Mortality, Outcomes, and Healthcare Costs in T2DM Patients at Risk for Cardiovascular Disease

March 18, 2015
David M. Kern, MS

Stephanie DeVore, MSPH

Jennifer Kim, PharmD

Ozgur Tunceli, PhD

Bingcao Wu, MS

Boaz Hirshberg, MD

The American Journal of Accountable Care, March 2015, Volume 3, Issue 1

Targeting glucose control and managing cardiovascular (CV) risk factors may prevent future CV events, and have positive downstream impact by reducing costs to healthcare stakeholders.

ABSTRACTObjectives: To evaluate cardiovascular outcomes and economic burden among type 2 diabetes mellitus (T2DM) patients at high risk for adverse cardiovascular events in a managed care setting.

Study Design: Retrospective cohort study.

Methods: Patients 40 years and older diagnosed with T2DM between January 1, 2007, and April 30, 2011, were identified with medical and pharmacy claims in a large US managed care database and followed for up to 5 years. T2DM diagnoses were based on 2 medical claims with diabetes-related International Classification of Diseases, Ninth Revision, Clinical Modification codes, or on 1 medical claim plus 1 pharmacy claim for diabetic medication. Patients with established cardiovascular disease (CVD) were classified as the secondary prevention cohort while those without established CVD but with CVD risk factors constituted the primary prevention cohort. Outcomes of myocardial infarction (MI), stroke, and death were evaluated using Kaplan-Meier survival analysis.

Results: Of the overall study population (N = 368,581), the secondary prevention cohort consisted of 177,140 patients, and the primary prevention cohort had 191,441 patients. Fifty-seven percent of patients were male and the mean age was 67.1 years. Both cohorts had high rates of mortality, stroke, and hospitalization for MI during follow-up. Mortality rates were 20% and 7% in the secondary and primary prevention groups, respectively, after 4 years of follow-up. Healthcare costs for the overall study population were $12,962 per patient-year of follow-up.

Conclusions: T2DM patients with established CVD or CVD risk factors are at significant risk for mortality and CVD events. They incur high healthcare utilization and costs, driven by inpatient hospitalizations and outpatient visits.Diabetes mellitus affects an estimated 25.8 million adults in the United States, and prevalence is expected to more than double by 2050, with the largest increases in older age groups.1,2 Type 2 diabetes mellitus (T2DM) accounts for up to 95% of all diagnosed cases of diabetes in adults.2 The increasing prevalence of T2DM imposes substantial mortality, morbidity, and healthcare cost burdens on society.

Patients with diabetes are at risk for both microvascular and macrovascular complications; macrovascular events (eg, stroke and myocardial infarction [MI]) are the leading cause of death in this population.2 The risk of stroke and death is 2 to 4 times greater for adults with T2DM than for the general population.2,3 Diabetes was an underlying or contributing cause of death for more than 230,000 individuals in the United States in 2007.1 T2DM patients often have comorbid conditions, including hypertension and dyslipidemia, which confer additional risk of cardiovascular (CV) events.4,5

The management of T2DM includes the use of insulin and several classes of oral medication along with appropriate diet, more exercise, and other lifestyle changes.6,7 Antihyperglycemic agents have shown to be effective in controlling blood glucose in the short term and in reducing the risk of the microvascular sequelae of diabetes. Furthermore, a meta-analysis of randomized clinical trials concluded that intensive glycemic control reduces coronary events compared with standard glycemic control, without increasing mortality risks.8

The evidence is inconclusive, however, on the ability of specific antihyperglycemic treatments to reduce the risk of major CV events and death in this patient population.9 Treatment effects appear to vary by drug class. A meta-analysis on the thiazolidinedione drug, rosiglitazone, raised concerns over increased risk of death from CV causes,10 and a nationwide observational cohort study suggested increased CV risk with sulfonylureas compared with metformin.10,11 Such evidentiary variations and knowledge gaps about treatment outcomes indicate an urgent need for rigorous long-term evidence for decision making.11,12

A handful of clinical trials—including a few evaluating dipeptidyl peptidase 4 (DPP-4) inhibitors—are under way to assess the efficacy and safety of T2DM treatment interventions on CV end points.13,14 One prominent study, the Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus (SAVOR)—Thrombolysis in Myocardial Infarction (TIMI) 53, is a phase 4, randomized, double-blind, placebo-controlled trial. With an expected duration of 5 years, the study enrolled 16,500 patients in 25 countries to evaluate the safety of saxagliptin treatment and its effect on CV events in T2DM patients.13 In an update, which assessed 16,492 patients after a median time of 2.1 years, the authors reported that saxagliptin treatment neither increased nor decreased the rate of ischemic events, although hospitalization and heart failure rates increased. They concluded that even though saxagliptin was associated with better glycemic control, other approaches were needed to reduce CV risk among T2DM patients.14

The purpose of the present study was to understand and quantify real-world outcomes for patients with T2DM and high CV risk, and to assess the cost burden of these outcomes in a US-based private payer setting.

MethodsData Source

All study data were queried from the HealthCore Integrated Research Environment (HIRE), a repository of medical and pharmacy claims data on approximately 30 million health plan members (at the time of the study), drawn from 14 geographically dispersed US commercial health plans. Researcher access was limited to de-identified data to ensure patient privacy and confidentiality. Strict measures were observed to ensure full compliance with the Health Insurance Portability and Accountability Act. Study data query dates included the period from January 1, 2007, through April 30, 2012, and to ensure a minimum follow-up time of 12 months for all patients, the study intake period was set at January 1, 2007, through April 30, 2011. The length of follow-up varied for the patients. Mortality data was obtained from the Social Security Death Index (SSDI) and matched with data from the HIRE repository to determine the date of death as applicable. All analyses were conducted under a pre-specified research plan, in accordance with current best practices for retrospective research.15,16

Patient Selection

The study population consisted of adults 40 years and older with T2DM, which was defined as having at least 2 claims with an International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis indicating T2DM (ICD-9-CM 250.x0 or 250.x2) or 1 T2DM diagnosis plus a prescription fill for an antidiabetic medication. The index date was defined as the earliest date during the study intake period at which the criteria for T2DM diagnosis were satisfied. The study population was further stratified into 2 cohorts: secondary and primary prevention. The secondary prevention cohort had established CV disease (CVD), defined as a history of MI, ischemic stroke/transient ischemic attack, peripheral vascular disease, unstable angina, congestive heart failure, other coronary heart disease, or revascularization procedures within the 12 months preceding the index date. The primary prevention cohort consisted of patients at risk for CVD and included patients aged 55 years or more (men) or aged 60 years or more (women) who had a diagnosis of hypertension, dyslipidemia, or both. Hypertension and dyslipidemia were each defined as having at least 1 ICD-9-CM diagnosis code or at least 2 medication prescription fills for the condition, within the 12 months prior to the index date. Those without established CVD and no CVD risk factors were excluded from the study.


The primary outcome was a composite event of all-cause death,stroke, or MI. All-cause deaths, as opposed to cardiovascular-related deaths, were used to measure mortality in this analysis because the actual cause of death is not reported in the SSDI. For stroke and MI, any hospitalization with a primary diagnosis code for stroke or MI and a minimum 1-night stay was defined as an event. The time to first event was captured on a per patient basis and analyzed via Kaplan-Meier analysis. Event rates were calculated over total patient-years of follow up, both for the entire study population and for each risk cohort. Healthcare resource utilization and costs were summarized, and consisted of both plan-paid and patient-paid costs. The Consumer Price Index for medical care, provided by the US Bureau of Labor Statistics, was used to convert prices to 2011 constant US dollars.17 All analyses were conducted using SAS version 9.2 (SAS Institute, Cary, North Carolina); descriptive statistics were reported.

ResultsPatient Characteristics

Appendix Figure

Of the more than 21 million enrollees in HIRE between January 1, 2007, and April 30, 2011 (the intake period), more than 1 million met the criteria for T2DM diagnosis. A total of 604,882 T2DM patients aged 40 years or more had continuous health plan enrollment for at least 12 months prior to the index date identified (). Of these patients, 368,581 with either established CVD or CV risk factors were identified and included in the analysis; 236,301 who did not have CVD or CV risk factors were eliminated from the analysis. The 368,581 patients were placed in the secondary prevention (n = 177,140) or primary prevention cohorts (n = 191,441) and were followed from first diagnosis (index date) until loss of eligibility, end of study period, or death.

Table 1

Table 2

A majority of patients (57.3%) were male; overall, 71.1% received coverage under preferred provider organization health plans. The mean age was 67.1 years for the total study population, and was slightly higher in the secondary prevention cohort despite the application of stricter age requirements for the primary prevention group (). Patients were treated at sites in all regions of the United States, with the largest representation from the midwest and south. More than 80% of the population had dyslipidemia and nearly 90% had hypertension (). A prescription for antidiabetic medication was filled by 43.8% of the sample in the 12 months preceding their index T2DM diagnosis, including metformin (24.5%), sulfonylureas (17.4%), and thiazolidinediones (14.1%). Other medical comorbidities were frequent—particularly among patients in the secondary prevention cohort. Of the conditions utilized to identify the cohort with prior CVD, other coronary heart disease conditions (including atherosclerosis, aortic aneurysm, post MI syndrome, and revascularization procedures, among others) were the most common (66.8%), followed by ischemic stroke/transient ischemic attack (30.5%), congestive heart failure (28.1%), and peripheral vascular disease (26.5%); MI (13.9%) and unstable angina (7.3%) were less common. Nearly half (49.7%) of patients with CVD had a Deyo Charlson Comorbidity Index (DCCI) score of 3 or higher. The DCCI is based on 17 diagnoses, identified by ICD-9-CM codes, each with a weight of 1 to 6. The final score is derived from the sum of the weighted values of the comorbidities present; higher scores signify greater comorbidity burden.18


Over the follow-up period—which averaged 2.78 years—14.3% of all patients experienced at least 1 of the primary outcome events: all-cause death, hospitalization for MI, or hospitalization for stroke.

Table 3

Figure 1

Table 4

In the secondary prevention cohort, 21.1% of patients had at least 1 event, while 8% of patients in the primary prevention cohort experienced an event. The incidence rate of these events, considering only the first event for each patient, was 53.3 per 1000 patient-years of observation in the overall population; 81.8 per 1000 patient-years in the secondary prevention cohort; and 28.8 per 1000 patient-years in the primary prevention cohort (). The majority of the observed events were deaths from any cause (56.2% of all events—34,961 deaths out of 62,219 total events). Stroke presented agreater risk as a cause of hospitalization than did MI. Kaplan-Meier estimates of the time to first event showed that more than 1 in 4 patients with diabetes and established CVD in this sample could be expected to experience a stroke or MI, or die from any cause, during the 4 years of follow-up (). Secondary outcomes, including hospitalizations due to other CV causes, were also common, with hospitalizations due to heart failure and revascularization procedures especially prevalent among patients in the secondary prevention cohort ().

Figure 2

Figure 3

Hospitalization rates were 701 and 290 events per 1000 patient-years in the secondary prevention and primary prevention cohorts, respectively (). The average cost per patient per year among all patients was $12,962 ($16,227 in the secondary and $10,059 in the primary prevention cohorts) (). Costs per patient-year were driven primarily by inpatient days, outpatient visits, and pharmacy costs in both cohorts.


The study cohort—primary prevention plus secondary prevention—represents an older population (mean age 67.1 years) with T2DM, and thus, an elevated risk of stroke, MI, and death. Over the follow-up period, patients in this population experienced high rates of hospitalization and outpatient visits, which contributed to their substantial healthcare costs. Comorbid conditions were prevalent, including chronic renal dysfunction, cancer, and respiratory conditions. All-cause mortality was especially high in the group with established CVD. More than 1 in 4 patients in this subgroup died or were hospitalized for MI or stroke within 4 years after the index date. Event rates in this study are consistent with those in prior studies, which demonstrated substantially higher CV event rates in populations of T2DM patients with CVD compared with those without CVD.19,20 These findings underscore the importance of treatment options for T2DM patients at elevated risk of CV events.

Other studies of diabetic patients have assessed similar outcomes. The Action to Control Cardiovascular Risk in Diabetes (ACCORD) blood pressure trial also examined rates of MI, stroke, and death (due to any cause and cardiovascular-related) in T2DM patients with high CVD risk.21,22 The ACCORD trial studied the effect of different therapeutic interventions (including lipid-lowering and blood-pressure—lowering medications) on each of these outcomes, while our study did not require any type of medical intervention. Outcomes in the ACCORD were less common than what was found in the current study, possibly due to therapeutic intervention, with event rates per 1000 person years ranging from 11.9 to 16.1, 11.3 to 14.4, and 3.0 to 4.7 for all-cause mortality, nonfatal MI, and nonfatal stroke, respectively.

The Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) trial also studied cardiovascular outcomes within a T2DM patient population.23,24 All patients in the ADVANCE trial were assigned to a therapeutic group for intensive blood glucose control or standard blood glucose control. Of the patients in the standard control group, 2.8% had a nonfatal MI, 3.8% had a nonfatal stroke, and 9.6% died from any cause. Those results were similar to the findings in this study—2.5% with MI, 3.9% with stroke, and 9.5% all-cause death—although median follow-up time in ADVANCE was roughly twice as long (5 years vs 2.6 years in our study).

The use of a large real-world existing data resource allowed for the reporting of timely estimates of event rates and healthcare costs in this population, with a search cutoff date of April 2012. In addition, this study included patients in a real-world setting, without having to restrict patients due to clinical trial criteria or requiring therapeutic intervention.


The limitations of the study include reliance on medical claims data, which lack clinical detail and are subject to potential coding, billing, and recording errors, as well as other inconsistencies. Claims databases also include only patients with specific healthcare coverage—in this case, coverage provided by private commercial health plans. Notably, patients without health insurance and a substantial portion of those who are covered under federal plans (Medicare or Medicaid) are not part of the analyzed population; therefore, the ability to generalize these results to the entire US population may be limited. The requirement for continuous medical and pharmacy coverage prior to the index date, which resulted in exclusion of one-third of otherwise eligible patients, could have introduced further bias. In addition, due to the lack of availability of certain clinical data, inclusion criteria and outcome definitions commonly used in clinical trials were not able to be exactly replicated. Specifically, blood glucose, lipids, and blood pressure levels, and the use of smoking status as a risk factor for CVD were not used; nor was there any requirement that patients have a minimum life expectancy at time of enrollment. Ideally, we would also have data on CV-related death, however, the SSDI does not specify the cause of death, and thus, only all-cause death was analyzed in this study.


Analysis of this large US database indicates that T2DM patients with either CV risk factors or previous CV disease experience high morbidity and mortality and incur substantial healthcare-related costs. Treatment goals that focus on glucose control and management of CV risk factors may prevent future CV events, thereby reducing costs to the healthcare system.Author Affiliations: HealthCore, Inc (DK, OT, BW), Wilmington, DE; AstraZeneca Pharmaceuticals (SD, BH, JK), Wilmington, DE.

Source of Funding: AstraZeneca LP sponsored this study. The researchers had complete access to the de-identified data set and formulated the protocol, study design, and statistical analysis. The researchers had full authority over the administration of the study and over the decision to publish their findings. Researchers from both AstraZeneca and HealthCore were involved in the interpretation of results, preparation and review of the manuscript prior to submission.

Author Disclosures: Dr Tunceli and Mssrs Kern and Wu are employees of HealthCore, Inc, which received funding from AstraZeneca for this work. Ms DeVore was an employee of AstraZeneca during the time the study was conducted and the manuscript was written. Drs Hirshberg and Kim are employees of AstraZeneca.

Address correspondence to: David M. Kern, MS, HealthCore, Inc, 800 Delaware Ave, Wilmington, DE 19801. Tel: 302-230-2102. E-mail:

1. National diabetes fact sheet, 2011. CDC website. Published 2011. Accessed February 21, 2014.

2. Roger VL, Go AS, Lloyd-Jones DM, et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics—2012 update: a report from the American Heart Association. Circulation. 2012;125(1):e2-e220.

3. Preis SR, Hwang SJ, Coady S, et al. Trends in all-cause and cardiovascular disease mortality among women and men with and without diabetes mellitus in the Framingham Heart Study, 1950 to 2005. Circulation. 2009;119(13):1728-1735.

4. Giorda CB, Avogaro A, Maggini M, et al; DAI Study Group. Incidence and risk factors for stroke in type 2 diabetic patients: the DAI study. Stroke. 2007;38(4):1154-1160.

5. van Dieren S, Beulens JW, Kengne AP, et al. Prediction models for the risk of cardiovascular disease in patients with type 2 diabetes: a systematic review. Heart. 2012;98(5):360-369.

6. American Diabetes Association. Executive summary: standards of medical care in diabetes—2012. Diabetes Care. 2012;35(suppl 1):S4-S10.

7. Qaseem A, Humphrey LL, Sweet DE, Starkey M, Shekelle P; Clinical Guidelines Committee of the American College of Physicians. Oral pharmacologic treatment of type 2 diabetes mellitus: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2012;156(3):218-231.

8. Ray KK, Seshasai SR, Wijesuriya S, et al. Effect of intensive control of glucose on cardiovascular outcomes and death in patients with diabetes mellitus: a meta-analysis of randomised controlled trials. Lancet. 2009;373(9677):1765-1772.

9. Cobble ME, Frederich R. Saxagliptin for the treatment of type 2 diabetes mellitus: assessing cardiovascular data. Cardiovasc Diabetol. 2012;11:6.

10. Jørgensen CH, Gislason GH, Andersson C, et al. Effects of oral glucose-lowering drugs on long-term outcomes in patients with diabetes mellitus following myocardial infarction not treated with emergent percutaneous coronary intervention—a retrospective nationwide cohort study. Cardiovasc Diabetol. 2010; 9:54.

11. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med. 2007;356(24):2457-2471.

12. Guidance for industry: diabetes mellitus—evaluating cardiovascular risk in new antidiabetic therapies to treat type 2 diabetes. FDA website. Published December 2008. Accessed February 21, 2014.

13. Scirica BM, Bhatt DL, Braunwald E, et al. The design and rationale of the saxagliptin assessment of vascular outcomes recorded in patients with diabetes mellitus-thrombolysis in myocardial infarction (SAVOR-TIMI) 53 study. Am Heart J. 2011;162(5):818-825.

14. Scirica BM, Bhatt DL, Braunwald E, et al; SAVOR-TIMI 53 Steering Committee and Investigators. Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med. 2013;369(14):1317-1326.

15. The GRACE initiative: GRACE principles. GRACE website. Published April 2010. Accessed February 21, 2014.

16. Motheral B, Brooks J, Clark MA, et al. A checklist for retrospective database studies—report of the ISPOR Task Force on Retrospective Databases. Value Health. 2003;6(2):90-97.

17. Bureau of Labor Statistics, US Department of Labor, Consumer Price Index for medical care. Accessed February 24, 2015.

18. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613-619.

19. Avogaro A, Giorda C, Maggini M, et al; Diabetes and Informatics Study Group; Association of Clinical Diabetologists; Istituto Superiore di Sanita. Incidence of coronary heart disease in type 2 diabetic men and women: impact of microvascular complications, treatment, and geographic location. Diabetes Care. 2007;30(5):1241-1247.

20. Juhaeri J, Gao S, Dai WS. Incidence rates of heart failure, stroke, and acute myocardial infarction among type 2 diabetic patients using insulin glargine and other insulin. Pharmacoepidemiol Drug Saf. 2009;18(6):497-503.

21. Cushman W. Effects of intensive blood pressure control on cardiovascular events in type 2 diabetes mellitus: the action to control cardiovascular risk in diabetes (ACCORD) blood pressure trial. Slides for American College of Cardiology March 14, 2010. Accessed February 24, 2015.

22. Ginsberg H. Effects of combination lipid therapy on cardiovascular events in type 2 diabetes mellitus: the action to control cardiovascular risk in diabetes (ACCORD) lipid trial. Slides for American College of Cardiology March 14, 2010. Accessed February 24, 2015.

23. ADVANCE Management Committee. Study rationale and design of ADVANCE: action in diabetes and vascular disease—preterax and diamicron MR controlled evaluation Diabetologia. 2001;44(9):1118-1120.

24. ADVANCE Collaborative Group; Patel A, MacMahon S, Chalmers J, et al. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008;358(24):2560-2572.