Intended for healthcare professionals


Hypoglycemia as an indicator of good diabetes care

BMJ 2016; 352 doi: (Published 07 March 2016) Cite this as: BMJ 2016;352:i1084
  1. Rene Rodriguez-Gutierrez, postdoctoral researcher12,
  2. Kasia J Lipska, assistant professor of medicine3,
  3. Rozalina G McCoy, assistant professor of medicine4 5,
  4. Naykky Singh Ospina, endocrinology fellow1,
  5. Henry H Ting, senior vice president6,
  6. Victor M Montori,, professor of medicine1,
  7. Hypoglycemia as a Quality Measure in Diabetes Study Group
  1. 1Knowledge and Evaluation Research Unit, Division of Endocrinology, Diabetes, Metabolism and Nutrition, Department of Medicine, Mayo Clinic, Rochester, MN 55905, USA
  2. 2Division of Endocrinology, University Hospital Dr Jose E Gonzalez, Autonomous University of Nuevo Leon, Monterrey, Mexico
  3. 3Section of Endocrinology, Yale School of Medicine, New Haven, CT, USA
  4. 4Division of Primary Care Internal Medicine, Mayo Clinic, Rochester, MN, USA
  5. 5Division of Health Care Policy and Research, Mayo Clinic, Rochester, MN, USA
  6. 6New York-Presbyterian Hospital, New York, USA
  1. Correspondence to: V M Montori montori.victor{at}
  • Accepted 5 February 2016

Rene Rodriguez-Gutierrez and colleagues argue that more attention should be paid to hypoglycemia when assessing management of diabetes

The goals of diabetes care are to reduce the risk of short and long term complications, increase longevity, and improve health related quality of life. Tight glycemic control—aiming for a hemoglobin A1c (HbA1c) concentration below 6.5-7.0%—has been the cornerstone of diabetes care based on the results of early randomized clinical trials that suggested a reduction in microvascular and macrovascular complications.1 2 These trials also found a twofold to threefold increase in the risk of severe hypoglycemia among patients randomized to intensive glycemic control.3 4 However, given the long term benefits of intensive glycemic control, this risk seemed justified and some have suggested a target for HbA1c “as close to normal as possible (<6%).”5 6 For many patients, hypoglycemia is the limiting factor to achieving stringent glycemic goals.5 7 It is also often perceived an unavoidable burden that needs to be accepted in order to accrue the potential long term benefits of glycemic control.

But this model of care is now changing, particularly in type 2 diabetes. The effect of tight glycemic control on microvascular and macrovascular outcomes that are important to patients (end stage renal disease, dialysis, blindness, clinical neuropathy, stroke, and death) remains uncertain,8 9 10 11 making it increasingly difficult for many patients and clinicians to accept the risk of severe hypoglycemia. Furthermore, even mild episodes of hypoglycemia (those that are self treated) impose a burden, cause distress, and disrupt the ability of patients to perform everyday activities.12 In addition, severe hypoglycemia is associated with adverse outcomes such as death, cardiovascular events, cognitive impairment, dementia, impaired autonomic function, fall related fractures, poor quality of life, and increased cost (table).12 13 14 15 16 17 18 19 20 21 Recent clinical diabetes guidelines advocate assessment and prevention of hypoglycemia, including reducing stringency of glycemic targets among patients at risk of severe hypoglycemia.2 22 We propose that hypoglycemia be included as a counterbalance measure of quality as an effort to fully optimize and individualize glycemic control and thereby improve the care of patients with diabetes.

Adverse outcomes (implications) of hypoglycemia that are important to patients

View this table:

Current quality measures in diabetes care

A key effort to standardize measurement and reporting of the quality of care in diabetes came with the Diabetes Quality Improvement Project (DQIP) in the early 2000s. The DQIP proposed HbA1c <9.5% (80 mmol/mol); low density lipoprotein (LDL) cholesterol ≤3.36 mmol/l (130 mg/dL); blood pressure control (<140/90 mm Hg); annual foot, eye, and nephropathy examination; smoking cessation counseling; annual HbA1c testing; and biennial lipid profiles. These measures were later used in the US Healthcare Effectiveness Data and Information Set and have been widely adopted worldwide,23 including by the National Committee for Quality Assurance and National Quality Forum in the United States, and the National Institute for Health and Care Excellence (NICE) in the United Kingdom.24 25 26 The guidelines give little attention to hypoglycemia. The International Diabetes Federation (IDF), an umbrella organization of 230 national diabetes associations, proposes more than 20 potential quality indicators, but none targets hypoglycemia.27 NICE simply adds that patients who have experienced severe hypoglycemia should be referred to a specialist diabetes team.26 Only the Department of Veterans Affairs has specifically focused on hypoglycemia with its recent initiative promoting the formulation of a personal plan for managing blood glucose.28

Towards hypoglycemia as a quality measure

Quality measures are used for both accountability (such as pay for performance programs) and to drive quality improvement. Quality measures should be evidence based, reflect processes or outcomes important to patients, have sufficient validity and reliability, be feasible to collect and report, and be usable by clinicians and patients.29

Most current quality measures in diabetes are based on processes of care (eg, periodic measurement of HbA1c; annual screening for nephropathy, neuropathy, and retinopathy; and aspirin use) and intermediate or surrogate clinical outcomes (eg, achieving prespecified thresholds of HbA1c, LDL cholesterol, and blood pressure). Though such targets are easy to standardize, measure, and report they neglect a key aspect of patient centered care: safety.

Accordingly, hypoglycemia can be a measure of the safety of optimizing glycemic control. Multiple factors suggest that hypoglycemia is a useful measurement of quality. Hypoglycemia is common and can affect most patients with type 1 diabetes and many patients with type 2 diabetes treated with insulin or insulin secretagogs. On average, patients have two to three episodes of mild hypoglycemia and one episode of severe hypoglycemia per year.30 31 These averages, however, do not reflect the fact that most episodes occur in a few high risk patients who tend to have more and more frequent episodes.32 Furthermore, although the incidence of chronic complications of diabetes seems to be plateauing, hypoglycemia continues to be a concern.30 33 Hypoglycemia can be prevented by increasing the awareness of patients and clinicians, teaching patients self management skills, and tailoring glycemic control targets and tactics to reduce the risk of both hypoglycemia and hyperglycemia.34 35 Its use as a quality measure could therefore lead to improvements in safety. The outcome is also important to patients, both acutely12 19 and in the long run.15 1617

An important criterion for a quality outcome is that it is measurable. We can measure blood sugar levels, patients can report symptoms, and episodes that lead to medical attention can be captured through electronic medical records or claims. Although no consensus on the definition of severe hypoglycemia exists, most guidelines use an agreed definition of needing third party assistance (not necessarily a medical professional).36 However, severe hypoglycemia episodes may be under-reported by patients, who may risk losing their driving license or their job.37 Milder episodes also represent a challenge. Blood glucose meters and continuous glucose monitoring devices can capture hypoglycemia in patients who self monitor. Nevertheless, wide consensus on evidence based definitions for mild and severe hypoglycemia is needed.38

Although it is premature to specify which quality measures will best address hypoglycemia, processes related to its prevention could serve as quality measures. These include assessment for hypoglycemia at every visit, change in the diabetes treatment program in response to a hypoglycemic episode, education for patients to prevent hypoglycemia, and prompt detection and treatment of events. Actions could also include education of patients on how to respond to hypoglycemia, use of a medical alert bracelet or similar identification, and prescription of glucose tablets or glucagon. Arguably, current quality measures do not promote improvements in the implementation of these measures. While not all of these measures are pertinent to every patient with type 1 and type 2 diabetes, some should be focused on the care of people at high risk of hypoglycemia—that is, patients who have experienced hypoglycemia.

Unintended consequences

Although the rationale to implement hypoglycemia as a quality measure is compelling, its implementation could be associated with unintended consequences that could affect the quality of care. As an example, in a misguided attempt to prevent hypoglycemia, HbA1c levels could rise in some patients to the level of symptomatic hyperglycemia. The cost of care may be increased by shifting to newer and more expensive drugs that are less likely to cause hypoglycemia, although there is no definitive evidence that such agents are truly better.39 Insulin use could be delayed or complex insulin regimens used less frequently. Self monitoring and the use of expensive continuous glucose devices may increase (or decrease if they “worsen” a quality measure used to document hypoglycemia), increasing the burden of treatment for some. In addition, the burden of measurement on practice and on the workload of clinicians may increase.

The risk of unintended consequences could be reduced by educating patients and clinicians that optimal glycemic control requires balancing hyperglycemia and hypoglycemia. As an example, loosening target goals in clinical practice guidelines to an HbA1c of 7.0-8.5%, rather than picking a tight cut-off may be helpful in optimising glycemic control. In future, the burden of measurement could be reduced by using automated data from electronic health records, rather than requiring manual extraction from charts. Nevertheless, the selection of quality measures for hypoglycemia processes will require careful research and judicious implementation. Appropriate measures could advance quality measurement in diabetes and improve the care for millions of patients with diabetes.

Key messages

Quality measures for diabetes have focused on hyperglycemia at the expense of hypoglycemia

Hypoglycemia disrupts patients’ lives and is associated with adverse health outcomes

Measures designed to detect, prevent, and promptly treat hypoglycemia are needed


  • Other members of the Hypoglycemia as a Quality Measure in Diabetes Study Group are: Yogish C Kudva, Mark E McConnell, and Nilay D Shah.

  • Contributors and sources: RRG, KJL, RGM, NSO, and VMM are endocrinologists and health services researchers with special interest in patient centered diabetes care. HHT is a cardiologist with extensive experience in quality of care. The article resulted from discussions about the burden of living with treatment-induced hypoglycemia and its noticeable absence from most quality improvement initiatives. RRG and VMM conceived the idea. RRG drafted the manuscript. All authors (KJL, RGM, NSO, HHT, and VMM) contributed to manuscript critical appraisal and review. All authors had full access to all of the data and take responsibility for the integrity of the data and the accuracy of data analysis.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare KJL receives support from CMS to develop and maintain publicly reported quality measures.

  • Provenance and peer review: Not commissioned; externally peer reviewed.


View Abstract