February 2007: Volume 33, Number 2
Tight Glycemic Control
Overcoming Barriers to Implementation In the Intensive Care Unit
by Frederick L. Kiechle, MD, PhD, Susana León, BS, MT, and Walter Severyn, MD
Hyperglycemia can result from defects in insulin secretion, insulin action, or both (1). Although these metabolic defects may result in diabetes mellitus Type 1 (autoimmune etiology) or Type 2 (intracellular signal transduction errors), life-threatening complications of insulin-dependent Type 1 or Type 2 diabetes mellitus are reduced if the patient’s glucose is maintained in the reference range. This therapeutic goal is usually achieved by intensive insulin therapy or by injecting insulin subcutaneously three to four times per day (2).
Regardless of a previous history of diabetes mellitus, patients may become hyperglycemic secondary to stress induced by critical illness, a condition which is referred to as diabetes of injury (3). The hyperglycemia associated with diabetes of injury has several detrimental effects, including impairments in immune response, gastrointestinal mobility, wound healing, and endothelial cell function. Clinically, the combination of these impairments in metabolic function leads to increased mortality and morbidity for patients suffering from acute stress (3,4). For example, the mortality rate of patients in a mixed surgical-medical intensive care unit (ICU) increases as the mean glucose concentration increases; patients with glucose in the range of 120 mg/dL–139 mg/dL have only a 15.1% mortality rate, while those with glucose levels above 300 mg/dL have a 42.5% mortality rate (4).
Recently, the remarkable improvement in patient outcomes achieved by tight glycemic control protocols in medical and/or surgical intensive care units has led to a revolution in patient care. The American College of Endocrinology has recommended that a preprandial target glucose level of 110 mg/dL and maximum glucose level of 180 mg/dL be established for all hospitalized patients, regardless of the presence or absence of a prior diagnosis of diabetes mellitus (5).
Other organizations have made similar recommendations (6). These recommendations, if adopted, would eliminate the target glucose range of 80 mg/dL–200 mg/dL used by most nursing units following admission of a previous ICU patient (6, 7).
However, barriers to the implementation of a tight glycemic control protocol include nurses’ fear of hypoglycemia; nursing unit costs (7); the need for a suitable method for glucose measurement; and lack of integrated computer software to monitor glycemic control, insulin drip rate, and hyperglycemic events. This article will discuss these challenges and provide suggestions for addressing them.
Insulin Therapy: Conventional vs. Tight Glycemic Control
Multiple studies now demonstrate that using aggressive insulin therapy decreases the total time a critically ill patient experiences hyperglycemia, remarkably reducing mortality and/or morbidity (3, 8, 9). For example, a randomized, prospective trial published in 2001 compared conventional insulin therapy with tight glycemic control in 1,548 patients admitted to a surgical ICU (SICU) (8). In this study, conventional insulin therapy using subcutaneous injections was initiated after the blood glucose was greater than 215 mg/dL and aimed to maintain a target glucose range of 80 mg/dL–200 mg/dL. The tight glycemic control protocol was more aggressive and initiated the use of an intravenous insulin drip after blood glucose was greater than 110 mg/dL, with a target glucose range of 80 mg/dL–110 mg/dL. An algorithm was used to adjust the insulin drip rate dependent on the glucose value. For example, if the glucose concentration was in the range of 120 mg/dL–140 mg/dL, the insulin infusion rate would be increased.
Compared to those on conventional insulin therapy, the patients treated with the tight glycemic protocol had reduced mortality (4.6% vs. 8.0%) and reduced morbidity (sepsis, length of stay, etc.). In a similar study conducted in a medical ICU (MICU), all patients on the tight glycemic control protocol demonstrated a reduction in morbidity but not in mortality, with the exception of patients who were treated for 3 or more days (52.5% for conventional insulin therapy vs. 43% for tight glycemic control) (9). Typically, after the patient is discharged from the ICU to a surgical or medical nursing unit, the tight glycemic control protocol is discontinued, and conventional insulin therapy is initiated using the insulin infusion rate to select the dose of insulin to be injected subcutaneously.
Many nurses fear, however, that potentially lethal hypoglycemia may occur during the intravenous infusion of insulin. This perception represents one of the primary challenges that institutions face during the initial implementation of a tight glycemic control protocol. Studies have demonstrated that hypoglycemia does occur more often in tight glycemic control patients than in patients treated with conventional insulin therapy; however, most patients with hypoglycemia experience only one episode (Figure 1) (8, 9). In our experience, once staff is comfortable with the orders and how to follow the protocol, they become more receptive. Seeing the improvement in patient outcomes especially helps staff accept the protocol as a positive change that is worth the effort. As new nurses join the staff, however, they many initially resist following the tight glycemic protocol.
Another potential barrier to implementing a tight glycemic control protocol is the cost of implementation. Aragon (7) has evaluated nurses’ perceptions and costs related to a variety of tight glycemic control protocols implemented in the ICU. The annual costs for performing 77,954 glucose measurements were $58,500 for supplies and $182,488 for nurses’ salaries, with the mean time to complete a glucose measurement and adjust the insulin drip rate being 4.72 minutes. The nurses felt an easier method for monitoring glucose levels was needed. A closed-loop control system that automatically regulates the dose of insulin based on glucose measurements could permit tight glycemic control without increasing the workload of the ICU nurses.
The costs of providing the direct patient care for tight glycemic control must be compared to the impact the program has on the major components of the cost of care (10). When the total costs for 800 patients on conventional insulin therapy were compared to those for 800 patients on a tight glycemic control protocol in a mixed medical-surgical ICU, net annualized costs decreased by $1,580 per patient if a tight glycemic control protocol was used (10), suggesting that the implementation of a tight glycemic control protocol can result in financial savings. Cost savings included a decrease in patient days in the ICU, reduced laboratory and radiology costs, and fewer days in the hospital following release from the ICU.
Glucose Measurement Method
In addition to overcoming nurses’ concerns about hypoglycemia and the financial costs of implementing a tight glycemic control protocol, establishing such a protocol requires a suitable method for measuring glucose. During the protocol, glucose may be measured by three potential methods: satellite laboratory analyzer, waived glucose analyzer, or continuous glucose monitoring.
Although glucose measurements can also be performed in the central lab, the requirement for consistent rapid turnaround time essentially eliminates this option for use as part of a tight glycemic control protocol. Van den Berghe et al. (8) used a table-top whole blood glucose analyzer in the SICU study, but in a later study the group switched to a waived point-of-care glucose analyzer for the MICU (9). The majority of hospitals in the U.S. that have implemented tight glycemic control protocols also use waived glucose meters to measure glucose in finger stick or arterial specimens.
The third method for measuring glucose is continuous glucose monitoring systems that measure glucose in interstitial fluid in the range of 40 mg/dL–400 mg/dL using a needle sensor or electrical transdermal extraction. These methods must be periodically calibrated by using a waived glucose meter at least four times a day. Also, the data must be downloaded to a laptop computer before evaluation, which eliminates real-time review. Finally, there is a 5–10 minute lag in the interstitial glucose value as compared to the plasma glucose value. Due to these multiple limitations, continuous glucose monitoring devices are rarely used during protocols for tight glycemic control. The optimal measurement solution would employ a non-invasive, precise, and accurate transdermal glucose determination; however, currently this type of device is not available.
Lack of Integrated Computer Software for Protocol Monitoring
Finally, a tight glycemic control protocol should ideally include computer software for protocol monitoring. The ideal features and functionality of such software would include real time tracking of multiple values—including glucose concentration and associated insulin drip rate, hypoglycemic events, and frequency of glucose values in defined intervals—and would sort data by individual nurse, patient diagnosis, specific nursing unit, primary physician for a specific hospital, or hospital system program. Peer group sorting using similar parameters would permit evaluation of a specific program compared with programs in hospitals of a similar size or similar geographic area, as well as teaching vs. non-teaching institutions and for-profit vs. nonprofit institutions, etc.
Medical Automation Systems, Inc. (MAS, Charlottesville, Va.) has designed a new module for its Remote Automated Laboratory Systems (RALS) to provide some of these features and functionalities. Called the RALS Tight Glycemic Control Module (RALS-TGCM), the module is designed to provide caregivers and administrators the ability to view results by patient, by unit, and by hospital, using graphs and tables that illustrate the adherence to the tight glycemic control protocol (6). The RALS-TGCM software provides the means to track the tight glycemic control protocol within a health care institution and between institutions; however, it does not include a method for monitoring morbidity and/or mortality.
MAS has recently conducted a study that included the collection of aggregated glucose data from 26 hospitals which together had 326,745 glucose values. The data was collected from units using RALS software from April to June of 2005. The results revealed variation in the success of the 26 hospitals in achieving patient glucose values in the 80–110 mg/dL range.
Overcoming Challenges Offers Rewards
While implementing a tight glycemic control protocol presents some challenges, overcoming these issues can prove rewarding. Figure 2 (below) illustrates the mean blood glucose in four geographic regions defined by the Centers for Disease Control and Prevention (CDC): Northeast, Midwest, South, and West. The graph also shows the total aggregate mean glucose for all regions and for our institution in the South region, Memorial Regional Hospital in Hollywood, Fla. This data demonstrates that the program at Memorial Regional Hospital is performing better in terms of glycemic control than the hospitals in each of the four geographic regions, as well as the aggregate performance of all 26 hospitals included in the study. This type of peer review data is very useful in monitoring program-to-program performance.
Planning, Education Are Key
In our experience, the tight glycemic control protocol is a relatively inexpensive method for improving patient outcomes. Implementation of the protocol at our hospital has led not only to a reduction in morbidity, but also mortality, in the medical/surgical ICU (5-10). Successful implementation of the protocol requires careful planning by all the stakeholders—laboratorians, endocrinologists, intensivists, nurses, and pharmacists. Ongoing education is also important to achieving compliance with the insulin drip algorithm.
- Kiechle FL, Moore KH. Insulin action and the clinical laboratory. J Clin Ligand Assay 2001; 24: 217–228.
- LeRoith D, Smith DO. Monitoring glycemic control: the cornerstone of diabetes care. Clin Therap 2005; 27: 1489–1499.
- Van den Berghe G. How does blood glucose control with insulin save lives in intensive care? J Clin Invest 2004; 114: 1187–1195.
- Krinsley JS. Association between hyperglycemia and increased hospital mortality in a heterogeneous population of critically ill patients. Mayo Clin Proc 2003; 78: 1471–1478.
- Garber AJ, Moghissi ES, Bransome ED, Clark NG, Clement S, Cobin RH, et al. American College of Endocrinology Task Force on Inpatient Diabetes Metabolic Control: American College of Endocrinology position statement on inpatient diabetes and metabolic control. Endocr Pract 2004; 10: 77–82.
- Zito D, Kongable G, Anderson M. The impact of intensive insulin protocols on the clinical laboratory. J Clin Ligand Assay 2005; 28: 202–206.
- Aragon D. Evaluation of nursing work effort and perceptions about blood glucose testing in tight glycemic control. Am J Crit Care 2006; 15: 370–377.
- Van den Berghe G, Wouters P, Wackers F, Verwaest C, Bruyninckx F, Schetz M, et al. N Engl J Med 2001; 345: 1359–1367.
- Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006; 354: 449–461.
- Krinsley JS, Jones RL. Cost analysis of intensive glycemic control in critically ill adult patients. Chest 2006; 129: 644–650.
Frederick L. Kiechle, MD, PhD, is the Medical Director of Clinical Pathology for the five-hospital Memorial Healthcare System in Hollywood, Fla., and is also employed by Pathology Consultants of South Broward, LLP. He is responsible for medical oversight of the point-of-care testing for the health care system. His email address is firstname.lastname@example.org.
Susana León, BS, MT, has been the Point-of-Care Coordinator at Memorial Regional Hospital and various satellite clinics since 1993. The point-of-care program includes more than 15 different assays, both waived and moderately complex. She oversees more than 600,000 tests per year at over 50 locations.
Walter Severyn, MD, is the Medical Director for Critical Care Medicine at Memorial Regional Hospital.