Impact of Dyspnea on Medical Utilization and Affiliated Costs in Patients with Acute Coronary Syndrome

Acute coronary syndrome (ACS) was responsible for > 1.1 million hospitalizations in 2007.¹ Acute coronary syndrome can be treated with a combination of antiplatelet and anticoagulant therapies. Current clinical practice guidelines recommend dual antiplatelet therapy with aspirin and clopidogrel or prasugrel for patients with ACS. A new antiplatelet therapy, ticagrelor, is being assessed and shows promise in clinical trials, but was not yet available in clinical practice during the execution of this research.² Ticagrelor is currently an experimental agent, pending a US Food and Drug Administration decision on July 20, 2011.

Despite the treatability of ACS, it remains costly, with a mean total cost of $22 538 for the first year.³ Treatment costs are greater due to the frequent ischemic events and bleeding complications caused by certain antiplatelet therapies. Analyzing the costs associated with adverse events caused by these therapies is important when making treatment choices.²

In the Platelet Inhibition and Patient Outcomes (PLATO) clinical trial, ticagrelor was found to have significantly lower rates of myocardial infarction and death compared with clopidogrel and placebo. In addition to bleeding, ticagrelor’s increased risk of dyspnea is well documented.^4-6 The PLATO trial found that dyspnea occurred in 13.8% of patients on ticagrelor compared with 7.8% of patients on clopidogrel.⁵ A multicenter, randomized, double-blind, double-dummy, parallel group study of the onset and offset of antiplatelet effects of ticagrelor compared with clopidogrel and placebo with aspirin as background therapy in patients with stable coronary artery disease (the ONSET/OFFSET trial) reported higher rates, with 38.6% of ticagrelor patients, 9.3% of clopidogrel patients, and 8.3% of patients on a placebo having a dyspnea episode.⁶ A 2008 study found a low incidence of dyspnea in patients receiving clopidogrel and/or aspirin. It also suggested that in the majority of patients with dyspnea was an underlying disease or condition that was not treatment related.⁷ Patients with a history of ACS presenting to the emergency room (ER) with dyspnea should receive immediate medical attention and are generally evaluated to determine the underlying cause.³ While dyspnea is often associated with heart failure, a number of other cardiac-, pulmonary-, and oncology-related diagnoses that cause dyspnea must be investigated as potential causes or complicating factors.⁷ Common procedures for evaluating dyspnea include chest radiograph, electrocardiogram (ECG), and B-type natriuretic peptide (BNP) test to assess dyspnea patients for possible heart failure.^3,8 The cost of dyspnea in patients with a history of ACS, including the cost of procedures used to assess heart failure, has not been fully investigated and may be important to health plan decision-makers when evaluating costs associated with each antiplatelet therapy.

The objective of this study was to determine the direct medical cost of dyspnea among a cohort of patients with a history of ACS using real-world data.

A cohort of patients who presented to the ER with a primary diagnosis of dyspnea were identified in a retrospective, claims-based study using the 2008 to 2009 Thomson Reuters MarketScan^® Commercial Claims and Encounters (Commercial) and Medicare Supplemental and Coordination of Benefits (Medicare) Databases. The Commercial Database contains the health care experience of patients covered by a variety of health plan types, over half of which are large, self-insured, employer-based plans. The Medicare Database contains health care claims for individuals with Medicare supplemental insurance paid for by employers. Combined, these databases contain detailed cost, use, and outcomes data from inpatient, outpatient, and outpatient prescription claims for > 73 million individuals in 2008 or 2009, prior to applying utilization or continuous-enrollment requirements. All database records were compliant with the Health Insurance Portability and Accountability Act and de-identified for research purposes.

Patients meeting the criteria for an episode of dyspnea from 2008 through 2009 were selected. The index date was defined as the first ER-originating outpatient claim for dyspnea (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code, 786.0x). Patients were also required to have 180 days of continuous medical enrollment before (designated the “baseline period”) index and ≥ 1 inpatient claim, ≥ 1 ER visit, or ≥ 2 outpatient claims with an ICD-9-CM diagnosis code for ACS (410.xx or 411.1x) in any position on the claim during either the baseline period or on the index date (Figure 1). There was no minimum period of eligibility post index; costs and procedures were only measured during the episode of dyspnea. An episode of dyspnea was defined as all ER and outpatient services on the day of an ER claim with a primary diagnosis of dyspnea, and any subsequent inpatient admissions on the day of or day following the ER visit.

View: Figure 1
Sample selection.

Procedure frequency was expressed as the percentage of eligible patients who received relevant procedures in the ER or other outpatient setting on the same day as the ER visit. Relevant procedures included ECG (ICD-9-CM codes, 89.51–89.53; HCPCS G0366-G0368, G0403-G0405; or CPT [Copyright 2010 American Medical Association. All rights reserved. CPT is a registered trademark of the American Medical Association] codes, 93000, 93005, 93010, 93012, 93014, 93040–93042, 93278, 0178T-0180T), chest radiograph (ICD-9-CM codes, 87.44, 87.49, 87.39; or CPT codes, 71010, 71015, 71020–71023, 71030, 71034–71035, 71040, 71060, 711101, 71111, 7600), BNP test (CPT code, 83880), and all other procedures occurring frequently among eligible patients. Inpatient admissions were defined by the percentage of eligible patients with an inpatient admission on the day of or day following their initial ER visit.

Health care costs were taken directly from adjudicated administrative claims in the Commercial and Medicare Databases. Health care expenditures captured both the patient and plan portions (including coordination of benefits, patient co-payments, deductibles, and co-insurance). Costs were expressed in 2009 US constant dollars, adjusted using the Medical Care component of the Consumer Price Index.⁹

Of 254 615 patients with an ER visit for dyspnea between 2008 and 2009 and 6 months of continuous enrollment prior to index, 8433 patients had been diagnosed with ACS within the past 180 days, 3504 in the Commercial population and 4929 in the Medicare population. Commercial patients were 44% female and had a mean age of 54.5 years (standard deviation [SD], 8.1 years). Medicare patients included slightly more females (49%) and, as expected, were older (mean age, 79.0 years). The majority of both populations (79.7% of Commercial and 87.8% of Medicare) were enrolled in noncapitated plans. More than half (62.4%) of Commercial patients were enrolled in preferred provider organizations. The majority of Medicare patients were divided between preferred provider organizations (41.4%) and comprehensive insurance (44.0%).

The average cost per dyspnea episode was $6958, of which $1621 were outpatient costs associated with an ER visit (SD, $3269) (Table 1). Along with physician services, assessment of dyspnea often included ECG (71.3%), chest radiograph (75.9%) and, in some cases, a BNP test (14.9%) or chest computed axial tomography (CAT) scan (12.2%). These procedures constituted 35.1% of the average outpatient costs. Major cardiovascular procedures were rare (2.1%), with mean payments of $1905 per episode. More than one-fourth (25.8%) of dyspnea ER visits led to an inpatient stay; the average cost per patient of an inpatient stay was $20 693.

Results were similar between the Commercial and Medicare populations. Total health care cost for the Commercial population was higher than that for the Medicare population ($9426 vs $5213, respectively; Tables 2, 3). A greater proportion of ER visits led to an inpatient stay for the Commercial population than for the Medicare population (28.3% vs 24.0%, respectively), and those visits were more costly ($25 279 vs $16 886, respectively; Tables 2, 3). Costs of receiving ECG, chest radiograph, and CAT scan imaging of the heart were more than twice as expensive in the Commercial population as in the Medicare population (Figure 2), while prevalence of these procedures remained similar (Figure 3).

View: Figure 2
Total costs (insurer and out-of-pocket) of recommended procedures rid="IFN1" >a among ACS patients presenting with dyspnea.
^aRecommended procedures are those defined in the literature as baseline procedures that should be presented for patients with a history of ACS presenting to the ER with dyspnea.3 ^,8 Abbreviations: ACS, acute coronary syndrome; BNP, B-type natriuretic peptide; CAT, computerized axial tomography; ER, emergency room.

View: Figure 3
Prevalance of recommended procedures rid="IFN3" >a among ACS patients presenting with dyspnea.
^aRecommended procedures are those defined in the literature as baseline procedures that should be presented for patients with a history of ACS presenting to the ER with dyspnea.3 ^,8 Abbreviations: ACS, acute coronary syndrome; BNP, B-type natriuretic peptide; CAT, computerized axial tomography; ER, emergency room.

Antiplatelet options for the treatment of ACS are expanding. The treatment choice should be based on clinical practice guidelines and provider choice, as well as consideration of the impact on total health care cost. The adverse event profiles of new ACS treatments should also be taken into account when making ACS treatment decisions, including how adverse events factor into the final medical- and pharmacy-related costs of the treatments.² It is well documented that ticagrelor has an increased risk of dyspnea compared with clopidogrel.^4-6 Dyspnea is a serious adverse event requiring immediate medical attention, including additional evaluation, as it can be viewed as an indicator for further cardiac or pulmonary complications in patients with a history of ACS.^3,7

While the PLATO study found that ticagrelor offered many benefits over clopidogrel or placebo, including reduction in all-cause mortality and key outcomes (cardiovascular death and myocardial infarction), it also found that ticagrelor’s rate of dyspnea is 6.0% greater than that of clopidogrel (13.8% vs 7.8%, respectively).⁵ A previous European Union economic analysis found ticagrelor to be cost-effective compared with clopidogrel; however, this analysis included only efficacy data and excluded safety data from PLATO.¹⁰ While it is unknown if incorporating this additional safety data would have reversed the cost-effectiveness of ticagrelor, it is clear that the safety profile of ticagrelor is a necessary component of its economic evaluation. Pairing PLATO’s findings with the results of this current analysis allows for assessment of the impact of using ticagrelor in place of clopidogrel. Assuming that a health plan has 1 million members, of whom 3.1% have a recent history of ACS¹ and 36% are treated with either ticagrelor or clopidogrel following the ACS event,¹¹ treating 11 160 patients with ticagrelor instead of clopidogrel will result in an additional 670 cases of dyspnea, resulting in a total additional cost of $4 660 278. It is unclear if this additional burden due to dyspnea would be offset by the superior efficacy profile of ticagrelor. Further economic analysis of ticagrelor that incorporates safety data is warranted alongside additional clinical studies on the causes and severity of dyspnea among patients using ticagrelor.¹² Similarly, future economic analysis of ticagrelor may consider subgroup analysis using efficacy and safety data, as well as costs within the North American population, given the lack of efficacy of ticagrelor in this population.⁵ Due to the future loss of US market exclusivity for Plavix^® (Bristol-Myers Squibb/Sanofi Pharmaceuticals, Bridgewater, NJ) the generics entry will lead to a price reduction, which should be included in economic evaluations of ACS therapies.

Although prevalence of dyspnea is generally rare and likely due to underlying disease characteristics rather than ACS treatments, it is a serious condition warranting immediate attention. However, dyspnea does not have a widely accepted treatment algorithm.¹³ Wang et al³ discussed the difficulties that physicians face when evaluating dyspnea patients for heart function and suggested that physicians frequently use chest radiograph, ECG, and BNP testing to identify heart failure. The current study results are consistent with this suggestion and found that among the study’s sample of 8433 patients, 71.3% received the recommended ECG and 75.9% received chest radiograph in addition to their physical examination. The suggested BNP test was administered less frequently, in only 14.9% of the study sample. In addition, 12.2% of patients received CAT scan imaging of the heart and 79.8% received other ER procedures, even though these procedures were not mentioned in prior literature.

This study is a retrospective claims-based study and has several limitations. First, episodes of ACS and all events during ER or outpatient admission were identified by ICD-9-CM diagnosis and procedure codes, which are recorded to support providers’ claims for reimbursement and may be subject to classification error. There is no way to evaluate the potential impact of this limitation. However, claims databases are frequently used in economic analysis and provide a valuable source of data regarding the payment for services.

Second, the MarketScan^® Databases represent a nonprobability convenience sample of individuals with either employer-sponsored health insurance or Medicare that form the basis of the Commercial and Medicare Databases. Thus, this study may not be generalizable to all individuals, particularly uninsured patients or Medicaid beneficiaries.

Third, because during the execution of this research ticagrelor was not available in clinical practice, there is no way to assess the costs of dyspnea associated with a patient prescribed ticagrelor compared with other antiplatelet therapies in a real-world setting. Additionally, this study did not include an active comparator in the analysis. This study instead generalized its population to all patients with a history of ACS, regardless of the treatments that they had been prescribed. Future analysis regarding the real-world impact of ticagrelor on patient health care costs will be important.

The management of dyspnea is expensive, with costs per patient averaging $6958 per dyspnea episode. As more treatment options become available for ACS in the United States, further research is necessary to assess all potential costs of these treatment options to allow health plan payers to assess all risks and benefits of these regimens.