You are here
Evaluation of an Adult Insulin Infusion Protocol At an Academic Medical Center
Acknowledging evidence of possible detrimental effects of tightly controlled blood glucose levels, the American Association of Clinical Endocrinologists and the American Diabetes Association published a consensus statement recommending less strict control for most diabetic patients. As a result of these recommendations, our academic center at Creighton University Medical Center revised its adult insulin infusion protocol to target blood glucose levels ranging from 120 to 180 mg/dL for regular (standard) glycemic control and 80 to 120 mg/dL for tight control; previous targets had ranged from 80 to 180 mg/dL and 70 to 110 mg/dL, respectively. The primary objective was to evaluate the time that blood glucose values were within the target range for patients receiving the new protocol, compared with patients receiving the previous protocol.
Our study was designed to evaluate the effectiveness and safety of the revised protocol. Using a retrospective chart review, we collected data for 4 months from patients on the old insulin protocol (May to August 2009) and for 4 months from patients on the new protocol (September to December 2009). Secondary endpoints included the number of hypoglycemic episodes (blood glucose below 70 mg/dL) and severe hypoglycemic episodes (blood glucose 40 mg/dL or lower) experienced by patients receiving the new insulin protocol compared with those receiving the former protocol.
Patient characteristics were similar at baseline. Blood glucose values stayed within the target range for a significantly shorter time with the new protocol than with the former protocol (44.6% vs. 56.8%, respectively;
Patients’ blood glucose levels were within the target range for a shorter time with the new protocol. Fewer episodes of hypoglycemia were recorded with the new protocol, but rates of severe hypoglycemia were similar with both protocols.
Hyperglycemia, whether a result of stress or illness, is common among hospitalized patients and has been associated with adverse outcomes.1 Early studies suggested that interventions to reduce blood glucose levels resulted in improved clinical outcomes, but these studies were weakened by retrospective designs and other methodological concerns.2
In a landmark study published by Van den Berghe et al. in 2001, a 42% reduction in the relative risk of mortality was observed in surgical intensive-care unit (ICU) patients who were treated with intensive rather than conventional glucose control (
Primarily on the basis of these two Van den Berghe trials, many hospitals pursued intensive glucose control as a standard of care to improve patient outcomes. Professional organizations also recommended glycemic control not only in critically ill patients but in all hospitalized patients as well.5
More recently, two studies have raised questions about the results of the Van den Berghe trials. Davos et al.6 and Brunkhorst et al.7 reported unacceptably high rates of hypoglycemia and did not observe reduced mortality rates with intensive insulin therapy. In addition, two meta-analyses of intensive insulin therapy, published by Pittas et al.8 and Wiener et al.,9 indicated considerable variability in target blood glucose concentrations and in the investigational methods used to evaluate the usefulness of strict blood glucose control.
Adding to the controversy was the conclusion of the multinational Normoglycemia in Intensive Care Evaluation–Survival Using Glucose Algorithm Regulation (NICE–SUGAR) trial.10 Investigators randomly assigned 6,104 ICU patients to receive either intensive insulin therapy (target blood glucose range, 81–108 mg/dL) or conventional insulin therapy (target blood glucose goal, below 180 mg/dL). The results contrasted with those of preceding studies. An absolute increase in the rate of the primary endpoint (death at 90 days) was observed with intensive glycemic control (27.5%), compared with conventional control (24.9%) (odds ratio, 1.14; P = 0.02). Moreover, severe hypoglycemia occurred in significantly more intensive-control patients than in conventional-control patients (6.8% vs. 0.5%, respectively; P < 0.001).
In light of the evidence suggesting a possible detrimental effect of tight glycemic control, the American Association of Clinical Endocrinologists (AACE) and the American Diabetes Association (ADA) agreed on less strict control for most patients. These organizations recommended initiating insulin when blood glucose levels exceeded 180 mg/dL and titrating the dose to maintain a target glucose range of 140 to 180 mg/dL for most critically ill patients.11
Following the AACE/ADA consensus statement, our institution (Creighton University Medical Center) revised its adult insulin infusion protocol to require a blood glucose target range of 120 to 180 mg/dL for regular glycemic control and a target range of 80 to 120 mg/dL for tight control. The former protocol was originally adapted from the University of Washington in Seattle, with goals of 80 to 180 mg/dL for regular control and 70 to 110 mg/dL for tight control.
The new lower limit, 120 mg/dL, was chosen over 140 mg/dL because earlier experience with the insulin infusion protocol at our institution suggested that mean blood glucose levels tended to be near the middle of the target range. In addition, the target patient population for tight glucose control was changed to consist of selected surgical patients only, and regular control was recommended for most other patients. Although the titration algorithm was not modified, additional instructions for nurses to adjust for changes in dietary status were added.
Our study sought to compare the revised insulin infusion protocol with the original protocol in terms of efficacy and safety. The primary objective was to evaluate the percentage of time patients’ blood glucose levels remained in the target range with the previous insulin protocol compared with the revised protocol. Secondary endpoints included the number of hypoglycemic episodes (blood glucose below 70 mg/dL) and severe hypoglycemic episodes (blood glucose 40 mg/dL or below) that patients experienced in the old and new protocols.
Our study was a retrospective chart review, and the protocol was approved by our institutional review board. Data were retrieved from the hospital’s electronic medical records, and patients were identified by screening records for insulin charges. We collected information from May 1 to August 31, 2009, for patients on the old protocol and from September 1 to December 31, 2009 for patients on the new protocol.
We included data for all adult patients (19 years of age or older) whose insulin infusion lasted for 12 hours or longer. Exclusion criteria included the addition of insulin to a concomitant total parenteral nutrition order, an addition of subcutaneous insulin, blood glucose target ranges modified by the physician, and incomplete or missing insulin infusion orders.
In addition to collating data on insulin infusion and hypoglycemia, we collected information related to patient demographics, comorbidities, and concurrent hospital medications and treatments. Medications that we considered to have an effect on blood glucose levels included corticosteroids, anti-psychotic agents, beta blockers, and hydrochlorothiazide. We calculated the total time that each patient spent receiving the insulin drip, the percentage of time during which blood glucose values were within the target range, the number of hypoglycemic episodes experienced, and the number of severe hypoglycemic episodes.
We used IBM SPSS software (version 18) to conduct the statistical analysis. Student’s t-test was used to compare the means of continuous variables for all endpoints. We used chi-square tests or Fisher’s exact test to compare categorical variables. A P value of less than 0.05 was considered to be statistically significant.
A total of 543 patients received an insulin infusion during the study period. Of these patients, 344 met the inclusion criteria. A total of 156 patients were treated with the new protocol, and 188 patients were treated with the old protocol. We identified these patients by screening for insulin infusion charges in the hospital billing rcords.
Baseline demographic characteristics were similar for the two groups (
With the revised protocol, blood glucose levels remained within their target range for a significantly shorter time compared with the previous protocol and the incidence of hypoglycemia (blood glucose, below 70 mg/dL) was reduced (
A separate analysis of tight glycemic control did not reveal any statistically significant differences for any endpoints (
Blood glucose levels were maintained within the newly specified target range for a significantly shorter time with the revised insulin infusion protocol than with the former protocol. Statistically significant differences were noted only with regular (standard) glycemic control, probably because of the narrower range on the revised regular-control protocol (120–180 mg/dL vs. 80–180 mg/dL). Because our study focused on the time that blood glucose levels remained within the target range, levels between 80 and 119 mg/dL were considered to be out of range, although these values are desirable in practice.
Severe hypoglycemia is a potentially devastating adverse event related to intensive insulin therapy. The revisions to the insulin infusion protocol resulted in a significant overall re duction in the number of hypoglycemic episodes. In fact, no episodes of severe hypoglycemia occurred with the revised tight-control protocol. The overall incidence of severe hypoglycemia was low (2.6%) for both old and new protocols, a substantially lower rate than the rates reported in larger, published trials.6–10
The total number of insulin infusions was lower with the new protocol, as expected, because infusions were not initiated until blood glucose levels exceeded 180 mg/dL. Previously, infusions were started when orders were written as long as glucose levels exceeded the lower limit of the target range. In addition, the percentage of insulin infusions that were ordered for tight control declined. The new protocol recommends that tight control be reserved for selected surgical patients, but a prescriber may still select tight control for critically ill patients if preferred. This suggests that prescribers accepted and adhered to the recommendations to maintain blood glucose levels below 180 mg/dL instead of trying to achieve super-tight glycemic control.
Our sample did not include patients with diabetic ketoacidosis (DKA). Patients with diabetic ketoacidosis receive a separate protocol with different therapy goals and instructions for insulin infusions.
Our study might have been limited by several factors, primarily because of its retrospective design. For example, we did not measure nurses’ execution of the insulin protocol. We also assumed that execution of the new protocol was consistent over time because the titration nomogram did not change. Therefore, adherence should have a minimal effect on the final results.
Our analysis focused on the number of hypoglycemic episodes that occurred with the revised protocol because of its clinical importance. Therefore, we did not collect data related to the time during which blood glucose levels exceeded the therapeutic range and the number of occurrences during which they exceeded 180 mg/dL. Having access to glucose data on a continuous basis would be ideal for calculating the time that these values are within the target range, but this is not a realistic expectation in practice. To minimize errors, we calculated the percentage of time that glucose values for all patients stayed consistently within the target range.
Patients receiving the revised insulin protocol had lower rates of hypoglycemia compared with those receiving insulin infusions via the former protocol. The overall incidence of severe hypoglycemia was low (2.6%) with the old and new protocols, a substantially lower rate than that in a number of trials. In our study, the revision of an existing protocol, which had been designed to maintain tighter blood glucose control, resulted in lower hypoglycemia rates while still providing acceptable glycemic control.
Randomized Clinical Studies of Intensive and Conventional Insulin Therapy
||1,548||Surgical||80–110||180–200||Death in ICU||4.6||8.0||0.58 (0.38–0.78)|
||1,200||Medical||80–110||180–200||Death in hospital||37.3||40.0||0.94 (0.84–1.06)|
||1,101||General||80–110||140–180||Death in ICU||16.7||15.2||1.10 (0.84–1.44)|
||537||General||80–110||180–200||Death at 28 days||24.7||26.0||Not reported|
||6,104||General||81–108||144–180||Death at 90 days||27.5||24.9||1.14 (1.02–1.28)|
CI = confidence interval; ICU = intensive care unit; NICE–SUGAR = Normoglycemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation; VISEP = Efficacy of Volume Substitution and Insulin Therapy in Severe Sepsis.
Baseline Demographic Characteristics of Patients Receiving Revised and Previous Insulin Infusion Protocols
|60.3 ± 14.7||61.6 ± 14.9||NS|
|93 (60.0)||120 (64.5)||NS|
|90.9 ± 27.2||92.7 ± 27.3||NS|
|12.7 ± 26.2||13.5 ± 13.8||NS|
|77 (50.3)||83 (45.8)||NS|
|3.3 ± 2.0||3.17 ± 1.9||NS|
|86 (55.8)||132 (71.4)||0.003|
|48 (31.4)||35 (19.2)||0.010|
|3 (1.95)||18 (9.73)||0.038|
|27 (18.1)||18 (10.2)||0.026|
NS = not significant; SD = standard deviation.
Overall Glucose Control and Hypoglycemia With Revised and Previous Insulin Infusion Protocols
|44.6 ± 21.4||56.8 ± 29.0||< 0.001|
|0.29 ± 0.71||0.54 ± 1.31||0.023|
|0.11 ± 0.31||0.15 ± 0.33||NS|
|0.03 ± 0.26||0.02 ± 0.14||NS|
|0.01 ± 0.10||0.004 ± 0.03||NS|
|5 (3.1)||4 (2.1)||NS|
NS = not significant; SD = standard deviation.
Tight Glucose Control in the Revised and Previous Insulin Infusion Protocols
|49.0 ± 19.4||46.1 ± 24.9||NS|
|4,376 ± 5,114.5||5,682.8 ± 6,428.1||NS|
|0.21 ± 0.53||0.57 ± 1.28||NS|
|0.07 ± 0.17||0.13 ± 0.32||NS|
|0 (0)||0.02 ± 0.137||0.03|
|0 (0)||0.002 ± 0.01||0.03|
NS = not significant; SD = standard deviation.
Regular Glucose Control in the Revised and Previous Insulin Infusion Protocols
|42.3 ± 22.1||70.5 ± 28.2||< 0.001|
|4,125 ± 3,965.1||4,911.4 ± 4,697.6||NS|
|0.34 ± 0.79||0.51 ± 1.06||NS|
|0.14 ± 0.36||0.17 ± 0.34||NS|
|0.05 ± 0.28||0.02 ± 0.15||NS|
|0.02 ± 0.13||0.008 ± 0.05||NS|
NS = not significant; SD = standard deviation.
- Kosiborod M, Rathore SS, Inzucchi SE, et al. Admission glucose and mortality in elderly patients hospitalized with acute myocardial infarction: Implications for patients with and without recognized diabetes. Circulation 2005;111:3078–3086.
- Furnary AP, Zerr KJ, Grunkemeier GL, et al. Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 1999;67:352–362.
- Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med 2001;345:1359–1367.
- Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354:449–461.
- Task Force on Inpatient Diabetes. American College of Endocrinology and American Diabetes Association consensus statement on inpatient diabetes and glycemic control. Endocr Pract 2006;12:458–468.
- Devos P, Preiser JC, Melot C. A prospective randomized multi-centre controlled trial on tight glucose control by intensive insulin therapy in adult intensive care units: The Glucontrol study. Intensive Care Med 2009;35:1738–1748.
- Brunkhorst FM, Engel C, Bloos F, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 2008;358:125–139.
- Pittas AG, Siegel RD, Lau J. Insulin therapy for critically ill hospitalized patients: A meta-analysis of randomized controlled trials. Arch Intern Med 2004;164:2005–2011.
- Wiener RS, Wiener DC, Larson RJ. Benefits and risks of tight glucose control in critically ill adults: A meta-analysis. JAMA 2008;300:933–944.
- The NICE–SUGAR Study Investigators. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283–1297.
- Moghissi ES, Korytkowsk MT, DiNardo M, et al. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Endocr Pract 2009;15;(4):353–369.