Economic impact of using inhaled corticosteroids without prior exacerbation among elderly patients with chronic obstructive pulmonary disorder

Abstract

Objective:

To assess the economic impact of initiating inhaled corticosteroids (ICS) without evidence of prior exacerbation among elderly patients with chronic obstructive pulmonary disease (COPD) in the US.

Methods:

This retrospective study used administrative claims to identify newly diagnosed COPD patients between 1/1/2005 and 6/30/2006 who were dispensed ICS. The dispense date of the first ICS was set as the index date. Patients with prior diagnoses for asthma, cystic fibrosis, or lung cancer were excluded. Cohorts were constructed based on whether ICS therapy was concordant with recommended guidelines of having prior COPD exacerbation. Each COPD patient with prior exacerbation was matched to four patients without exacerbation based on age, gender, Charlson Comorbidity Index, and whether COPD diagnosis code was not elsewhere specified (i.e., 496). Multivariate regressions were estimated to assess the association between use of ICS therapy without prior exacerbation and total healthcare costs, controlling for demographics and clinical characteristics.

Results:

The study included 3650 patients: 730 with prior exacerbation and 2920 without prior exacerbation. Patients were 76 years of age and 54% were male. Those with prior exacerbation were more likely to have inpatient stays both prior to (74.4 vs. 44.1%, p < 0.05) and following (37.0 vs. 33.1%, p < 0.05) the index date. Controlling for patient characteristics, patients who were dispensed ICS without prior exacerbation had $1859 higher in total costs (p < 0.05) compared to patients with prior exacerbation during the 12 months following ICS initiation.

Limitations:

The retrospective design of this study limits the interpretation of findings as association and not causality. This study is subject to selection bias due to unobservable confounders.

Conclusions:

Among COPD patients, initiation of ICS without prior exacerbation appears to be associated with increased healthcare costs. These findings suggest that ICS initiation without evidence of exacerbation as consistent with guidelines is associated with adverse economic consequences.

Key words::

• Chronic obstructive pulmonary disease
• Exacerbation
• Healthcare costs
• Inhaled corticosteroids

Introduction

Chronic obstructive pulmonary disease (COPD) includes both emphysema and chronic bronchitis, both of which cause irreversible damage to the lungs and decrease the ability of the lungs to take in oxygen and dispel carbon dioxide1. COPD results in difficulty breathing, and can cause heavy coughing, wheezing, shortness of breath, and chest tightness, particularly in current or past cigarette smokers, who make up the majority of COPD patients2. The condition develops over time, and is therefore typically diagnosed in middle‐aged or elderly adults. As COPD symptoms worsen, affected individuals may become unable to perform day‐to‐day activities such as walking2. COPD is currently the fourth leading cause of death in the United States, affecting approximately 30 million individuals and claiming the lives of over 127,000 people nationally in 20053,4. In 2006 alone, 9.5 million Americans reported having physician‐diagnosed COPD4. The total annual costs of COPD are currently estimated at $42.6 billion in 2007, with direct healthcare costs totaling $26.7 billion, indirect illness‐related costs totaling $8 billion, and indirect death‐related costs totaling $7.9 billion2.

Treatment goals for COPD include the management of respiratory exacerbations associated with the disease. US guidelines suggest that patients be initiated on short‐acting bronchodilators, followed by long‐acting bronchodilators, and then finally inhaled corticosteroids (ICS) if the previous therapies provide only temporary relief5. According to recent practice guidelines, the purpose of ICS therapy is to reduce frequent respiratory exacerbations, primarily among patients with moderate‐to‐severe COPD5–9. ICS should be prescribed as a maintenance therapy for COPD patients for whom bronchodilators or other therapeutic regimens are not adequate for disease management. Clinical practice guidelines for the management of COPD only recommend the use of ICS, in combination with a long‐acting beta agonist (LABA), for patients who have both severe COPD and frequent exacerbations, with severe COPD commonly defined as forced expiratory volume in one second (FEV₁) <50% predicted, and ‘frequent exacerbations’ defined as having at least one exacerbation in the past year5,6,8,10.

This study sought to assess the economic impact of initiating ICS therapy without evidence of prior exacerbation among elderly COPD patients. Although ICS treatment has become part of the standard of care approach for COPD patients, questions have arisen regarding whether or not they are appropriately prescribed in clinical practice. Published guidelines for COPD limit the use of ICS to patients who have severe COPD with frequent exacerbations5–9, but it appears that ICS therapy is often used in mild COPD, and is used inconsistently even in severe COPD11–13. The misuse or overuse of ICS could contribute to increased healthcare costs and utilization; thus, it is crucial to understand the available economic data regarding the use of ICS in COPD. The findings from this study will assist the medical community and policy makers in understanding the potential economic consequences of using ICS for COPD patients without evidence of prior exacerbations.

Methods

Data source

This study analyzed administrative claims data from the Thomson’s Medstat MarketScan Medicare Supplemental Insurance Claims database (Thomson Reuters Healthcare Inc., Ann Arbor, MI, USA) between 2004 and 2008. This database provides comprehensive information on enrollment records, medical service, and pharmacy claims for individuals enrolled in geographically distributed health plans that provide Medicare supplemental insurance in the US. All records can be linked through encrypted unique patient‐level identifiers, which allow for longitudinal tracking of individuals’ healthcare utilization. Demographic and health insurance plan characteristics including age, gender, US geographic region, health plan type, and enrollment status are available in the enrollment files. Detailed healthcare encounters with both inpatient and outpatient care are recorded in the medical services claims, which include date and place of services, provider type, payments, and up to 15 International Classification of Diseases, 9th Revision, Clinical Modification (ICD‐9‐CM) diagnosis and procedure codes. Pharmacy claims files contain National Drug Code (NDC), dispense date, quantity dispensed, days supplied, and plan‐ and patient‐paid amounts.

Study sample selection

Incident COPD patients were identified as having two or more medical claims at least 30 days apart with an associated diagnosis of COPD (ICD‐9‐CM: 491.xx, 492.x, 496) between January 1, 2005 and June 30, 2006 of the initial COPD diagnosis. Patients were considered as newly‐diagnosed if they had no prior COPD diagnosis or ICS use during the previous 12 months. Other inclusion criteria were: (1) 65+ years of age as of the COPD diagnosis; (2) had continuous health plan enrollment from 12 months prior through 12 months following the initial COPD diagnosis; (3) had no diagnosis of asthma (ICD‐9‐CM: 493.xx), cystic fibrosis (ICD‐9‐CM: 277.0x) or lung cancer (ICD‐9‐CM: 160.x–164.x, 231.xx) prior to the initial COPD diagnosis. We then selected COPD patients who were dispensed ICS therapy (minimum of 30 days of supply) after the first COPD diagnosis. The dispense date of the first prescription was denoted as the index date. Eligible study participants had continuous health plan enrollment from 12 months prior through 12 months following the ICS index date and had no diagnosis of asthma 12 months prior to and following the ICS index date.

Two study cohorts of ICS users were created based on whether or not use of ICS therapy was concordant with guidelines’ recommendation for having evidence of prior exacerbation. Exacerbations were identified based on inpatient admissions with COPD as primary diagnosis or pulmonary‐related primary diagnosis with a COPD secondary diagnosis, emergency room visits with shortness of breath (ICD‐9‐CM: 786.05) or COPD diagnosis, and office visits with COPD diagnosis. Patients with COPD‐related emergency room visits and office visits were also required to have oral corticosteroids or antibiotic medications dispensed within 3 days of the service.

Study measures

Patient demographic and clinical characteristics were evaluated during the 12 months prior to ICS use. Age, gender, and Charlson Comorbidity Index14 were assessed for each study patient. Dichotomous variables were created for respiratory‐related comorbidities (i.e., bronchitis, acute upper respiratory tract infection, pneumonia, pleural effusion, and respiratory tuberculosis) based on ICD‐9‐CM diagnosis codes recorded in the medical claims. Additionally, resource utilization was examined during the 12 months prior to and following ICS use. The proportion of patients with any inpatient hospital stays or outpatient ER visits was assessed, as well as the number of outpatient hospital visits and number of physician office visits.

Healthcare costs during the 12 months prior to and following the ICS index date were examined, overall and by type of service (i.e., inpatient, outpatient, and pharmacy). All costs were calculated as the sum of the paid amounts (reimbursed amount by health plans as well as patient out‐of‐pocket payment) from both the medical and pharmacy claims. All costs were adjusted to 2008 US dollars using the medical care component of the US Consumer Price Index15.

Analysis

Each patient with prior COPD exacerbation was matched to four patients without any exacerbation based on age, gender, Charlson Comorbidity Index, and whether COPD diagnosis was not elsewhere specified (i.e., 496). Descriptive analyses were conducted to examine cross‐cohort differences in patient demographics, comorbid medical conditions, and resource utilization over the 12 months prior to and following the ICS index date. Mean and standard deviation were reported for continuous variables (i.e., age), and Student t‐tests were used to detect significant differences between the cohorts. Percentages were reported for categorical variables (i.e., gender, comorbidities, any inpatient stays, and any ER visits), and statistical differences between cohorts were assessed using chi‐square tests. Non‐parametric Wilcoxon rank‐sum tests were used for count variables (i.e., Charlson Comorbidity Index, healthcare costs, number of outpatient hospital visits, and number of physician office visits).

Because the distribution of healthcare costs are highly skewed, the adjusted cost differences between cohorts in the 12 months following the ICS index date were estimated using generalized linear regressions with a log‐link function and a gamma distribution16,17. Explanatory variables included patient demographics, clinical characteristics, and prior hospitalization. Adjusted differences were calculated as the mean of differences among all individuals when their cohort variable was alternatively coded as ‘0’ and ‘1’16. All analyses were conducted using SAS version 9.2 (SAS Institute, Inc., Cary, NC, USA), and findings of p‐values <0.05 were considered statistically significant.

Results

After applying inclusion/exclusion criteria and matching, this study identified 3650 patients: 730 with prior exacerbation and 2920 without prior exacerbation (Table 1). Due to the matched study design, baseline characteristics were similar between cohorts. Patients were 76 years of age, had a mean Charlson Comorbidity Index of 3, and approximately 54% were male. Those with prior exacerbation had significantly higher rates of bronchitis (34.8 vs. 26.3%, p < 0.05) but lower rates of pleural effusion (10.1 vs. 6.7%, p < 0.05). Prior to ICS initiation, patients with prior exacerbation were more likely to have any inpatient hospital stays (74.4 vs. 44.1%, p < 0.05) but had significantly fewer physician office visits (13.3 vs. 14.4, p < 0.05). Patients with prior exacerbation were also more likely to have any hospital stays during the 12 months following ICS initiation (37.0 vs. 33.1%, p < 0.05). Total healthcare costs during the pre‐index period were significantly greater for patients with prior exacerbation ($19,954 vs. $19,401, p < 0.05), although patients without prior exacerbation incurred greater outpatient costs ($8087 vs. $7637, p < 0.05). During the 12 months following ICS initiation, healthcare costs were similar between cohorts. It appeared that, for both cohorts, the major attribute of healthcare costs shifted from inpatient to outpatient from the pre‐index period to the post‐index period.

Table 1.  Demographics, clinical characteristics, resource use and healthcare costs.

	ICS cohorts
	With prior exacerbation	Without prior exacerbation
Number of patients	730	2,920
Percent of patients (%)	20.0	80.0
Age, years: mean (SD)	76.0 (6.1)	76.0 (6.2)
Male (%)	54.3	54.3
Charlson Comorbidity Index: mean (SD)	2.9 (2.5)	3.0 (2.6)
Respiratory‐related comorbidities (%)
Bronchitis	34.8	26.3*
Acute upper respiratory tract infection	16.0	14.5
Pneumonia	8.8	7.1
Pleural effusion	6.7	10.1*
Respiratory tuberculosis	0.1	0.1
Pre‐index healthcare utilization and costs:
Any hospital stays (%)	74.4	44.1*
Any ER visits (%)	20.7	19.6
Number of outpatient hospital visits: mean (SD)	4.4 (4.9)	4.7 (6.2)
Number of physician office visits: mean (SD)	13.3 (12.1)	14.4 (11.2)*
Total healthcare costs: mean (SD)	$19,954 ($39,395)	$19,401 ($31,476)*
Inpatient costs: mean (SD)	$8,370 ($20,078)	$7,477 ($23,033)*
Outpatient costs: mean (SD)	$7,637 ($24,379)	$8,087 ($16,420)*
Pharmacy costs: mean (SD)	$3,947 ($3,479)	$3,837 ($3,646)
Post‐index healthcare utilization and costs:
Any hospital stays (%)	37.0	33.1*
Any ER visits (%)	22.7	20.0
Number of outpatient hospital visits: mean (SD)	4.8 (6.1)	5.3 (7.2)
Number of physician office visits: mean (SD)	15.7 (13.3)	16.6 (12.8)
Total healthcare costs: mean (SD)	$20,667 ($29,151)	$20,839 ($29,851)
Inpatient costs: mean (SD)	$5,929 ($18,203)	$5,626 ($19,300)
Outpatient costs: mean (SD)	$9,150 ($16,663)	$9,687 ($17,580)
Pharmacy costs: mean (SD)	$5,589 ($4,423)	$5,526 ($4,510)

Chi‐square test was used for categorical variables (gender, any resource utilization). Student t‐test was used for continuous variables (age), and Wilcoxon rank‐sum test was used for count variables (CCI, healthcare costs, and number of office visits).

Costs reported in 2008 US dollars.

*Statistically significant at p < 0.05.

SD, standard deviation.

Controlling for demographic and clinical characteristics (Table 2), patients who were prescribed ICS without prior exacerbation had significantly higher total costs compared to patients with prior exacerbation (marginal effect = $1859, p < 0.05) during the 12 months following ICS initiation. Having one or more Charlson comorbidities was associated with significantly higher healthcare costs (p < 0.05) and this association was stronger as the Charlson score increased further. In addition presence of acute upper respiratory tract infection, pleural effusion or any hospital admission prior to ICS initiation were all associated with significantly higher healthcare costs (p < 0.05).

Table 2.  Results from generalized linear regression model of total healthcare costs.

	Parameter estimate	p‐value
Without prior exacerbation	0.09	0.01
Age, years
65–74	Reference
75–84	0.06	0.06
85+	−0.03	0.54
Gender
Male	Reference
Female	0.01	0.74
Clinical characteristics 1 year pre ICS
CCI = 0	Reference
CCI = 1	0.23	<0.01
CCI = 2	0.34	<0.01
CCI = 3+	0.60	<0.01
Pleural effusion	0.13	0.01
Acute upper respiratory tract infection	0.15	<0.01
Bronchitis	0.03	0.36
Pneumonia	0.01	0.91
Congestive heart failure	0.06	0.09
Any pre‐index hospital admission	0.32	<0.01

Model also controlled for region.

Discussion

This retrospective cohort database study analyzed administrative claims and eligibility records of a large US elderly population who had Medicare Part A and B coverage, as well as supplemental insurance through their former employers to examine the economic impact associated with initiating ICS therapy without evidence of prior exacerbation among COPD patients. The use of ICS for the treatment of COPD is highly debated18, given the evidence to support this was limited. Recent meta‐analyses also showed the increased risk of pneumonia associated with ICS therapy in COPD19,20. Although guidelines suggest the use of ICS therapy only for patients with severe COPD and frequent exacerbations, it was found to be used frequently among less severe patients11–13. From empirical data, this study found that COPD patients who initiated ICS therapy without prior exacerbation had higher total healthcare costs. Inappropriate use of ICS in treating COPD without evidence of prior exacerbation may have significant economic impact.

This study had several limitations. First, due to the retrospective cohort design, study findings can only be interpreted as association and not causality. The study can only describe the association between healthcare costs and ICS use as according to guidelines, and the differences observed cannot be determined as purely due to whether or not the patient used ICS according to guideline recommendations. Second, all medical conditions were identified based solely on administrative claims data using ICD‐9‐CM diagnosis codes which are subject to coding inaccuracy since no validation with patient medical records was performed. Medical problems not captured in the database therefore were not recognized. It is also methodologically challenging to separate COPD patients from asthmatic patients who would benefit from ICS therapy21. Although we excluded patients who had observed medical services associated with asthma, the study sample might include some asthmatic patients. Finally, only direct medical payments for covered services were examined, as indirect costs such as opportunity costs and loss of productivity were not available in the administrative claims database.

Conclusions

Among COPD patients with Medicare supplemental insurance, patients who initiated ICS therapy without prior exacerbation appear to incur higher healthcare costs. These findings suggest that ICS initiation in COPD without evidence of exacerbation as consistent with evidence‐based guidelines is associated with adverse economic consequences.

Transparency

Declaration of funding

The funding of this project was provided by Novartis Pharmaceuticals. The sponsor had no role in the design, methods, data collection, analysis, or preparation of the manuscript. The statements contained in this paper are solely those of the authors and no endorsement by Novartis should be inferred or implied.

Declaration of financial/other relationships

C.A.P. is employed by Novartis and owns company stocks. S.C., N.W., K.F., and L.B. are employed by United BioSource Corporation, and no financial conflicts need to be declared.

Acknowledgments

No assistance in the preparation of this article is to be declared.