Quality Measure Improvement Strategies for Elderly Patients With Diabetes

May 23, 2016
John J. Sheehan, PhD, RPh

Joanna P. MacEwan, PhD

Evidence-Based Diabetes Management, May 2016, Volume 22, Issue SP7

The authors discuss a simple strategy for payers to ensure more patients with type 2 diabetes achieve control of A1C.

Possibly the first treatment guidelines and the precursors to quality metrics survive in one of the oldest known medical texts. In 1862, Edwin Smith discovered an ancient Egyptian papyrus with medical instruction predating Hippocrates by at least a millennia. The papyrus included sound medical advice. For example, after head trauma it instructed the physician to remove splinters from the brain before bandaging. It also included some questionable advice—the papyrus recommended the physician apply a crushed ostrich egg topically after skull fracture. Notably, prior to describing the recommended treatment for each of the 48 listed conditions, the papyrus instructed the physician to determine which of 3 prognoses best suited the patient. It advised that, depending on the condition, physicians should tell patients they will either: (1) treat the condition, (2) contend with the condition, or (3) not treat the condition. In the first case, the physician practically guarantees success, and in the second case, the physician offers his best effort. The third option, which appeared here for the first time, is the most foreign—why would a physician not treat a patient? Scholars believe the papyrus advises against the treatment of certain difficult conditions in an effort to protect the reputation of the physician.1 After all, a physician who takes on many hopeless cases may struggle to build a successful track record.

Today, moral, legal, and geographic constraints forbid the denial of care to needy patients, and assessment of care quality has evolved from informal provider reputations to a rigorously-quantified system. Medicare quality measures are arguably the most important quality measure systems among elderly patients in the United States. These measures assess quality using a 5-star rating system based on prior years’ performance. The program applies material incentives to improve performance: reputational incentives around the actual star rating, financial incentives around bonus payments, and recruitment incentives around expanded enrollment windows.2,3 Diabetes ranks among—if not the most—important conditions in the assessment of quality performance. The Medicare Shared Savings Program (MSSP) for accountable care organizations includes diabetes on a short list of chronic conditions, particularly important to the Medicare population. When simply counting the number of measures for various conditions, diabetes ties for first with ischemic heart disease for most measures in the MSSP.4 Moreover, when Consumer Reports ranked health insurance plans nationwide, the authors used how well the plan managed patients with diabetes as the sole measure of plan performance, based on the assumption that diabetes management served as a valid proxy for overall plan performance.5

Diabetes likely ranks high because of the prevalence of the condition, the cost of treatment, and the holistic treatment approach to the disease that quality measure creators seek to encourage. Additionally, the HHS considers good performance measures relevant, measurable, accurate, and feasible; the treatment of diabetes lends itself to measures that meet these 4 criteria. The quality measures themselves are drawn from treatment guidelines and are components of the transition toward value-based care and alternative payment models.

As part of a comprehensive approach to minimize the risk of complications of diabetes, treatment guidelines feature glucose control as a major focus in diabetes management. The joint American Diabetes Association/European Association for the Study of Diabetes recommend a glycemic target of glycated hemoglobin (A1C) 7% for most patients.6 However, the guidelines recommend flexibility around this target depending on patient and disease features including disease duration, life expectancy, and risks associated with hypoglycemia. Therefore, an older patient with long-standing diabetes may pursue a less-stringent target. The American Association of Clinical Endocrinologists and American College of Endocrinology also recommend individualization of targets, but list an A1C level of less than or equal to 6.5% as the typical target.7

Specific to A1C, CMS translates these guideline-based targets into the quality measure: Diabetes HbA1c {Poor Control}.8 This measure scores a plan on the number of patients whose most recent HbA1c level during the year was greater than 9% among the total number of patients with diabetes between 18 and 75 years old. The lower the share of patients who measure above 9%, the better. As an outcome measure, the HbA1c Poor Control score contributes triple weight compared with a process measure, such as Adult Body Mass Index Assessment. The selection of this modest A1C goal by CMS likely reflects a do-no-harm perspective, considering the vulnerability of the Medicare population and the recommendation to set less stringent glycemic targets for these patients. However, given the growing number of classes of medication available to treat diabetes with a low intrinsic risk of hypoglycemia, and that every 1 percentage point reduction in A1C reduces the risk of developing eye, kidney, and nerve disease by 40%, it is not clear that setting a more stringent population-level goal for Medicare patients would not ultimately generate net benefit.9 We also note that other programs (eg, HEDIS) include more stringent A1C targets of 7% and 8%. How, for example, should a plan approach benefit design if it seeks to improve performance on the quality measure HbA1c Poor Control?

Shortly after launch of a new medication, pharmacy decision makers rely on clinical trial data and cost-effectiveness models. These economic models extrapolate clinical trial results and include outcomes of interest to decision makers: the predicted health and economic outcomes of different treatment pathways.10 However, these models do not completely capture the short- to mid-term needs of a US pharmacy director seeking to put together a competitive benefit package at a competitive price, in a marketplace where patients may change insurers. Predicted quality measure achievement of different pharmacotherapy treatment pathways, over a shorter time-horizon, may also be relevant to US decision makers. This methodology was used to compare canagliflozin and sitagliptin in a “cost-efficiency” model.11 Importantly, current, or likely future, quality measures for patients with diabetes go beyond A1C and include medication adherence, blood pressure, hypoglycemia, body weight, and blood lipids. In general, quality measure creators should consider tailoring diabetes quality measures and measures of holistic care to provide appropriate overall incentives in the management of patients. The pharmacologic effects of different classes of diabetes medications on these additional outcomes vary. Fortunately, cost-effectiveness models already account for these additional outcomes allowing smooth incorporation of these quality measure-based outcomes into cost-efficiency models.

As a simple stylized example, how might an insurer institute a program to improve performance on the quality measure HbA1c Poor Control? Start with the low-hanging fruit: patients who do not test A1C by definition do not meet the treatment goal. Reaching out to untested patients with diabetes at mid-year to encourage A1C testing is a simple strategy that could prove hugely effective (FIGURES 1 AND 2). By waiting until mid-year, the plan avoids the cost of the intervention among patients who would otherwise receive a test and reach target on their own, but the intervention remains early enough so that improvement is possible. Ideally, an outreach program would be designed to enhance shared decision making and target those patients who otherwise would not seek testing. The importance of targeting the program appropriately increases as the cost of the intervention increases (eg, a pharmacist phone call compared with an automated email). If things go well, patients above goal receive an intervention (diet, exercise, and/or pharmacotherapy) to improve A1C. If the clinician selects pharmacotherapy, the patient’s benefit design steers the patient to the treatment most likely to improve overall quality measure performance. Finally, the same outreach program could target those patients above goal at mid-year to return for a second test 3 months later, so that those patients, who were above goal initially, receive a documented final A1C that, ideally, meets goal. The significant improvements in A1C in the Mobile Diabetes Intervention Study, among others, suggest that outreach programs that enable shared decision making, patient education/coaching, and provider communication have real potential to improve performance on quality measures and patient health outcomes.12,13

This strategy would improve quality measure performance while benefiting patients. We concede the program described above certainly does not constitute all that could be done to help patients with diabetes, and falls short of guideline—recommended monitoring. However, in the absence of any program from the insurer, this intervention will likely improve quality measure performance and benefit patients. Moreover, the plan could expand the program further to include other quality measures among patients with diabetes. Overall, quality measures derive from ever-evolving treatment guidelines and seek to assure quality healthcare and improved health outcomes. Quality measures and incentives are increasingly important. The management of patients with diabetes constitutes a particularly important component of overall quality measure performance. To maximize performance on diabetes-related quality measures, plans should factor quality measure performance when implementing outreach programs and planning formulary design.

RELATED COVERAGE: Links Found Among Higher Copayments, Lower Adherence, Higher Medical Costs in Medicare Patients with Type 2 Diabetes

John J. Sheehan, PhD, RPh, is a Senior Director, Health Economics and Outcomes Research at AstraZeneca, Fort Washington, PA.

Joanna P. MacEwan, PhD, is a Research Economist at Precision Health Economists, LLC, Los Angeles, CA. The contents of this paper do not necessarily represent the views of AstraZeneca Pharmaceuticals or Precision Health Economics.

We thank Ross MacLean for his contribution to the development of this article. References

1. Brier B. History of Ancient Egypt: Medicine the Necessary Art. Chantilly, VA: The Teaching Company; 2001.

2. Layton TJ, Ryan AM. Higher incentive payments in Medicare Advantage’s Pay-for-Performance Program did not improve quality but did increase plan offerings. Health Serv Res. 2015;50(6):1810-1828. doi:10.1111/1475-6773.12409.

3. 5-star special enrollment period. Medicare.gov website. https://www.medicare.gov/sign-up-change-plans/when-can-i-join-a-health-or-drug-plan/five-star-enrollment/5-starenrollment- period.html. Accessed February 5, 2016.

4. McClellan M, Penso J, Valuck T, Dugan D, Dubois RW, Westrich K. Accountable care measures for high-cost specialty care and innovative treatment: you get what you pay for— improving measures for accountable care. National Pharmaceutical Council website, 2014. http://www.npcnow.org/system/files/research/download/accountable-care-measures-forhigh- cost-specialty2014-final.pdf.

5. Find a ‘Best Value’ health plan: new rankings of 987 health insurance plans nationwide can help you chose the best one for you and your family. ConsumerReports website. http:// www.consumerreports.org/cro/magazine/2013/11/find-a-best-value-health-plan/index.htm?_sm_au_=iQHjHJNkTWPJDMqM. Accessed March 4, 2016.

6. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycemia in type 2 diabetes, 2015: a patient-centered approach. Diabetes Care. 2015;38(1):140-149. doi:10.2337/ dc14-2441.

7. Garber AJ, Abrahamson MJ, Barzilay JI, et al. Consensus statement by the American Association of Clinical Endocrinologists and American College of Endocrinology on the comprehensive type 2 diabetes management algorithm: 2016 executive summary. Endocr Pract. 2016;22(1):84-113. doi:10.4158/EP151126.CS.

8. Diabetes HbA1c {Poor Control}. HHS website. http://www.hrsa.gov/quality/toolbox/measures/diabetes/. Accessed February 5, 2016.

9. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34): UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352(9531):854-865. doi:http://dx.doi.org/10.1016/S0140-6736(98)07037-8.

10. Charokopou M, Sabater FJ, Townsend R, Roudaut M, McEwan P, Verheggen BG. Methods applied in cost-effectiveness models for treatment strategies in type 2 diabetes mellitus and their use in Health Technology Assessments: a systematic review of the literature from 2008 to 2013. Curr Med Res Opin. 2016;32(2):207-218. doi:10.1185/03007995.2015.1102722.

11. Ektare VU, Lopez JM, Martin SC, Patel DA, Rupnow MF, Botteman MF. Cost efficiency of canagliflozin versus sitagliptin for type 2 diabetes mellitus. Am J Manag Care. 2014;20 (suppl10):S204-S215.

12. Quinn CC, Clough SS, Minor JM, Lender D, Okafor MC, Gruber-Baldini A. WellDoc mobile diabetes management randomized controlled trial: change in clinical and behavioral outcomes and patient and physician satisfaction. Diab Technol Ther. 2008;10(3):160-168. doi:10.1089/dia.2008.0283

13. Izquierdo RE, Knudson PE, Meyer S, Kearns J, Ploutz-Snyder R, Weinstock RS. A comparison of diabetes education administered through telemedicine versus in person. Diabetes Care. 2003;26(4):1002-1007. doi:10.2337/diacare.26.4.1002.