Glycemic variability in hospitalized patients: choosing metrics while awaiting the evidence.

Hyperglycemia, hypoglycemia, preexisting diabetes, and glycemic variability each may affect hospital outcomes. Observational findings derived from randomized trials or retrospective studies suggest that independent of hypoglycemia and hyperglycemia, a relationship exists between variability and hospital outcomes. A review of studies conducted in diverse hospital populations is reported here, showing a relationship between measures of variability and nonglycemic outcomes, including ICU and hospital mortality and length of stay. "Glycemic variability" has an intuitive meaning, understood as a propensity of a single patient to develop repeated episodes of excursions of BG over a relatively short period of time that exceed the amplitude expected in normal physiology. It is proposed that each of 3 dimensions of variability should be separately studied: (1) magnitude of glycemic excursions during intervals of relative stability of the moving average of BG, (2) frequency with which a critical magnitude of excursion is exceeded, and (3) presence or absence of fine tuning. Multiple hospital studies have found that the standard deviation (SD) of the data set of blood glucose values (BG) of individual patients predicts outcomes. An appropriate refinement would be to report the "Reverse-transformed group mean of the SD of the logarithmically transformed BG data set of each patient," with confidence intervals. In logarithmic space, group means of the SD of BGs of each patient may be compared, using an appropriate parametric test. Upon reverse transformation, the upper and lower bounds of the confidence intervals become asymmetric about the reverse-transformed group mean of the SD. There is a need to understand what patterns of dispersion of BG over time are captured by SD as a predictor of outcomes. Among the causes of high SD, a subgroup may consist of patients having frequent oscillations of BG. Another subgroup may consist of patients experiencing a major change of overall glycemia during the timeframe of data collection. Appropriate metrics should be developed to recognize both variability in the sense of recurrent large oscillations of BG, and separately to recognize any time-dependent change of overall glycemia during hospitalization. Especially in relation to uncontrolled diabetes, there is a need to know whether rapid correction of chronic hyperglycemia adversely affects hospital outcomes. We have some understanding of how to control or prevent change of overall glycemia, and less understanding of how to control variability. Each may be associated with outcomes, and each may be detected by a high SD, but it remains uncertain whether intervention to prevent either pattern of changing glycemia would affect outcomes.

Guidelines for the use of an insulin infusion for the management of hyperglycemia in critically ill patients.

OBJECTIVE: To evaluate the literature and identify important aspects of insulin therapy that facilitate safe and effective infusion therapy for a defined glycemic end point. METHODS: Where available, the literature was evaluated using Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) methodology to assess the impact of insulin infusions on outcome for general intensive care unit patients and those in specific subsets of neurologic injury, traumatic injury, and cardiovascular surgery. Elements that contribute to safe and effective insulin infusion therapy were determined through literature review and expert opinion. The majority of the literature supporting the use of insulin infusion therapy for critically ill patients lacks adequate strength to support more than weak recommendations, termed suggestions, such that the difference between desirable and undesirable effect of a given intervention is not always clear. RECOMMENDATIONS: The article is focused on a suggested glycemic control end point such that a blood glucose ≥ 150 mg/dL triggers interventions to maintain blood glucose below that level and absolutely <180 mg/dL. There is a slight reduction in mortality with this treatment end point for general intensive care unit patients and reductions in morbidity for perioperative patients, postoperative cardiac surgery patients, post-traumatic injury patients, and neurologic injury patients. We suggest that the insulin regimen and monitoring system be designed to avoid and detect hypoglycemia (blood glucose ≤ 70 mg/dL) and to minimize glycemic variability.Important processes of care for insulin therapy include use of a reliable insulin infusion protocol, frequent blood glucose monitoring, and avoidance of finger-stick glucose testing through the use of arterial or venous glucose samples. The essential components of an insulin infusion system include use of a validated insulin titration program, availability of appropriate staffing resources, accurate monitoring technology, and standardized approaches to infusion preparation, provision of consistent carbohydrate calories and nutritional support, and dextrose replacement for hypoglycemia prevention and treatment. Quality improvement of glycemic management programs should include analysis of hypoglycemia rates, run charts of glucose values <150 and 180 mg/dL. The literature is inadequate to support recommendations regarding glycemic control in pediatric patients. CONCLUSIONS: While the benefits of tight glycemic control have not been definitive, there are patients who will receive insulin infusion therapy, and the suggestions in this article provide the structure for safe and effective use of this therapy.

Stress hyperglycaemia.

Results of randomised controlled trials of tight glycaemic control in hospital inpatients might vary with population and disease state. Individualised therapy for different hospital inpatient populations and identification of patients at risk of hyperglycaemia might be needed. One risk factor that has received much attention is the presence of pre-existing diabetes. So-called stress hyperglycaemia is usually defined as hyperglycaemia resolving spontaneously after dissipation of acute illness. The term generally refers to patients without known diabetes, although patients with diabetes might also develop stress hyperglycaemia-a fact overlooked in many studies comparing hospital inpatients with or without diabetes. Investigators of several studies have suggested that patients with stress hyperglycaemia are at higher risk of adverse consequences than are those with pre-existing diabetes. We describe classification of stress hyperglycaemia, mechanisms of harm, and management strategies.

Treatment of Diabetes in Older Adults: An Endocrine Society* Clinical Practice Guideline.

OBJECTIVE: The objective is to formulate clinical practice guidelines for the treatment of diabetes in older adults. CONCLUSIONS: Diabetes, particularly type 2, is becoming more prevalent in the general population, especially in individuals over the age of 65 years. The underlying pathophysiology of the disease in these patients is exacerbated by the direct effects of aging on metabolic regulation. Similarly, aging effects interact with diabetes to accelerate the progression of many common diabetes complications. Each section in this guideline covers all aspects of the etiology and available evidence, primarily from controlled trials, on therapeutic options and outcomes in this population. The goal is to give guidance to practicing health care providers that will benefit patients with diabetes (both type 1 and type 2), paying particular attention to avoiding unnecessary and/or harmful adverse effects.

Patient-level glucose reporting: averages, episodes, or something in between?

The article by Van Herpe and colleagues in the previous issue of Critical Care describes the glycemic penalty index (GPI), which weights both hyperglycemic and hypoglycemic blood glucose measurements commensurate to their clinically significant difference from target. Although certain adverse consequences result from isolated severe hyperglycemic episodes, several specific outcomes depend upon overall hyperglycemia. In contrast, although mortality has been related epidemiologically to overall low blood glucose, specific negative outcomes may depend upon isolated episodes. Capturing both hypoglycemia and hyperglycemia in a single index will be shown to be useful if the GPI enables us to better define insulin strategies, outcomes, and targets.

Managing hyperglycemia in hospitalized patients.

Insulin infusion is used in the critical care setting for prevention of hyperglycemia and is administered most safely under a structured, dynamic, dose-defining algorithm. The ordering of basal-prandial-correction SC insulin therapy, appropriate for most hospitalized patients who are eating, is simplified and standardized to excellence by the development of institutional order sets or computerized order entry templates. Basal insulin therapy is prescribed as intermediate-acting insulin or long-acting insulin analogue. Prandial insulin therapy is delivered with meals to prevent excessive glycemic excursions from occurring after ingestion of meals and is prescribed as rapid-acting insulin analogue. Correction-dose insulin therapy is ordered as small doses of rapid-acting insulin analogue delivered to correct hyperglycemia and is prescribed with appropriate timing so as to avoid stacking with previously administered doses of rapid-acting insulin analogue. Patients knowledgeable in diabetes self-management will experience satisfaction under an institutional policy that allows self-management to continue under appropriate conditions during hospitalization. To craft appropriate institutional tools for patient care, the input and consensus of a multidisciplinary group of health care professionals, including primary care providers and hospitalists, as well as specialists in diabetes with backgrounds in endocrinology, nutrition and dietetics, nursing, pharmacy, laboratory sciences, and quality assurance, is required.

Effects of outcome on in-hospital transition from intravenous insulin infusion to subcutaneous therapy.

It is widely accepted by medical practitioners that diabetes is a major independent risk factor for the development of cardiovascular disease. However, less attention has been directed toward elevated blood glucose as a predictor of poor outcomes in hospitalized patients in cardiac critical care. This has occurred despite documentation of hyperglycemia in a significant proportion of patients admitted for cardiac care and considerable data supporting the use of intravenous (IV) insulin to achieve glycemic control. The increased risk for mortality due to hyperglycemia provides a strong rationale for an intensive approach using insulin to control blood glucose levels in cardiac patients being treated in acute care and surgical settings. IV insulin infusion is the therapy of choice for patients in cardiac critical care units, with transition to a subcutaneous insulin therapy regimen when appropriate. The timing of this transition can be critical. Strong evidence from studies on patients who have undergone cardiac surgery suggests that glycemic control by insulin infusion should be maintained for > or =3 postoperative days. Nonetheless, transition from IV to subcutaneous therapy must occur at some point during the hospital stay. In conclusion, the implementation of measures to achieve glycemic control in acute cardiac care hospital settings can significantly reduce morbidity and mortality and can substantially decrease the costs associated with prolonged hospital stays. This report reviews recent clinical data on the benefits of IV insulin infusion in cardiac patients in critical care and provides recommendations on transitioning patients from IV insulin infusion to subcutaneous therapy.

Performance of a dose-defining insulin infusion protocol among trauma service intensive care unit admissions.

BACKGROUND: Among critically ill patients, glycemic control reduces mortality and morbidities, but the use of intravenous insulin infusion is complicated by hypoglycemia. Having a standardized algorithm increases the likelihood of effective and safe utilization of intravenous insulin therapy. A tabular dose-defining protocol for intravenous insulin infusion is described, containing design elements intended to minimize risk for hypoglycemia while seeking control in a narrow target range, and performance is evaluated among critically ill trauma service patients. METHODS: The protocol assigns insulin infusion rate (IR) for ranges of blood glucose (BG). The columns are arranged in order of increasing maintenance rate (MR) for insulin infusion. Patient column assignment is determined according to rate of change of BG. During stable column assignment, the IR is a function of column MR and BG. Within-column, the protocol formula provides that (a) for BG between 70 mg/dL and target BG, the IR increases exponentially to the column MR; and (b) for BG above upper target BG range, the IR increases linearly as an adaptation of the rule of 1800, with slope determined by the column MR. Values for IR calculated by formula are rounded to correspond to BG ranges of the table. Performance was assessed in 27 sequential runs among 24 trauma service patients admitted to a surgical intensive care unit (2004-2005). RESULTS: Using point-of-care measurements, mean preinfusion BG was 230.0 +/- 67.9 mg/dL. BG < 140 mg/dL was reached during all 27 runs (median time 5.0 h), and target BG was < 110 mg/dL during 25 runs (median time 11.0 h). For the group of runs attaining target before interruption of insulin infusion, the average +/- SD of the principal measure of glycemic control, the within-run mean BG, was 113.7 +/- 14.8 mg/dL (coefficient of variation 13%, n = 25 runs). After attaining target, the average within-run SD for BG was 22.9 mg/dL. The within-run frequency of hypoglycemic measurements (BG < 70 mg/dL) as a percentage of BG determinations was 2.4%. In this series, no instance of BG <50 mg/dL was seen. CONCLUSIONS: This report describes a nurse-implemented tabular protocol for intravenous insulin infusion having the advantages of efficacy, safety, and simplicity of use. Wide variability of IR in the neighborhood of BG 110 mg/dL is associated with stable BG response, and protection against hypoglycemia is achieved by rapid decline of IR at BGs in or below the target range.

The transition from insulin infusions to long-term diabetes therapy: the argument for insulin analogs.

After cardiac surgery, it is medical mismanagement to place an order for sliding scale insulin at the time of transitioning from intravenous insulin. Use of basal-prandial-correction therapy with insulin analogs constitutes a suitable transitioning regimen for inpatient management of hyperglycemia after heart surgery, to be ordered before interruption of intravenous insulin infusion, in conjunction with a program of blood glucose monitoring before meals, at bedtime, and midsleep. In the ambulatory setting, in comparison to neutral protamine Hagedorn, long-acting insulin analogs reduce hypoglycemia. In comparison to regular insulin, rapid-acting insulin analogs reduce hypoglycemia and improve postprandial control. A standardized approach to order entry for basal-prandial-correction therapy enhances safety and staff familiarity while preserving individualization of patient care. Proposed predictors of successful transition are described. Dose requirements during intravenous insulin infusion can be used to guide initial dose assignments of basal insulin therapy. As the patient approaches discharge, the total daily doses of subcutaneous insulin and basal insulin dose are decreased, and the proportion of prandial insulin approaches or exceeds 50% of the total daily dose as the absolute amount of prandial insulin increases. Before discharge, hyperglycemic patients not known to have diabetes should be advised of the need for outpatient reassessment, and those known to have diabetes but requiring intensification of therapy should participate in decision-making concerning their options for intensified treatment.

No patient left behind: evaluation and design of intravenous insulin infusion algorithms.

OBJECTIVE: To define the characteristics of performance evaluation, algorithm design, and regulation of insulin delivery by which professionals and the healthcare system might differentiate between methodologies for intravenous insulin infusion. METHODS: Published performance criteria used in the assessment of intravenous insulin infusion algorithms are classified. The structure of intravenous insulin infusion formulae is reviewed, as are technologies that might lead to future improvement. RESULTS: Among published reports, no standardization was discernable for description of algorithm characteristics or performance. Except for time-to-target and hypoglycemic episodes, measures using the patient as unit of observation are not employed consistently. CONCLUSION: The healthcare system needs criteria for evaluation and minimal acceptable standards for assessing performance of any algorithm, decision support system, or closed-loop system for intravenous insulin infusion. Inclusion of patient-based measures is necessary to assess the ability of an algorithm to control variability between patients and within a given run. Standardization of performance reporting will help users to select appropriate methodologies.

Intravenous insulin infusion therapy: indications, methods, and transition to subcutaneous insulin therapy.

OBJECTIVE: To describe indications for intravenous (IV) insulin infusion therapy and glycemic thresholds, discuss methods and protocols, and promote use of and access to IV insulin infusion therapy for all appropriate patients in the hospital setting. RESULTS: Randomized, prospectively designed trials support the use of IV insulin infusion therapy for patients in the surgical intensive-care unit, including postoperative cardiac patients and patients having myocardial infarction. Among patients in the surgical intensive-care unit, reanalysis of the data suggested no threshold at which benefit occurred above the blood glucose level of 110 mg/dL. In another study, retrospective analysis of data among critically ill medical and surgical patients suggested a target blood glucose level of 145 mg/dL or less. In other populations, the threshold or ideal target blood glucose range has not been determined. Three protocols for IV insulin infusion are described that maintain blood glucose levels safely below the upper limit of their respective target ranges without substantial risk of hypoglycemia. CONCLUSION: The threshold for initiation of IV insulin infusion is 110 mg/dL for critically ill surgical patients, 140 mg/dL for other medical or surgical patients, 180 mg/dL for patients in whom subcutaneous insulin regimens fail, and 100 mg/dL for pregnant women. The blood glucose target range is 80 to 110 mg/dL for selected critically ill surgical patients, 70 to 100 mg/dL for pregnant women, and 90 to 140 mg/dL for all other patients. Hospitals should develop procedures to make IV insulin infusion therapy available to all appropriate patients.

Description and evaluation of a glycemic management protocol for patients with diabetes undergoing heart surgery.

OBJECTIVE: To evaluate a protocol for glycemic management in the treatment of postoperative heart patients with diabetes in the setting of a community hospital. METHODS: The protocol included a standardized guideline in tabular form for nurse-implemented insulin infusion ("drip") therapy for postoperative glycemic control. At the time of discontinuation of the insulin drip, the glycemic status of patients with diabetes was managed by the endocrine department. Overall, 29 patients were assessed without and 29 patients with use of the protocol in a community hospital. RESULTS: From postoperative days 0 through 4, use of the protocol resulted in a greater number of blood glucose determinations, a trend toward greater utilization of insulin drip therapy without a significant increase in the number of patients treated with insulin drip, and no change in the frequency of hypoglycemic episodes. During the same time interval, the percentages of postoperative days during which at least one blood glucose value equaled or exceeded 250 mg/dL were 27.5% without the protocol and 16.8% with use of the protocol (P = 0.0318). The principal finding of the study was reduction in the percentage of postoperative days during which mean blood glucose values equaled or exceeded 200 mg/dL to less than half the previously observed frequency-from 38.4% without the protocol to 16.8% with the protocol (P = 0.0001). The effectiveness of the insulin drip component of the protocol is suggested by a trend, shown on postoperative days 2 through 4, of 70 patient days with mean blood glucose levels <200 mg/dL (58 of these days without insulin drip therapy) and 15 patient days with mean blood glucose values > or =200 mg/dL (none of these days associated with same-day insulin drip therapy). CONCLUSION: A standardized approach to insulin drip therapy, in combination with subspecialty consultation for follow-up glycemic management with use of subcutaneous administration of insulin, can improve glycemic control in postoperative heart patients without continuation of insulin drip therapy outside the critical-care unit. The trends observed on postoperative days 2 through 4, that most patients maintained glycemic control without insulin drip therapy and that all failures of glycemic control occurred among patients who no longer received insulin drip therapy, suggest the possibility of developing criteria for selection of patients for continuation of insulin infusion outside the critical-care unit.

Hospital hypoglycemia: not only treatment but also prevention.

OBJECTIVE: To propose a strategy, applicable on general hospital wards, for prevention of hypoglycemia in hospitalized patients. RESULTS: Although the mortality rate among hospitalized patients with hypoglycemia has been shown to be 22.2 to 27% in series that included patients with diabetes, some investigators have shown that hypoglycemia is not an independent predictor of mortality. Outside the critical care setting, the comparative risks of hyperglycemia and hypoglycemia and the relationship of hospital hypoglycemia to intensification of glycemic control have not been determined. The reported incidence of hospital hypoglycemia ranges from 1.2% for hospitalized adults to 20% for nonpregnant patients with diabetes admitted without a metabolic emergency. Among patients receiving antihyperglycemic therapy, the literature describes precipitating events--usually a sudden change of caloric exposure-- and predisposing conditions for hypoglycemic episodes. CONCLUSION: Hospital hypoglycemia is predictable, and it is preventable by measures other than undertreatment of hyperglycemia. Physician orders for antihyperglycemic therapy should be written and, if necessary, be revised so as to respond to the presence of predisposing conditions for hypoglycemia. A ward-based protocol or hospital-wide policy should establish the appropriate response to triggering events. Within the time frame of action of previously administered antihyperglycemic drugs (after abrupt interruption of caloric exposure), the threshold for preventive intravenous administration of dextrose is a glucose concentration of 120 mg/dL.

Anticruelty care: commentary.

The anticruelty policy is a best-interests test for treatment plans including decisions to forgo life-sustaining therapy for certain incompetent patients. In connection with specific proposed therapy, the policy requires no reference to the patient's unknowable values, subjective experiences, or quality of life. The decision to undertake a treatment plan derives from the caregiver's knowledge of burdens and benefits of that treatment when used in caring for the competent or for those incompetents capable of growth or repair. The caregiver should weigh the potentially cruel effects of treatment against the likelihood of reducing suffering or encumbrance with the treatment. The terms "burden" and "benefit," in fact, are replaced by the terms "cruelty" and "beneficence," as the relevant opposing outcomes that must be weighed. Thus, the anticruelty policy shifts our scrutiny from experiences of the patient that we cannot evaluate to the proposed actions of the competent decision makers and caregivers. Notably, it is a protreatment policy when the goals of medicine are attainable; and it is an anticruelty policy when they are not. The policy does evaluate the world of the patient to the extent that it requires a judgment based upon external appearances about patient pleasure or happiness in living. It presumes to universalize larger societal values about cruelty, beneficence, compassionate concern for the helpless, and certain rights of individuals. And it presumes to universalize on the patient's behalf specific medical values about hopeless injury, timely death, the goals of medicine, and cruelty, which should remain open to societal discussion and revision. The presented definition of hopeless injury does not require brain death, coma, or persistent vegetative state. Specifically, the policy holds that death is timely for a patient with hopeless injury, and that prevention of death for such patients is not a goal of medicine but a cruelty.

A quadruply-asymmetric sigmoid to describe the insulin-glucose relationship during intravenous insulin infusion.

For hospitalized patients requiring intravenous insulin therapy, an objective is to quantify the intravenous insulin infusion rate (IR) across the domain of blood glucose (BG) values at a single timepoint. The algorithm parameters include low BG (70 mg/dL), critical high BG, target range BG limits, and maintenance rate (MR) of insulin infusion, which, after initialization, depends on rate of change of blood glucose, previous IR, and other inputs. The restraining rate (RR) is a function of fractional completeness of ascent of BG (FCABG) from BG 70 mg/dL to target. The correction rate (CR) is a function of fractional elevation of BG (FEBG), in comparison to elevation of a critical high BG, above target. IR = RR + CR. The proposed mathematical model describing a sigmoidal relationship between IR and BG may offer a safety advantage over the linear relationship currently employed in some intravenous glucose management systems.