SEARCH

SEARCH BY CITATION

Keywords:

  • depression;
  • chronic obstructive pulmonary disease;
  • maintenance medication adherence;
  • maintenance medication;
  • Medicare beneficiaries

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

Objective

Depression is a significant comorbidity in patients with chronic obstructive pulmonary disease (COPD). Although comorbid depression is associated with low use and poor adherence to medications treating other chronic conditions, evidence of the relationship between depression and COPD management is limited. This study estimated the association between depression and COPD maintenance medication (MM) adherence among patients with COPD.

Methods

This cross-sectional study used a 5% random sample of 2006–2007 Chronic Condition Warehouse data. Medicare beneficiaries enrolled in Parts A, B, and D plans with diagnosed COPD who survived through 2006 were included (n = 74,863). COPD MM adherence was measured as medication discontinuation and proportion of days covered (PDC). Depression was identified through the International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis codes. Multivariable models with modified generalized estimating equations were used to estimate adjusted association between depression diagnosis and medication adherence, controlling for sociodemographics, comorbidities, and disease severity.

Results

Among the sample, about one third (33.6%) had diagnosed depression. More than half (61.8%) of beneficiaries with COPD filled at least one COPD MM prescription. Depressed beneficiaries had a higher likelihood of using COPD MM than non-depressed beneficiaries (adjusted prevalence ratios [PR] = 1.02; 95% confidence intervals [CI] = 1.01, 1.03). Among COPD MM users, depressed beneficiaries were more likely to discontinue medications (PR = 1.09; 95% CI = 1.04, 1.14) and less likely to exhibit PDC ≥ 0.80 (PR = 0.89; 95% CI = 0.86, 0.92) than non-depressed beneficiaries.

Conclusions

Depression is prevalent in Medicare beneficiaries with COPD and independently associated with lower COPD MM adherence. Interventions to improve medication adherence for COPD patients may consider management of comorbidities such as depression. Copyright © 2013 John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

Chronic obstructive pulmonary disease (COPD) is a chronic condition associated with significant morbidity and mortality. It has become the third leading cause of death among US adults (Centers for Disease Control and Prevention, 2008). Approximately one in eight community-dwelling Medicare beneficiaries is diagnosed with COPD (Stuart et al., 2007). Preventing exacerbations is an important goal of COPD management. A cornerstone of COPD management is pharmacotherapy, which aims to reduce symptoms, decrease the frequency and severity of exacerbations, and improve health status and exercise tolerance (American Thoracic Society, 2004, Global Initiative for Chronic Obstructive Pulmonary Disease, 2008). However, COPD maintenance medication use among COPD patients remains low: only 40% of Medicare beneficiaries with COPD fill one or more COPD maintenance medications annually (Stuart et al., 2010).

Among individuals who take at least one COPD maintenance medication, many fail to adhere to prescribed therapy. In the literature, adherence to COPD medications ranges from 40% to 60% (Rand, 2005, Restrepo et al., 2008, Charles et al., 2010). One observational study, using national Veteran's Affairs data, found only half of COPD patients used any maintenance medication. Among users, the overall medication adherence, measured by mean Medication possession ratio (MPR [standard deviation]), was 0.44 [0.32] during the last year of life (Jung et al., 2009). Another observational study estimated adherence to inhalation medications through refill rates among stable COPD patients in Belgium. This study found almost half (48%) of COPD patients were under-adherent annually (Mehuys et al., 2010).

There is empirical evidence demonstrating the beneficial influence of good adherence on clinical outcomes among COPD patients (Vestbo et al., 2009, Toy et al., 2011). One study, using administrative claims data from US employees and retirees and measuring proportion of days covered (PDC) to assess COPD medication adherence, concluded that good adherence was associated with a lower annual rate of inpatient and emergency room visits (Toy et al., 2011). Another clinical study found association between good medication adherence and improved mortality and reduction in hospital admission among moderate and severe COPD patients (Vestbo et al., 2009). Despite the importance of adherence to maintenance medications in preventing COPD exacerbations, adherence to these medications remains a problematic area where improvements could result in significant benefits.

Depression remains an important and overlooked risk factor for medication non-adherence. In other chronic conditions, including diabetes and heart failure, comorbid depression has been found to be associated with lower medication use and adherence (Ciechanowski et al., 2000, DiMatteo et al., 2000, Lin et al., 2004, Morgan et al., 2006, Hansen et al., 2009). Depression occurs commonly in COPD patients. Prevalence estimates of comorbid depression range between 10% and 42% in stable COPD patients (Yohannes et al., 2000, Lacasse et al., 2001, van Manen et al., 2002, Kunik et al., 2005, Maurer et al., 2008, Yohannes et al., 2010). The wide range of estimates is due to differences in study populations, depression assessment measures, and illness severity. Increased evidence has shown depression places COPD patients at higher risks for hospitalizations, readmissions, and increased hospital days (Almagro et al., 2002, Ng et al., 2007, Xu et al., 2008, Stuart et al., 2010). However, despite the high prevalence of comorbid depression in COPD patients, little is known about the role of depression on COPD maintenance medication use or adherence.

The objective of this study is to estimate the association between comorbid depression diagnosis and COPD maintenance medication adherence among a nationally representative sample of Medicare beneficiaries with COPD. This study is among the first to date to explore the relationships among diagnosed depression and COPD maintenance medication use and adherence in Medicare beneficiaries with COPD.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

Data source

This retrospective cross-sectional study used a random 5% sample of 2006–2007 Centers for Medicare and Medicaid Services Chronic Condition Warehouse (CCW) data. The CCW files include Medicare Parts A (inpatient), B (outpatient), and D (prescription) claims data plus 27 chronic condition flags (including COPD and depression). These flags are operationalized on the basis of the relevant International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes associated with Medicare claims from 1999 forward (Chronic Condition Data Warehouse, 2012). Personal identifying information (such as name, address, and Medicare claim number) has been stripped and replaced with encrypted beneficiary IDs. The researchers had no access to the encryption links or to personal identifying information of beneficiaries.

Study population

Medicare beneficiaries in stand-alone Part D plans (PDPs) through the 2-year study period with a COPD diagnosis between 1 July 2005 and 30 June 2006 were identified. Individuals with a COPD diagnosis were identified using the CCW COPD algorithm (Chronic Condition Data Warehouse, 2012), which uses at least one inpatient, skilled nursing home, home health, or two hospital outpatient carrier claims with any of the following ICD-9-CM codes: 491.xx, 492.xx, or 496.xx. The study sample was further restricted to those enrolled in PDPs between 1 January 2006 and 30 June 2006. Subjects were required to be continuously enrolled in Medicare Parts A, B, and stand-alone PDPs through the 2-year study period in order to observe all medical and prescription events.

Individuals enrolled in Medicare Advantage plans or Health Maintenance Organizations were excluded because of inability to elucidate diagnoses (n = 10,447), as were those who died prior to 1 January 2007 (all beneficiaries were required to live through 2006 to have 1-year observation window for baseline sociodemographic and clinical factors). Beneficiaries with selected respiratory conditions (cystic fibrosis, alpha-1 antitrypsin, respiratory cancer, fibrosis due to tuberculosis, bronchiecstasis, pneumoconiosis, pulmonary fibrosis, pulmonary tuberculosis, and sarcoidosis) were also excluded as their use of COPD maintenance medications is markedly different from those with COPD (n = 9412). After applying inclusion and exclusion criteria, the study cohort included a total of 74,863 Medicare beneficiaries. The conduct of this study was approved by the institutional review board of the University of Maryland Baltimore, Maryland, and the requirement for written informed consent was waived.

Measures

Depression status

Individuals with depression were identified using the CCW depression algorithm, which included the following ICD-9-CM codes for depression (ICD-9-CM codes: 296.20–296.26, 296.30–296.36, 296.50–296.56, 296.60–296.66, 296.89, 298.0, 300.4, 309.1, and 311.x) in at least one inpatient, skilled nursing home, home health, hospital outpatient, or carrier claim (Chronic Condition Data Warehouse, 2012). Because the CCW depression definition includes a broad array of conditions with a depression component, such as bipolar, schizophrenia, and other psychotic disorders, sensitivity analyses excluding beneficiaries with these comorbid conditions were performed to test the robustness of study results. Depression diagnoses were assessed during the 2-year study period.

Chronic obstructive pulmonary disease maintenance medication use and adherence

On the basis of the Global Initiative for Chronic Obstructive Pulmonary Disease guidelines, the following therapeutic classes were included as COPD maintenance medications: inhaled corticosteroids (ICS) (alone or in combination with long-acting β2-agonists [LABA]), LABA, anticholinergics, and methylxanthines. Use of any of these medications was measured as a binary variable based on one or more prescription claims in the 2-year study period. Adherence measures included medication discontinuation (defined as a binary variable of medication stopped 3 months or more prior to the end of the observation period or death, whichever came first) and PDC. PDC values range from 0 to 1 and are calculated as the number of days with any COPD maintenance medication divided by duration of therapy with these agents. PDC has been widely used to measure adherence to medication in observational studies (Choudhry et al., 2009, Toy et al., 2011), and it was calculated for COPD maintenance medications prescribed to each beneficiary from the date of the first to the last prescription during the follow-up period. PDC was calculated at the therapeutic class level (ICS, LABA, ICS/LABA combination, anticholinergics, and methylxanthines) and combined across classes. In this study, the cut point for PDC was ≥ 0.80, considered “good” adherence (Choudhry et al., 2009, Zeng et al., 2010). Measurement of adherence was limited to those beneficiaries who had evidence of any COPD maintenance medication prescriptions with at least two Medicare Part D claims in order to calculate medication adherence (Peterson et al., 2007, Karve et al., 2008).

Covariates

All covariates were measured in baseline year 2006. Demographic characteristics included age, gender, race, and geographic region. Medicare coverage variables included original social security disability insurance Medicare entitlement and low-income subsidy (LIS) status. General health indicators included evidence of specific comorbid conditions. Among those, the CCW condition flags were used to identify some comorbidities (diabetes, ischemic heart disease, congestive heart failure, and Alzheimer's disease and related dementias); other conditions of interest without CCW condition flags (asthma, hypertension, other cardiovascular diseases, and anxiety) were identified by the presence of any claim in 2006 with relevant ICD-9-CM codes. Any all-cause hospitalization was identified as a general health status measure. Two other indicators in claims data were used to assess COPD severity—any evidence of supplemental oxygen use and COPD rescue medication use (short-acting β2-agonists) (Yohannes et al., 2002).

Statistical analyses

Univariable and bivariable analyses were conducted to describe COPD cohort characteristics and COPD maintenance medication use and adherence. Chi-square tests (for differences in distribution of categorical variables) and t tests (for differences in values of continuous variables) were used to estimate the unadjusted relationships between depression status and medication use and adherence. Multivariable models using modified Poisson regression and log link with generalized estimating equations were applied to estimate the association between depression and COPD maintenance medication use and adherence, controlling for covariates (Spiegelman and Hertzmark, 2005). All multivariable analyses reported adjusted prevalence ratios (PR) with 95% confidence intervals (CI) to reflect the marginal differences. Significance was set a priori at p < 0.05. All analyses were performed using sas 9.2 (SAS Institute Inc., Cary, NC, USA).

In order to test the robustness of findings and to identify possible disease misclassification bias, a series of sensitivity analyses were performed: (i) exclusion of beneficiaries with evidence of both COPD and asthma (ICD-9-CM code 493.xx) to examine potential respiratory disease misclassification bias; (ii) exclusion of beneficiaries with evidence of bipolar (ICD-9-CM codes 296.5x, 296.6x, 296.7x, 296.8x, and 296.9x) and/or psychotic disorders (ICD-9-CM codes 293.81, 293.82, 295.xx, 297.1x-297.3x, 297.8x, 297.9x, and 298.x) to study potential psychiatric disease misclassification bias; (iii) comparison of different PDC cut points (e.g., PDC = 0.90 and 0.70 relative to PDC ≥ 0.80); and (iv) exclusion of beneficiaries who died in 2007 to examine potential bias on medication adherence measures before death.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

Among the COPD cohort (n = 74,863), 33.6% had diagnosed depression. The mean age within the cohort was 72 years and almost two thirds (63.2%) were women (Table 1). Over 85% of the sample was identified as White and over three fifths (61.0%) were LIS eligible for Part D benefits. Relative to non-depressed COPD beneficiaries, depressed beneficiaries were younger, more likely to be women and White, and more likely to be LIS eligible (all p < 0.0001). A higher proportion of those with depression had other conditions, including asthma, diabetes, ischemic heart disease, congestive heart failure, hypertension, Alzheimer's disease and related dementias, and anxiety (all p < 0.0001). Compared with non-depressed COPD beneficiaries, those with diagnosed depression had a higher proportion of baseline all-cause hospitalization and rescue medication use (all p < 0.0001) but similar supplemental oxygen use (p = 0.4492).

Table 1. Characteristics of beneficiaries with COPD enrolled in Medicare Part D plans, stratified by depression status (n = 74,863)
CharacteristicsAll COPD beneficiariesDepressiona
YesNo
N%N%N%
  • COPD, chronic obstructive pulmonary disease; SD, standard deviation.

  • a

    2006 and/or 2007 depression annual flag in the Chronic Condition Warehouse data.

  • b

    SSDI, social security disability insurance.

  • p values < 0.0001 from Chi-square tests for categorical variables or t tests for continuous variables.

Sample size74,863100.025,13010049,733100
Age (years)      
Mean (SD)72.212.070.013.673.410.9
<6515,17520.37,42629.67,74915.6
65–7424,95033.37,17028.517,78035.8
75–8424,16332.37,06728.117,09634.4
8510,57514.13,46713.87,10814.3
≥65 years were SSDIb eligible10,28613.73,68614.76,60013.3
Sex      
Female47,33763.217,65270.229,68559.7
Male27,52636.87,47829.820,04840.3
Race      
White64,28385.921,93887.342,34585.1
Black6,5908.81,9417.74,6499.4
Hispanic1,9532.68033.21,1502.3
Asian1,0711.41420.69291.9
Other9661.33061.26601.3
Region      
Northeast12,83517.14,34717.38,48817.0
North central18,39024.66,35425.312,03624.2
West10,60014.23,22312.87,37714.8
South32,93744.011,20644.621,83243.9
Low-income subsidy status      
Yes45,67961.018,47173.527,20854.7
No29,18439.06,65926.522,52545.3
Comorbid medical conditions      
Asthma21,46628.78,24432.813,22226.6
Diabetes26,63635.69,93739.516,69933.6
Ischemic heart disease41,48155.414,68658.426,79553.9
Congestive heart failure32,77243.812,14648.320,62641.5
All other cardiovascular disease51,70869.117,12468.134,58469.5
Hypertension60,71381.120,93583.339,77880.0
Alzheimer's disease and related dementias14,66219.67,51929.97,14314.4
Anxiety14,17218.99,76638.94,4068.9
COPD severity indicators      
Had supplemental oxygen claims14,69919.64,97319.89,72619.6
Had any COPD rescue medication23,13830.98,73634.814,40229.0
Hospitalization at baseline      
Had any hospitalization34,90546.614,68358.420,22240.7
# of hospitalization (mean, SD)2.21.82.52.11.91.4

Overall, more than three fifths (61.8%) of the cohort received at least one COPD maintenance medication prescription during the 2-year study period (Table 2). The proportion of COPD maintenance medication use was higher in depressed than in non-depressed beneficiaries (64.1% vs. 60.6%). Among COPD maintenance medication users with at least two Part D prescription claims (n = 35,985), depressed patients were more likely to discontinue (23.9% vs. 20.6%) and were less likely to exhibit PDC ≥ 0.80 (31.4% vs. 38.5%) than non-depressed patients (all p < 0.0001).

Table 2. Two-year COPD maintenance medication use and adherence in Medicare Part D beneficiaries with COPD, stratified by depression status
Drug use and hospitalization measuresAll COPDDepressiona
BeneficiariesYesNo
N%N%N%
  • COPD, chronic obstructive pulmonary disease.

  • a

    2006 and/or 2007 depression annual flag in the Chronic Condition Warehouse data.

  • b

    Among the total sample (n = 74,863).

  • c

    Limited to who had any COPD maintenance medication use and at least two Part D claims (n = 35,985).

  • p values < 0.0001 from Chi-square tests or t tests.

Sample size74,863100.025,13010049,733100
Any COPD maintenance medication      
Any useb46,27161.816,11564.130,15660.6
Sample sizec35,985 12,574 23,411 
Discontinuerc7,83521.83,00923.94,82620.6
Proportion of days covered ≥ 0.80c12,95636.03,95331.49,00338.5

In multivariable models, depression was associated with a higher likelihood of COPD maintenance medication use (adjusted PR = 1.02; 95% CI = 1.01, 1.03) (Table 3). However, among users with two or more Part D claims, depressed patients were 9% more likely to discontinue medication (PR = 1.09; 95% CI = 1.04, 1.14) and 11% less likely to exhibit PDC ≥ 0.80 (PR = 0.89; 95% CI = 0.86, 0.92) than non-depressed patients, controlling for all covariates.

Table 3. Association between depression diagnosis and 2-year COPD maintenance medication use and adherence in Medicare Part D beneficiaries with COPD
PredictorsMedication use and adherence
Any useaDiscontinuationbPDC (≥ 0.8) b
PRLower 95% CIUpper 95% CIPRLower 95% CIUpper 95% CIPRLower 95% CIUpper 95% CI
  1. PR, prevalence ratios; CI, confidence intervals; COPD, chronic obstructive pulmonary disease; PDC, proportion of days covered. Multivariable models used modified Poisson regression and log link with generalized estimating equations. Bold numbers indicate p < 0.05.

  2. a

    Among the total sample (n = 74,863).

  3. b

    Limited to who had any COPD maintenance medication and had at least two Part D claims (n = 35,985).

  4. c

    SSDI, social security disability insurance.

Crude depression1.061.051.071.161.121.210.820.790.84
Adjusted depression1.021.011.031.091.041.140.890.860.92
Age (years)         
<65ReferenceReferenceReference
65–741.081.061.090.820.770.871.091.041.13
75–841.061.041.080.810.760.861.040.991.08
85+0.970.950.990.830.770.891.000.951.06
≥65 years were SSDIc eligible0.990.971.001.061.001.131.010.971.05
Female1.000.991.011.020.981.070.930.900.96
Race         
WhiteReferenceReferenceReference
Black0.900.880.911.191.121.280.790.750.84
Hispanic1.010.981.051.231.091.400.930.831.03
Asian0.880.830.921.451.251.680.970.861.10
Other0.960.911.001.291.111.510.870.761.00
Region         
NortheastReferenceReferenceReference
North central0.970.960.991.071.011.140.970.931.01
West0.920.900.941.131.061.210.950.900.99
South0.970.960.991.141.081.200.930.900.97
Low-income subsidy status         
NoReferenceReferenceReference
Yes1.061.051.070.870.830.910.970.931.00
Comorbid medical conditions         
Asthma1.391.381.410.920.880.961.010.981.04
Diabetes0.980.970.991.121.071.170.950.920.98
Ischemic heart disease0.970.960.991.131.081.180.920.890.95
Congestive heart failure1.031.021.041.040.991.090.980.951.01
All other cardiovascular disease1.000.981.011.071.021.120.970.941.00
Hypertension0.990.981.011.040.981.100.960.931.00
Alzheimer's disease and related dementias0.920.910.941.101.041.160.840.810.88
Anxiety1.011.001.031.020.971.070.990.951.03
Had any baseline hospitalization0.940.930.951.201.151.250.860.830.89
COPD severity indicators         
Had supplemental oxygen claims1.391.381.410.760.730.801.331.301.37
Had any COPD rescue medication1.441.431.460.930.890.971.101.071.13

In addition, differences in COPD maintenance medication use and adherence also were found by covariates (Table 3). For example, compared with beneficiaries less than 65 years old, those 65 years and older (except for the subgroup of 85 years and older) were more likely to fill a COPD maintenance medication and less likely to discontinue their treatment. Women were less likely to exhibit PDC ≥ 0.80 (PR = 0.93; 95% CI = 0.90, 0.96) than men. Compared with Whites, Blacks were less likely to fill a COPD maintenance medication (PR = 0.90; 95% CI = 0.88, 0.91), more likely to discontinue their treatment (PR = 1.19; 95% CI = 1.12, 1.28), and less likely to exhibit PDC ≥ 0.80 (PR = 0.79; 95% CI = 0.75, 0.84). Geographically, beneficiaries residing in north central, south, and west regions were less likely to fill a COPD maintenance medication and adhere to treatment than those in the northeast region. Beneficiaries who were eligible for LIS were more likely to fill a COPD maintenance medication and less likely to discontinue treatment than non-LIS beneficiaries. Further, beneficiaries with worse health status (having any all-cause hospitalization at baseline) and selected comorbidities (including diabetes, ischemic heart disease, and Alzheimer's disease) were less likely to fill a COPD maintenance medication and adhere to treatment. However, beneficiaries with severe COPD were more likely to use COPD maintenance medication and adhere to treatment.

Sensitivity analyses confirmed the robustness of study results (Table 4). First, after restricting the sample to beneficiaries without comorbid asthma (excluding 28.7% of study sample with both COPD and asthma), depression diagnosis was still associated with a 7% higher likelihood of discontinuation (PR = 1.07; 95% CI = 1.01, 1.14) and 9% lower likelihood of PDC ≥ 0.80 (PR = 0.91; 95% CI = 0.87, 0.95). Second, after excluding beneficiaries with comorbid bipolar and/or schizophrenia (n = 2919), depressed beneficiaries still were 8% more likely to discontinue medication (PR = 1.08; 95% CI = 1.03, 1.13) and 10% less likely to exhibit PDC ≥ 0.80 (PR = 0.90; 95% CI = 0.87, 0.93). Third, using different PDC cut points (e.g., 0.90 and 0.70) to define “good” adherence did not affect the findings—those with depression were less likely to exhibit good adherence. Finally, after excluding beneficiaries who died in 2007 (n = 9287) to avoid bias on medication adherence measures before death, depression was still independently associated with 10% higher likelihood of discontinuation (PR = 1.10; 95% CI = 1.05, 1.16) and 13% lower likelihood of PDC ≥ 0.80 (PR = 0.87; 95% CI = 0.84, 0.91).

Table 4. Influence of depression on 2-year COPD maintenance medication use and adherence in Medicare Part D beneficiaries with COPD sensitivity analysisa
ModelsMedication adherence
 DiscontinuationPDC (≥ 0.8)
 PRLower 95% CIUpper 95% CIPRLower 95% CIUpper 95% CI
  1. PR, prevalence ratios; CI, confidence intervals; COPD, chronic obstructive pulmonary disease; PDC, proportion of days covered. Multivariable models used modified Poisson regression and log link with generalized estimating equations. Bold numbers indicate p < 0.05.

  2. a

    Limited to who had any COPD maintenance medication and had at least two Part D claims (n = 35,985).

I: Excluding asthma1.071.011.140.910.870.95
II: Excluding bipolar and/or other psychotic disorders1.081.031.130.900.870.93
IIIa: Using PDC cut point 0.90   0.830.790.87
IIIb: Using PDC cut point 0.70   0.920.890.94
IV: Excluding beneficiaries who died in 20061.101.051.160.870.840.91

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

Findings from this study demonstrate COPD maintenance medication use is low in Medicare beneficiaries with COPD, regardless of the presence of depression. However, among those COPD patients who filled COPD maintenance medications, depression diagnosis is associated with lower COPD maintenance medication adherence. This study is among the first to date to explore the relationships among diagnosed depression and COPD maintenance medication use and adherence in Medicare beneficiaries with COPD. Findings indicate that interventions to improve medication adherence for COPD patients may consider management of comorbidities such as depression.

Indeed, the finding that 40% of Medicare beneficiaries diagnosed with COPD failed to use any COPD maintenance medication in 2 years indicates the challenges clinicians face in managing the care of their COPD patients. Our finding is consistent with previous studies demonstrating that less than half of COPD patients were prescribed maintenance medication (Jung et al., 2009, Stuart et al., 2010). Sub-analyses from the Understanding Potential Long-term Impacts on Function with Tiotropium trial indicated that initiating maintenance treatment at early stages of COPD can alter progression of the disease and maximize patient benefit at later stages of the disease (Troosters et al., 2010). Therefore, clinicians might focus on improving maintenance medication use to reduce symptoms.

Consistent with studies demonstrating depression as an independent risk factor for reduced adherence to treatments for other chronic conditions (Ciechanowski et al., 2000, DiMatteo et al., 2000, Lin et al., 2004), this study found that depression is associated with lower COPD maintenance medication adherence. Depression may cause patients to neglect medical care and their prescribed regimens so that, in turn, being non-adherent may increase respiratory symptoms (Restrepo et al., 2008). In addition, COPD patients who report poor quality of life are more likely to be depressed, feel unsupported by clinic staff, and be poorly adherent to treatment (Bosley et al., 1996). One study investigating psychosocial factors on medication adherence among COPD patients in outpatient clinics also found that the presence of depressed mood was associated with medication non-adherence (Khdour et al., 2012). Future research is warranted to investigate causal inferences and explore mechanisms to explain this relationship.

Limited evidence has demonstrated that patients who are most severely ill with chronic conditions may be at greatest risk for treatment non-adherence (DiMatteo et al., 2007). However, in this study, COPD severity indicators (supplemental oxygen use and COPD rescue medication use) were associated with better adherence to COPD maintenance medications. In sensitivity analyses estimating adherence among beneficiaries requiring supplemental oxygen (considered the most severe stage of COPD), the association between depression diagnosis and lower medication adherence still presented (data not shown but available from the authors).

The independent association between comorbid depression and COPD maintenance medication adherence has implications for both research and practice. It suggests that interventions to improve COPD medication use and adherence are needed for this particular population. Growing evidence suggests antidepressant medications may be effective in treating depression and depressive symptoms in COPD patients (Borson et al., 1992, Smoller et al., 1998, Yohannes et al., 2001, Lacasse et al., 2004). Thus, interventions to better manage depression in COPD population are needed.

This study has several limitations. First, as in all analyses conducted using administrative claims data, measurements and disease ascertainment are limited to information available in such data. Thus, it was unable to ascertain the severity of COPD or depression afforded from laboratory data and other clinical measures (e.g., forced expiratory volume in 1 s measures and clinical depression scales). Although we controlled COPD severity in multivariable analysis and conducted a series of sensitivity analyses to examine potential disease misclassification (e.g., COPD and asthma, or depression and schizophrenia/bipolar disorders), the unobservable variation in depression severity and potential duration may affect study outcomes. In addition, depression diagnosis regardless of treatment effects on depression was assessed. Beneficiaries with evidence of treatment for depression might exhibit improved COPD maintenance medication adherence. Future investigations on the relationships between treatment effects for depression on COPD medication adherence and clinical outcomes are needed.

Second, measures were determined by refill patterns available in prescription drug event claims data and do not reflect actual patient consumption. However, adherence measured by prescription refill patterns in administrative claims data is widely accepted and shown to be valid in observational studies (Choudhry et al., 2009, Toy et al., 2011). Different from two other studies examining COPD medication adherence in one year (Jung et al., 2009, Toy et al., 2011), this study assessed COPD maintenance medication adherence over 2 years, which achieves more stable adherence measurements. Furthermore, because COPD patients can be on multiple inhalation medications concurrently, PDC was used instead of MPR to avoid overestimation of adherence behaviors (Choudhry et al., 2009). Sensitivity analyses using varying PDC levels other than 0.80 obtained similar results on the associations among depression diagnosis and COPD maintenance medication adherence. However, reasons for medication discontinuation such as intolerance or inadequate response to the regimens were not taken into account in the multivariate analysis because of the lack of information used to identify reasons of discontinuation in claims data.

Third, because both depression diagnosis and COPD maintenance medication adherence were measured during the same 2-year study period, findings are associational rather than causal. At the time of this analysis, only 2006–2007 CCW data were available. Future research incorporating additional years of data will allow construction of longitudinal analytic files and time-to-event analyses to confirm the associations reported in this study. Additionally, study results might not be generalizable to newly approved COPD maintenance medications after 2007.

Finally, even though a wide range of potential confounders were controlled, unobservable factors were not. Findings therefore may not be generalizable to Medicare Advantage plans and/or Health Maintenance Organization enrollees and those with characteristics excluded from the study sample and other non-Medicare population.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

In conclusion, findings demonstrate Medicare beneficiaries with COPD and comorbid depression have lower COPD maintenance medication use and adherence. An implication of the findings of this study is that healthcare providers should focus on improving maintenance medication use and adherence, especially for patients with chronic comorbidities such as depression. Depression should be screened among COPD population and considered as a potential risk factor for low COPD maintenance medication adherence. Further research evaluating whether interventions treating depression improve COPD maintenance medication use, adherence, and/or outcomes is warranted.

Conflicts of interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

Drs. Simoni-Wastila and Stuart received grants support from GlaxoSmithKline. Dr. Rattinger was supported by NIH Institutional Career Development Grant (K12 HD043489).

Key points

  • COPD maintenance medication use is low in Medicare beneficiaries with COPD, regardless of the presence of comorbid depression.
  • Among COPD patients who filled COPD maintenance medications, depression diagnosis is associated with lower COPD maintenance medication adherence.
  • Healthcare providers should focus on improving maintenance medication use and adherence, as well as consider the presence of comorbid depression in their COPD patients as a potential risk factor for low COPD maintenance medication adherence.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References

The study was funded by a grant from the Commonwealth Fund (contract number 20080306) and conducted at the University of Maryland, Baltimore, Maryland, USA. The authors assumed full responsibility for the accuracy and completeness of the ideas presented.

The authors acknowledge the valuable input from the staff of Pharmaceutical Research Computing, University of Maryland School of Pharmacy.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Conflicts of interest
  9. Acknowledgements
  10. References
  • American Thoracic Society. 2004. American Thoracic Society and European Respiratory Society standards for the diagnosis and management of patients with COPD. Accessed August 25, 2009.
  • Centers for Disease Control and Prevention. 2008. Deaths: preliminary data. Available at http://www.cdc.gov/nchs/data/nvsr/nvsr59/nvsr59_02.pdf. Accessed February 15, 2011.
  • Chronic Condition Data Warehouse. 2012. 27 chronic condition alogrithms. Buccaneer Computer Systems & Services, Inc., Available at http://www.ccwdata.org/cs/groups/public/documents/document/ccw_conditioncategories2011.pdf. Accessed August 15, 2012.
  • Global Initiative for Chronic Obstructive Pulmonary Disease (GOLD). 2008. Global strategy for diagnosis, management, and prevention of chronic obstructive pulmonary disease (updated 2008). Accessed August 25, 2009.
  • Almagro P, Calbo E, Ochoa De Echaguen A, et al. 2002. Mortality after hospitalization for COPD. Chest, 121, 14418.
  • Borson S, Mcdonald GJ, Gayle T, et al. 1992. Improvement in mood, physical symptoms, and function with nortriptyline for depression in patients with chronic obstructive pulmonary disease. Psychosomatics, 33, 190201.
  • Bosley CM, Corden ZM, Rees PJ, Cochrane GM. 1996. Psychological factors associated with use of home nebulized therapy for COPD. Eur Respir J, 9, 234650.
  • Charles MS, Blanchette CM, Silver H, et al. 2010. Adherence to controller therapy for chronic obstructive pulmonary disease: a review. Curr Med Res Opin, 26, 24219.
  • Choudhry NK, Shrank WH, Levin RL, et al. 2009. Measuring concurrent adherence to multiple related medications. Am J Manag Care, 15, 45764.
  • Ciechanowski PS, Katon WJ, Russo JE. 2000. Depression and diabetes: impact of depressive symptoms on adherence, function, and costs. Arch Intern Med, 160, 327885.
  • Dimatteo MR, Haskard KB, Williams SL. 2007. Health beliefs, disease severity, and patient adherence: a meta-analysis. Med Care, 45, 5218.
  • Dimatteo MR, Lepper HS, Croghan TW. 2000. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Arch Intern Med, 160, 21017.
  • Hansen RA, Dusetzina SB, Song L, Gaynes BN, et al. 2009. Depression affects adherence measurement but not the effectiveness of an adherence intervention in heart failure patients. J Am Pharm Assoc (2003), 49, 7608.
  • Jung E, Pickard AS, Salmon JW, Bartle B, Lee TA. 2009. Medication adherence and persistence in the last year of life in COPD patients. Respir Med, 103, 52534.
  • Karve S, Cleves MA, Helm M, et al. 2008. An empirical basis for standardizing adherence measures derived from administrative claims data among diabetic patients. Med Care, 46, 112533.
  • Khdour MR, Hawwa AF, Kidney JC, et al. 2012. Potential risk factors for medication non-adherence in patients with chronic obstructive pulmonary disease (COPD). Eur J Clin Pharmacol, 68, 136573.
  • Kunik ME, Roundy K, Veazey C, et al. 2005. Surprisingly high prevalence of anxiety and depression in chronic breathing disorders. Chest, 127, 120511.
  • Lacasse Y, Beaudoin L, Rousseau L, Maltais F 2004. Randomized trial of paroxetine in end-stage COPD. Monaldi Arch Chest Dis, 61, 1407.
  • Lacasse Y, Rousseau L, Maltais F. 2001. Prevalence of depressive symptoms and depression in patients with severe oxygen-dependent chronic obstructive pulmonary disease. J Cardiopulm Rehabil, 21, 806.
  • Lin EH, Katon W, Von Korff M, et al. 2004. Relationship of depression and diabetes self-care, medication adherence, and preventive care. Diabetes Care, 27, 215460.
  • Maurer J, Rebbapragada V, Borson S, et al. 2008. Anxiety and depression in COPD: current understanding, unanswered questions, and research needs. Chest, 134, 43S-56S.
  • Mehuys E, Boussery K, Adriaens E, et al. 2010. COPD management in primary care: an observational, community pharmacy-based study. Ann Pharmacother, 44, 25766.
  • Morgan AL, Masoudi FA, Havranek EP, et al. 2006. Difficulty taking medications, depression, and health status in heart failure patients. J Card Fail, 12, 5460.
  • Ng TP, Niti M, Tan WC, et al. 2007. Depressive symptoms and chronic obstructive pulmonary disease: effect on mortality, hospital readmission, symptom burden, functional status, and quality of life. Arch Intern Med, 167, 607.
  • Peterson AM, Nau DP, Cramer JA, et al. 2007. A checklist for medication compliance and persistence studies using retrospective databases. Value Health, 10, 312.
  • Rand CS. 2005. Patient adherence with COPD therapy. Eur Respir Rev, 14, 97101.
  • Restrepo RD, Alvarez MT, Wittnebel LD, et al. 2008. Medication adherence issues in patients treated for COPD. Int J Chron Obstruct Pulmon Dis, 3, 37184.
  • Smoller JW, Pollack MH, Systrom D, Kradin RL. 1998. Sertraline effects on dyspnea in patients with obstructive airways disease. Psychosomatics, 39, 249.
  • Spiegelman D, Hertzmark E. 2005. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol, 162, 199200.
  • Stuart B, Simoni-Wastila L, Zuckerman I, et al. 2007. Medication use by age and disabled Medicare beneficiaries across the spectrum of morbidity—a chartbook. The Peter Lamy Center on Drug Therapy and Aging, University of Maryland School of Pharmacy, Baltimore, MD.
  • Stuart BC, Simoni-Wastila L, Zuckerman IH, et al. 2010. Impact of maintenance therapy on hospitalization and expenditures for Medicare beneficiaries with chronic obstructive pulmonary disease. Am J Geriatr Pharmacother, 8, 44153.
  • Toy EL, Beaulieu NU, Mchale JM, et al. 2011. Treatment of COPD: relationships between daily dosing frequency, adherence, resource use, and costs. Respir Med, 105, 43541.
  • Troosters T, Celli B, Lystig T, et al. 2010. Tiotropium as a first maintenance drug in COPD: secondary analysis of the UPLIFT trial. Eur Respir J, 36, 6573.
  • Van Manen JG, Bindels PJ, Dekker FW, et al. 2002. Risk of depression in patients with chronic obstructive pulmonary disease and its determinants. Thorax, 57, 4126.
  • Vestbo J, Anderson JA, Calverley PM, et al. 2009. Adherence to inhaled therapy, mortality and hospital admission in COPD. Thorax, 64, 93943.
  • Xu W, Collet JP, Shapiro, S, et al. 2008. Independent effect of depression and anxiety on chronic obstructive pulmonary disease exacerbations and hospitalizations. Am J Respir Crit Care Med, 178, 91320.
  • Yohannes AM, Baldwin RC, Connolly M. 2002. Mortality predictors in disabling chronic obstructive pulmonary disease in old age. Age Ageing, 31, 13740.
  • Yohannes AM, Baldwin RC, Connolly MJ. 2000. Depression and anxiety in elderly outpatients with chronic obstructive pulmonary disease: prevalence, and validation of the BASDEC screening questionnaire. Int J Geriatr Psychiatry, 15, 10906.
  • Yohannes AM, Connolly MJ, Baldwin RC. 2001. A feasibility study of antidepressant drug therapy in depressed elderly patients with chronic obstructive pulmonary disease. Int J Geriatr Psychiatry, 16, 4514.
  • Yohannes AM, Willgoss TG, Baldwin RC, Connolly MJ. 2010. Depression and anxiety in chronic heart failure and chronic obstructive pulmonary disease: prevalence, relevance, clinical implications and management principles. Int J Geriatr Psychiatry, 25, 120921.
  • Zeng F, Patel BV, Andrews L, Frech-Tamas F, Rudolph AE. 2010. Adherence and persistence of single-pill ARB/CCB combination therapy compared to multiple-pill ARB/CCB regimens. Curr Med Res Opin, 26, 287787.