Management of Inpatient Hyperglycemia and Diabetes in Older Adults

Introduction

Over 450 million people worldwide are living with diabetes, leading to an associated estimated global healthcare expenditure of $850 billion in the year 2017.¹ By the year 2030, global costs of diabetes are estimated exceed 2015 levels by 88%, accounting for 2.2% of the global GDP (compared with 1.8% in 2015).² In addition to the socioeconomic burden on national healthcare systems and global economies, diabetes disproportionally affects older adults. It is estimated that one in five individuals over the age of 65 years has diabetes in the United States (U.S.).³ Not only are older individuals more likely to have diabetes, but those with diabetes are three times more likely to require hospital admission compared to younger individuals.⁴

Glycemic control in the hospital setting is important for reducing complications and mortality.^5–7 However, maintaining strict or tight glycemic control in the inpatient setting has been associated with poor outcomes related to iatrogenic hypoglycemia.⁸ The balance between achieving glycemic control and avoiding hypoglycemia is especially challenging in older individuals with longer diabetes duration, diabetes-related complications, multiple medical comorbidities and functional decline. It is important to consider the following factors when determining glycemic targets and diabetes treatment regimens in elderly patients, including: functional status, life expectancy, level of frailty, and cognitive impairment.

Research focused on inpatient diabetes management in older patients has been limited.⁹ We review the prevalence, clinical presentation, the assessment of older adults with diabetes in the hospital. We also explore the role of multidisciplinary team approach, diabetes education, glycemic control strategies, and recommendations for discharge regimens in this population.

Prevalence and Economic Burden

From the years 2000 to 2010, the prevalence of diabetes in the U.S. increased from 39% to 42% according to data from the National Inpatient Sample Database.¹⁰ The majority of reported diabetes in older adults is attributed to type 2 diabetes, which will be the focus of this review. However, type 1 diabetes has been estimated to account for between 5–10% of diabetes cases worldwide, with most cases diagnosed in childhood and young adulthood, though several studies suggest an additional rise in diagnosis rates later in life (age >50 years).^11,12 A total of 7.2 million hospital discharges included diabetes as a hospital diagnosis among U.S. adults in 2014.¹³ Between 2007 and 2014 there were more than 32 million hospitalizations of patients with diabetes older than 65 years,¹⁴ representing 57% of all admissions of patients with diabetes. Over the same time period, there was also an increase in hospital admissions for hyperglycemic crises (0.8% in 2007 to 1.2% in 2014), with older adults experiencing a disproportionally high mortality rate (12.8%) in this setting.¹⁴

The total costs of diagnosed diabetes have increased to $327 billion in 2017 from $245 billion in 2012.¹⁵ The continued increase in diabetes costs in the U.S. is primarily related to the growing diabetes prevalence among the older adult population ≥ 65 years. Approximately 61% of all diabetes healthcare expenditures are attributed to health resources for older adults.¹⁵ The average annual excess expenditures for those aged <65 years and >65 years is estimated at $6,675 and $13,239, respectively. About 30% of total diabetes care costs represent inpatient care.¹⁵

Hyperglycemia, defined as a blood glucose (BG) > 140 mg/dL, commonly occurs in hospitalized older adult patients. Approximately one-third of general medicine and surgery patients over the age of 65 years develop hyperglycemia during hospital admission, with hyperglycemia rates increasing to over 70% in critically ill and cardiac surgery patients of similar age.^16–19 Additionally, older adults with diabetes have higher hospitalization rates compared to patients without diabetes. Older adults with uncontrolled diabetes [hemoglobin A1c (HbA1c) >7%] have a disproportionally higher hospital admission rates compared to those with controlled diabetes, prediabetes or without a diagnosis of diabetes. ²⁰ The presence of diabetes and hyperglycemia in an aging population are often associated with poor outcomes, highlighting the importance of determining effective inpatient and outpatient treatment regimens.

Pathophysiology

Aging itself has a complex effect on glucose metabolism including beta-cell dysfunction, changes in insulin secretion and insulin resistance, as well as fat regulation.²¹ A decrease in beta-cell function and insulin secretion is known to occur with aging.²² The etiology of hyperglycemia and diabetes in older adults is thought to be multifactorial, and includes genetic factors, diet, lifestyle, body composition, and peripheral insulin sensitivity. Studies comparing hepatic glucose production in younger and older adults have been mixed, and appears that hepatic glucose output may not be affected by age as much as by differences in body composition.^21,23–25 Beta-cell function and insulin secretion in older adults have been shown to be decreased, accompanied by worsening insulin resistance compared to younger adults. ^{26,27 28} In addition, changes in body fat mass with a reduction in lean muscle mass that occur with age, as well as adipose tissue dysfunction [a pro-inflammatory state] increases insulin resistance in older adults.

Acute illness is associated with increased cortisol and catecholamine release, along with gluconeogenesis and decreased peripheral tissue glucose uptake, all contributing to hyperglycemia during stress. Therefore, older adults in the hospital are particularly susceptible to worsening glucose homeostasis.

Assessment of older adults with diabetes in the inpatient setting

Diabetes in older adults is accompanied by elevated risk of diabetes complications and, as expected, is associated with an increased incidence and prevalence of geriatric syndromes. ^29,30 Geriatric syndromes are multifactorial clinical conditions in older adults; they can greatly impact diabetes self-care, quality of life, and health outcomes, (rehospitalization, mortality). ^31,32 These include cognitive or psychological impairments, including depression, delirium, and dementia; also more functional disabilities can be present including falls, malnutrition, and urinary incontinence.^31,33

These types of geriatric syndromes are highly prevalent in the inpatient setting. A prospective study of hospitalized older adults >70 years with diabetes reported high rates of cognitive impairment including dementia (27.8%), delirium (21.1%), depression (38.9%), and dependency in activities of daily living (91.1%) ³⁴. Cognitive evaluation with mini-mental status examination found low number of patients with normal results (12.2%), and a high number of patients who screened positive for dementia (43.8%) without a prior diagnosis ³⁴.

Care for older adults with diabetes and hyperglycemia requires assessment of comorbid conditions and geriatric syndromes with a comprehensive geriatric assessment (CGA), Figure 1.³¹ The CGA is a multimodal, multidisciplinary evaluation comprised of validated screening tools and assessments targeting medical conditions, functional capacity, cognitive, psychological and social domains for older adults ^33,35. The individual components of the CGA are required by the centers for Medicare and Medicaid services, and are typically conducted by primary care physicians ³⁶. This evaluation specifically aims to identify existing geriatric syndromes to assess risk, while comprehensively reviewing patients’ overall health and values ^37,38. In hospitalized and community dwelling older adults, routine CGA is associated with improved outcomes, including lower rates of institutionalization, slower decline in quality of life and lower mortality rates; it is included in most guidelines for management of diabetes in older adults in both the inpatient and outpatient setting ^{30,31,37,39–42}.

Figure 1.

Comprehensive geriatric assessment throughout the hospital stay

The CGA can provide a framework for assessment and development of care plans in the inpatient setting. Older adults with diabetes are at risk of progression of such geriatric syndromes ⁴³. Hospitalization provides an opportunity for the primary team to identify these risks, by careful evaluation of pertinent medical conditions, assessment of cognition, polypharmacy, functional status and goals of care.⁴⁴ The multidisciplinary team for older adults with hyperglycemia/diabetes typically can include (but not limited to) consultation services (endocrinology, geriatrics, neurology, psychiatry, and palliative care teams), diabetes education, physical and occupational therapy, registered dietitians, pharmacy, nursing, social work and care management. Using this model, components rather than a full CGA can be best utilized in the inpatient setting. On discharge, evaluation of functional capacity, social and financial support can help formulate a safe inpatient to outpatient care plan, Figure 1.

Expert consensus panels have developed targeted geriatric assessment tools; a one page assessment specifically based on the CGA was recently validated at a geriatric day hospital in Brazil (mean age 79.5±8.4 years) ⁴⁵. It included 10 domains (social support, recent hospital admissions, falls, number of medications, ADL, cognitive performance, self-rated health, depressive symptoms, nutritional status, and gait speed) ⁴⁵. Further, this instrument has been used to predict 1-year mortality in hospitalized older adults (79.4±8.4 years) ⁴⁶. This screening tool requires additional testing to validate its widespread use, as efficient CGA models are essential for hospital practice.

Two specific domains – cognitive impairment and frailty – may be most beneficial to assess in a targeted CGA for hospitalized older adults with diabetes. Cognitive impairment is seen in up to half of all older adults with diabetes and is associated with increased risk of hypoglycemia ⁴⁷. The validated mini-cog test is used for screening cognitive impairment in older adults in the outpatient setting, taking only about 3 minutes to perform ^48,49. Diabetes in older adults is associated with reduced muscle strength, sarcopenia, accelerated loss of muscle mass, poor muscle quality and represents an independent risk factor for frailty, which in turn, is associated with increased mortality ^41,50,51. Several frailty-screening tools have been validated and can be utilized in older adults with diabetes, but there is currently no inpatient gold standard for their use ^52–55.

Multidisciplinary CGA offered in an inpatient geriatric evaluation and management unit (GEMU), may be beneficial for the care for frail older persons admitted to the hospital and lead to less functional decline at discharge and a lower rate of institutionalization one year post discharge.⁴³ Another successful program is the Acute Care for Elders (ACE) model, developed to prevent functional decline and adverse outcomes after hospitalization in older adults. It is characterized by patient-centered care, and includes frequent medical review, early rehabilitation and discharge planning. A systematic review and meta-analysis of older adults (average age 81 years) admitted to ACE units (N = 6,839) demonstrated fewer falls, less delirium, less functional decline, shorter length of stay, fewer discharges to nursing home, and more discharged to home.⁵⁶ These inpatient models highlight the significance of a multidisciplinary team approach and use of the CGA as components of specialized inpatient care for older adults and may play a significant role in care of older adults with diabetes.

Multidisciplinary teams and diabetes education in older adults

Hospitalization offers a unique opportunity for multidisciplinary care in diabetes management and education with specialty services available if needed, Figure 2. Retrospective studies highlight the significance of specialized diabetes consultation team participation in inpatient glucose management, as demonstrated by reduction in 30-day readmission rates and length of stay ^57,58. Additionally, formal diabetes education in hospitalized patients with poorly controlled DM is independently associated with lower frequency of hospital readmission at 30 days ⁵⁹. A study evaluating the role of a multidisciplinary team focused on inpatient glycemic control, diabetes education, and discharge planning showed improved glycemic control at 1-year post discharge with significant improvement in glycemic control in patients newly started on insulin with an average HbA1c reduction of 2.4%. ⁶⁰ A multidisciplinary team approach with an individualized diabetes care plan in older adults provides the initial framework for a successful long-term diabetes care in this high-risk population.

Figure 2.

Components of a comprehensive geriatric assessment. ADL, activities of daily living; IADL, instrumental activities of daily living; MMSE,Mini-Mental State Examination; MOCA, Montreal Cognitive Assessment; GDS, Geriatric Depression Scale; PHQ-9, patient health questionnaire 9.

From DeCarlo K, Wallia A. Inpatient management of T2DM and hyperglycemia in older adults. Curr Diab Rep. 2019;19(10):104; with permission.

Glycemic Targets

Data from observational studies have shown a consistent association between both hypoglycemia and hyperglycemia with increased complications and mortality in hospitalized patients.^61,62 Although clinical trials in the hospital have included a broad range of participants including older patients, the most frail and elderly populations have often been excluded from these trials. ^7,63–65

Recommendations for glycemic targets in hospitalized patients have historically been divided between populations with critical vs. non-critical illness. Even though older adults have been included in these trials, no specific recommendations have been made regarding glycemic target goals based on age and age-related factors alone. In the absence of specific recommendations for inpatient glycemic targets in older adults, we recommend following general clinical guidelines for adult patients, but with an emphasis on hypoglycemia prevention.

The 2019 American Diabetes Association (ADA) Standards of Medical Care in Diabetes on inpatient glycemic control recommends targeting glucose levels between 140 and 180 mg/dL (7.8 – 10 mmol/L) for most hospitalized patients (in both critical and non-critical care settings) to improve outcomes associated with uncontrolled hyperglycemia while minimizing the risk of hypoglycemia.⁶⁶ Treatment of inpatient hyperglycemia in those with and without diabetes is typically recommended once glucose levels reach 180mg/dL, including in surgical patients during the perioperative period. ^66,67 More stringent goals may be appropriate for select hospital populations (i.e. cardiac surgery, acute stroke), provided the targets can be achieved without significant hypoglycemia. However, in agreement with less aggressive glycemic targets, two recent randomized controlled trials in cardiac surgery patients failed to demonstrate that intensive insulin therapy targeting glucose levels between 100–140 mg/dL reduced hospital complications when compared to a target glucose range of 140–180 mg/dL. ^68,69

Admission HbA1c concentration is a good predictor of inpatient glycemic control and hypoglycemia risk in insulin treated patients with type 2 diabetes ⁷⁰. Patients with admission HbA1c levels < 7% are more likely to achieve target glucose control during admission, compared to those with HbA1c levels between 7–9% and >9%, but are also at higher odds for hypoglycemia. ⁷⁰ To decrease the risk of iatrogenic hypoglycemia, it has been recommended that daily insulin dosage adjustments occur when glucose values fall below 100 mg/dL (5.6 mmol/L)⁷¹. Less stringent target glucose ranges are needed to reduce the risk of hypoglycemia in terminally ill patients, in patients with severe comorbidities, and in care settings where frequent glucose monitoring or close nursing supervision is not feasible.

Inpatient Hypoglycemia

Patients most likely to develop hypoglycemia during admission tend to be older, have more comorbidities, and are receiving insulin therapy.⁷² The most common risk factor for inpatient hypoglycemia is treatment with insulin, though other risk factors include use of sulfonylureas, insulin administration errors, poor nutrition intake, and changes to hospital routine. ^72–74 Elderly people are also more prone to hypoglycemia due to a higher rate of comorbidities such as renal failure, malnutrition, malignancies, dementia and frailty ^75,76.

Several observational and clinical trials have reported that hypoglycemia in hospitalized older adults is associated with longer length of hospital stay and higher mortality ^75,77,78. A case-control study showed that older patients admitted to acute medical and geriatric wards with hypoglycemia had 3.7 higher odds of dying in the hospital. This association was independent from other risk factors.⁷⁷ Kagansky et al, ⁷⁵ in a study including over five thousand hospitalized patients 70 years or older, found that sepsis, albumin level, malignancy, sulfonylurea and insulin treatment, alkaline phosphatase level, female sex, and creatinine level were all independent predictors of hypoglycemia in the hospital. Hypoglycemia was associated with mortality, however, this association was lost after adjusting for multiple covariates.⁷⁵ Additional studies have suggested that in patients with and without diabetes, mortality is associated with spontaneous hypoglycemia and not with iatrogenic-induced hypoglycemia (i.e., insulin therapy).^72,79,80 Establishing an independent causal relationship between hypoglycemia and mortality among elderly individuals with multiple comorbidities is complex in observational studies. Clinical trials in non-ICU settings have reported an incidence of hypoglycemia around 12%–38% for patients with type 2 diabetes receiving insulin therapy. ^{7,64,65,81–83} These trials, however, have reported a very low incidence of severe hypoglycemia in a controlled setting and have not been powered to assess mortality. However, in critically-ill patients, the presence and degree of hypoglycemia are related to increasing mortality risk. In The Normoglycemia in Intensive Care Evaluation–Survival Using Glucose Algorithm Regulation (NICE-SUGAR) trial reported a mortality rate of 35.4% in those experiencing severe hypoglycemia (< 40 mg/dL [2.2 mmol/L]) with an associated hazard ratio for death of 3.21 (95% CI, 2.49 to 4.15) compared to patients without hypoglycemia ⁸.

Management of Hyperglycemia in the ICU

Intravenous (IV) insulin is the preferred method for treatment of hyperglycemia in the critical care setting. Various continuous insulin infusion (CII) protocols for the treatment of medical and surgical ICU patients have been reported in the literature, ⁸⁴ and more recently computer-based algorithms to guide clinical staff in adjusting insulin infusion have become commercially available. Although the use of computer-based algorithms has been associated with lower rates of hypoglycemia, glycemic variability and a higher percentage of glucose readings within target range, no studies have evaluated their impact on hospital complications or mortality compared to use of standard paper-based algorithms ^5,85.

Management of Hyperglycemia in Non–Critical Care Settings

Basal and rapid-acting subcutaneous insulin

Subcutaneous (SC) insulin has been considered the preferred agent for glycemic control in non-ICU settings, typically using basal insulin alone, or in combination with prandial insulin.⁷¹ Selecting insulin treatment regimens in older adult patients offers more complexity related to medical comorbidities, variable nutrition status or presence of frailty, as well as the associated hypoglycemia risk with increased potential for hypoglycemia unawareness.

For most patients with adequate oral intake with uncontrolled hyperglycemia a basal-bolus approach may be considered, starting total daily dose (TDD) of 0.2–0.3 units per kilogram, with half of the TDD given as basal insulin and the other half as rapid-acting insulin divided between meals.⁹ Variable nutritional status and oral intake may be a more common chronic issue compounded by the effects of acute illness in an older adult population, leading to an increased risk of iatrogenic hypoglycemia with certain insulin regimens. Several studies have shown higher rates of hypoglycemia with use of human insulin (NPH and Regular), particularly with pre-mixed formulations (3-fold increase in hypoglycemia) in older adults compared to treatment with a basal bolus regimen using insulin analogs.^83,86 We recommend against the use of premixed insulin in the hospital.

Several modifications to insulin regimens have been suggested to reduce hypoglycemia in at-risk populations. The Basal Plus⁶⁵ trial evaluated treatment with basal and prandial insulin (basal bolus regimen) compared to basal insulin with only SSI at mealtimes (basal plus regimen) or SSI alone in patients with type 2 diabetes. The total daily dose of insulin was reduced for those with age ≥70 years and/or serum creatinine ≥2.0 mg/dL with a starting basal dose of 0.15 units/kg/day. Results showed comparable glycemic control and frequency of hypoglycemia between the basal bolus and basal plus groups, suggesting that the use of basal insulin with SSI at meals combined with a lower TDD of insulin may be preferred in older patients with poor oral intake.⁶⁵ A retrospective study evaluating use of the GesTIO hospital insulin protocol (lower TDD of insulin at 0.2–0.3 units/kg/day and reduction of prandial insulin dose for pre-meal BG 70–90 mg/dL) in older frail patients showed improvement in mean daily BG, a 9.1% incidence of mild hypoglycemia and no severe episodes of hypoglycemia.⁸⁷

Supplemental/correction doses of short-acting insulin (“sliding scale”)

The use of sliding scale insulin (SSI), regular or rapid-acting insulin administered at scheduled intervals, continues to be a common method for treatment of hyperglycemia in the hospital, including in older patients. The reactive nature of SSI use alone to correct hyperglycemia often predisposes patients to wide glycemic variability with episodes of both hyper- and hypoglycemia, as well as increased risk of complications, ^7,64 though some patients with mild hyperglycemia may achieve adequate glycemic control with the use of SSI alone.

Inpatient use of Non-insulin Therapies for Management of Hyperglycemia

Due to concerns regarding the safe use of older oral agents in the acute setting, professional associations have recommended the use of insulin regimens as the preferred approach to treat hyperglycemia in the hospital. Safety concerns have been related mainly to the fear of lactic acidosis with metformin, delayed onset of action and volume retention with thiazolidinediones, and sustained hypoglycemia with sulfonylureas.^71,88 Despite such recommendations the use of OADs is common in clinical practice in the US and worldwide.⁸⁸ No studies, however, have been conducted to prospectively evaluate the safety and efficacy of older OADs in the hospital. In a randomized controlled trial (RCT) enrolling residents at long term facilities (LTC) or skilled nurse facilities, we observed that the continuation of oral antidiabetic agents (OADs) resulted in similar glycemic control compared to low dose insulin therapy (glargine, starting dose 0.1 U/kg/day).⁸⁹ Nevertheless, we observed a high rate of hypoglycemia in both groups (~30 over a 26 week period).⁸⁹

More recently, agents associated with a low risk of hypoglycemia (i.e. incretin based therapies) have been prospectively tested in hospitalized patients (including older adults) with diabetes and residents in long term care and skilled nurse facilites.^88,90,91

Incretin-Based Therapies

Incretin agents provide a normal physiologic response in the presence of glucose with a low risk of hypoglycemia.^92,93 Large cardiovascular outcomes trials (CVOT) testing DPP-4 inhibitors have confirmed a safe cardiovascular risk profile for this drug class. CVOT have also confirmed the safety of GLP-1 receptor analogs (GLP-1 RA) and have reported a reduction in major cardiovascular events.^94–97 In the inpatient setting, clinical trials with DPP-4 inhibitors, native GLP-1, and GLP-1 RA have shown promising results in acutely ill patients.

Dipeptidylpeptidase-4 inhibitors

Multiple RCTs and observational studies have reported that DPP4 inhibitors alone or in combination with basal insulin are safe and effective for the management of general medicine and surgery patients with type 2 diabetes.^98–102 An initial pilot study suggested sitagliptin alone or in combination with basal insulin was effective in patients with type 2 diabetes with mild to moderate hyperglycemia.⁹⁹ Participants were treated at home with either diet alone, oral agents or low doses of insulin (<0.4 units/kg/day). Ninety participants were randomized to sitagliptin, sitagliptin plus basal insulin, or basal-bolus insulin. No differences were observed in the primary outcome of mean daily blood glucose between groups. The TDD of insulin and the number of insulin injections were lower in those receiving sitagliptin. A sub-analysis showed, however, that patients with randomization BG >180 mg/dL assigned to sitagliptin alone had higher mean daily BG, suggesting monotherapy with a DPP-4 inhibitor may not be as effective for patients with moderate to severe hyperglycemia. In a follow up larger study, we compared sitagliptin with a single dose of basal insulin with basal-bolus regimen in a broad population of hospitalized patients with insulin home doses up to 0.6 units/kg/day.⁹⁸ Both groups achieved similar glycemic control. Treatment failure occurred in 16% patients assigned to sitagliptin–basal and 19% of those receiving basal–bolus. Treatment failure was independent of group assignment but was associated with higher A1c levels, as previously described.⁷⁰ The odds of failure were 30% higher per one unit change in HbA1c (OR 1.3, 95% CI 1.2–1.5).

In a RCT enrolling only surgical patients we compared linagliptin vs basal-bolus in patients treated with diet, oral agents or total insulin dose (TDD) ≤0.5 units/kg/day prior to admission.¹⁰² We observed a mean daily BG difference of 10.8 mg/dl (95% confidence interval 0.72, 22 mg/dl). Among participants with randomization BG<200 mg/dl (63% of cohort), glycemic control was similar in linagliptin and basal-bolus groups (160±41 vs. 157±41mg/dL, p=0.43); however, patients with BG≥200 mg/dl assigned to linagliptin had worse glycemic control compared to basal-bolus (196±47 vs. 165±47 mg/dl, p<0.001). Linagliptin resulted a remarkable reduction in the incidence of hypoglycemic events (1.6% vs. 11%, p=0.001, 86% relative risk reduction). Similarly, Garg et al compared the use of saxagliptin with a basal bolus regimen and observed no difference in glycemic control in patients with very mild hyperglycemia (admission BG ~150 mg/dl and mean A1c<7% ).¹⁰¹

In addition, we recently reported the results of a RCT comparing linagliptin with low dose of basal insulin in residents of long-term care (LTC) and skilled nursing facilities.⁹¹ Participants had a known history of T2DM, BG > 180 mg/dL, and/or with HbA1c >7.5% while receiving treatment with diet, OADs, or insulin at a total daily dose ≤0.1 U/kg. Treatment with linagliptin resulted in noninferior glycemic control compared to insulin glargine. Furthermore, linagliptin resulted in fewer hypoglycemic events <70 mg/dL (3% vs. 37%) or <54 mg/dL (7% vs 0%) compared to glargine.⁹¹ These findings cannot necessarily be translated to the hospital setting; however, they clearly show insulin use in frail patients is associated with a high rate of hypoglycemia even at very low doses.

The results of these trials indicate that treatment with DPP-4 inhibitors is associated with lower rates of hypoglycemia and similar glycemic control compared to more complex insulin regimens, particularly among patients with mild to moderate hyperglycemia, Figure 3.

Figure 3. Recommendations to start anti-hyperglycemic therapy in older adults with type 2 diabetes in the hospital.

TDD: total daily dose, AC: before meals, TDD: total daily dose.

* Adjust dose according to GFR (sitagliptin or saxagliptin), no adjustment is needed with linagliptin.

** Antidiabetic agents: oral agents and GLP1-RA

No prospective studies have determined the efficacy of other oral antidiabetic drugs in the hospital setting.

Adapted from Pasquel FJ, Fayfman M, Umpierrez GE. Debate on insulin vs non-insulin use in the hospital setting-is it time to revise the guidelines for the management of inpatient diabetes? Curr Diab Rep. 2019;19(9):65; with permission.

Glucagon-like peptide-1 receptor analogs

Pilot studies have reported that infusion of native GLP-1 or GLP-1RA may lead to improved endothelial function,¹⁰³ reduced infarct size following myocardial infarction,¹⁰⁴ increased left ventricular function.^105–108 GLP-1RA and native GLP-1 have also been tested in critically ill and surgical patients^109–112 In a non-randomized study in cardiac surgery patients with hyperglycemia suggested intravenous exenatide was effective, and associated with a low-risk of hypoglycemia compared to historical controls.¹⁰⁹ Similarly, in a randomized trial, Besch et al reported 72% of patient treated with exenatide and 80% of the insulin-treated group (p=0.3) met target glucose concentration. Subjects in the exenatide group received less insulin overall and had a longer time interval to initiation of insulin. ¹¹³

Perioperative treatment with liraglutide, administered before non-cardiac surgery was compared with insulin infusion, or subcutaneous insulin with 50% of their home insulin dose given the morning of surgery.¹¹¹ Treatment with liraglutide was associated with lower glucose levels one hour after surgery with no differences in hypoglycemia or postoperative complications. Liraglutide use was associated with increased pre-operative nausea rates. Similarly, preliminary results of the GLOBE trial ¹¹⁴ presented at the American Diabetes Association Scientific Sessions in 2019, showed that liraglutide administered before surgery resulted in significant improvement in glycemic control with lower insulin requirements after cardiac surgery ¹¹⁵.

We recently compared subcutaneous exenatide 5 mcg administered twice daily both with and without basal insulin with a standard basal-bolus insulin regimen in general medicine and surgery patients with diabetes.¹¹⁶ Patients receiving exenatide plus basal insulin in the hospital achieved similar mean BG compared to those on basal-bolus insulin (154±39 vs 166±40 mg/dL, p=0.31) and lower mean BG compared to exenatide treatment alone (177±4 mg/dL, p=0.02). A higher proportion of participants in the exenatide plus basal insulin group had blood glucose readings within target range of 70–180 mg/dL (78%) compared to exenatide alone (62%) or basal-bolus insulin (63%). Patients receiving exenatide experienced higher rates of nausea.¹¹⁶

These studies suggest GLP-1 RA may be efficacious in controlling glucose levels without increasing the risk of hypoglycemia. Its use, however, may be limited by potential risk of gastrointestinal side effects and may need to be avoided in people who have active gastrointestinal conditions or those who are frail and have poor oral intake.

We conducted a pooled analysis of incretin-based therapies in inpatient older adults withT2D including a total of 192 elderly patients treated incretin therapy (DPP-4 inhibitor or GLP-1 RA) alone, incretin therapy plus basal insulin, or basal bolus. Incretin therapy was associated with similar glycemic control and lower rates of hypoglycemia compared to a complex basal-bolus regimen. ¹¹⁷

Sodium glucose −2 inhibitors

SGLT-2 inhibitors are a class of oral anti-hyperglycemic agents that prevent glucose reabsorption by the kidneys leading to glycosuria¹¹⁸. In addition to improving glycemic control, large CVOT have shown that SGLT-2 inhibitors have significant cardiovascular benefit with large randomized controlled trials showing reduced CVD-related mortality, fewer hospitalizations for heart failure, as well as a reduction in renal outcomes in patients with diabetic kidney disease.^119–121 These cardiorenal outcomes have also been observed in older adults.¹²² Despite these benefits, SGLT-2i can lead to side effects that may limit their potential for use in the hospital, such as euglycemic diabetic ketoacidosis, acute kidney injury, an genitourinary infections. These potential side effects make the use of SGLT2 inhibitors less attractive older patients with acute illnesses requiring hospitalizations.

New Diabetes Technologies in the Hospital

The use of diabetes technology is emerging in the inpatient setting and includes the use of continuous glucose monitoring (CGM) and insulin pump technology. CGM has allowed for more complete characterization of glycemic control by providing estimated blood glucose values every 5–15 using a small subcutaneous sensor. Although CGM use in the hospital remains investigational, a recent pilot study explored its use in an older adult population to alert nursing staff to hypoglycemic events.¹²³ Patients were randomized to standard-of care capillary glucose monitoring vs real-time CGM monitoring with data transmitted to nursing personnel in a telemetry-type method with an alert for sensor glucose values 85 mg/dL or lower. In this study, overall rates of hypoglycemia rates were lower in the intervention group, and no patient in this group experienced severe hypoglycemia. This study supports continued research into use of CGM for prevention of hypoglycemia in older adult patients with risk factors for hypoglycemia and hypoglycemia unawareness.

Leading edge diabetes technology also includes the use of hybrid closed-loop, or artificial pancreas technology. This refers to the use of CGM and insulin pump therapy integrated with and algorithm for insulin infusion adjustments made based on CGM glucose values. A recent study investigated the use of hybrid closed-loop therapy compared to standard basal bolus insulin administration in non-critically ill hospitalized patients with type 2 diabetes (mean age 67 years).¹²⁴ Time in target glucose range (100–180 mg/dL) was higher in the closed-loop group compared to conventional insulin therapy (65.8 ± 16.8% vs. 41.5 ± 16.9%, p <0.001) without increased rates of hypoglycemia. Further studies are needed to assess feasibility, patient-centered outcomes, and cost-effectiveness of hybrid-closed loop therapy, especially in an older adult inpatient population.

Discharge Planning

Hospital hyperglycemia management is anchored on insulin therapy; however, not all patients require the same home regimen. Discharge for older adults can be clinically complex, given co-morbidities, complex drug regimens, and assessment of geriatric needs (such as the components of the comprehensive geriatric assessment). The Endocrine Society recently published guidelines for the care of diabetes in older adults in which a tiered framework was reviewed to categorize patients by level of health and mortality risk. In categorizing healthy, intermediate health, and poor health by evaluation of chronic conditions, visual impairment and functional status, one can appropriately recommend goals of care. ^37,125 It is recommended that the outpatient goal HbA1c for the healthy elderly patients be 7.0–7.5%, intermediate health group 7.5–8.0%, and those with poor health 8.0–8.5%.³⁷

While inpatient or recent HbA1c can help guide discharge therapy, there have been no validated algorithms or studies evaluating discharge recommendations in older adults only. However, we have previously recommended the following discharge DM therapy in older adults: admission HbA1c range <7.5–8%, re-start home regimen (oral agents with insulin if necessary), HbA1c 8.0–10.0%, consider oral agents plus basal insulin (50% of hospital insulin dose), and for HbA1c >10%, patients should be discharged on a basal bolus insulin regimen or on a combination of pre-admission oral agents with approximately 80% of hospital basal insulin dose ⁹; this is less aggressive than our treatment algorithm in a broad inpatient population. ¹²⁶ It is prudent to evaluate at discharge the ability of a patient to perform self-care, such as glucose monitoring, ability to self-inject, and oral intake and/or ability complete activities of daily living. Based on this assessment, the diabetes plan may need to be tailored to include modified regimens (e.g. use of long acting insulin + oral) or utilization of caregivers to help with diabetes self-care.

In the outpatient setting, simplifying insulin regimens can be markedly beneficial in this population. Munshi et. al noted reduction in hypoglycemia, and reduced DM-related distress, with no change in HbA1c with simplification; simplification included moving bed time to morning, changing from mixed to basal, and reduction and/or switching of prandial insulin with oral medications.¹²⁷

Metformin should be considered a first line agent at discharge, assuming no contraindications (severe renal disease, GI distress). DPP-4 inhibitors have been shown to be safe and efficacious in older adults with less adverse events compared to sulfonylureas.¹²⁸ A recent RCT showed a twofold increase in the probability of achieving glycemic targets without an excess of hypoglycemia after adding a DPP-4 inhibitor to older patients treated with basal insulin.¹²⁹ The addition of sitagliptin and metformin at discharge following an HbA1c based algorithm was efficacious in a broad inpatient population.¹³⁰ We suggest caution when adding new OADs to older adults treated with insulin or sulfonylureas. A careful reduction in insulin doses or discontinuation of sulfonylureas (particularly glyburide) when replacing with agents with low risk of hypoglycemia is recommended. More data has been forthcoming on the safety and use of SGLT2s in older adults.¹²² Clinical trials evaluating the role of these agents in older patients with heart failure at discharge are needed.

Discharge planning is key, and follow up should be individualized based on the patient’s needs and capabilities. An outpatient follow-up visit for diabetes is generally recommended within 2–4 weeks of discharge if possible. Discharge recommendation should include medication reconciliation, structured communication in regards to medication changes, follow up testing and follow up clinic appointment and contact information. ¹³¹ Home health referral can be utilized for those patients who are eligible, and if the referral lists DM as a concern in the consultation, glucose monitoring and DM care can be included. Patients under Medicare coverage are eligible if they are homebound and require skilled nursing care, physical/occupational therapy, or speech services. ¹³² If a patient requires full time or extended skilled nursing care, home health is not applicable, and care would be taken over at the facility based on discharge recommendations. ^132,133 In a recent RCT enrolling long-term care residents (~70 years), we observed that the use of a DPP-4 inhibitor can result in similar glycemic control along with a marked reduction in hypoglycemia when comparing to low doses of basal insulin.⁹¹

Conclusions

The number of adults living with diabetes has more than tripled during the past two decades, representing one of the fastest growing health challenges of the 21^st century.¹³⁴ Patients with diabetes are more likely to be hospitalized compared to patients without diabetes. Among them, the most vulnerable are older patients with uncontrolled diabetes. In the hospital more than half of all admitted patients with diabetes are older than 65 years, however, limited research has focused specifically on this population. Older patients with diabetes are commonly frail and have multiple comorbidities, therefore, an individualized patient-centered treatment approach is needed to avoid dangerous hypo- and hyperglycemic events. Approaches with simplified regimens utilizing agents with lower risk of hypoglycemia should be preferred for older patients with diabetes in the inpatient setting. Insulin therapy remains a useful drug for many elderly patients, such as those with moderate to severe hyperglycemia, history of hyperglycemic emergency, and those who fail to maintain glucose control with oral agents. Recent clinical trials have consistently shown that DPP-4 inhibitors alone or in combination with low doses of basal insulin have safety advantages and comparable efficacy to complex insulin therapy regimens.

KEY POINTS.

• Diabetes in older adults is a growing public health concern in the outpatient and inpatient settings.

• Most hospitalized patients with diabetes are older (age > 65 years), frail, and have multiple comorbidities.

• A comprehensive geriatric assessment and a team-based approach are recommended to tailor an individualized care plan, with consideration of the patients’ personal goals, comorbidities, functional status, life expectancy, and risk of clinically significant hypo- and hyperglycemia.

• Cautious use of insulin therapy is recommended in older adults with risk factors for hypoglycemia, including decreased renal function and poor nutritional status.

• The results of recent clinical trials suggest simplified therapeutic regimens using agents with a lower risk of hypoglycemia are preferable for most patients with mild to moderate hyperglycemia.

SYNOPSIS.

Diabetes is one of the fastest growing health challenges in the world. With the increase in life expectancy during the last 50 years, the proportion of older adults with diabetes continues to increase. Older adult patients with diabetes are more likely to be hospitalized than patients without the disease, and most hospitalized patients with a diagnosis of diabetes are over the age of 65 years. Older patients with diabetes are commonly frail and have multiple comorbidities, necessitating an individualized patient-centered treatment approach to avoid dangerous fluctuations in blood glucose levels. Insulin therapy remains a useful regimen for many elderly patients, such as those with moderate to severe hyperglycemia, type 1 diabetes, hyperglycemic emergencies, and those who fail to maintain glucose control on oral agents alone. Recent clinical trials, however, have consistently shown that several non-insulin agents alone, or in combination with low doses of basal insulin, have comparable efficacy and potential safety advantages to complex insulin therapy regimens. Determining the most appropriate diabetes management plan for older hospitalized patients requires consideration of many factors to prevent poor outcomes related to dysglycemia.