ABSTRACT
Background
Acute cardiovascular (CV) events have been evaluated in patients with specific comorbidities but have not focused on patients with hyperlipidemia or on the their long‐term costs.
Objectives
To evaluate incidence of CV events, costs, and resource utilization among patients with hyperlipidemia and baseline risk of CV disease (CVD).
Methods
Patients (age 18 to 64 years) diagnosed with hyperlipidemia or using lipid‐modifying medications were identified from administrative claims. Patients were categorized into 3 cohorts based on pre‐index clinical characteristics—secondary prevention (SP; history of CV event, n = 15 613); high risk (HR; CVD, n = 47 600); and primary prevention (PP; no CV event history or CVD, n = 60 637)—and followed up to 2 years after the CV event.
Results
During follow‐up, ≥1 new CV event occurred in 43.0% of the SP cohort, 33.9% of HR, and 20.9% of PP; and ≥3 new events occurred in 19.8% of the SP cohort, 12.9% of HR, and 5.5% of PP. Incremental total costs were $19 320 for SP, $20 003 for HR, and $17 650 for PP. Compared with patients with only 1 CV event, the mean 2‐year cost was 30% higher in patients with 2 CV events and 48% higher in patients with 3 CV events. Only 50% of HR patients (with or without CV events) received statins.
Conclusions
Patients with recurrent CV events had higher total health care costs during 24‐month follow‐up for each type of CV event. Total health care costs among patients with a CV event were higher for the initial as well as subsequent events. Statins and lipid‐modifying medications were significantly underutilized in all cohorts, despite the presence of CVD.
Introduction
As many as 84 million Americans have cardiovascular disease (CVD). Even among those without CVD at age 50 years, the lifetime risk of developing CVD exceeds 50% in men and 39% in women.1 Patients with a history of coronary heart disease (CHD) or CHD risk equivalents (eg, diabetes mellitus [DM], peripheral arterial disease, symptomatic carotid artery disease, 10‐year Framingham risk >20%) are considered at high risk for CV events.2
Current information concerning the long‐term costs of CV events in commercially insured patients with varying risk is limited.3, 4 Previous studies evaluated acute‐CV‐event costs in patients with hypertension, acute coronary syndrome, DM, or atherosclerosis, but not among patients with hyperlipidemia.4, 5, 6, 7 Although short‐term costs have been examined,3, 4, 5, 6, 7, 8, 9 the extent to which costs and resource utilization vary among subgroups of patients at risk for CVD has not been studied, nor have long‐term (up to 2 years post‐CV event) costs or resource utilization associated with recurrent CV events been explored. Existing studies estimating costs over a 24‐month period are now outdated.6 To fill this research gap, the current study used administrative claims data from a large US commercial health plan to analyze subsequent CV events, resource utilization, and long‐term costs among patients with hyperlipidemia and CVD or CHD risk equivalents.
Methods
Data Source and Patient Identification
This retrospective observational cohort study used administrative claims from the HealthCore Integrated Research Database (HIRD). The HIRD contains longitudinal medical and pharmacy claims data for approximately 33 million members from 14 commercial health plans across the United States. All claims data were from a limited dataset with de‐identified patient information. No patients were directly involved in the study; therefore, this study was exempt from an institutional review board review.
The study sample included adults age 18 to 64 years with a diagnosis of hyperlipidemia or use of lipid‐modifying medications (LMT; statins, ezetimibe, bile acid sequestrants, fibric acid derivatives, nicotinic acid derivatives, and other lipid‐modifying agents) from January 1, 2007, to December 31, 2008 (intake period). The index date was defined as the first occurrence of a hyperlipidemia diagnosis or use of any LMT during the intake period. The baseline period was 12 months prior to the index date, and patients were followed until the end of the study period, end of the eligibility period, or death, whichever occurred first. The follow‐up period for the occurrence of a new CV event (CV event date) was set from the index date to the end of study follow‐up. The follow‐up period for health care resource utilization and costs ran from the CV event date until the end of the study follow‐up period. Patients were followed for 2 years for the assessment of new CV events, health care resource utilization, and costs.
Patients were categorized into the following 3 cohorts based on CV risk level during the baseline period: (1) secondary CVD prevention (SP; patients with a history of a CV event, including myocardial infarction [MI], stroke, unstable angina [UA], coronary artery bypass graft [CABG], or percutaneous coronary intervention [PCI]); (2) high risk (HR; patients not in the secondary CVD prevention cohort but who had CVD or risk‐equivalent conditions, including chronic ischemic heart disease, stable angina, peripheral arterial disease, abdominal aortic aneurysm, transient ischemic attack [TIA], or type 2 DM); and (3) primary prevention (PP; patients with no CV event history or CHD risk equivalents).
In each cohort, patients with CV‐event‐related hospitalizations for MI, UA, CABG, PCI, ischemic stroke, TIA, or heart failure (HF) after the index dates were included in the case group; those with no CV‐event‐related hospitalization were in the control group. Cases and controls were matched within each cohort for age, sex, geographic region, comorbidities, baseline health care resource utilization, and length of continuous health plan eligibility during follow‐up using propensity score matching. Patients in each cohort were required to have ≥24 months of continuous follow‐up after the CV event date for evaluation of long‐term costs.
Outcome Measures
A CV event during follow‐up was defined as hospitalization for MI, ischemic stroke, CABG, PCI, UA, TIA, or HF. If a patient was hospitalized for MI or UA and also had a CABG or PCI on the same date, the CV event was identified as MI or UA. Patient characteristics were captured during baseline. All health care resource utilization, including LMT use, and total health care costs were assessed during baseline and follow‐up. Resource utilization included inpatient hospitalizations (including length of stay), emergency department (ED) visits, office visits, outpatient facility visits, and pharmacy prescription fills. Utilization of LMT included the number of patients with prescription fills for each drug in that group. For statins, patients were also stratified by each specific statin and by statin intensity levels (based on dosages and as defined in the 2013 American College of Cardiology/American Heart Association [ACC/AHA] guideline).10 Total costs included both medical (inpatient services, ED, office visit, and outpatient services) and pharmacy costs. Costs were defined as the sum of the amount paid by the health plan and the patient out‐of‐pocket. Costs were adjusted to 2013 dollars based on the Consumer Price Index.11 Total medical costs were further categorized by patients with 1, 2, or 3 CV events during follow‐up for each type of CV event. The incremental cost of a new CV event was calculated as the total cost of care for the case group compared with the control group within each CV risk cohort. The burden of subsequent CV events during follow‐up included the number of patients with 1, 2, or 3 subsequent CV events and mean time to subsequent CV events. Study outcomes were assessed for each type of CV event and for a composite endpoint of any CV event.
Statistical Analysis
Baseline demographic and clinical characteristics, resource utilization and cost, and medication use were described for all patients in each cohort; burden of subsequent CV events was described only for cases in each cohort. Descriptive analyses included means and standard deviation and relative frequencies for continuous and categorical variables, respectively. Continuous outcomes were compared using independent t tests and categorical outcomes were compared using χ2 tests. Propensity score matching using a 1:1 greedy algorithm was used to balance the differences in baseline demographic and clinical characteristics, resource utilization and cost, and medication use between cases and controls within each CV risk cohort.12 Balance after matching was assessed using standardized differences.13 A standardized difference of 10% indicated negligible correlation between the cases and controls.14
Results
Patient Characteristics
A total of 123 850 patients were included: 15 613 (12.6%) in the SP cohort, 47 600 (38.4%) in the HR cohort, and 60 637 (49.0%) in the PP cohort (Table 1). The cases and controls within each cohort were well matched with standardized differences well below 10% for all propensity score matching variables. The mean age at baseline was 55.9 years, with the PP cohort 1.7 years younger than the SP cohort. Approximately two‐thirds of the overall population was male (65.4%).
Table 1.
Baseline Demographics and Clinical Characteristicsa
| SP Cohort | HR Cohort | PP Cohort | ||||
|---|---|---|---|---|---|---|
| Control Group, n = 7814 | Case Group, n = 7799 | Control Group, n = 23 776 | Case Group, n = 23 824 | Control Group, n = 30 388 | Case Group, n = 30 249 | |
| Age (on index date), y, mean (SD) | 55.3 (6.6) | 54.9 (6.6) | 55.3 (6.3) | 54.9 (6.5) | 53.6 (7.1) | 53.2 (7.2) |
| Age category, n (%) | ||||||
| 18–39 y | 201 (2.6) | 199 (2.6) | 499 (2.1) | 628 (2.6) | 1337 (4.4) | 1504 (5.0) |
| 40–64 y | 7613 (97.4) | 7600 (97.5) | 23 277 (97.9) | 23 196 (97.4) | 29 051 (95.6) | 28 745 (95.0) |
| Female sex, n (%) | 2629 (33.6) | 2519 (32.3) | 8113 (34.1) | 7848 (32.9) | 11 530 (37.9) | 11 148 (36.9) |
| Geographic region, n (%) | ||||||
| Northeast | 1628 (20.8) | 1653 (21.2) | 4905 (20.6) | 5139 (21.6) | 5641 (18.6) | 5858 (19.4) |
| Midwest | 1960 (25.1) | 1993 (25.6) | 5937 (25.0) | 6035 (25.3) | 7229 (23.8) | 7206 (23.8) |
| South | 2388 (30.6) | 2372 (30.4) | 7283 (30.6) | 7278 (30.6) | 9422 (31.0) | 9317 (30.8) |
| West | 1285 (16.4) | 1228 (15.8) | 3962 (16.7) | 3872 (30.6) | 6258 (20.6) | 6204 (20.5) |
| Unknown/missing | 553 (7.1) | 553 (7.1) | 1689 (7.1) | 1500 (6.3) | 1838 (6.1) | 1664 (5.5) |
| Type of health plan, n (%) | ||||||
| HMO | 2328 (29.8) | 2492 (32.0) | 7036 (29.6) | 7403 (31.1) | 7658 (25.2) | 7791 (25.8) |
| PPO | 5411 (69.3) | 5235 (67.1) | 16 588 (69.8) | 16 264 (68.3) | 22 402 (73.7) | 22 122 (73.1) |
| CDHP | 75 (1.0) | 72 (0.9) | 152 (0.6) | 157 (0.7) | 328 (1.1) | 336 (1.1) |
| Medicare Advantage | 354 (4.5) | 432 (5.5) | 1129 (4.8) | 1251 (5.3) | 816 (2.7) | 912 (3.0) |
| Comorbidities, n (%) | ||||||
| Hypertension | 6108 (78.2) | 6088 (78.1) | 17 209 (72.4) | 17 228 (72.3) | 13 468 (44.3) | 13 302 (44.0) |
| Metabolic syndrome | 137 (1.8) | 135 (1.7) | 394 (1.7) | 405 (1.7) | 215 (0.7) | 214 (0.7) |
| Liver disease | 435 (5.6) | 454 (5.8) | 1165 (4.9) | 1217 (5.1) | 858 (2.8) | 847 (2.8) |
| Renal disease | 502 (6.4) | 542 (7.0) | 1478 (6.2) | 1507 (6.3) | 478 (1.6) | 528 (1.8) |
| Quan‐Charlson comorbidity index, mean (SD)§ | 2.5 (1.9) | 2.5 (1.9) | 1.7 (1.6) | 1.7 (1.6) | 0.4 (1.0) | 0.4 (1.1) |
| Any dyslipidemic medications,b n (%) | 5013 (64.2) | 4992 (64.0) | 15 293 (64.3) | 15 199 (63.8) | 11 247 (37.0) | 11 014 (36.4) |
| Statinsb | 4377 (56.0) | 4347 (55.7) | 12 645 (53.2) | 12 578 (52.8) | 9188 (30.2) | 9019 (29.8) |
| Bile acid sequestrantsb | 104 (1.3) | 96 (1.2) | 250 (1.1) | 277 (1.2) | 236 (0.8) | 229 (0.8) |
| Fibric acid derivativesb | 630 (8.0) | 657 (8.4) | 2546 (10.7) | 2490 (10.5) | 1305 (4.3) | 1288 (4.3) |
| Ezetimibeb | 1294 (16.6) | 1293 (16.6) | 3737 (15.7) | 3738 (15.7) | 1894 (6.2) | 1854 (6.1) |
| Nicotinic acid derivativesb | 416 (5.3) | 434 (5.6) | 968 (4.1) | 999 (4.2) | 400 (1.3) | 377 (1.3) |
| Other antihyperlipidemicsb | 126 (1.6) | 128 (1.6) | 326 (1.4) | 317 (1.3) | 160 (0.5) | 154 (0.5) |
| Inpatient hospitalizations, n (%) | 4594 (58.8) | 4623 (59.3) | 4321 (18.2) | 4537 (19.0) | 2363 (7.8) | 2433 (8.0) |
| No. of hospitalizations, mean (SD) | 0.9 (1.1) | 0.9 (1.1) | 0.3 (0.9) | 0.3 (0.8) | 0.1 (0.7) | 0.1 (0.5) |
| Length of stay, d, mean (SD) | 4.0 (9.1) | 4.4 (9.8) | 1.1 (5.3) | 1.4 (6.1) | 0.3 (2.3) | 0.4 (2.9) |
| Total costs, mean (SD) | $35 255 ($58 631) | $35 300 ($51 281) | $15 814 ($30 368) | $16 038 ($29 048) | $7542 ($16 637) | $7707 ($17 393) |
Abbreviations: CDHP, consumer‐driven health plan; HMO, health maintenance organization; HR, high risk; PP, primary prevention; PPO, preferred provider organization; SD, standard deviation; SP, secondary prevention.
After matching.
Patients with ≥1 fill, n (%).
Quan H, Sundararajan V, Halfon P, et al.: Coding algorithms for defining comorbidities in ICD‐9‐CM and ICD‐10 administrative data. Med Care. 2005; 43:1130–1139.
A similar proportion of patients in the SP and HR cohorts (64.1%, both cohorts) received LMT at baseline. Slightly more than half of patients in the SP (55.9%) and HR (53.2%) cohorts received a statin at baseline. A sizeable proportion of patients in the SP (43.6%) and HR cohorts (42.7%) were using nonstatin LMT (eg, ezetimibe, bile acid sequestrants, fibric acid derivatives, nicotinic acid derivatives, omega‐3 fatty acids). In the PP cohort, 36.7% received lipid‐lowering medication at baseline, with 30.0% using statins and 16.7% using nonstatin LMT. At baseline, 7.2% in the SP cohort, 6.4% in the HR cohort, and 2.7% in the PP cohort received combination therapy. Patterns of statin use were similar across all 3 cohorts, with atorvastatin used most commonly (SP cohort, 33.6%; HR cohort, 31.2%; PP cohort, 17.9%). In the SP and HR cohorts, 8.2% and 5.6% of patients received high‐intensity statins, respectively, and 2.8% from the PP cohort received low‐intensity statins (Table 1).
Burden of Cardiovascular Events
In the SP cohort, 43.0% of cases had ≥1 new CV event, with 19.8% having ≥3 new events over 2 years of follow‐up (Table 2). In the HR cohort, 33.9% of cases had ≥1 new CV event, with 12.9% having ≥3 new events. In the PP cohort, 20.9% of cases had ≥1 CV event, with 5.5% having ≥3 events.
Table 2.
CV‐Related Event Risk
| SP Cohort, n = 8218 | HR Cohort, n = 23 959 | PP Cohort, n = 30 250 | |
|---|---|---|---|
| Any CV event (composite endpoint), n (%) | |||
| ≥1 events | 3532 (43.0) | 8123 (33.9) | 6320 (20.9) |
| Only 1 event | 1152 (14.0) | 3040 (12.7) | 2952 (9.8) |
| Only 2 events | 756 (9.2) | 1984 (8.3) | 1695 (5.6) |
| ≥3 events | 1624 (19.8) | 3099 (12.9) | 1673 (5.5) |
| Events per patient for all patients, mean (SD) | 1.56 (3.3) | 1.00 (2.3) | 0.46 (1.2) |
| Events per patient for patients with ≥1 events, mean (SD) | 3.64 (4.1) | 2.94 (3.0) | 2.19 (1.9) |
| Time to first event from index date, d, mean (SD) | 870 (562) | 1009 (565) | 1007 (555) |
| Time between first and second events, d, mean (SD) | 299 (339) | 274 (324) | 272 (345) |
| Time between second and third events, d, mean (SD) | 225 (266) | 205 (258) | 213 (290) |
Abbreviations: CV, cardiovascular; HR, high‐risk; PP, primary prevention; SD, standard deviation; SP, secondary prevention.
In the SP cohort, among 3532 patients who had ≥1 CV event, 52.6% had ≥1 HF event, followed by UA (36.9%), ischemic stroke (33.5%), and PCI (26.0%). Among 8123 patients with ≥1 CV event in the HR cohort, 59.3% had ≥1 HF event, followed by UA (28.1%), ischemic stroke (24.5%), and PCI (21.5%). Lastly, among 6230 patients with ≥1 CV event in the PP cohort, 43.9% had ≥1 HF event, followed by UA (26.5%), ischemic stroke (26.2%), and MI (21.7%).
The mean time to the first new CV event was shortest for the SP cohort (870 days) compared with the HR (1009 days) and PP (1007 days) cohorts (Table 2). The time between the first and second CV event was 299 days in the SP cohort, 274 days in the HR cohort, and 272 days in the PP cohort. The time between the second and third CV event was 225 days in the SP cohort, compared with 205 days in the HR cohort and 213 days in the PP cohort.
Long‐Term Health Care Resource Utilization and Costs
Health care resource utilization at 24 months of follow‐up was higher among cases than controls in each of the CV risk cohorts (Table 3). Differences between cases and controls were significant in each cohort for inpatient hospitalizations, ED visits, and outpatient services. In the HR cohort, controls had significantly higher utilization than cases for office visits and prescriptions. Although the proportion of patients with office visits and prescriptions was not significantly different between cases and controls in the SP and PP cohorts, the mean number of office visits and prescriptions was significantly higher among cases compared with controls.
Table 3.
Overall All‐Cause Resource Utilization and Costs
| SP Cohort | HR Cohort | PP Cohort | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Control Group, n = 6903 | Case Group, n = 7024 | p‐valuea | Incremental Costsb | Control Group, n = 20 516 | Case Group, n = 26 640 | p‐valuea | Incremental Costsb | Control Group, n = 26 567 | Case Group, n = 26 337 | p‐valuea | Incremental Costsb | |
| Inpatient hospitalizations, n (%) | 883 (12.8) | 5550 (79.0) | <0.0001 | 1928 (9.4) | 15198 (73.6) | <0.0001 | 1513 (5.7) | 18590 (70.6) | <0.0001 | |||
| Mean no. of hospitalizations per patient (SD) | 0.2 (0.58) | 1.4 (1.4) | <0.0001 | 0.1 (0.5) | 1.3 (1.4) | <0.0001 | 0.1 (0.5) | 1.07 (1.1) | <0.0001 | |||
| Mean length of stay, d, (SD) | 0.7 (4.0) | 5.8 (12.8) | <0.0001 | 0.6 (4.0) | 6.1 (14.0) | <0.0001 | 0.3 (2.4) | 4.4 (10.7) | <0.0001 | |||
| Mean cost, $ (SD) | 3298 (19 556) | 31 320 (60 027) | <0.0001 | 15 077 | 2062 (13 917) | 31 252 (64 632) | <0.0001 | 15 245 | 1200 (8397) | 27 816 (59 213) | <0.0001 | 13 599 |
| ED visits, n (%) | 1108 (16.1) | 1850 (26.3) | <0.0001 | 2756 (13.4) | 4773 (23.1) | <0.0001 | 2606 (9.8) | 5107 (19.4) | <0.0001 | |||
| Mean visits per patient (SD) | 0.2 (1.0) | 0.5 (1.4) | <0.0001 | 0.2 (1.0) | 0.4 (1.0) | <0.0001 | 0.1 (0.9) | 0.3 (0.9) 0 | <0.0001 | |||
| Mean cost, $ (SD) | 393 (1812) | 757 (2349) | <0.0001 | 359 | 282 (1941) | 600 (2036) | <0.0001 | 286 | 201 (1485) | 526 (2069) | <0.0001 | 251 |
| Office visits, n (%) | 5836 (84.5) | (85.7) | 0.0568 | 17 139 (83.5) | 16 788 (81.3) | <0.0001 | 21 117 (79.5) | 20 951 (79.6) | 0.8556 | |||
| Mean visits per patient (SD) | 7.8 (8.5) | 11.0 (10.1) | <0.0001 | 7.3 (8.2) | 10.7 (10.3) | <0.0001 | 5.4 (7.4) | 9.0 (9.2) | <0.0001 | |||
| Mean cost, $ (SD) | 1128 (2554) | 1592 (3365) | <0.0001 | 498 | 1107 (4349) | 1591 (3674) | <0.0001 | 527 | 783 (2121) | 1439 (4008) | <0.0001 | 516 |
| Outpatient office visits, n (%) | 5690 (82.4) | 5949 (84.7) | 0.0003 | 16 805 (81.9) | 16 639 (80.6) | 0.0008 | 20 390 (76.8) | 20 640 (78.4) | <0.0001 | |||
| Mean visits per patient (SD) | 9.0 (13.5) | 16.4 (23.3) | <0.0001 | 9.0 (14.4) | 17.1 (24.6) | <0.0001 | 6.3 (10.6) | 13.7 (19.4) | <0.0001 | |||
| Mean cost, $ (SD) | 4066 (10 253) | 7641 (19 180) | <0.0001 | 3063 | 3680 (10 522) | 8090 (21 604) | <0.0001 | 3408 | 2472 (8971) | 6378 (18 161) | <0.0001 | 2726 |
| Prescriptions, n (%) | 5819 (84.3) | 5976 (85.1) | 0.1994 | 17 038 (83.1) | 16 600 (80.4) | <0.0001 | 20 864 (78.5) | 20 736 (78.7) | 0.575 | |||
| Mean no. of prescriptions per patient (SD) | 40.8 (36.9) | 52.2 (41.4) | <0.0001 | 39.9 (36.5) | 52.0 (43.4) | <0.0001 | 22.3 (25.9) | 37.1 (34.0) | <0.0001 | |||
| Mean cost, $ (SD) | 4209 (6838) | 5246 (6860) | <0.0001 | 1673 | 4259 (7127) | 5205 (6889) | <0.0001 | 1643 | 2157 (4926) | 3514 (6465) | <0.0001 | 1284 |
| Total medical costs, $, mean (SD) | 8884 (24 188) | 41 309 (67 779) | <0.0001 | 18 117 | 7131 (20 994) | 41 534 (73 203) | <0.0001 | 18 824 | 4655 (14 334) | 36 159 (66 368) | <0.0001 | 16 638 |
| Total costs,c $, mean (SD) | 13 094 (26 050) | 46 555 (69 423) | <0.0001 | 19 320 | 11 389 (23 403) | 46 739 (74 939) | <0.0001 | 20 003 | 6813 (16 004) | 39 673 (68 039) | <0.0001 | 17 650 |
Abbreviations: ED, emergency department; HR, high risk; PP, primary prevention; SD, standard deviation; SP, secondary prevention.
P values are from t test for mean values.
Incremental cost equals all‐cause cost for the case − all‐cause cost for the matched control.
Sum of medical and pharmacy costs.
Total health care costs were higher among cases compared with controls in all 3 cohorts, and costs remained higher throughout the follow‐up period for each type of CV event. The incremental total costs over 2 years of follow‐up were highest for the HR cohort ($20 003) compared with the SP ($19 320) and PP ($17 650) cohorts (Table 4). Mean 2‐year total costs among patients with only 1 CV event ranged from $46 133 to $76 239 in the SP cohort and from $54 011 to $79 442 in the HR cohort. Across all cohorts, CABG was associated with the highest incremental costs (SP cohort, $76 958; HR cohort, $73 318; PP cohort, $70 416); TIA was associated with the lowest incremental costs (SP cohort, $37 478; HR cohort, $36 186; PP cohort, $26 723). Forty‐three percent of first‐year costs associated with TIA events occurred during the first 30 days post‐CV event among cases in the SP cohort; 48% occurred during the first 30 days in the HR cohort. Across all cohorts, 63% to 73% of first‐year costs associated with CABG occurred during the first 30 days post‐CV event (SP cohort, 63%; HR cohort, 69%; PP cohort, 73%).
Table 4.
Overall Follow‐up Total All‐Cause Costsa
| SP Cohort | HR Cohort | PP Cohort | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | Control Group,Mean (SD) | Case Group, Mean (SD) | p‐valueb | Incremental Costsc | n | Control Group, Mean (SD) | Case Group, Mean (SD) | p‐valueb | Incremental Costsc | n | Control Group, Mean (SD) | Case Group, Mean (SD) | p‐valueb | Incremental Costsc | |
| MI total cost, $ | 1830 | 12 535 (25 112) | 58 720 (78 345) | <0.0001 | 49 504 | 4868 | 9890 (18 410) | 61 303 (97 892) | <0.0001 | 54 082 | 6651 | 5975 (12 968) | 51 058 (80 138) | <0.0001 | 46 626 |
| For patients with only 1 event | 580 | 13 443 (23 276) | 66 633 (71 796) | <0.0001 | 55 734 | 1220 | 10 559 (19 041) | 75 310 (88 360) | <0.0001 | 66 979 | 1092 | 6893 (14 050) | 76 100 (95 386) | <0.0001 | 70 614 |
| For patients with only 2 events | 103 | 17 369 (48 136) | 96 642 (106 992) | <0.0001 | 84 548 | 218 | 12 305 (21 487) | 104 709 (242 503) | <0.0001 | 95 543 | 143 | 6453 (8528) | 98 554 (207 302) | <0.0001 | 92 865 |
| For patients with only 3 events | 45 | 10 525 (11 855) | 128 938 (174 925) | <0.0001 | 118 751 | 69 | 16 865 (27 083) | 124 552 (144 163) | <0.0001 | 109 650 | 19 | 13 297 (29 805) | 133 705 (173 904) | <0.0001 | 127 380 |
| UA total cost, $ | 2785 | 12 234 (23 548) | 47 520 (61 901) | <0.0001 | 38 672 | 7247 | 10 223 (20 486) | 46 945 (71 085) | <0.0001 | 39 580 | 8252 | 6207 (14 531) | 39 863 (57 284) | <0.0001 | 35 364 |
| For patients with only 1 event | 768 | 13 441 (25 749) | 56 039 (58 364) | <0.0001 | 46 133 | 1698 | 11 213 (27 483) | 64 409 (103 689) | <0.0001 | 55 880 | 1266 | 6681 (16 328) | 62 855 (93 300) | <0.0001 | 57 398 |
| For patients with only 2 events | 240 | 13 982 (30 035) | 66 558 (53 590) | <0.0001 | 54 665 | 295 | 11 091 (17 725) | 68 901 (63 113) | <0.0001 | 59 594 | 225 | 8334 (15 966) | 72 224 (60 022) | <0.0001 | 65 068 |
| For patients with only 3 events | 65 | 22 586 (42 403) | 68 917 (61 004) | <0.0001 | 53 766 | 76 | 9520 (13 110) | 86 962 (78 016) | <0.0001 | 78 217 | 56 | 6960 (9552) | 81 153 (51 829) | <0.0001 | 74 210 |
| Ischemic stroke total cost, $ | 2330 | 11 794 (20 477) | 50 658 (76 577) | <0.0001 | 42 128 | 5033 | 9592 (17 909) | 53 983 (103 521) | <0.0001 | 47 247 | 6964 | 6345 (14 622) | 43 456 (87 001) | <0.0001 | 39 381 |
| For patients with only 1 event | 711 | 12 619 (23 036) | 63 076 (85 956) | <0.0001 | 53 046 | 1342 | 10 759 (20 846) | 74 773 (136 656) | <0.0001 | 66 293 | 1203 | 7008 (14 285) | 70 270 (119 946) | <0.0001 | 64 984 |
| For patients with only 2 events | 186 | 12 711 (15 929) | 85 678 (114 452) | <0.0001 | 74 749 | 311 | 10 367 (15 360) | 89 122 (128 049) | <0.0001 | 80 429 | 207 | 6960 (14 021) | 92 298 (127 481) | <0.0001 | 86 613 |
| For patients with only 3 events | 60 | 11 421 (11 363) 7887 | 85 728 (80 322) 58 700 | <0.0001 | 75 833 | 93 | 7941 (10 104) | 106 255 (128 189) | <0.0001 | 99 059 | 59 | 5845 (6082) | 119 760 (174 762) | <0.0001 | 114 968 |
| CABG total cost, $ | 693 | 13 270 (27 939) | 87 348 (107 744) | <0.0001 | 76 958 | 2602 | 9814 (17 394) | 81 448 (91 101) | <0.0001 | 73 318 | 2726 | 6965 (14 280) | 76 035 (78 195) | <0.0001 | 70 416 |
| For patients with only 1 event | 229 | 14 028 (35 727) | 86 943 (94 871) | <0.0001 | 76 329 | 668 | 9839 (19 626) | 87 406 (94 244) | <0.0001 | 79 442 | 495 | 7537 (15 386) | 90 895 (96 337) | <0.0001 | 84 911 |
| For patients with only 2 events | 4 | 17 853 (27 237) | 174 856 (92 124) | <0.0001 | 157 002 | 8 | 12 133 (17 811) | 174 647 (174 419) | <0.0001 | 164 666 | 5 | 13 038 (18 438) | 201 245 (129 253) | <0.0001 | 188 207 |
| For patients with only 3 events | 0 | 0 | 0 | ||||||||||||
| PCI total cost, $ | 2185 | 12 084 (23 226) | 49 624 (55 336) | <0.0001 | 40 438 | 5670 | 9900 (20 679) | 49 841 (62 458) | <0.0001 | 42 337 | 6475 | 6281 (13 847) | 46 045 (60 895) | <0.0001 | 41 158 |
| For patients with only 1 event | 601 | 13 260 (29 009) | 58 943 (68 049) | <0.0001 | 48 474 | 1319 | 9633 (13 715) | 62 240 (77 698) | <0.0001 | 54 011 | 1017 | 6308 (11 524) | 66 936 (83 694) | <0.0001 | 61 421 |
| For patients with only 2 events | 144 | 16 030 (32 957) | 69 829 (53 671) | <0.0001 | 58 581 | 224 | 11 175 (19 678) | 83 760 (75 464) | <0.0001 | 74 519 | 144 | 7515 (13 331) | 74 917 (59 475) | <0.0001 | 68 583 |
| For patients with only 3 events | 43 | 11 883 (13 806) | 73 371 (63 880) | <0.0001 | 62 713 | 62 | 12 213 (21 914) | 116 029 (105 987) | <0.0001 | 106 109 | 27 | 7662 (12 432) | 91 527 (66 572) | <0.0001 | 83 899 |
| TIA total cost, $ | 996 | 12 134 (22 654) | 46 183 (64 114) | <0.0001 | 37 478 | 2508 | 9312 (17 766) | 42 897 (99 019) | <0.0001 | 36 186 | 4138 | 6016 (14 100) | 30 525 (68 767) | <0.0001 | 26 723 |
| For patients with only 1 event | 326 | 13 327 (28 721) | 65 131 (83 869) | <0.0001 | 54 677 | 652 | 10 961 (23 197) | 74 284 (172 482) | <0.0001 | 65 818 | 602 | 6948 (14 187) | 63 335 (105 228) | <0.0001 | 58 256 |
| For patients with only 2 events | 66 | 13 398 (19 164) | 61 893 (71 100) | <0.0001 | 51 141 | 74 | 13 482 (30 615) | 76 061 (104 643) | <0.0001 | 66 248 | 66 | 8295 (15 051) | 55 348 (67 532) | <0.0001 | 48 486 |
| For patients with only 3 events | 16 | 13 886 (17 644) | 72 647 (72 967) | <0.0001 | 63 748 | 16 | 6502 (5042) | 75 826 (45 384) | <0.0001 | 69 746 | 11 | 2954 (6159) | 123 247 (155 736) | <0.0001 | 120 521 |
| HF total cost, $ | 2627 | 12 335 (21 780) | 65 357 (97 613) | <0.0001 | 56 078 | 8769 | 10 402 (18 690) | 60 539 (100 804) | <0.0001 | 53 247 | 7893 | 6757 (15 465) | 54 205 (97 553) | <0.0001 | 49 637 |
| For patients with only 1 event | 739 | 12 844 (25 296) | 66 087 (78 464) | <0.0001 | 55 996 | 2126 | 10 926 (19 511) | 63 390 (85 300) | <0.0001 | 55 451 | 1541 | 7705 (18 000) | 69 771 (103 004) | <0.0001 | 64 167 |
| For patients with only 2 events | 285 | 13 078 (17 770) | 76 136 (99 449) | <0.0001 | 65 473 | 848 | 10 526 (20 368) | 85 251 (114 790) | <0.0001 | 76 976 | 515 | 7532 (15 733) | 96 620 (153 105) | <0.0001 | 90 921 |
| For patients with only 3 events | 174 | 13 821 (19 010) | 92 482 (92 496) | <0.0001 | 80 646 | 453 | 11 650 (19 353) | 101 438 (162 312) | <0.0001 | 92 392 | 204 | 7983 (15 475) | 95 393 (95 319) | <0.0001 | 89 004 |
Abbreviations: CABG, coronary artery bypass graft; HF, heart failure; HR, high risk; MI, myocardial infarction; PCI, percutaneous coronary intervention; PP, primary prevention; SP, secondary prevention; TIA, transient ischemic attack; UA, unstable angina.
All‐cause total costs includes both medical and pharmacy costs for all medical events during the follow‐up period.
P values are from t test for mean values.
Incremental cost equals all‐cause cost for the case − all‐cause cost for the matched control.
Across all cohorts, the mean total cost among cases with only 1 CV event was lower than that for those who had 2 or 3 CV events, except for TIA. Compared with patients with only 1 CV event, the mean 2‐year cost was 30% higher in patients with 2 CV events and 48% higher in patients with 3 CV events.
Discussion
Cardiovascular disease is a major cause of morbidity and mortality in the United States, not only in terms of the immediate acute phase for CV events, but also in subsequent years. This retrospective, real‐world analysis demonstrated that utilization and health care costs were higher among patients with a CV event for the initial event as well as for subsequent events during 2 years of follow‐up. A higher proportion of patients who had a CV event at baseline (SP cohort) had a recurrent CV event during the study period, compared with patients who had no CV event at baseline (HR and PP cohorts). Secondary prevention cohort patients also had shorter time to the first recurrent CV event compared with other cohorts. A relatively large number of patients in each cohort had secondary or tertiary events during the 2‐year follow‐up period. Heart failure was the most frequent type of CV event among all cohorts. Acute MI is associated with myocardial necrosis and fibrosis. The incremental loss of myocardial tissue mass from sequential acute coronary syndromes and the progression of atherosclerotic disease increase risk for developing left ventricular systolic dysfunction and ischemic cardiomyopathy, respectively. Both of these clinical sequelae increase risk of the development of HF and can be directly related to dyslipidemia because they correlate with severity of atherosclerotic disease. However, we were unable to ascertain subtypes of HF in this study.
Despite the fact that this study was conducted 5 years after the National Cholesterol Education Program Adult Treatment Panel III guidelines were introduced into clinical care,15 a surprisingly low percentage of patients with CVD or CHD risk equivalents received statin therapy, leaving these patients at high risk of CV events.
Additionally, across all cohorts, the percentage of patients using high‐intensity statin therapy was considerably lower than in other recently published US studies. For example, among Medicare beneficiaries discharged following a CHD event, 27% of first‐fill statin prescriptions were for high‐intensity statins,16 whereas among veterans with CVD, 36.5% received high‐intensity statins.17 Both studies acknowledged these rates of high‐intensity statin use were low. However, the rates of high‐intensity statin use reported in the current study were more closely aligned with those reported in recent European studies. In those studies, high‐intensity statin use ranged from 7.4% among unselected patients to 9.4% in high‐risk patients with controlled low‐density lipoprotein cholesterol (LDL‐C) levels.18, 19
The ACC/AHA blood cholesterol guidelines advocate for use of risk‐appropriate statin dose and potency (as well as adjuvant LMY as indicated) to optimally reduce risk for cardiovascular events.10 There is a clear continued need to encourage clinicians to treat patients at risk more aggressively and with higher doses of statins than is currently observed. It is also important to emphasize that among high‐risk patients, if a >50% reduction in LDL‐C cannot be achieved, then use of adjuvant LMT is indicated as promulgated by the guideline.
Total health care resource utilization was higher among cases compared with controls in all 3 cohorts, and the cost difference was maintained throughout the follow‐up period. The costs incurred beyond the acute phase are significant, because a substantial number of patients had second and third CV events during this 24‐month period. Costs increased with the number of subsequent events, with mean 2‐year costs nearly 50% higher for a third CV event compared with the first event. Avoidance of subsequent CV events through more effective management of hyperlipidemia could reduce this current cost burden. Incremental costs remained higher for cases than controls throughout the entire follow‐up period; costs for the case groups did not fall to the levels observed in the control group at any point, which adds to the economic burden.
The health care costs associated with CVD can be staggering. Although the initial hospitalization for a CV event can range from $7000 to > $56 000, the follow‐up costs can be just as burdensome, averaging $16 600 in the first year following the CV event and $34 000 3 years later.8 The costs in the current study were greater, with the 2‐year total ranging from $46 133 to $79 442 for the first CV event.
Study Limitations
The claims database used for this study is large and geographically diverse, with long follow‐up data available. However, a limitation of claims analyses is that the data are intended for reimbursement purposes, not research, so the coding for CV events and comorbid conditions may contain undetected errors or omissions. All patients included in the study were members of US commercial health plans; the results may not be generalizable to patients with other types of health insurance or living outside the United States. Lack of clinical and health behavior information (eg, blood pressure, smoking status) limits the determination of specific risk level for the study population. Additionally, because laboratory results were unavailable, the success of statin therapy in lowering LDL‐C levels to treatment goals was uncertain.
Conclusion
The results of this retrospective study show a substantial proportion of patients experience subsequent CV events both near term and 2 years after an initial CV event. Patients with recurrent CV events had higher total health care costs for the initial as well as subsequent events, compared with patients who did not have subsequent events. Statins and LMT were significantly underutilized in all cohorts, despite the presence of CVD. Future research is needed to thoroughly examine patient‐physician behavior and treatment patterns to improve patient outcome. A better understanding of the reasons for underutilization of current LMT and availability of new therapies to reduce LDL‐C may provide clinicians and patients an opportunity to reduce the burden of CV events among a high‐CVD‐risk population.
Funding for this study was provided by Amgen, Inc., Thousand Oaks, California. Maxine Fisher, PhD; Rajeshwari S. Punekar, MPH, PhD; and Mark J. Cziraky, PharmD, are employees of HealthCore, Inc., an independent research organization that received funding from Amgen, Inc., for the conduct of the study. Kathleen M. Fox, MHS, PhD, is an independent consultant who received research funds from Amgen, Inc. for this study. Akshara Richhariya and Shravanthi R. Gandra are employees and stock owners of Amgen, Inc. Peter P. Toth, MD, PhD, is a member of the speakers bureau for Amarin, AstraZeneca, GlaxoSmithKline, Kowa, and Merck and Co. and is a consultant/advisory board member for Amgen, AstraZeneca, Atherotech, Kowa, LipoScience, and Merck and Co.
The authors have no other funding, financial relationships, or conflicts of interest to disclose.
References
- 1. Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics—2014 update: a report from the American Heart Association. Circulation. 2014;129:e28–e292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. National Cholesterol Education Program . Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). National Heart, Lung, and Blood Institute. NIH publication no. 01–3670. [Google Scholar]
- 3. Duh MS, Fulcher NM, White LA, et al. Costs associated with cardiovascular events in patients with hypertension in US managed care settings. J Am Soc Hypertens. 2009;3:403–415. [DOI] [PubMed] [Google Scholar]
- 4. O'Sullivan AK, Rubin J, Nyambose J, et al. Cost estimation of cardiovascular disease events in the US. Pharmacoeconomics. 2011; 29:693–704. [DOI] [PubMed] [Google Scholar]
- 5. Eisenstein EL, Shaw LK, Anstrom KJ, et al. Assessing the clinical and economic burden of coronary artery disease: 1986–1998. Med Care. 2001:39:824–835. [DOI] [PubMed] [Google Scholar]
- 6. Straka RJ, Liu LZ, Girase PS, et al. Incremental cardiovascular costs and resource use associated with diabetes: an assessment of 29 863 patients in the US managed‐care setting. Cardiovasc Diabetol. 2009;8:53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ohlsfeldt RL, Gandhi SK, Fox KM, et al. Medical and cost burden of atherosclerosis among patients treated in routine clinical practice. J Med Econ. 2010;13:500–507. [DOI] [PubMed] [Google Scholar]
- 8. Chapman RH, Liu LZ, Girase PG, et al. Determining initial and follow‐up costs of cardiovascular events in a US managed‐care population. BMC Cardiovasc Disord. 2011;11:11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Etemad LR, McCollam PL. Total first‐year costs of acute coronary syndrome in a managed care setting. J Manag Care Pharm. 2005;11:300–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Stone NJ, Robinson JG, Lichtenstein AH, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines [published correction appears in Circulation. 2014;129(25 suppl 2):S46–S48]. Circulation. 2014;129(25 suppl 2):S1–S45. [DOI] [PubMed] [Google Scholar]
- 11. US Department of Labor, Bureau of Labor Statistics . Series CUUR0000SAM and CUUR0000SAM2. http://www.bls.gov/cpi/home.htm. Accessed November 8, 2014.
- 12. Parsons LS. Reducing bias in a propensity score matched‐pair sample using greedy matching techniques. Paper 214–26. http://www2.sas.com/proceedings/sugi26/p214‐26.pdf. Accessed July 14, 2014.
- 13. Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity‐score matched samples. Stat Med. 2009;28:3083–3107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38:1228–1234. [Google Scholar]
- 15. Executive summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486–2497. [DOI] [PubMed] [Google Scholar]
- 16. Rosenson RS, Kent ST, Brown TM, et al. Underutilization of high‐intensity statin therapy after hospitalization for coronary heart disease. J Am Coll Cardiol. 2015;65:270–277. [DOI] [PubMed] [Google Scholar]
- 17. Virani SS, Woodard LD, Akeroyd JM, et al. Is high‐intensity statin therapy associated with lower statin adherence compared with low‐ to moderate‐intensity statin therapy? Implications of the 2013 American College of Cardiology/American Heart Association Cholesterol Management Guidelines. Clin Cardiol. 2014;37:653–659. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Vodonos A, Ostapenko I, Toledano R, et al. Statin adherence and LDL cholesterol levels. Should we assess adherence prior to statin upgrade? Eur J Intern Med. 2015;26:268–272. [DOI] [PubMed] [Google Scholar]
- 19. Halcox JP, Tubach F, Lopez‐Garcia E, et al. Low rates of both lipid‐lowering therapy use and achievement of low‐density lipoprotein cholesterol targets in individuals at high‐risk for cardiovascular disease across Europe. PloS One. 2015;10:e0115279. [DOI] [PMC free article] [PubMed] [Google Scholar]