COPD Treatment Ratio: A Measure for Improving COPD Population Health

ABSTRACT

Objectives: Despite chronic obstructive pulmonary disease (COPD) being a leading cause of death in the US, there are few COPD measures in current quality programs. The objective of this study was to assess the validity and applicability of the COPD treatment ratio (CTR) as a surrogate marker of COPD exacerbation risk for use in quality measurement. CTR is defined as the ratio of COPD maintenance medications to all COPD medications (maintenance and rescue).

Study Design: This retrospective cohort study used 2016-2019 administrative claims from Optum Clinformatics Data Mart to evaluate CTR values over a 12-month baseline period, with exacerbations measured the following year. Patients 40 years or older with Medicare Advantage or commercial insurance and with a COPD diagnosis were included.

Methods: Logistic regression models were used to examine relationships between CTR values and COPD exacerbations. Prediction model performance was evaluated using C statistics, and receiver operating characteristics were used to determine the optimal cut point for CTR.

Results: Of 132,960 patients included in the analysis, 79.5% were Medicare Advantage beneficiaries, and the mean age was 69.6 years. Higher CTR values were significantly associated with reduced risk of any, moderate, and severe exacerbations in the total population and when stratified by insurance type. CTR performed fairly to moderately well in predicting COPD exacerbations. The optimal cut point for COPD exacerbation prediction was 0.7.

Conclusions: Study results substantiated CTR as a valid measure of COPD exacerbation risk and support the use of CTR in quality improvement to drive evidence-based care for individuals with COPD.

Am J Manag Care. 2024;30(9):In Press

Takeaway Points

Our results support the validity of the chronic obstructive pulmonary disease (COPD) treatment ratio (CTR) as a predictor of COPD exacerbation risk within commercial and Medicare Advantage lines of business.

• The use of CTR in population health and quality initiatives could decrease costs for health plans and patients, support health outcomes, improve medication management, and increase the use of evidence-based practices in COPD care.
• CTR is a suitable metric to monitor progress on national and global COPD policies, action plans, and initiatives.
• Our results add to the evidence base supporting the development of a CTR quality measure that can be implemented to advance COPD value-based care programs.

Chronic obstructive pulmonary disease (COPD) is a serious public health issue affecting millions of people globally. Despite being the third leading cause of mortality worldwide, COPD is underdiagnosed and undertreated.^1,2 COPD is characterized by exacerbations (ie, acute worsening of respiratory symptoms, or flare-ups) leading to a decline in lung function, hospitalization, impaired quality of life, long-term disability, and death.

In 2018, COPD was the fourth leading cause of death in the US, with 38.2 deaths per 100,000 population.^3,4 In 2019, it accounted for approximately 163.3 hospitalizations and 401.9 emergency department (ED) visits per 100,000 population.⁵ Annual costs attributed to COPD have been estimated at $50 billion, with $29.5 billion in direct medical costs.⁵ Notably, the majority (85%) of COPD direct medical costs are due to exacerbations and subsequent complications, events that could be reduced through optimal medication management.⁶

The Global Initiative for Chronic Obstructive Lung Disease (GOLD) was initiated in 1998 to provide recommendations for the management of COPD based on the best scientific information available. According to the 2024 GOLD report, pharmacologic therapy of COPD can reduce symptoms and the frequency and severity of exacerbations and provide beneficial effects on lung function decline and mortality.⁷

Quality measures are a vital component of value-based payment models, but there are few COPD measures in current US quality reporting programs, despite the established burden of COPD. The COPD treatment ratio (CTR) is a promising measure of COPD exacerbation risk for medication use quality assessment. Prior studies showed CTR to be a reliable predictor of moderate or severe exacerbation risk using several sources of claims data.^8-10 CTR is defined as the ratio of COPD maintenance medications to all COPD medications (maintenance and rescue). Maintenance medications used on a routine basis are a critical part of a COPD treatment regimen to reduce flare-ups and the need for rescue inhaler use. In contrast, rescue medications are recommended for use as needed by patients to alleviate acute respiratory symptoms (eg, dyspnea).¹¹

CTR can be used to inform public health and quality improvement efforts by identifying patients who may benefit from interventions to optimize COPD medication use.^12,13 Prior studies developed exacerbation risk models or relied on exacerbation history, both of which use medical record data not easily accessible for population-based measurement.¹⁴ CTR, in contrast, is an easily calculable and interpretable measure using prescription claims data.¹⁵ Incremental increases in CTR have been associated with reductions in the risk of severe COPD exacerbations among commercial and Medicare beneficiaries.^8-10 However, CTR evidence has not been stratified by insurance type in past studies. Given that measures used in US quality programs require supporting evidence relevant to specific populations of interest, stratification of insurance types is important to support real-world implementation for performance measurement and improvement. Therefore, this study aimed to further test and validate CTR as a measure of moderate and severe COPD exacerbation risk, stratified by commercial and Medicare populations.

METHODS

Study Objectives

The primary objective of this study was to validate the CTR as a measure of COPD exacerbation risk separately within Medicare Advantage and commercially insured populations. The secondary objective was to examine optimal CTR values in subcohort sensitivity analyses to further evaluate CTR as a predictor of COPD exacerbation risk.

Study Design

This was a retrospective cohort study using Optum Clinformatics Data Mart research data sets, with data from January 1, 2016, to December 31, 2019, inclusive of commercial and Medicare Advantage patients with a COPD diagnosis.

Patients were followed for 2 years after their first diagnosis of COPD (index date) within the identification period (January 1, 2016-December 31, 2017). Criteria for COPD diagnosis are described in the study population section. Patient risk factors were assessed during the 12-month baseline period starting at the index date, and moderate and severe COPD exacerbations were assessed in the following 12-month follow-up period (Figure 1).

Study Population

This study included patients 40 years and older who had at least 1 inpatient hospitalization with a COPD diagnosis or at least 2 outpatient visits, which included physician, urgent care, or ED visits, with a COPD diagnosis between January 1, 2016, and December 31, 2017. These inclusion criteria align with the CMS Chronic Conditions Data Warehouse.16 The second outpatient visit must have occurred within 365 days of the first outpatient visit. Among those with a COPD diagnosis, patients were included in the study if they had at least 1 prescription claim for a maintenance COPD medication during the 12-month baseline period. Patients were required to be continuously enrolled in prescription drug and medical coverage for 2 years following their index date. Patients with any diagnosis of cancer during the baseline period were excluded (Figure 2).

COPD diagnoses were identified using International Statistical Classification of Diseases, Tenth Revision (ICD-10) codes and International Classification of Diseases, Ninth Revision, Clinical Modification codes that were converted to ICD-10 codes using general equivalence mappings and also manually reviewed by specialists for accuracy (eAppendix Table 1 [eAppendix available at ajmc.com]).

Data Source

The Optum Clinformatics Data Mart is an administrative health claims database consisting of data from commercial and Medicare Advantage health plans. The data include member enrollment, medical claims, and pharmacy claims from 2016 to 2019. The data are statistically deidentified under the Expert Determination method consistent with the Health Insurance Portability and Accountability Act. Institutional review board approval was not obtained because this study did not involve human participants and used a deidentified data set.

CTR Calculation

CTR was calculated as the total units of maintenance medications appearing on prescription claims during a 12-month period divided by the total units of maintenance and rescue medications appearing on prescription claims during the same 12-month period. Maintenance medications were defined as inhaled corticosteroids, long-acting β₂-adrenergic agonists, long-acting muscarinic antagonists, long-acting fixed-dose combinations, phosphodiesterase-4 inhibitors, and methylxanthines. Rescue medications were defined as short-acting β₂-adrenergic agonists, short-acting muscarinic antagonists, and short-acting fixed-dose combinations.

A unit of maintenance or rescue medication was defined as up to a 30-day supply for oral medications, a single canister for nonnebulized inhaled medications, and the adjusted number of ampules or unit doses for inhaled nebulized medications (based on the formula provided below). Oral medication units were calculated using the days’ supply field of the prescription claim (1 oral medication unit equals ≤ 30-day supply of medication). Inhaled medication units were calculated using the quantity dispensed field of the prescription claim and information regarding the product package sizes and package quantities that describe the volume and/or number of unit doses (1 inhaled medication unit = quantity dispensed / package size / package quantity). This formula applied to both nebulized and nonnebulized medications.

Outcomes

The outcomes of interest in this study were the occurrence of any, moderate, or severe COPD exacerbation during the 1-year follow-up period. A severe COPD exacerbation was defined as an inpatient hospitalization with either a primary diagnosis of COPD or a secondary diagnosis of COPD with a primary diagnosis of respiratory failure (ICD-10 codes in eAppendix Table 2). Moderate exacerbations were defined as outpatient visits or ED visits that did not result in a hospital admission with either primary diagnosis of COPD or a secondary diagnosis of COPD with a primary diagnosis of respiratory failure and a dispensing for an oral corticosteroid within 7 days of the visit.^12,17 Any exacerbation was also used as an end point, defined as any moderate or severe COPD exacerbation.

Covariates

Demographic and clinical risk factors used in previous CTR studies were included in the adjusted model and assessed during the baseline period.8 These characteristics have been shown to be associated with the occurrence of a COPD exacerbation. Demographic characteristics included age at index date. Clinical risk factors included a measure of disease burden (presence of comorbidities), concomitant medication use (β-blockers, diuretics, statins, antidepressants), COPD medication use (oral corticosteroids, theophylline), and occurrence of moderate or severe baseline exacerbations.

Sensitivity Analyses

Patients with a history of asthma or other severe lung disease were anticipated to have different treatment and exacerbation patterns compared with the general COPD population and thus to have a potentially different relationship between CTR and exacerbations. Sensitivity analyses were completed to determine whether subpopulations with a history of asthma or other severe lung disease should be excluded when using CTR as an indicator of COPD exacerbation risk. The first analysis evaluated the sensitivity of the optimal CTR cut point in a subpopulation without a diagnosis of asthma (ICD-10 codes in eAppendix Table 3) at baseline. A second sensitivity analysis examined the frequency of severe lung conditions in the baseline period and evaluated mean CTR values in a subpopulation without a diagnosis of severe lung disease (ICD-10 codes in eAppendix Table 4).

Statistical Analysis

Descriptive statistics were assessed by insurance type, comparing patients with commercial and Medicare Advantage insurance during the baseline period using Wilcoxon tests for continuous measures and χ² tests for categorical measures. Statistical significance was assessed at a P value less than .05.

Adjusted and unadjusted logistic regression models were used to examine the relationship between dichotomized CTR values (eg, ≥ 0.1, ≥ 0.2, ≥ 0.3) and COPD exacerbations in the total population and by insurance type. Separate logistic regression models were used for each type of exacerbation as the outcome variable: none vs moderate exacerbation, none vs severe exacerbation, and none vs any exacerbation. All associations are presented as an OR or adjusted OR (AOR) with a 95% CI.

Performance of the prediction models was evaluated using the C statistic, a goodness-of-fit measure used to assess the ability of a risk factor to predict an outcome in logistic regression models. Receiver operating characteristic (ROC) curves and the Youden index were used to determine the optimal CTR value by determining the point at which the sensitivity and the specificity of the CTR measure in predicting an exacerbation were optimized. Analyses were performed within the full cohort and stratified by insurance type (commercial, Medicare Advantage). All statistical analyses were conducted using SAS 9.4 (SAS Institute Inc).

RESULTS

A total of 132,960 patients were included in the eligible study population, of whom 27,280 (20.5%) were enrolled in commercial insurance and 105,680 (79.5%) were enrolled in Medicare Advantage (Table). The mean (SD) age of the patients was 69.6 (9.8) years, and 60.8% of patients were women. During the baseline period, 15.0% (n = 19,996) experienced a moderate or severe exacerbation. Medicare Advantage beneficiaries were older and had more exacerbations in both the baseline and follow-up periods, more comorbidities, and more concomitant medications than patients with commercial insurance.

Model fit statistics indicated that unadjusted CTR performed fairly in predicting moderate exacerbations in the commercial and Medicare populations, with C statistics of 0.588 and 0.584, respectively. Model fit was slightly better in predicting severe exacerbations, with C statistics of 0.632 in the commercial population and 0.619 in the Medicare Advantage population. The adjusted model fit statistics indicated better performance in predicting both moderate and severe exacerbations than the unadjusted model in both populations, with C statistics ranging from 0.658 to 0.757.

The ROC analysis of the dichotomized CTR values suggested 0.8 and 0.7 as the optimal cut points for all 3 outcomes (any, moderate, or severe exacerbation) in the commercial insurance population and the Medicare Advantage population, respectively (Figure 3). In the total population, the optimal CTR value for any exacerbation and severe exacerbation was 0.7, and for moderate exacerbation was 0.8. The plurality of ROC analysis results indicated an optimal prediction value of 0.7, which was selected for further evaluation.

In the total population, comparing patients with a CTR of at least 0.7 vs patients with a CTR less than 0.7, the OR of any exacerbation was 0.56 (95% CI, 0.54-0.58), the OR of moderate exacerbation was 0.61 (95% CI, 0.59-0.63), and the OR of severe exacerbation was 0.47 (95% CI, 0.45-0.49) in the unadjusted model (Figure 4). In the adjusted model in the total population, a CTR of at least 0.7 was associated with a 25.0% reduction in any exacerbation (AOR, 0.75; 95% CI, 0.73-0.75), a 21.6% reduction in moderate exacerbation (AOR, 0.78; 95% CI, 0.76-0.81), and a 32.0% reduction in severe exacerbation (AOR, 0.68; 95% CI, 0.65-0.71) compared with those with a CTR less than 0.7. Similar results were found by insurance type in the unadjusted and adjusted models (Figure 4).

A sensitivity analysis within the subpopulation without asthma found an optimal cut point of 0.8 for any, moderate, and severe exacerbations within the total population and similar C statistics for the unadjusted and adjusted models in the study population (eAppendix Table 5). Additionally, the subpopulation with severe lung disease had a similar mean CTR value to the total study population (eAppendix Table 6).

DISCUSSION

CTR performed well in this study in predicting any, moderate, and severe exacerbation risk. These results support the validity of CTR as a predictor of COPD exacerbation risk for both commercial and Medicare Advantage populations. Sensitivity analyses and stratification by insurance type, which better reflect real-world implementation of quality measures, represent a valuable addition to the body of knowledge surrounding CTR as a predictor of COPD exacerbation risk.

The model fit in this study was similar to that derived by Stanford et al in 2019, which used Humana’s Medicare and commercial database but did not stratify results by insurance type.8 The optimal cut point of 0.7 in our study was also consistent with the findings of Stanford et al. However, our study found that patients with a CTR of at least 0.7 had considerably reduced odds of COPD exacerbations, with a reduction of 25%, 22%, and 32% in any, moderate, and severe exacerbations, respectively, compared with those with a CTR less than 0.7. In comparison, Stanford et al found a modest reduction of 8% in severe exacerbations in those with a CTR of at least 0.7 and did not find a significant reduction in any or moderate exacerbations.⁸ These differences in reduced likelihood of an exacerbation may be explained by the prior study’s inclusion of patients without any maintenance medications at baseline and exclusion of patients with asthma. However, consistent with prior studies, we found that CTR performed best in predicting severe exacerbations.

Approximately 40% of our study population did not have a claim for maintenance medications during the baseline period. Although prior studies included this subpopulation in their eligible study samples,^8,9,15 the CTR value for these patients was, by definition, zero. We therefore diverged from prior studies to exclude this subpopulation for assessing CTR cut points. Our finding that a large portion of the study population with an eligible COPD diagnosis had no claims for a maintenance medication indicates a potential opportunity to align with COPD clinical guideline treatment recommendations more closely. Although some patients in our study may have recently received a diagnosis of COPD and thus did not have an opportunity to fill a maintenance medication, poor patient adherence to treatment for COPD is not uncommon.¹⁸ Approximately 40% to 60% of patients with COPD adhere to their prescribed regimen,^19,20 and nonadherence to COPD medications is associated with increased rates of morbidity and health care expenditures, largely due to exacerbations.^21-23 The complexity of COPD treatment likely contributes to nonadherence because it requires patients to have knowledge about the disease, various medications, and devices.¹⁹ Furthermore, high inhaler out-of-pocket costs are associated with COPD medication nonadherence, and patients may opt for lower-cost rescue inhalers in lieu of higher-cost inhalers for long-term symptom control.²⁴ As a result, insurance benefit design that reduces patient out-of-pocket costs or improves patient education may be useful to increase use of controller medications, improve CTR, and drive decreases in exacerbations and overall costs.

Our study also confirmed results from a prior study that examined the association between CTR and exacerbations in an unadjusted model.¹⁰ Although the risk factors in previous studies were carefully chosen to include important factors associated with COPD exacerbation, quality measures are frequently calculated as unadjusted measures. Furthermore, although the model fit was modest in the unadjusted model, CTR remained a robust predictor, and a CTR value above 0.7 was associated with significant reduction in any, moderate, and severe exacerbation risk. This study confirmed that CTR is an independent predictor of COPD exacerbation, although factors included in the adjusted model, such as a patient’s previous history of exacerbation, strongly predicted the occurrence of future exacerbations.^25,26

Existing COPD quality measures focus on pharmacotherapy after a COPD exacerbation or use of spirometry, but not on prevention of exacerbations.^27,28 The CTR provides a benchmark for COPD care, allowing health care providers for commercial and Medicare Advantage populations to monitor and improve the quality of care they provide to ensure medication therapies are optimized to reduce exacerbation risk. By identifying and validating an optimal CTR value, our findings add to the evidence base for a population-level (eg, health plan) quality measure to examine the proportion of patients meeting the optimal CTR cut point of 0.7. For individuals with COPD who are not receiving evidence-based care, this measure would help inform targeted interventions to improve quality of care and patient outcomes.

Further research on CTR may focus on specific subpopulations of interest to better target groups that may benefit most from intervention. The interplay between sociodemographic variables and CTR may also have important implications for health equity.

Limitations

The first study limitation is that the definition of COPD and algorithms used to identify exacerbation events were designed based on medical claims data and ICD-10 coding that may be incomplete or contain errors. This may have led to misclassification of COPD diagnoses and outcome variables. However, the methods used in this study to identify patients with COPD and exacerbations in administrative data are highly specific in a US managed care population and commonly used in research analyzing administrative claims data.^29,30 Second, although prescription claims are often used as a reliable source of information to evaluate medication utilization,³¹ this study does not provide direct evidence that medication dispensed was taken by a patient. Third, we note that the severe exacerbation C statistic may be overestimated by nature of comparing the severe exacerbation group vs the neither moderate nor severe group (which inherently excludes the moderate exacerbation group). Finally, the CTR measure and 0.7 cut point are appropriate for population-level efforts to improve medication therapy quality for patients with COPD but should not be used for individual clinical decisions.

CONCLUSIONS

These study results support CTR as a suitable tool to monitor the quality of medication use for managing COPD. Use of CTR in population health quality initiatives could increase use of evidence-based medication management of COPD and ultimately support improved health outcomes. Given existing disparities in COPD prevalence, treatment, and outcomes,^32,33 additional work may need to be done to assess whether this measure functions differently within various subgroups defined by characteristics such as race/ethnicity, sex, and dual-eligibility status. 

Acknowledgments

The authors thank Lynn Pezzullo and Joel Montavon for their contributions.

Author Affiliations: Pharmacy Quality Alliance (MAP, MHG, BES, PJC), Alexandria, VA; now with Merck & Co, Inc (PJC), Rahway, NJ; University of North Carolina Eshelman School of Pharmacy (SCB), Chapel Hill, NC.

Source of Funding: This study was funded by GSK. The study sponsor funded all aspects of the study, except manuscript preparation and submission. The sponsor was not involved in the study design, data analysis, data interpretation, or writing of the report, and the decision to submit the manuscript for publication.

Author Disclosures: Dr Parikh, Dr Gabriel, Mr Shirley, and Dr Campbell completed this work as employees of the Pharmacy Quality Alliance, which received funding from GSK for this study. Dr Campbell is now an employee of Merck & Co, Inc. Dr Burbage 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 (MAP, PJC); acquisition of data (MAP); analysis and interpretation of data (MHG, BES, PJC); drafting of the manuscript (MAP, SCB, BES); critical revision of the manuscript for important intellectual content (MAP, SCB, MHG, PJC); statistical analysis (MAP, MHG); obtaining funding (PJC); administrative, technical, or logistic support (SCB, BES, PJC); and supervision (PJC).

Address Correspondence to: Ben E. Shirley, BSPH, Pharmacy Quality Alliance, 5911 Kingstowne Village Pkwy, Ste 130, Alexandria, VA 22315. Email: bshirley@pqaalliance.org.

REFERENCES

1. Chronic obstructive pulmonary disease (COPD). World Health Organization. Updated March 16, 2023. Accessed December 22, 2022. https://www.who.int/news-room/fact-sheets/detail/chronic-obstructive-pulmonary-disease-(copd)

2. Lamprecht B, Soriano JB, Studnicka M, et al; BOLD Collaborative Research Group, the EPI-SCAN Team, the PLATINO Team, and the PREPOCOL Study Group. Determinants of underdiagnosis of COPD in national and international surveys. Chest. 2015;148(4):971-985. doi:10.1378/chest.14-2535

3. Kochanek KD, Xu J, Arias E. Mortality in the United States, 2019. NCHS Data Brief. 2020;(395):1-8.

4. Wheaton AG, Cunningham TJ, Ford ES, Croft JB; CDC. Employment and activity limitations among adults with chronic obstructive pulmonary disease—United States, 2013. MMWR Morb Mortal Wkly Rep. 2015;64(11):289-295.

5. COPD trends brief: burden. American Lung Association. Accessed November 28, 2023. https://www.lung.org/research/trends-in-lung-disease/copd-trends-brief/copd-burden

6. Blanchette CM, Gross NJ, Altman P. Rising costs of COPD and the potential for maintenance therapy to slow the trend. Am Health Drug Benefits. 2014;7(2):98-106.

7. Global Strategy for the Diagnosis, Management, and Prevention of Chronic Obstructive Pulmonary Disease (2024 Report). Global Initiative for Chronic Obstructive Lung Disease; 2023. Accessed November 28, 2023. https://goldcopd.org/wp-content/uploads/2023/11/GOLD-2024_v1.0-30Oct23_WMV.pdf

8. Stanford RH, Lau MS, Li Y, Stemkowski S. External validation of a COPD risk measure in a commercial and Medicare population: the COPD treatment ratio. J Manag Care Spec Pharm. 2019;25(1):58-69. doi:10.18553/jmcp.2019.25.1.058

9. Stanford RH, Nag A, Mapel DW, et al. Claims-based risk model for first severe COPD exacerbation. Am J Manag Care. 2018;24(2):e45-e53.

10. Stanford RH, Korrer S, Brekke L, Reinsch T, Bengtson LGS. Validation and assessment of the COPD treatment ratio as a predictor of severe exacerbations. Chronic Obstr Pulm Dis. 2020;7(1):38-48. doi:10.15326/jcopdf.7.1.2019.0132

11. COPD: treatment. National Heart, Lung, and Blood Institute. Updated October 25, 2023. Accessed January 24, 2023. https://www.nhlbi.nih.gov/health/copd/treatment

12. Mapel DW, Schum M, Lydick E, Marton JP. A new method for examining the cost savings of reducing COPD exacerbations. Pharmacoeconomics. 2010;28(9):733-749. doi:10.2165/11535600-000000000-00000

13. French J. Making merit-based payment meaningful: the next step in healthcare quality measurement. Healthcare Information and Management Systems Society. April 21, 2021. Accessed June 1, 2021. https://www.himss.org/resources/making-merit-based-payment-meaningful-next-step-healthcare-quality-measurement

14. Bertens LCM, Reitsma JB, Moons KGM, et al. Development and validation of a model to predict the risk of exacerbations in chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis. 2013;8:493-499. doi:10.2147/COPD.S49609

15. Stanford RH, Nag A, Mapel DW, et al. Validation of a new risk measure for chronic obstructive pulmonary disease exacerbation using health insurance claims data. Ann Am Thorac Soc. 2016;13(7):1067-1075. doi:10.1513/AnnalsATS.201508-493OC

16. Chronic obstructive pulmonary disease and bronchiectasis. Chronic Conditions Data Warehouse. Updated February 2022. Accessed September 20, 2023. https://web.archive.org/web/20210805194410/https:/www2.ccwdata.org/documents/10280/19139608/ccw-cond-algo-copd.pdf

17. Abudagga A, Sun SX, Tan H, Solem CT. Exacerbations among chronic bronchitis patients treated with maintenance medications from a US managed care population: an administrative claims data analysis. Int J Chron Obstruct Pulmon Dis. 2013;8:175-185. doi:10.2147/COPD.S40437

18. Bourbeau J, Bartlett SJ. Patient adherence in COPD. Thorax. 2008;63(9):831-838. doi:10.1136/thx.2007.086041

19. Restrepo RD, Alvarez MT, Wittnebel LD, et al. Medication adherence issues in patients treated for COPD. Int J Chron Obstruct Pulmon Dis. 2008;3(3):371-384. doi:10.2147/copd.s3036

20. Ágh T, Inotai A, Mészáros A. Factors associated with medication adherence in patients with chronic obstructive pulmonary disease. Respiration. 2011;82(4):328-334. doi:10.1159/000324453

21. Cote C. Pharmacoeconomics and the burden of chronic obstructive pulmonary disease. Clin Pulm Med. 2005;12(4):S19-S21. doi:10.1097/01.cpm.0000170110.31441.29

22. Mäkelä MJ, Backer V, Hedegaard M, Larsson K. Adherence to inhaled therapies, health outcomes and costs in patients with asthma and COPD. Respir Med. 2013;107(10):1481-1490. doi:10.1016/j.rmed.2013.04.005

23. Davis JR, Wu B, Kern DM, et al. Impact of nonadherence to inhaled corticosteroid/LABA therapy on COPD exacerbation rates and healthcare costs in a commercially insured US population. Am Health Drug Benefits. 2017;10(2):92-102.

24. Castaldi PJ, Rogers WH, Safran DG, Wilson IB. Inhaler costs and medication nonadherence among seniors with chronic pulmonary disease. Chest. 2010;138(3):614-620. doi:10.1378/chest.09-3031

25. Al-ani S, Spigt M, Hofset P, Melbye H. Predictors of exacerbations of asthma and COPD during one year in primary care. Fam Pract. 2013;30(6):621-628. doi:10.1093/fampra/cmt055

26. Montserrat-Capdevila J, Godoy P, Marsal JR, Barbé F. Predictive model of hospital admission for COPD exacerbation. Respir Care. 2015;60(9):1288-1294. doi:10.4187/respcare.04005

27. HEDIS and performance measurement. National Committee for Quality Assurance. Accessed February 1, 2023. https://www.ncqa.org/hedis/

28. Quality measures: traditional MIPS requirements. Quality Payment Program. Accessed February 1, 2023. https://qpp.cms.gov/mips/quality-requirements

29. Gothe H, Rajsic S, Vukicevic D, et al. Algorithms to identify COPD in health systems with and without access to ICD coding: a systematic review. BMC Health Serv Res. 2019;19(1):737. doi:10.1186/s12913-019-4574-3

30. Annavarapu S, Goldfarb S, Gelb M, Moretz C, Renda A, Kaila S. Development and validation of a predictive model to identify patients at risk of severe COPD exacerbations using administrative claims data. Int J Chron Obstruct Pulmon Dis. 2018;13:2121-2130. doi:10.2147/COPD.S155773

31. Sikka R, Xia F, Aubert RE. Estimating medication persistency using administrative claims data. Am J Manag Care. 2005;11(7):449-457.

32. Pleasants RA, Riley IL, Mannino DM. Defining and targeting health disparities in chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis. 2016;11:2475-2496. doi:10.2147/COPD.S79077

33. Eisner MD, Blanc PD, Omachi TA, et al. Socioeconomic status, race and COPD health outcomes. J Epidemiol Community Health. 2011;65(1):26-34. doi:10.1136/jech.2009.089722