• Center on Health Equity and Access
  • Clinical
  • Health Care Cost
  • Health Care Delivery
  • Insurance
  • Policy
  • Technology
  • Value-Based Care

Quality Measure Attainment in Patients With Type 2 Diabetes Mellitus

Publication
Article
Supplements and Featured PublicationsOpportunities for Improving Attainment of Quality Measures in Patients With Type 2 Diabetes Mellitus
Volume 20
Issue 1 Suppl

Objectives: This study examined the demographics, comorbidities, clinical characteristics, and treatments of people with type 2 diabetes mellitus (T2DM) treated with metformin and sulfonylurea as well as an elderly subgroup. Achievement of predefined quality measure goals (glycated hemoglobin [A1C], blood pressure [BP], low-density lipoprotein cholesterol [LDL-C], body mass index [BMI]) and their association with diabetes-related healthcare costs were assessed.

Study design: The study applied a retrospective longitudinal cohort design.

Methods: Health insurance claims and electronic medical records from 14,532 adults with T2DM (2007- 2011) were used to identify a sample receiving metformin and sulfonylurea (MET+SU) concomitantly. The index date was the first dispensing of MET+SU after 6 months of eligibility. Clinical characteristics were assessed during baseline. Quality measure attainment (A1C <8%, BP <140/90 mm Hg, LDL-C level <100 mg/dL, BMI <30 kg/m2), was evaluated during the 12 months following the index date. Association between attainment and diabetes-related costs was evaluated using non-parametric bootstrap methods adjusting for imbalance in baseline characteristics between cohorts.

Results: Among 2044 patients, including 1283 patients 65 years and older, hyperlipidemia, hypertension, and cardiovascular disease were the most common baseline comorbidities. Quality measure goal attainment was 63.9% for A1C, 33.1% for BP, 68.2% for LDL-C level, and 34.4% for BMI, and was associated with significantly lower diabetes-related costs per patient per year compared with nonattainment (adjusted mean cost differences: —$1445 for A1C; –$1218 for BMI; –$2029 for A1C and BMI; –$2073 for A1C, BMI, and BP; all P <.05).

Conclusion: This study highlights the high incidence of comorbidities and potential financial implications of attaining T2DM quality outcomes.

(Am J Manag Care. 2014;20:S5-S15)The total economic burden of diabetes in the United States in 2012 was estimated at $245 billion, a 21% increase from 2007.1 The prevalence of diabetes increases with age and reaches its peak in the population aged 60 to 74 years.2 Management of diabetes is challenging, especially in an elderly population due to complex comorbid medical issues and a generally lower functional status.3

According to guidelines for diabetes monitoring and treatment, glycated hemoglobin (A1C), blood pressure (BP), lowdensity lipoprotein cholesterol (LDL-C) level, and weight/body mass index (BMI) have been identified as important interrelated quality measures in the treatment and monitoring of type 2 diabetes mellitus (T2DM).4 The Centers for Medicare & Medicaid Services (CMS) has set specific treatment goals for particular clinical parameters that accountable care organizations participating in its Medicare Shared Savings Program have to achieve when treating patients with diabetes: A1C: 8%, BP: 140/90 mm Hg, LDL-C level: 100 mg/dL.5-7 The Medicare Shared Savings Program is an initiative under the Patient Protection and Affordable Care Act. CMS records a provider’s score for achieving these measures to calculate the part of cost savings that it will share with the provider. A BMI less than 30 kg/m2 is an emerging quality measure for diabetes, because a higher BMI indicates obesity, a major risk factor for T2DM.8 This threshold is already being applied by the Clinical Advisory Committee of Better Health Greater Cleveland, a network of primary care practice partners responsible for Cuyahoga County in northeastern Ohio.9

In addition to improving diabetes control to minimize the risk of complications and to potentially improve the well-being of patients with T2DM, achievement of such predefined quality measure goals may also have important economic implications since improved diabetes control leads to fewer disease complications, which can be costly. A study found that the average lifetime medical costs of treating T2DM, including diabetic complications, totaled $85,200, with 53% due to treating diabetic complications.10 To the authors’ knowledge, despite the importance of these quality measures for diabetes control and treatment and their potential importance for the management of the economic burden of T2DM, there are no studies that use real-world data from patients with T2DM to assess the achievement of these quality measure goals and its association with diabetes-related costs.

The objective of this study was to examine the demographic and clinical characteristics and the achievement of predefined quality measure goals in a T2DM population receiving metformin and sulfonylurea, as well as in an elderly subgroup of this population, and to assess the association of goal attainment with diabetes-related healthcare costs.

Metformin and sulfonylurea are well-established core therapies for controlling hyperglycemia in patients with T2DM.11 Adding a sulfonylurea is one of the treatment options available for metabolic management if metformin and lifestyle changes alone fail to achieve or sustain glycemic control. Hence, patients who receive this combination therapy can be considered as having faced challenges in controlling their hyperglycemia. Achieving predefined quality measure goals may therefore have particular effects on clinical outcomes and healthcare cost for this patient population.

MethodsData Source

The study used data from health insurance claims and electronic medical records from the Reliant Medical Group (RMG) database covering the 5-year period from January 1, 2007, through December 31, 2011. RMG is a large nonprofit multispecialty group practice that provides comprehensive care for patients; it has a total of over 1 million patient visits each year to more than 250 physicians practicing across 20 locations throughout central Massachusetts.

The data were de-identified in compliance with the Health Insurance Portability and Accountability Act of 1996 to preserve patient confidentiality. Database elements used for this study included longitudinal, memberlinked medical claims, pharmacy claims, enrollment records (including patient demographics), laboratory results (eg, A1C), and clinical measures (eg, weight and BP).

Study Design and Patient Selection

A retrospective longitudinal landmark cohort design was used (Figure 1). Adult patients with at least 1 diagnosis of T2DM (International Classification of Diseases, 9th Revision [ICD-9]= 250.x0 or 250.x2) who were being treated concomitantly with metformin and a sulfonylurea (eg, chlorpropamide, glimepiride, glipizide, glyburide, tolazamide, and tolbutamide) after at least 6 months of continuous eligibility (baseline period) were included in the study population. The index date was defined as a patient’s first day of a metformin and sulfonylurea prescription. Patients had to be continuously eligible for at least 12 months after the index date (landmark period). Patients who were diagnosed with type 1 diabetes mellitus or treated with insulin (with or without other oral antihyperglycemic agents in combination) during the baseline period were excluded from the study population.

From the real-world sample of 14,532 patients with T2DM, 2666 were being treated with a combination of metformin and sulfonylurea. Applying subsequent inclusion and exclusion criteria resulted in an overall study population of 2044 patients, including 1283 patients 65 years and older.

Study Outcomes

Evaluation of Quality Measures and Goal Attainment

Quality measures (A1C, BP, LDL-C level, BMI) were evaluated during the landmark period and reported for the overall population and the elderly subgroup. Quality measure goal attainment was defined as having no values equal to or higher than the predefined threshold during the landmark period. The individual thresholds were:

  • A1C: 7%, 8%, and 9%
  • BP: 140/90 mm Hg
  • LDL-C level: 100 mg/dL
  • BMI: 30 kg/m2

The combined thresholds were:

  • A1C: 8%; BP: 140/90 mm Hg; LDL-C level: 100 mg/dL
  • A1C: 7%; BP: 140/90 mm Hg; LDL-C level: 100 mg/dL
  • A1C: 8%; BMI: 30 kg/m2
  • A1C: 8%; BMI: 30 kg/m2; BP: 140/90 mm Hg

Quality Measure Goal Attainment and Healthcare Costs

Evaluated during the observation period, diabetes-related medical healthcare costs were identified through claims associated with a diagnosis for diabetes (ICD-9 = 250.x0 or 250.x2), whereas diabetes-related pharmacy costs were defined as claims for antihyperglycemic agents. Medical costs were further broken down into emergency department costs, inpatient costs, and outpatient/other visits, according to their recorded medical cost category.

Statistical Analyses

Univariate descriptive statistics were generated to describe the baseline demographics, clinical characteristics, and costs of the overall population and an elderly subgroup (≥65 years of age), as well as the evaluation of quality measures during the landmark period, including mean ± standard deviation for continuous data and relative frequencies for categorical data.

Costs were reported in US dollars (2011) per patient per year (PPPY) for quality measure goal achievers versus non— goal achievers. Unadjusted and adjusted cost differences were estimated using (a) generalized linear models (GLMs) with a log link and gamma distribution; or (b) 2-part models (for cost components with a portion of zero values greater than 5%), where the first part was a logistic model and the second part was a GLM model with a log link and a gamma distribution. The gamma distribution was chosen, as it is recognized to fit well-skewed healthcare cost data.12 Controlling factors included age, sex, year of index date, race, payer type, and Charlson Comorbidity Index. As baseline diabetes-related comorbidities may contribute to goal achievement, it was decided not to adjust for them in the multivariate analysis to avoid removing the explanatory effect of the goal achievement variable. Statistical differences between groups (P values) and 95% confidence intervals were calculated using nonparametric bootstrap re-sampling techniques with 499 replications to ensure parameter stability.13 Significance level was set at a 2-sided α value of 0.05.

Results

Baseline Population Characteristics

Demographics, clinical characteristics, and costs at index date/baseline are presented in Table 1.

Mean (median) age at index date was 66.6 (69.0) years in the overall study population and 74.6 (74.0) years in the elderly population subgroup. The incidence of macrovascular comorbidities, including cardiovascular disease and chronic heart failure, was higher than that of microvascular comorbidities, such as nephropathy, neuropathy, and retinopathy, in both the overall study population and the elderly population subgroup. The most common comorbidities were hyperlipidemia (overall population, 73.9%; elderly subgroup, 78.1%), hypertension (66.5% and 74.2%, respectively), and cardiovascular disease (25.5% and 33.4%, respectively). Compared with the overall population, the elderly population subgroup had higher incidence of most of the macrovascular and microvascular complications and other comorbidities. Antihypertensives were prescribed to 67.2% of all patients and 75.1% of those 65 years and older. Loop diuretics accounted for 10.5% and 13.6% and non-loop diuretics accounted for 21.1% and 24.5% in the total and elderly populations, respectively.

Evaluation of Quality Measures and Goal Attainment

Figures 2A and 2B present the evaluation of quality measures and goal attainment. Figure 2A shows that in the overall study population, mean A1C was 7.5%, mean BP was 131.8/72.8 mm Hg, mean LDL-C level was 85.2 mg/dL, and mean BMI was 32.7 kg/m2. In the overall study population, 63.9% achieved the A1C goal of less than 8%, 27.2% the A1C goal of less than 7%, 33.1% the BP goal (<140/90 mm Hg), 68.2% the LDL-C level goal (<100 mg/dL), and 34.4% the BMI goal (<30 kg/m2).

Figure 2B illustrates that in the elderly subgroup, mean values were similar to those in the overall population, with 7.4% for A1C, 133.5/70.5 mm Hg for BP, 82.2 mg/dL for LDL-C level, and 31.1 kg/m2 for BMI. In this population, 70.7% achieved the A1C goal of less than 8%, 29.8% the A1C goal of less than 7%, 29.1% the BP goal, 73.3% the LDL-C level goal, and 42.1% the BMI goal.

With respect to achievement of various composite measure goals, 15.9% of the overall population and 16.3% of the elderly subgroup achieved the combined A1C less than 8%, BP, and LDL-C level goal; 7.6% and 8.2% achieved the combined A1C less than 7%, BP, and LDL-C level goal; 21.9% and 28.2% achieved the combined A1C and BMI goal; and 7.4% and 8.0% achieved the combined A1C, BMI, and BP goal, respectively.

Quality Measure Goal Attainment and Healthcare Costs

For the overall study population, quality measure goal attainment was associated with significantly lower total diabetes-related healthcare costs PPPY (Table 2 and Figure 3). For the individual goals, the adjusted mean cost difference between achievers and non-achievers was —$992 (P = .060) for A1C less than 7%, —$1445 (P = .016) for A1C less than 8%, —$749 (P = .112) for BP, —$37 (P = .844) for LDL-C level, and —$1218 (P = .032) for BMI. For the combined quality measure goals, the adjusted mean cost difference was —$595 (P = .460) for A1C less than 8%, BP, and LDL-C level; —$1044 (P = .176) for A1C less than 7%, BP, and LDL-C level; —$2029 (P <.001) for A1C less than 8% and BMI; and —$2073 for A1C less than 8%, BMI, and BP (P <.001).

The largest significant contributor to cost savings was outpatient costs for the A1C goal (adjusted mean cost difference: —$969, P <.001), the combined A1C less than 8%, BP, and LDL-C level goal (—$582, P <.001) and the combined A1C less than 7%, BP, and LDL-C level goal (—$580, P = .028). Inpatient costs were the largest significant contributor to cost savings for BMI (—$1319, P <.001), the combined A1C and BMI goal (—$1270, P = .012), and the combined A1C, BMI, and BP goal (—$1370, P = .028).

Discussion

This study examined baseline demographic and clinical characteristics as well as the achievement of predefined quality measure goals (A1C <7%, A1C <8%, BP <140/90 mm Hg, LDL-C level <100 mg/dL, and BMI <30 kg/m2) in a population with T2DM and an elderly subgroup treated with metformin and sulfonylurea. Since this patient population can be considered as having faced challenges in controlling their hyperglycemia, achieving predefined quality measure goals may have particular effects on clinical outcomes and healthcare costs. The study further assessed how achieving these goals was associated with diabetes-related healthcare costs.

Previously published results support the findings of this study with respect to the cost impact of individual quality measures. Improved glycemic control in patients with type 1 diabetes mellitus and T2DM has been associated with lower direct medical and pharmacy costs in studies using health plan administrative data.14,15 A study using survey data combined with cost data from a large health maintenance organization found that future healthcare costs were higher for persons who are overweight, especially for those with BMI of at least 30 kg/m2.16 Similarly, an observational study in patients with T2DM from Spain found that reductions in BMI were associated with appreciable short-term economic benefits, especially in obese patients.17 A recently published study found that patients with TD2M who achieved the LDL-C level goal of <100 mg/dL incurred significantly lower diabetes-related medical service costs compared with non-goal achievers (—$246, P <.001). However, a statistical difference was not observed with respect to costs for patients who achieved only an A1C goal of less than 7% compared with patients who achieved both the LDL-C and the A1C goal (—$57, P = .404).18 Both the study population (US military veterans, predominantly male) and the definition of diabetes-related costs (medical service costs associated with a diagnosis of diabetes or microvascular or macrovascular complications) in the cited study differed from the one in the study presented here, potentially explaining the difference in economic and statistical significance while supporting the consistency of the negative sign of the association.

The small difference between the cost savings associated with the combined A1C/BMI goal ($2029, P <.001) and those related to the combined A1C/BMI/BP goal ($2073, P <.001) in the present study should not be interpreted as indicating that BP is insignificant for reducing overall healthcare costs. First, there may be a maximum effect; the cost savings that can be realized through achieving additional quality measures may be marginally decreasing. Second, this study only evaluated the impact of quality measure goal achievement on diabetes-related costs. BP control is likely to have an important influence on other healthcare costs, especially cardiovascular-related costs. Since these costs are primarily associated with cardiovascular complications, they may not be linked to a diabetes diagnosis in a claims database, even if the patient has diabetes. Hence, the potential importance of BP control for overall healthcare costs may be underestimated when only diabetes-related costs are considered.

The present study adds to this literature by showing that the achievement of a composite quality measure goal is associated with greater cost savings than the attainment of individual goals alone. Only a few other studies analyze the achievement of multiple goals simultaneously. Eddy and Shah (2012) evaluated the effects of the Medicare Shared Savings Program quality measures and performance targets on Medicare costs in patients with T2DM and reported only minor cost savings, after taking the cost of performance improvement into account.7 However, their study differs from the one described here in several aspects. First, Eddy and Shah simulated the study population and the behavior of healthcare providers, instead of obtaining it from administrative real-world data, as the present study did. Hence, unobservable interactions between treatment, quality measure attainment, and healthcare costs that take place in the real world may not be incorporated in the simulation model. To determine the cost impact of quality measure attainment, Eddy and Shah calculated the effect of treatment that brings measures to exactly the thresholds specified, but not beyond. In reality, improved monitoring and control of these quality measures may often lead to patients not only achieving, but exceeding, treatment goals. The cost of performance improvement was also estimated based on assumptions about the number of additional tests and visits required to reach the performance targets. The study presented here did not impose how the improvements in quality measures can be achieved.

Apart from analyzing the association between quality measure attainment and healthcare costs, the present study also confirmed the high incidence of comorbidities in a T2DM population, especially in the elderly subgroup. This highlights the need for integrated management of T2DM that considers a variety of quality measures and interventions (including lifestyle changes and pharmacologic treatments) simultaneously to improve patients’ outcomes and quality of life.19,20

Limitations

This study has several limitations. It was designed to investigate quality measure attainment in a study population of patients with T2DM receiving metformin and sulfonylurea, defined as those with at least 1 day of treatment with both metformin and sulfonylurea during the baseline period. Since the study did not intend to attribute quality measure achievement to a particular treatment regimen, it did not require patients to be new users of metformin and sulfonylurea or to continue treatment throughout the landmark period.

Administrative claims data may also be subject to inaccuracies and miss some specific clinical information. For example, incorrect diagnoses recorded on claims may hamper the correct attribution of healthcare costs to clinical conditions. Also, claims do not contain information on other unrecorded factors that may affect healthcare costs, such as smoking habits and over-the-counter drug use (eg, aspirin). Nevertheless, the use of linked, electronic medical record data made it possible to provide unique insights into the association between quality measure attainment and healthcare costs for a real-world population. To assess this association, the study excluded patients without A1C, BP, LDL-C level, or BMI assessment during the landmark period, thus potentially resulting in a subpopulation different from the general T2DM population treated with metformin and sulfonylurea. The strict definition of goal achievement in this study (no values above a specific threshold) may have further selected a population that is relatively stable and may have influenced the associated economic burden.

The study assessed the cost impact of quality measure achievement, controlling for demographic and clinical baseline characteristics. It could not, however, account for unobserved factors that may affect healthcare costs, such as general lifestyle changes or improved adherence to treatment regimens. If these factors are correlated with quality measure achievement, the estimates for the cost impacts of these quality measures may be biased.

Because patients from the RMG are predominantly located in central Massachusetts, the study results may not necessarily be generalizable to other populations. The limited external validity may, however, be overweighed by the high internal validity that stems from the large sample size of the data set used; this is the current trade-off with all data sets.

Conclusion

This study confirms the high incidence of comorbidities in a population with T2DM, particularly in elderly patients. Using real-world data, the study demonstrated that the achievement of quality measure goals in terms of A1C, BP, and BMI is significantly associated with important reductions in diabetes-related healthcare costs in a population of patients with T2DM exposed to metformin and sulfonylurea.Author affiliations: Health Economics and Outcomes Research, Janssen Scientific Affairs, LLC, Raritan, NJ (RAB, SCM); Analysis Group, Inc, Boston, MA (MSD); Reliant Medical Group, Worcester, MA (LG); Groupe d’Analyse, Ltée, Analysis Group, Inc, Montreal, QC, Canada (JG, AMG, MHL, PL).

Funding source: This supplement was supported by Janssen Pharmaceuticals, Inc.

Author disclosures: Dr Bailey reports employment with Janssen Scientific Affairs, LLC (a Johnson & Johnson company), and reports stock ownership with Johnson & Johnson. Dr Duh, Mr Gravel, Ms Grittner, Ms Lafeuille, and Mr Lefebvre report employment with Analysis Group, Inc (recipient of grants from Janssen Pharmaceuticals, Inc). Mr Martin reports employment with Janssen Scientific Affairs, LLC (a Johnson & Johnson company) and stock ownership with Johnson & Johnson. Dr Garber reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this supplement.

Authorship information: Concept and design (RAB, MSD, LG, JG, MHL, PL, SCM); acquisition of data (RAB, LG, MHL, PL, SCM); analysis and interpretation of data (RAB, MSD, JG, AMG, MHL, PL, SCM); drafting of the manuscript (RAB, AMG, MHL, PL, SCM); critical revision of the manuscript for important intellectual content (RAB, MSD, AMG, MHL, PL, SCM); statistical analysis (JG, AMG, MHL, PL); obtaining funding (MSD, MHL, PL); administrative, technical, or logistic support (LG); and supervision (RAB).

Address correspondence to: Marie-Hélène Lafeuille, MA, Groupe d’Analyse, Ltée, 1000 De La Gauchetiere West, Suite 1200, Montreal, QC, Canada H3B 4W5. E-mail: mlafeuille@analysisgroup.com.

  1. American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. Diabetes Care. 2013;36(4):1033-1046.
  2. Cowie CC, Rust KF, Ford ES, et al. Full accounting of diabetes and pre-diabetes in the U.S. population in 1988-1994 and 2005- 2006. Diabetes Care. 2009;32(2):287-294.
  3. Chau D, Edelman SV. Clinical management of diabetes in the elderly. Clinical Diabetes. 2001;19(4):172-175.
  4. American Diabetes Association. Standards of medical care in diabetes — 2011. Diabetes Care. 2011;34(suppl 1):S11—S61.
  5. Centers for Medicare & Medicaid Services. Medicare program; Medicare Shared Savings Program: accountable care organizations: final rule. Fed Regist. 2011;76(212):67802—67990.
  6. RTI International, Telligen. Accountable care organization 2012 program analysis, quality performance standards narrative measure specifications, final report. Research Triangle Park, NC: RTI International; December 12, 2011. RTI Project No. 0213195 .000.004.
  7. Eddy DM, Shah R. A simulation shows limited savings from meeting quality targets under the Medicare Shared Savings program. Health Aff (Millwood). 2012;31(11):2554-2562.
  8. Obesity and Overweight. World Health Organization website. www.who.int/dietphysicalactivity/media/en/gsfs_obesity.pdf. Accessed October 8, 2013.
  9. Cebul RD, Love TE, Jain AK, Hebert CJ. Electronic health records and quality of diabetes care. N Engl J Med. 2011;365(9): 825-833.
  10. Zhuo X, Zhang P, Hoerger TJ. Lifetime direct medical costs of treating type 2 diabetes and diabetic complications. Am J Prev Med. 2013;45(3):253-261.
  11. Nathan DM, Buse JB, Davidson MB, et al; American Diabetes Association; European Association for Study of Diabetes. Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. 2009;32(1): 193-203.
  12. Manning WG, Basu A, Mullahy J. Generalized modeling approaches to risk adjustment of skewed outcomes data. J Health Econ. 2005;24(3):465-488.
  13. Davidson R, MacKinnon JG. Econometric Theory and Methods. New York, NY: Oxford University Press; 2004.
  14. Manning WG, Basu A, Mullahy J. Generalized modeling approaches to risk adjustment of skewed outcomes data. J Health Econ. 2005;24(3):465-488.
  15. Shetty S, Secnik K, Oglesby AK. Relationship of glycemic control to total diabetes-related costs for managed care health plan members with type 2 diabetes. J Manag Care Pharm. 2005;11(7): 559-564.
  16. Thompson D, Brown JB, Nichols GA, Elmer PJ, Oster G. Body mass index and future healthcare costs: a retrospective cohort study. Obes Res. 2001;9(3):210-218.
  17. Dilla T, Valladares A, Nicolay C, Salvador J, Reviriego J, Costi M. Healthcare costs associated with change in body mass index in patients with type 2 diabetes mellitus in Spain: the ECOBIM study. Appl Health Econ Health Policy. 2012;10(6):417-430.
  18. Shi L, Ye X, Lu M, et al. Clinical and economic benefits associated with the achievement of both HbA1c and LDL cholesterol goals in veterans with type 2 diabetes mellitus. Diabetes Care. 2013;36(10):3297-3304.
  19. Kalda R, Rätsep A, Lember M. Predictors of quality of life of patients with type 2 diabetes. Patient Prefer Adherence. 2008; 2:21-26.
  20. Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH, Pedersen O. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med. 2003;348(5): 383-393.
© 2024 MJH Life Sciences
AJMC®
All rights reserved.