System/Current Procedural Terminology codes (eAppendix D).

Statistical Analysis

All data analyses were conducted using SAS version 9.3 and SAS Enterprise Guide version 7.1 (SAS Institute, Cary, North Carolina). The a priori α level for all inferential analyses was set at .05, and all statistical tests were 2-tailed, unless otherwise specified. Data were evaluated for violations of assumptions underlying the associated statistical tests as appropriate.

Cohort 1. The proportion of patients with COPD receiving triple therapy was calculated using the number of patients prescribed triple therapy in the 24-month postindex period in the numerator and all patients with COPD meeting the inclusion and exclusion criteria for this cohort in the denominator.

Cohort 2. Descriptive analyses were conducted to summarize the baseline demographic, clinical, and HRU characteristics. Mean, median, SD, and range were reported for continuous variables and frequency (number and percentage) reported for categorical variables. According to 2017 GOLD guidelines, triple therapy is not recommended for patients with COPD who have no or 1 exacerbation (not leading to hospitalization). The proportion of patients receiving triple therapy who experienced no or 1 exacerbation (not leading to hospitalization) was quantified to estimate the extent to which triple therapy was prescribed outside of the GOLD guidelines.

In addition, the demographic profiles of patients receiving triple therapy who experienced no or 1 exacerbation (not leading to hospitalization) and those who experienced more than 1 exacerbation (not leading to hospitalization) or any exacerbations resulting in hospitalization were described. Because these data were examined for descriptive purposes only, no statistical testing was performed.

RESULTS

Cohort 1

A total of 597,284 patients with medical claims for COPD were identified between January 1, 2012, and December 31, 2014. Of these, 230,442 met the inclusion and exclusion criteria and were included in the study (Table 1). Further, 92,248 patients received any maintenance COPD medications in the 24-month postindex period. Among all patients with a COPD diagnosis (n = 230,442), 6.9% (n = 15,797) received triple therapy during the study period of January 2012 to December 2014. Of the 92,248 patients with COPD receiving any maintenance medications in the 24-month postindex period, the proportion of patients receiving triple therapy was 17.1% (n = 15,797). Most patients’ triple therapy regimen was LAMA + LABA/ICS (90.1%; n = 14,307).

Cohort 2

A total of 78,872 patients receiving triple therapy for at least 30 days were identified between January 1, 2013, and November 30, 2016. Of these, 19,645 patients met the inclusion and exclusion criteria and were included in the study (Table 1). Similar to cohort 1, 88.4% (n = 17,361) of patients’ triple therapy was LAMA + LABA/ICS.

Demographic characteristics of patients with COPD receiving triple therapy are presented in Table 2. The mean (SD) age of the patients was 70.71 (8.39) years, and 49% were women. A majority (65.0%) resided in the geographic South. Baseline clinical characteristics of patients with COPD receiving triple therapy are presented in Table 3. Mean (SD) DCI score was 2.51 (2.15). The most frequent comorbidities were hypertension (78.2%), hypercholesterolemia/hyperlipidemia (68.3%), and ischemic heart disease (38.7%). A majority of patients (65.3%) had COPD of mixed etiology. Nearly half of patients (46.5%) did not experience COPD exacerbations during the preindex period, whereas 13.5% experienced 1 mild exacerbation and 12.8% had more than 1 mild exacerbation. Moderate and severe exacerbations were experienced by 14.1% and 13.1% of patients, respectively. Antibiotics were prescribed in 50.8% of patients, and 50.2% were prescribed oral corticosteroids. Of the COPD rescue or maintenance medications, the most frequently used were SABA (62.8%), LABA + ICS (56.9%), LAMA (55.3%), and systemic corticosteroids (49.8%). Antidiabetic medications were used by 21.0% of patients, and 67.1% used cardiovascular medications.

Baseline all-cause and COPD-related HRU of patients with COPD receiving triple therapy is presented in Table 4. Of these, 41.7% experienced hospitalizations, 98.8% experienced outpatient visits, and 51.7% experienced ED visits. Further, 54.6% had spirometry assessments and 37.8% received oxygen therapy. Counting only utilization due to COPD, 22.3% experienced hospitalizations, 72.9% experienced outpatient visits, and 17.4% experienced ED visits.

Table 5 reports baseline demographic characteristics of patients with no exacerbations or 1 mild exacerbation (ie, GOLD groups A and B) (n = 11,790) and of patients experiencing more than 1 mild exacerbation or any moderate/severe exacerbations (ie, GOLD groups C and D) (n = 7855). The mean (SD) age of patients in GOLD groups A and B was 70.45 (8.46) years, and 47.9% were women. A majority (65.1%) resided in the geographic South. These characteristics were very similar for patients in GOLD groups C and D: The mean (SD) age of these patients was 71.09 (8.28) years, and 50.7% were women. A majority (64.8%) of them resided in the geographic South.

DISCUSSION

In a predominantly Medicare Advantage population, this study showed that among patients with a diagnosis of COPD and no concomitant diagnosis of asthma, cystic fibrosis, or lung cancer, 6.9% received triple therapy during the period of January 2012 through December 2014. Among the subset of patients with COPD who were treated with any maintenance medication, the proportion receiving triple therapy was 17.1%. These estimates were slightly lower than those reported by Simeone et al, which were 7.5% and 25.5%, respectively.5 Their study sample was drawn from the 2009-2013 IMS PharMetrics Plus database, which was composed of administrative claims for mostly commercial plan enrollees, who had a mean age of 65 years. As expected with a largely Medicare population, the average age of our study cohorts was higher, and multiple comorbid conditions were common. HRU was also higher than has been reported elsewhere (ie, Simeone et al), especially for hospitalization and ED encounters. Spirometry testing and oxygen therapy were also more frequently observed in the present study than in that of Simeone et al.5 With a mean age approximately 6 years older, it is likely that the increased utilization seen in the present study reflects the differing age profile in this study population compared with that of Simeone et al.5 Health plan characteristics and geographic variations may also have contributed to these differences.

According to 2017 GOLD guidelines, triple therapy is indicated only for patients with very severe COPD (GOLD group D). These are patients who (1) experience multiple exacerbations or at least 1 exacerbation that leads to hospitalization and (2) score 2 or more on the modified Medical Research Council breathlessness scale or above 10 on the COPD Assessment Test. In this study, GOLD groups were approximated using baseline exacerbation frequency and severity. However, as data for this study did not include symptomology information, patients categorized in group A could not be further differentiated from group B, and similarly patients in group C could not be distinguished from group D. Among patients treated with triple therapy, only 13% were classified as having experienced a severe COPD exacerbation. Based on claims data, 60% of patients receiving triple therapy in this study did not experience any exacerbations or experienced only 1 mild exacerbation that did not result in a hospitalization, prior to initiation of triple therapy. This suggests that, according to the GOLD guidelines, these patients may not have been appropriate candidates for triple therapy. It is worth noting that because we could not distinguish GOLD group C from group D because of the lack of patient symptomology information, likely not all patients in the remaining 40% should have progressed to triple therapy.

The present study supports findings from previous research that have reported potential overuse of triple therapy among patients with COPD for whom current GOLD guidelines do not recommend the triple therapy regimen.3-5 It is important to note that this evidence is based solely on administrative data. Clinical assessment and symptom evaluation at the point of service could very well have justified triple therapy treatment based on clinician knowledge and other information not captured in claims, such as self-management of acute episodes with antibiotics and oral corticosteroids. Nonetheless, attention and further investigation of the appropriate use of triple therapy in the context of COPD management is warranted.

It should be noted that this study was conceived at the time of the 2016 GOLD guidelines. Since then, the guidelines have been updated annually, with the most recent release of 2019 updates. While largely maintaining the ABCD assessment for initial therapy, newer guidelines offer recommendations for therapy adjustments that are no longer dependent on the ABCD assessment at treatment initiation.10 Our study contributes to the literature by providing real-world data on ABCD grouping at the time of triple therapy initiation and assessing the appropriateness of such therapies against GOLD recommendations. In light of the newer guidelines, future research examining progression to triple therapy and therapy adjustments in follow-up care would be fruitful.

Limitations

There are certain limitations associated with this study that should be considered when interpreting the study findings. The results of this study were based on administrative claims data of individuals enrolled with Humana; therefore, they may not be generalizable to the US population due to variations in geographic coverage and/or penetration. Retrospective studies using administrative claims are prone to coding errors and/or incomplete claims information, which may lead to misclassification in the variables of interest. Estimates of triple therapy initiation in cohort 1 may have been influenced by formulary variations and by requiring 24 months of postdiagnosis survival. In the absence of medical records, this study relied on a claims-based algorithm to classify COPD exacerbation. Although the algorithm was used in a previous study,5 it may not fully capture the information obtained from a clinical assessment. Preindex prescription of antibiotics and oral corticosteroids may have indicated a patient self-management strategy intended to enable patients to respond to acute episodes without triggering a claim record of the event. In such cases, and those in which antibiotics may have been obtained via “cash-only” pharmacy discount programs that also did not generate a claim record, the full extent of exacerbations may have been underestimated.

CONCLUSIONS

Using administrative claims data from a Humana database, this study showed that 6.9% of all patients with COPD and 17.1% of patients with COPD treated with any maintenance therapy received triple therapy. Furthermore, 60% of patients who received triple therapy either did not experience an exacerbation based on claims data or had only 1 mild exacerbation that did not result in hospitalization during the study period (ie, GOLD groups A and B). Although the full extent of illness severity may have been underestimated, this study’s findings suggest that the real-world use of triple therapy may not be in concordance with GOLD guidelines, possibly resulting in avoidable risk exposure and increased HRU burden. Due to the limitation of administrative claims data in the assessment of patients’ symptom burden, further studies using alternative data sources are required to categorize patients receiving triple therapy into the respective GOLD groups and to better assess concordance with GOLD guidelines in this population.Author Affiliations: Humana Healthcare Research Inc (YL, SS), Louisville, KY; Boehringer-Ingelheim Pharmaceuticals Inc (JL, SK, AS), Ridgefield, CT; Humana Inc (AR), Louisville, KY.

Source of Funding: Boehringer-Ingelheim Pharmaceuticals Inc.

Author Disclosures: Drs Li and Stemkowski are employed by Humana Healthcare Research. Mr Lim and Drs Kaila and Shaikh are employed by Boehringer-Ingelheim Pharmaceuticals, which manufactures several drugs for the treatment of chronic obstructive pulmonary disease. Dr Renda reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (YL, JL, SS, SK, AR, AS); acquisition of data (YL, SS); analysis and interpretation of data (YL, JL, SS, SK, AR, AS); drafting of the manuscript (YL, JL, SS, AR, AS); critical revision of the manuscript for important intellectual content (YL, JL, SS, SK, AR, AS); statistical analysis (YL, SS); obtaining funding (JL, SK); administrative, technical, or logistic support (JL, SS); and supervision (YL, JL, SK).

Address Correspondence to: Yong Li, PhD, Humana Healthcare Research Inc, 515 W Market St, Louisville, KY 40202. Email: yli4@humana.com.REFERENCES

1. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: 2017 report. Global Initiative for Chronic Obstructive Lung Disease website. goldcopd.org/wp-content/uploads/2017/02/wms-GOLD-2017-FINAL.pdf. Published 2017. Accessed September 20, 2018.

2. Patel AR, Patel AR, Singh S, Singh S, Khawaja I. Global Initiative for Chronic Obstructive Lung Disease: the changes made. Cureus. 2019;11(6):e4985. doi: 10.7759/cureus.4985.

3. Price D, West D, Brusselle G, et al. Management of COPD in the UK primary-care setting: an analysis of real-life prescribing patterns. Int J Chron Obstruct Pulmon Dis. 2014;9:889-904. doi: 10.2147/COPD.S62750.

4. Asche CV, Leader S, Plauschinat C, et al. Adherence to current guidelines for chronic obstructive pulmonary disease (COPD) among patients treated with combination of long-acting bronchodilators or inhaled corticosteroids. Int J Chron Obstruct Pulmon Dis. 2012;7:201-209. doi: 10.2147/COPD.S25805.

5. Simeone JC, Luthra R, Kaila S, et al. Initiation of triple therapy maintenance treatment among patients with COPD in the US. Int J Chron Obstruct Pulmon Dis. 2016;12:73-83. doi: 10.2147/COPD.S122013.

6. Mannino DM, Yu TC, Zhou H, Higuchi K. Effects of GOLD-adherent prescribing on COPD symptom burden, exacerbations, and health care utilization in a real-world setting. Chronic Obstr Pulm Dis. 2015;2(3):223-235. doi: 10.15326/jcopdf.2.3.2014.0151.

7. Lipson DA, Barnhart F, Brealey N, et al; IMPACT Investigators. Once-daily single-inhaler triple versus dual therapy in patients with COPD. N Engl J Med. 2018;378(18):1671-1680. doi: 10.1056/NEJMoa1713901.

8. Make B, Dutro MP, Paulose-Ram R, Marton JP, Mapel DW. Undertreatment of COPD: a retrospective analysis of US managed care and Medicare patients. Int J Chron Obstruct Pulmon Dis. 2012;7:1-9. doi: 10.2147/COPD.S27032.

9. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613-619. doi: 10.1016/0895-4356(92)90133-8.

10. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: 2019 report. Global Initiative for Chronic Obstructive Lung Disease website. goldcopd.org/wp-content/uploads/2018/11/GOLD-2019-v1.7-FINAL-14Nov2018-WMS.pdf. Published 2019. Accessed March 22, 2019.