Healthcare resource utilization and costs in working-age patients with high-risk atherosclerotic cardiovascular disease: findings from a multi-employer claims database

Abstract

Objective:

Patients with coronary artery disease with diabetes, a history of acute coronary syndromes, cerebrovascular atherosclerotic disease, or peripheral arterial disease are at particularly high risk for a cardiovascular (CV) event and can be defined as having high-risk atherosclerotic cardiovascular disease (ASCVD). The objective of this study is to examine healthcare resource utilization (HRU) and total healthcare costs (THC) for patients with ASCVD in a commercially insured population.

Methods:

A retrospective cohort study was conducted using a large, US employer-based, claims database. Patients with an ASCVD diagnosis between October 1, 2008 to September 30, 2009 who met eligibility requirements were included. All-cause and ASCVD-related HRU and THC for the first and second year of follow-up were examined for all patients and by the number of arterial beds affected. Adjusted THC were compared across groups with and without polyvascular disease.

Results:

The analysis included 152,290 patients with ASCVD. Use of CV-related medications, hospitalizations, and office visits were highest among patients with three arterial beds affected. Mean all-cause THC for patients with ASCVD were ∼$19,000 per patient in Year 1 or Year 2, with medical costs as the main driver. ASCVD-related THC were also similar for Year 1 ($8699) and Year 2 ($7925) across all patients. Adjusted all-cause and ASCVD-related THC for both years were greatest for patients with three affected arterial beds compared with one or two affected beds (p < 0.001 for each comparison).

Conclusions:

This is the first study in a managed care setting to systematically estimate all-cause and ASCVD-related THC for an aggregated population of ASCVD patients at high risk for a CV event. The economic burden of ASCVD in working-age patients in the US is substantial. Significantly higher HRU and costs were found in patients with polyvascular disease compared with those with only one affected bed.

Keywords::

• Atherosclerotic cardiovascular disease
• Cost
• Resource use
• Commercial payer

Introduction

The presence of atherosclerosis and development of atherothrombosis is the underlying cause of atherosclerotic cardiovascular disease (ASCVD) including coronary artery disease (CAD), acute coronary syndrome, cerebrovascular atherosclerotic disease (CeVAD), or peripheral arterial disease (PAD), all of presumed atherosclerotic origin1^,2. ASCVD is the leading cause of morbidity and mortality in the industrialized world2. In the US, an estimated 15.4 million adults have CAD. Additionally, ∼8.5 million individuals suffer from PAD and 7 million individuals have had a stroke3. Despite treatment, ASCVD patients remain at high risk for cardiovascular (CV) events4^,5, which put significant economic burden on the US healthcare system. The US spends an estimated $126 billion per year direct costs for patients with CAD6, $21 billion for PAD7, $19 billion for stroke8, and $75 billion for ACS9.

Atherosclerosis is a disease that can affect multiple arterial beds (e.g. coronary, cerebral, or lower extremity), and atherothrombotic ischemic events may occur in one or more different arterial beds within an individual patient. Patients with ASCVD share risk factors and systemic inflammatory pathophysiology, as well as similar comorbidity profiles including high rates of hyperlipidemia or hypertension and use of guideline-recommended drug therapy10–13. The prevalence of atherosclerotic involvement in multiple arterial beds implies that atherothrombotic ischemic events are best treated as a systemic disease14–17. Thus, treating an ischemic event in one arterial bed may provide primary prevention benefits for other arterial beds that have not yet demonstrated a CV event17.

Diabetes is a frequent comorbidity in patients with ASCVD, with ∼44% also suffering from diabetes18. Individuals with diabetes mellitus are also considered to be at increased risk of ASCVD1^,19^,20 because diabetes is a significant predictor of future ischemic events and leads to increased ASCVD-related morbidity and mortality21–25. CAD patients with comorbid diabetes are considered to be at highest risk of a CV event20^,26–28. Thus, this study focuses on patients with established ASCVD at high risk for subsequent CV events, including coronary artery disease with diabetes (CADD), PAD, a history of ACS within the past year, or CeVAD.

While studies have evaluated costs associated with ACS, CAD, PAD, or CeVAD individually, no study has evaluated healthcare resource utilization (HRU) and costs of high-risk ASCVD patients as a group. Given that a high percentage of ASCVD patients have more than one affected arterial bed18, simply adding the costs associated with ACS, PAD, CAD, and CeVAD will increase the risk of double counting costs. Data systematically estimating the total direct costs associated with patients at high risk for a CV event are limited, particularly from the US managed care perspective or among working age patients. The objective of this study is to examine HRU and total healthcare costs (THC) associated with ASCVD in a commercially insured US population.

Patients and methods

A retrospective cohort study was conducted using a large, employer-based, US, administrative claims database. Truven Health Analytics MarketScan Commercial Claims and Encounters (CCAE) database captures person-specific clinical utilization, expenditures, and enrollment across inpatient, outpatient, prescription drug, and carve-out services from a selection of large employers and health plans for covered adults between the ages of 18–64 years29. This dataset links costs and encounter data to detailed patient information across sites and types of providers, and over time. The annual medical databases include private sector health data from ∼100 payers with, historically, more than 500 million claim records available. Data from the MarketScan CCAE database from October 1, 2007 to September 30, 2011 were included in the analysis.

Sample selection

Individuals were included in the study if they had a diagnosis of ASCVD and were between 18–65 years of age during the identification period from October 1, 2008 to September 30, 2009 (i.e., identification window). In this study, high-risk ASCVD was specifically defined as having a diagnosis with at least one of the following: (1) A history of ACS (i.e., ≥30 days through 365 days after discharge for ACS) based on International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) codes: 410.xx and 411.1; (2) CeVAD based on ICD-9-CM codes: 362.3, 433.00–433.91, 434.00–434.91, 435.0–435.9, 436, 437.x, excluding 437.2 and 437.8; (3) PAD based on ICD-9-CM codes: 440.2, 440.3, 440.4, and 443.9, or ICD-9-Procedure codes: 38.08, 38.13, 38.18, 39.25, 39.26, 39.29, 39.50, 39.90, 00.55, 84.10, 84.11, 84.12, and 84.15, or a pharmacy claim for cilostazol or pentoxifylline; or (4) Patients with diabetes mellitus (250.xx or with related diabetic medication) with known CAD (CADD) based on 410.xx–414.00, excluding ACS codes of 410.xx and 411.1, which fall under the diagnosis of history of ACS.

Index date was defined as the earliest diagnosis date of the above four diseases during the identification window (October 1, 2008 to September 30, 2009). Individuals with a minimum 12-month pre-index and 24-month post-index insurance eligibility were included in the analysis. After patients were grouped according to their index condition (ACS, CeVAD, PAD, or CADD), as observed during the identification window, a 1-year look-back period from the index claim was evaluated, and patients were categorized into three groups based on the number of areas of co-existent established arterial disease including coronary, cerebral, and/or lower extremity (i.e., one, two, or three affected arterial bed(s)) defined as patients with an additional, alternative ASCVD claim type during the 1-year look-back period:

• One affected arterial bed: defined as ACS-only, CeVAD-only, PAD-only, CADD-only, or CADD + ACS;

• Two affected arterial beds: defined as ACS + CeVAD, ACS + PAD, (ACS + CADD) + CeVAD, (ACS + CADD) + PAD, CeVAD + PAD, CeVAD + CADD, or PAD + CADD; and

• Three affected arterial beds: ACS + CeVAD + PAD, CADD + CeVAD + PAD, or (ACS + CADD) + CeVAD + PAD.

Study end-points

Data on patients’ baseline demographic characteristics, comorbidities, and evidence-based CV risk reduction medication use were presented for overall ASCVD patients and by the number of arterial beds affected. Comparisons were made between patients with more than one affected arterial bed (two or three beds) to those with one affected arterial bed. Healthcare resource utilization including frequency of hospitalization, physician visits, and emergency department (ED) visits was described for all patients with ASCVD and by the number of arterial beds affected for the first and second year of follow-up.

All-cause and ASCVD-related THC costs for the first and second year of follow-up were examined for patients with ASCVD overall and by the number of arterial beds affected. All-cause THC were defined as all healthcare costs associated with inpatient care, outpatient care, and pharmacy services. ASCVD-related THC were defined as the costs of services and/or medications linked to any ASCVD diagnosis as previously described. Specific ASCVD-related inpatient costs were estimated including hospitalizations for primary diagnoses of UA, MI, stroke, or PAD, and costs associated with ASCVD related procedures such as amputation or coronary, carotid, or peripheral revascularization. First- and second-year all-cause and ASCVD-related THC were compared across groups with one, two, or three affected arterial beds.

Statistical methods

For all unadjusted analyses, t-tests were used for comparison of continuous variables and Chi-square tests were used for comparison of categorical variables. A generalized linear model (GLM), which assumed a gamma distribution and log link function, was used for Year 1 and Year 2 all-cause or ASCVD-related THC analyses adjusting for baseline demographics, prior comorbidities (yes/no), prior CV risk reduction medication use (yes/no), and baseline HRU. Model-based adjusted mean costs and associated 95% confidence intervals (CIs) for different arterial beds and p-values from comparisons among one, two, and three affected arterial beds were also computed from the GLM models. A p-value <0.05 was considered statistically significant. All analyses were completed using SAS version 9.2 (SAS Institute, Inc., Cary, NC).

Results

A total of 24 million individuals between the ages of 18–64 years were included in the MarketScan Commercial database between October 1, 2008 and September 30, 2009. Among these, 525,893 (2.2%) had at least one ASCVD diagnosis including a history of ACS, CeVAD, PAD or CADD. Including only those with a minimum 12-month pre-index and 24-month post-index insurance eligibility, the final analytical sample consisted of 152,290 patients with ASCVD.

Demographic and clinical characteristics

Among the 152,290 patients with ASCVD, 89% had one arterial bed affected, 10% had two arterial beds affected, and 1% had three or more arterial beds affected (Figure 1). The mean age for ASCVD patients was 54.5 (SD = 7.5) years. Patients with one bed affected (mean = 54.3 (SD = 7.7) years) were likely to be younger compared with those with two (mean = 56.5 (SD = 5.7) years) or three (mean = 57.1 (SD = 5.3) years) beds affected (Table 1). Among all patients with ASCVD, hypertension and hypercholesterolemia were the most frequent comorbidities at baseline. The frequency of all comorbidities evaluated was significantly greater in patients with two or three arterial beds affected compared with those with only one arterial bed affected (p < 0.001 for each comparison).

Figure 1. Patient disposition. Patients with ASCVD at high risk most frequently presented with one arterial bed affected. ACS, acute coronary syndrome; CADD, coronary artery disease with diabetes; CeVAD, cerebrovascular atherosclerotic disease; PAD, peripheral arterial disease.

Table 1. Baseline demographic and clinical characteristics.

		One affected arterial bed
Characteristic	All ASCVD patients	CeVAD only	PAD only	CADD only	ACS only	CADD + ACS	Total	Two affected arterial beds*	Three affected arterial beds*
n	152,290	60,159	28,197	36,298	5,381	6,243	136,278	14,568	1444
Age, Mean (SD)	54.5 (7.5)	53.5 (8.4)	53.6 (8.0)	56.2 (5.8)	53.4 (7.0)	55.2 (6.1)	54.3 (7.7)	56.5 (5.7)†	57.1 (5.3)†
Male gender, %	54.4	44.3	48.9	67.4	74.5	68.6	53.7	60.1†	60.1†
Baseline comorbidities, %
Diabetes	47.1	17.4	28.8	100.0	2.9	100.0	45.0	63.5†	81.2†
Hypertension	56.8	49.6	50.9	63.6	57.2	73.4	55.0	71.2†	80.5†
Hypercholesterolemia	54.5	46.5	47.3	64.9	60.6	69.3	53.2	65.7†	69.9†
Renal Disease	4.4	2.3	4.1	5.6	2.6	7.4	3.8	8.8†	16.0†
Hepatic Impairment	3.5	3.1	3.4	3.7	2.7	3.8	3.3	4.7†	5.6†
Congestive Heart Failure	7.7	3.4	3.9	10.2	15.5	20.6	6.6	16.4†	24.9†
Myopathy	6.1	6.6	6.8	4.6	3.6	5.3	5.9	7.5†	9.7†
Baseline medications, %
Prescription Antiplatelets	19.5	8.7	13.8	21.8	71.1	47.6	17.5	35.7†	49.4†
Oral Anticoagulants	5.4	5.1	5.0	4.8	5.3	5.9	5.0	8.4†	8.9†
Prescription NSAIDs	22.2	21.9	25.5	21.6	17.3	20.9	22.3	21.0†	19.7^‡
Lipid Lowering Agents
Statins	47.7	34.2	36.5	64.9	77.4	68.2	46.1	60.3†	66.8†
Niacin	4.3	2.3	2.6	7.8	4.3	6.4	4.1	6.3†	7.6†
Fibrates	7.6	4.1	5.6	13.1	4.9	13.2	7.3	10.3†	12.8†
Ezetimibe	11.1	6.9	7.9	18.9	5.2	13.3	10.5	15.9	19.3†
Bile Acid Sequestrants	1.3	1.0	1.0	1.6	0.6	1.3	1.2	1.9†	2.4†
Any Lipid Altering Agent	51.3	37.4	40.1	69.4	78.8	71.7	50.0	64.4†	70.4†
Anti-Diabetic Drugs
Insulin	10.6	3.4	7.5	20.4	0.9	24.2	9.7	17.8†	27.1†
Oral Diabetic Medication	27.8	10.3	17.7	60.0	2.5	57.0	26.9	34.8†	42.4†
Cardiovascular Agents
CA Channel Blockers	17.6	14.5	16.4	20.9	14.0	21.1	16.9	23.2†	28.3†
Beta Blockers	35.1	20.8	21.9	51.3	76.6	63.9	33.3	49.8†	58.9†
ACE Inhibitors	29.7	19.7	23.1	41.2	52.6	48.2	28.8	37.3†	42.9†
ARBs	17.4	13.7	15.5	23.2	12.0	21.1	16.9	21.1†	25.6†
Diuretics	20.7	15.5	19.1	26.2	15.9	29.0	19.7	28.2†	35.7†
Nitrates	10.9	2.6	3.6	16.9	46.4	35.6	9.8	19.5†	27.6†
Peripheral arterial claudication medications
Cilostazol	1.7	0.0	6.5	0.0	0.0	0.0	1.4	4.6†	8.2†
Pentoxifylline	1.4	0.0	6.7	0.0	0.0	0.0	1.4	1.8†	2.4^‡
Mean # of CV-related medications  per patient (Mean, SD)	3.5 (3.2)	2.1 (2.3)	2.7 (2.7)	5.0 (3.3)	4.5 (2.5)	5.9 (3.8)	3.3 (3.1)	4.8 (3.7)†	6.0 (4.1)†

ACE, angiotensin converting enzyme; ACS, acute coronary syndrome; ARB, angiotensin II receptor blocker; CA, calcium; CADD, coronary artery disease with diabetes; CeVAD, cerebrovascular atherosclerotic disease; NSAID, non-steroidal anti-inflammatory drug; PAD, peripheral arterial disease; SD, standard deviation.

*p-values are for the comparison vs one arterial bed. T-tests were used to calculate p-values for continuous variables and chi-square tests were used to calculate p-values for categorical variables.

†p < 0.001.

‡p < 0.05.

Among all ASCVD patients with one arterial bed affected, the mean age was 54.3 (SD = 7.7) years and ranged from 53.4 (SD = 7.0) years in those with a history of ACS only to 56.2 (SD = 5.8) years in patients with CADD only. Similar to all patients with ASCVD, hypertension and hypercholesterolemia were the most frequent comorbidities at baseline among those with one arterial bed affected, and consistent across each of the types of atherosclerotic disease. The frequency of comorbid congestive heart failure (CHF) varied widely among patients with one arterial bed affected, ranging from 20.6% in those with a history of ACS and CADD to only 3.4% in those with CeVAD only.

At baseline, patients with ASCVD were most frequently prescribed CV agents, lipid-lowering therapies, and anti-diabetic agents (Table 1). Baseline use of guideline-directed therapies was significantly greater in patients with two or three arterial beds affected compared with those with one arterial bed affected (p < 0.05 for each comparison). Among the CV therapies prescribed in patients with ASCVD, beta blockers and ACE inhibitors were the most frequently used followed by diuretics. Statins (48%) were the most frequently used lipid-lowering therapy.

Moderate use of CV agents was seen across all ASCVD conditions at baseline. Use of statins varied with ACS only and CADD + ACS patients having the most frequent use, 77% and 68%, respectively, and patients with PAD only or CeVAD only having the lowest use, 37% and 34%, respectively. Baseline medication use for ACS patients is high because it reflects medications used within the first 30 days post-discharge based on the index date, as defined by the study inclusion criteria. Similarly, use of antidiabetic therapies varied, with highest use among those with a primary diagnosis of CADD.

Healthcare resource utilization

Both all-cause and ASCVD-related HRU among patients with ASCVD is substantial (Table 2). Among all ASCVD patients in Years 1 and 2, respectively, 23% and 15% of patients had an all-cause hospitalization and 17% and 10% of patients had an ASCVD-related hospitalization. There were a mean of 22 (SD = 21) and 18 (SD = 21) all-cause office visits per patient and six (SD = 7.6) and four (SD = 6.8) ASCVD-related office visits in Years 1 and 2, respectively. A total of 17% of ASCVD patients had an ASCVD-related ED visit in Year 1 and 11% had an ASCVD-related ED visit in Year 2. Overall, among patients with ASCVD, both all-cause and ASCVD-related HRU was high in the first and second year of follow-up.

Table 2. All-cause and ASCVD-related healthcare resource utilization by number of affected arterial beds and year of follow-up.

		One affected arterial bed
Resource utilization	All ASCVD patients	CeVAD only	PAD only	CADD only	ACS only	CADD + ACS	Total	Two affected arterial beds*	Three affected arterial beds*
n	152,290	60,159	28,197	36,298	5,381	6,243	136,278	14,568	1444
First year healthcare resource utilization
All-cause resource use
% patients with ≥ 1 hospitalization	22.7	21.9	19.5	20.4	18.4	38.7	21.6	31.8†	39.3†
# of hospitalizations per patient, mean (SD)	0.3 (0.8)	0.3 (0.8)	0.3 (0.8)	0.3 (0.7)	0.3 (0.7)	0.6 (1.1)	0.3 (0.8)	0.6 (1.1)†	0.8 (1.5)†
Hospitalization LOS (days), mean (SD)	1.6 (6.7)	1.5 (6.4)	1.5 (6.2)	1.2 (4.8)	1.0 (5.4)	2.3 (7.8)	1.4 (6.0)	2.8 (9.9)†	4.3 (18.5)†
# of doctor office visit, mean (SD)	22.0 (21.0)	20.7 (19.5)	21.6 (22.1)	21.6 (19.7)	18.4 (16.7)	24.7 (21.9)	21.2 (20.2)	27.5 (25.7)†	34.0 (31.5)†
% patients with ≥ 1 ED visit	33.4	36.8	26.5	28.7	32.3	47.8	32.9	38.1†	41.8†
# of ED visits per patient, mean (SD)	0.3 (0.5)	0.4 (0.5)	0.3 (0.4)	0.3 (0.5)	0.3 (0.5)	0.5 (0.5)	0.3 (0.5)	0.4 (0.5)†	0.4 (0.5)†
ASCVD-related resource use
% patients with ≥ 1 hospitalization	17.3	15.1	12.7	17.0	15.0	36.7	16.1	27.1†	35.7†
# of hospitalizations per patient, mean (SD)	0.3 (0.7)	0.2 (0.6)	0.2 (0.6)	0.2 (0.6)	0.2 (0.6)	0.6 (1.0)	0.2 (0.6)	0.4 (1.0)†	0.7 (1.4)†
Hospitalization LOS (days), mean (SD)	1.1 (5.7)	0.9 (4.8)	1.0 (4.9)	1.0 (4.3)	0.8 (5.0)	2.1 (7.4)	1.0 (4.9)	2.3 (8.9)†	3.8 (18.1)†
# of doctor office visit, mean (SD)	6.0 (7.6)	4.0 (6.1)	4.5 (7.2)	7.9 (6.7)	7.4 (9.1)	10.6 (9.3)	5.6 (7.1)	9.3 (10.1)†	13.1 13.4)†
% patients with ≥ 1 ED visit	17.0	17.1	8.9	16.9	18.2	36.6	16.3	23.2†	28.8†
# of ED visits per patient, mean (SD)	0.2 (0.4)	0.2 (0.4)	0.1 (0.3)	0.2 (0.4)	0.2 (0.4)	0.4 (0.5)	0.2 (0.4)	0.2 (0.4)†	0.3 (0.5)†
Second year healthcare resource utilization
All-cause resource use
% patients with ≥ 1 hospitalization	14.5	11.5	13.4	15.6	13.2	22.0	13.5	22.5†	29.8†
# of hospitalizations per patient, mean (SD)	0.2 (0.7)	0.2 (0.6)	0.2 (0.7)	0.2 (0.7)	0.2 (0.6)	0.4 (0.9)	0.2 (0.7)	0.4 (1.0)†	0.6 (1.2)†
Hospitalization LOS (days), mean (SD)	1.0 (5.7)	0.8 (5.0)	1.0 (5.6)	1.0 (5.2)	0.8 (4.8)	1.4 (5.6)	0.9 (5.2)	1.9 (8.3)†	3.0 (12.3)†
# of doctor office visit, mean (SD)	18.5 (20.7)	16.7 (18.6)	18.2 (21.5)	19.7 (20.5)	12.7 (14.8)	20.1 (21.9)	17.8 (19.8)	23.7 (25.4)†	29.2 (30.8)†
% patients with ≥ 1 ED visit	26.1	23.6	23.4	26.8	27.7	36.4	25.2	33.2†	38.2†
# of ED visits per patient, mean (SD)	0.3 (0.4)	0.2 (0.4)	0.2 (0.4)	0.3 (0.4)	0.3 (0.5)	0.4 (0.5)	0.3 (0.4)	0.3 (0.5)†	0.4 (0.5)†
ASCVD-related resource use
% patients with ≥ 1 hospitalization	9.8	5.4	7.8	12.8	10.2	19.5	8.7	18.2†	25.4†
# of hospitalizations per patient, mean (SD)	0.1 (0.5)	0.1 (0.4)	0.1 (0.5)	0.2 (0.6)	0.1 (0.5)	0.3 (0.8)	0.1 (0.5)	0.3 (0.8)†	0.5 (1.1)†
Hospitalization LOS (days), mean (SD)	0.7 (4.7)	0.4 (3.7)	0.7 (4.5)	0.8 (4.6)	0.5 (3.7)	1.2 (5.1)	0.6 (4.2)	1.5 (7.5)†	2.5 (10.6)†
# of doctor office visit, mean (SD)	4.0 (6.8)	2.1 (4.9)	3.1 (6.7)	6.4 (7.0)	3.1 (4.8)	7.1 (7.5)	3.7 (6.3)	6.8 (9.0)†	10.2 (11.9)†
% patients with ≥ 1 ED visit	10.6	5.5	7.2	15.2	13.1	24.4	9.6	18.3†	24.4†
# of ED visits per patient, mean (SD)	0.1 (0.3)	0.1 (0.2)	0.1 (0.3)	0.2 (0.4)	0.1 (0.3)	0.2 (0.4)	0.1 (0.3)	0.2 (0.4)†	0.2 (0.4)†

ACS, acute coronary syndrome; ASCVD, atherosclerotic cardiovascular disease; CADD, coronary artery disease with diabetes; CeVAD, cerebrovascular atherosclerotic disease; ED, emergency department; LOS, length of stay; PAD, peripheral arterial disease; SD, standard deviation.

*p-values are for the comparison vs one arterial bed. Chi-square tests were used to calculate p-values for percentage patients with ≥ 1 hospitalization and percentage patients with ≥ 1 ED visit. Generalized Linear Model with assumption of negative binomial and log link function were used to calculate p-values for number of hospitalizations per patient, hospitalization LOS (days), number of doctor office visit, and number of ED visits per patient.

†p < 0.001

Among patients with one arterial bed affected, all-cause and ASCVD-related HRU was generally similar across each type of ASCVD (Table 2). However, patients with CADD plus ACS consistently used more all-cause and ASCVD-related healthcare resources in both Years 1 and 2 of follow-up. Approximately 16% of patients with one arterial bed affected were hospitalized for ASCVD during Year 1. This rate was lowest for those with PAD only (13%) and highest in patients with CADD plus ACS, where 37% of patients were hospitalized for an ASCVD-related condition in Year 1. Similar findings were seen in ASCVD-related hospitalizations in Year 2; however, the overall frequency of ASCVD-related hospitalization among those with one arterial bed affected was lower (9%) than in Year 1.

Use of both all-cause and ASCVD-related healthcare resources was highest among patients with three arterial beds affected and lowest among patients with only one bed affected (p < 0.001 for each comparison) (Table 2). Compared with patients with one arterial bed affected, significantly more patients with two or three arterial beds affected had all-cause and ASCVD-related hospitalizations and ED visits in Year 1 and Year 2 (p < 0.001 for each comparison). Similarly, compared with patients with one arterial bed affected, patients with two or three arterial beds affected had significantly higher mean numbers of all-cause and ASCVD-related hospitalizations, office visits and ED visits per patient in each study year (p < 0.001 for each comparison). For both Years 1 and 2, all-cause and ASCVD-related hospital length of stay (LOS) was also significantly lower in patients with one arterial bed affected compared with those with two or three arterial beds affected (p < 0.001 for each comparison). The trend in greater HRU among patients with three arterial beds affected was seen for both Year 1 and Year 2 follow-up periods.

Total healthcare costs in patients with ASCVD

Mean all-cause THC for patients with ASCVD in both Year 1 and Year 2 were ∼$19,000 per patient with SDs for Year 1 of $39,350 and Year 2 of $47,533 (Table 3). In Year 1, outpatient care represented 51% of all-cause THC, while inpatient care contributed 33% followed by pharmacy costs (16%). The distribution of THC in Year 2 was similar to Year 1 (outpatient care 46%, inpatient care 34%), with a slightly greater proportion of THC represented by pharmacy costs (20%). The mean ASCVD-related THC were also similar for Year 1 ($8699; SD = $25,655) and Year 2 ($7925; SD = $31,309), and represented ∼ 46% of all-cause THC in Year 1 and 43% in Year 2. In both study years, inpatient care represented the majority (53% and 54%) of ASCVD-related total costs, followed by outpatient care (31% and 25%) and pharmacy costs (16% and 21%). ASCVD-related inpatient costs in Year 1 were greatest for hospitalizations associated with stroke or coronary revascularizations. Among all patients with ASCVD, in both the first and second year of follow-up, mean pharmacy costs per patient were greatest for lipid-lowering therapies followed by anti-hypertensives, anti-platelets, and anti-diabetic therapies (Table 4).

Table 3. All-cause and ASCVD-related total healthcare costs by number of affected arterial beds and year of follow-up.

		One affected arterial bed
Mean healthcare costs per patient	All ASCVD patients	CeVAD only	PAD only	CADD only	ACS only	CADD + ACS	Total	Two affected arterial beds*	Three affected arterial beds*
n	152,290	60,159	28,197	36,298	5,381	6,243	136,278	14,568	1444
First year healthcare costs ($), mean (SD)
All-cause costs
Outpatient cost	9698 (22,324)	8329 (17,571)	10,157 (24,687)	9479 (20,865)	7963 (14,873)	11,743 (24,893)	9155 (20,405)	13,885 (34,010)	18,687 (36,220)
Inpatient cost	6286 (25,843)	5056 (22,716)	5933 (24,381)	6130 (23,149)	5698 (23,153)	9365 (28,388)	5746 (23,504)	10,366 (35,349)	16,132 (73,526)
Pharmacy cost	3018 (4753)	2396 (4842)	2604 (4674)	3868 (4530)	2680 (3429)	4054 (4628)	2918 (4714)	3809 (4983)	4487 (5054)
Total healthcare cost	19,003 (39,350)	15,781 (33,244)	18,694 (40,437)	19,477 (35,809)	16,341 (31,963)	25,163 (43,614)	17,820 (36,062)	28,060 (55,674)*	39,306 (89,087)*
ASCVD-related costs
Outpatient cost	2727 (7862)	1464 (4673)	2507 (8914)	3484 (8447)	3744 (8998)	5041 (9769)	2472 (7331)	4635 (10,663)	7596 (15,295)
Inpatient cost	4587 (22,486)	2997 (18,167)	3805 (19,485)	5126 (21,408)	4651 (21,067)	8499 (26,798)	4049 (19,962)	8623 (31,765)	14,638 (72,477)
UA Hospitalization	499 (5941)	156 (3089)	209 (3628)	737 (7739)	1031 (7103)	2068 (10,765)	444 (5513)	945 (8850)	1187 (7783)
MI Hospitalization	449 (7202)	178 (3681)	328 (7645)	484 (7177)	908 (6589)	1599 (12,093)	384 (6349)	977 (12,525)	1203 (8911)
Stroke Hospitalization	1055 (12,338)	1689 (13,911)	373 (6608)	376 (7300)	305 (5085)	713 (9155)	967 (10,674)	1572 (15,084)	4087 (54,740)
PAD Hospitalization	691 (8727)	142 (5841)	1825 (11,891)	210 (4292)	140 (3446)	231 (3398)	512 (7118)	1936 (16,316)	5007 (22,968)
Coronary Revascularization	1051 (8846)	421 (5580)	537 (7728)	1699 (11,185)	1770 (9683)	3330 (14,869)	972 (8593)	1725 (10,775)	1728 (10,318)
Carotid Revascularization	118 (2155)	186 (2829)	38 (1055)	25 (842)	3 (181)	64 (1791)	100 (2026)	238 (2607)	646 (5757)
Peripheral Revascularization	326 (6958)	51 (1777)	895 (7612)	93 (2518)	45 (1277)	106 (2042)	239 (3929)	793 (7568)	3846 (55,296)
Amputation	55 (2900)	10 (1472)	105 (3259)	23 (1653)	15 (911)	29 (1208)	34 (1996)	225 (6975)	266 (4467)
Pharmacy cost	1385 (1838)	864 (1402)	1000 (1671)	2105 (2072)	1876 (1461)	2265 (2036)	1327 (1787)	1846 (2164)	2198 (2173)
Total healthcare cost	8699 (25,655)	5325 (19,973)	7312 (23,503)	10,715 (24,573)	10,270 (24,634)	15,805 (30,361)	7847 (22,926)	15,104 (36,078)*	24,432 (77,186)*
Second year healthcare costs ($), mean (SD)
All-cause costs
Outpatient cost	8531 (23,957)	7007 (19,226)	8675 (25,562)	8972 (23,812)	5946 (13,068)	10,636 (26,530)	8000 (22,128)	12,623 (34,455)	17,318 (45,526)
Inpatient cost	6220 (34,460)	4635 (30,982)	6288 (35,808)	6804 (35,027)	5720 (39,217)	8664 (34,939)	5782 (33,671)	9507 (38,663)	14,344 (54,511)
Pharmacy cost	3797 (6799)	3120 (7048)	3376 (6815)	4853 (6478)	3362 (5691)	4749 (6750)	3719 (6831)	4424 (6,569)	4851 (5611)
Total healthcare cost	18,547 (47,533)	14,762 (41,310)	18,339 (49,737)	20,628 (47,608)	15,028 (45,522)	24,048 (49,546)	17,501 (45,512)	26,554 (59,225)*	36,513 (79,663)*
ASCVD-related costs
Outpatient cost	1988 (6808)	923 (4115)	1657 (7132)	2876 (7815)	1886 (5831)	3844 (8983)	1767 (6336)	3682 (9530)	5694 (11,660)
Inpatient cost	4281 (29,120)	2284 (23,386)	4016 (28,727)	5496 (31,974)	4618 (37,106)	7507 (31,027)	3829 (28,024)	7720 (35,344)	12,222 (50,082)
UA Hospitalization	406 (5967)	164 (4158)	251 (5403)	543 (6342)	808 (8368)	1362 (10,459)	363 (5678)	746 (8079)	1029 (7180)
MI Hospitalization	529 (9528)	228 (5477)	471 (10,420)	662 (10,400)	1074 (12,904)	1293 (18,158)	476 (9288)	942 (11,215)	1391 (12,635)
Stroke Hospitalization	661 (11,550)	764 (13,822)	452 (8645)	494 (10,065)	293 (7477)	542 (7437)	599 (11,470)	1101 (11,835)	2088 (15,403)
PAD Hospitalization	573 (10,079)	148 (4471)	1306 (15,638)	349 (9463)	145 (6524)	321 (5582)	449 (9305)	1396 (12,688)	3987 (29,990)
Coronary Revascularization	741 (9952)	295 (6584)	566 (10,182)	1205 (12,622)	1165 (10,345)	1631 (11,591)	689 (9674)	1169 (12,287)	1294 (9370)
Carotid Revascularization	60 (2437)	63 (2924)	43 (1384)	43 (2526)	2 (138)	45 (1523)	50 (2445)	113 (1853)	418 (5268)
Peripheral Revascularization	270 (7294)	69 (3563)	681 (12,389)	168 (7498)	82 (2209)	167 (4478)	227 (7315)	563 (6560)	1400 (11,170)
Amputation	55 (3407)	12 (1132)	156 (6964)	26 (1391)	0 (0)	26 (1277)	46 (3346)	102 (2648)	501 (9851)
Pharmacy cost	1656 (2413)	1079 (1948)	1244 (2161)	2533 (2794)	2163 (2303)	2440 (2569)	1605 (2378)	2062 (2678)	2329 (2408)
Total healthcare cost	7925 (31,309)	4286 (24,662)	6918 (31,030)	10,905 (34,229)	8668 (38,600)	13,791 (33,660)	7202 (30,000)	13,464 (38,531)*	20,246 (54,400)*

ACS, acute coronary syndrome; ASCVD, atherosclerotic cardiovascular disease; CADD, coronary artery disease with diabetes; CeVAD, cerebrovascular atherosclerotic disease; MI, myocardial infarction; PAD, peripheral arterial disease; SD, standard deviation; UA, unstable angina.

*p < 0.0001 for the comparison vs one affected arterial bed computed based on genmod with gamma distribution and log link function. p-values were computed for all-cause and ASCVD-related total healthcare costs only, not for each cost component.

Table 4. Mean ASCVD-related pharmacy costs by medication class and year of follow-up.

		One affected arterial bed
Mean pharmacy costs per patient	All ASCVD patients	CeVAD only	PAD only	CADD only	ACS only	CADD + ACS	Total	Two affected arterial beds*	Three affected arterial beds*
n	152,290	60,159	28,197	36,298	5,381	6,243	136,278	14,568	1444
Mean first year pharmacy cost per patient ($)
Antiplatelets	257 (543)	139 (416)	142 (399)	296 (574)	902 (734)	686 (734)	237 (526)	415 (638)	544 (678)
Antihypertensives	274 (450)	207 (401)	218 (404)	386 (516)	242 (329)	356 (473)	265 (443)	343 (496)	398 (547)
Anti-anginals	123 (267)	89 (236)	93 (240)	175 (313)	135 (213)	168 (274)	119 (263)	159 (290)	186 (321)
Lipid lowering therapies	412 (681)	273 (537)	277 (551)	636 (830)	555 (673)	591 (756)	396 (667)	541 (770)	618 (829)
PAD therapies (cilostazol or pentoxifylline)	5 (42)	0 (7.1)	20 (78)	0 (9)	0 (6)	1 (20)	4 (37)	12 (64)	21 (97)
Anti-diabetics	245 (698)	86 (430)	155 (554)	559 (982)	12 (145)	416 (816)	239 (690)	292 (749)	376 (855)
Oral Anticoagulants	39 (685)	40 (653)	59 (984)	20 (291)	16 (310)	25 (243)	37 (646)	56 (996)	30 (242)
Mean second year pharmacy cost per patient ($)
Antiplatelets	300 (720)	173 (583)	181 (563)	366 (775)	942 (1134)	673 (918)	279 (706)	467 (809)	599 (811)
Antihypertensives	315 (566)	252 (524)	266 (523)	433 (640)	273 (449)	369 (556)	309 (562)	367 (600)	402 (579)
Anti-anginals	138 (333)	106 (303)	113 (311)	191 (382)	144 (277)	169 (322)	135 (330)	165 (364)	177 (330)
Lipid lowering therapies	510 (935)	352 (780)	359 (790)	776 (1123)	736 (1082)	673 (984)	496 (928)	624 (981)	682 (1023)
PAD therapies (cilostazol or pentoxifylline)	5 (41)	0 (12)	17 (77)	0 (12)	0 (9)	1 (18)	4 (37)	11 (61)	18 (82)
Anti-diabetics	311 (966)	118 (630)	212 (831)	702 (1365)	23 (290)	509 (1104)	307 (969)	335 (935)	394 (995)
Oral Anticoagulants	41 (858)	42 (891)	55 (1069)	27 (461)	25 (668)	21 (228)	39 (815)	64 (1216)	31 (300)

ACS, acute coronary syndrome; ASCVD, atherosclerotic cardiovascular disease; CADD, coronary artery disease with diabetes; CeVAD, cerebrovascular atherosclerotic disease; NSAIDS, non-steroidal anti-inflammatory drug; PAD, peripheral arterial disease; SD, standard deviation.

Among patients with one arterial bed affected, during the first year of follow-up, there was variation in all-cause and ASCVD-related THC across the primary ASCVD diagnoses, with CeVAD patients having the lowest costs and patients with CADD plus ACS having the highest costs (Table 3). All-cause THC in Year 2 followed a similar pattern to Year 1. Patients with diabetes (i.e., CADD alone or CADD plus ACS) had the highest all-cause and ASCVD-related total costs across both follow-up periods. Among patients with one arterial bed affected, ASCVD-related inpatient costs were greatest for hospitalizations associated with stroke or coronary revascularizations in Year 1. In the second year of follow-up, the mean costs per hospitalization were more evenly distributed with the highest ASCVD-related inpatient costs represented by coronary revascularization, stroke, PAD, and MI. CV-related pharmacy costs were also highest in patients with CADD (with or without ACS) in both follow-up periods.

For both Year 1 and 2 of follow-up, all-cause and ASCVD-related THC were significantly higher for patients with two or three arterial beds affected compared with those with one arterial bed affected (p < 0.0001 for each comparison) (Table 3). After controlling patient baseline clinical characteristics in the GLM model, adjusted means for all-cause and ASCVD-related THC in Years 1 and 2 were significantly highest for patients with three affected arterial beds and lowest for those with one arterial bed affected (Figure 2).

Figure 2. Adjusted mean all-cause and ASCVD-related healthcare cost per patient by number of affected arterial beds in year 1 and year 2. For years 1 and 2 of follow-up, adjusted mean all-cause and ASCVD-related total healthcare costs were lowest among patients with one arterial bed affected and highest for those with three arterial beds affected. Costs are adjusted for age, gender, baseline co-morbidity, baseline medication, baseline ED visit, ASCVD-related ER visit, baseline total number of hospital visits, baseline total number of ASCVD-related visits, total days in hospital, outpatient visit, and ASCVD-related outpatient visit in baseline period. ASCVD, atherosclerotic cardiovascular disease. *p < 0.0001 when compared with patients with one arterial bed affected.

Discussion

This study shows the high economic burden of high-risk ASCVD using a large-scale, real-world working age population in the US. To our knowledge, this is the first study to systematically estimate HRU and THC in US-managed care patients with high-risk ASCVD, a group of patients with CeVAD, PAD, a history of ACS, or CADD, who are at increased risk for a subsequent CV event. Regardless of the origin of the affected arterial beds (coronary, cerebral, or lower extremity), our study shows ASCVD patients share similar risk profiles (e.g., high rates of comorbid hypercholesterolemia and hypertension). Additionally, consistent with findings from the international Reduction of Atherothrombosis for Continued Health (REACH) registry18, our study also finds a substantial percentage of ASCVD patients have more than one affected arterial bed.

Our results also show that both all-cause and ASCVD-related HRU and THC are highest in those with multiple affected arterial beds compared with individuals with only one arterial bed affected. The significant increase in THC with each additional affected arterial bed emphasizes the importance of early control of risk factors and secondary prevention to prevent progression to polyvascular disease.

The current study shows an all-cause THC of ∼ $19,000 for ASCVD patients in each of the first and second years of follow-up. Most importantly, close to half of the all-cause THC are attributed to ASCVD-related THC. In the US in 2010, the average healthcare per-capita spending was $6125 per year for people of working age (19–64 years of age)30. Cost estimates from the present study demonstrate that ASCVD-related THC account for a large proportion of total healthcare expenditures at the national level.

To put these costs into perspective, using a prevalence of 2.2% based on the number of patients in the MarketScan Commercial database between October 1, 2008 and September 30, 2009 with at least one ASCVD diagnosis, in a hypothetical managed care plan of 1 million members, ∼22,000 members may have high-risk ASCVD, leading to an annual ASCVD-related THC of more than $191 million. Estimated costs could be greater depending on the proportion of ASCVD patients with more than one affected arterial bed, which may be particularly meaningful to the commercial, managed care population.

Limitations

There are limitations to the current study. Even though all-cause and ASCVD-related THC were adjusted for patient demographics, comorbid conditions, and illness severity, bias between comparison groups may still exist, as only factors that could be captured in a retrospective analysis of claims data were included in the GLM models to predict estimated mean costs. Additionally, the current patient population reflects working-age individuals in the US who are commercially insured by a managed care organization; thus, there may be a lack of generalizability of study results to individuals beyond the patient population represented in the claims database. The working age patient population analyzed in this study, however, is of particular importance as a large number of younger people suffer CV diseases. About 155,000 people (20%) who died from cardiovascular disease (CVD) in 2011 were younger than age 6531. Although studies have shown that direct medical costs of CVD were similar for patients younger than age 65 compared to their older counterparts, total costs in younger patients are much higher, due to the large impact of premature mortality31.

The analysis only included patients with at least 24 months of post-index insurance eligibility. Thus, there is a survivor bias, where patients who died before the end of the 24-month post-index period were excluded from the analysis. Although these patients may incur substantial end-of-life healthcare costs that were not captured in this analysis, they also have a shorter duration of survival and long-term treatment costs are also not included. Thus, the costs presented here may be a conservative estimate of first- and second-year all-cause and ASCVD-related THC.

There is a wealth of information to be gained from further evaluation of these data. Future analyses may include evaluating potential differences in THC due to ASCVD based on age or gender. Additionally, the presence of diabetes has been associated with higher HRU and costs related to CV events32–35. Additional analyses to further investigate the impact of comorbid diabetes on HRU and costs in this high-risk, real-world, working age population would help to provide additional clarity.

Recent American College of Cardiology (ACC)/American Heart Association (AHA) guidelines on the management of cholesterol call for expanded lipid-lowering treatment for all patients at high risk for a CV event with the intensity of therapy proportional to the level of CV risk. These new guidelines have expanded treatment recommendations for all ASCVD patients calling for use of high- or medium-intensity statins, regardless of LDL cholesterol level, which was previously recommended by the Third Report of the National Cholesterol Education Program Adult Treatment Panel (NCEP ATP III)36^,37. Future research is needed, particularly among working age patients with ASCVD at highest risk for a CV event, to understand the impact of these guideline changes on the rate of CV events and resulting HRU and treatment costs.

Conclusion

The economic burden of ASCVD in working-age patients in the US is substantial. Significantly higher resource use and costs were found in ASCVD patients with multiple arterial beds affected compared with those with only one affected arterial bed. Future studies of the effectiveness of novel therapeutic strategies aimed at reducing ASCVD-related events should focus specifically on controlling underlying risk factors, particularly the occurrence and severity of atherosclerosis. This is particularly important in patients who have already experienced an atherosclerotic event in whom early control of risk factors and secondary prevention are needed to prevent progression to polyvascular disease. Such studies are warranted because effective interventions may reduce the otherwise impending economic burden of ASCVD.

Transparency

Declaration of funding

This study was funded by Eli Lilly and Company.

Declaration of financial/other relationships

With the exception of YF, all authors are employed by Eli Lilly and Company.

Acknowledgments

The authors wish to acknowledge Dr Susan L Dennett of Strategic Health Outcomes, Inc. for her technical writing support, funded by Eli Lilly and Company. Portions of this study were presented as a poster at the International Society for Pharmacoeconomics and Outcomes (ISPOR) 18th Annual International Meeting, New Orleans, LA, May 2013 and at the ISPOR 19th Annual International Meeting, Montreal, QC, Canada, May 2014.