Printer Friendly

Glucose control in the medical patient: bolus insulin dosing compared to basal-bolus insulin dosing.

Hyperglycemia is a common occurrence among acutely ill patients, regardless of a previous diagnosis of diabetes. Stress of an illness alone can trigger a rise in blood glucose. Hyperglycemia is known to have long-term complications at both macro-vascular and micro-vascular levels (Buttaro, Trybulski, Bailey, & Sandberg-Cook, 2008). Patients with diabetes, which is characterized by hyperglycemia, have over twice the annual health care expense as those without diabetes (Songer, Ettaro, & Economics of Diabetes Project Panel, 2010). Additionally, patients with diabetes have up to four times the number of hospitalizations when compared to populations without diabetes (Coursin, Connery, & Ketzler, 2004).

Hyperglycemia has been linked increasingly to adverse outcomes in varied clinical settings and patient populations (Krinsley, 2006). Improvement of glycemic control has been shown to decrease cardiovascular complications related to diabetes as well as decrease nerve and kidney damage (Buttaro et al., 2008). Considering prevalence and potential complications, the prevention and management of hyperglycemia in the acute care setting has become a key issue.


For the purpose of this study, the following definitions are outlined:

* Hyperglycemia--blood glucose above 110 mg/dl

* Hypoglycemia--blood glucose below 70 mg/dl

* Bolus insulin--rapid acting insulin (insulin aspart [Novolog[R]]) used in response to fluctuation in blood glucose

* Basal-bolus insulin--combination of rapid-acting insulin and long-acting insulin (insulin glargine [Lantus[R]]) used to provide stability in blood glucose

Literature Review

The role of strict management of hyperglycemia in the critical care setting has been well defined. A frequently referenced study of hyperglycemia in the critical care unit (Van den Berghe et al., 2001) revealed specific benefits to patients whose blood glucose was maintained at 70-110 mg/dl. Patients maintained in the target range had 34% reduction in mortality, 46% reduced incidence of sepsis, 41% decreased need for dialysis, 50% decreased blood transfusion needs, and 44% decreased incidence of polyneuropathy.

Additional key research concerning control of blood glucose among the acutely ill is the DIAGAMI (Diabetes and Glucose Infusion in Acute Myocardial Infarction) study, which focused on patients who suffered from an acute myocardial infarction. This study noted a decrease in mortality following an acute myocardial infarction in patients who were managed aggressively to maintain normal glucose levels (Malmberg, Norhammar, Wedel, & Ryden, 1999). Other documented benefits to patients in the acute care setting included improved healing times, decreased rates of infection, decreased mortality, and decreased organ dysfunction, specifically in cardiovascular patient populations (Ghandi, Nuttall, & Abel, 2005).

Various methods are used to control blood glucose within acute care settings, including continuous insulin infusion, bolus (sliding scale) coverage, carbohydrate intake dosing, and basal-bolus combination dosing. The predominant method noted in reviewed literature for this project was continuous insulin infusion on patients within a critical care setting. Insulin infusion to manage hyperglycemia requires intense monitoring of blood glucose and frequent adjustments in the rate to control insulin delivery. Because of the intense monitoring required, this method is not managed easily outside critical care settings where nurses care for multiple patients at a time.

Bolus dosing of insulin remains common practice for managing hyperglycemia in the medical-surgical patient, despite documented increases in episodes of hypoglycemia and hyperglycemia (Molitch, 2007). Independent use of bolus dosing coverage should be removed from inpatient practice and replaced with combination of basal and bolus therapy (Moghissi, 2004). The American College of Endocrinology indicated effective therapy should provide both basal components as well as nutritional/corrective components in order to achieve optimal glucose control (Garber et al., 2004).


The purpose of this study was to compare the efficacy of two methods of managing hyperglycemia in patients hospitalized on general medical units. Blood glucose values of patients who received bolus-only insulin therapy were compared to glucose values for patients who received basal-bolus insulin therapy. Efficacy was evaluated based upon control of hyperglycemia, while also comparing rates of hypoglycemia between the two groups.

Research Design

With approval from the institutional review board, a retrospective review compared patient blood glucose for two treatment methods: bolus-only insulin therapy and basal-bolus insulin therapy. Review of the online medical record was conducted for patients who had been admitted to three medical units, hereafter named units A, B, and C, within a midwestern regional hospital. Study period was July 1-August 31, 2007.


The following was done for each medical record:

* A review of physician orders determined the method of management utilized to treat hyperglycemia. Patient records were divided into two groups based on ordered treatment for elevated blood glucose, and not randomly assigned.

* All actual blood glucose values were reviewed and recorded from laboratory results.

* Blood glucose values from patients treated with bolus-only insulin (Group 1) were compared to values for patients treated by basal-bolus insulin (Group 2).

* Potential confounding factors were reviewed to determine patients' risk categories. The medical director of the diabetes management center at the regional hospital broadly categorized the level of difficulty in glucose management (risk category) according to general disease processes. Low-risk patients had no infectious process and no history of diabetes. Moderate-risk patients had an infectious process but were without a previous history of diabetes. Lastly, high-risk patients had a previous diagnosis of diabetes regardless of infectious processes.


The sample was drawn from all patients admitted to adult medical inpatient units A, B, and C at the study hospital during July-August 2007. Unit A had 44-beds with a predominance of oncology and pulmonary diagnoses. The 36-bed unit B had a predominance of renal failure diagnoses. Unit C contained 27 beds, with a predominance of cardiac diagnoses.

Data collection was completed by the following procedure. Charts of all patients were reviewed in chronological order by admission date starting July 1, 2007. The patients meeting the criteria noted below were utilized in order of admission. This procedure was followed until at least 10 patients were included from each unit, with an equal number of male and female patients. The bolus treatment group (Group 1) resulted in 15 male and 15 female patients naturally. In order for the basal-bolus group (Group 2) to have the same balance, one female chart was excluded and the next male patient's was utilized. Thirty patients in each group were included in the study. Inclusion criteria required subjects to be over age 18; admitted to hospital units A, B or C; and receiving subcutaneous insulin to treat hyperglycemia. Excluded patients were under age 18; receiving steroids; in hospice care; in the hospital overnight only following surgery; and receiving insulin infusion. The populations of patients receiving steroids or under age 18 were too small to conduct meaningful statistical analyses. Hospice patients received comfort care and often did not have blood glucose testing. The low number of surgical patients had potential influence of other hospital initiatives that may have impacted outcomes. Patients receiving insulin infusions generally were managed by an endocrinologist, resulting in a more aggressive degree of management not comparable to the general medical population.


Instruments utilized as part of this project included the Abbott MediSense Precision PCX glucometers used within the hospital. Blood glucose testing for patients was performed on the medical units by staff who had completed training on correct utilization of glucometers. Critical components of training and testing included cleansing skin properly, removing excess alcohol, obtaining blood (including importance of wiping first drop of blood from the finger with a gauze pad), and utilizing additional blood for the glucose testing. Timely downloading of glucometers to the laboratory information system, which interfaces with the electronic medical record, also was reviewed. The glucometers used at the study hospital have built-in quality controls, with a safety measure preventing use until this quality check has been completed every 24 hours. In addition, laboratory personnel complete monthly cleaning and validity testing on point-of-care equipment and maintain manufacturer recommendations.

The data collected for each patient included type of treatment, unit name, coded ordering physician, diagnosis of diabetes, hospital diagnosis risk, length of stay, age, sex, and blood glucose values. During analysis, blood glucose results were grouped by ranges. The target range for blood glucose determined by the medical director of the diabetes management for medical patients was 70-150 mg/dl; however, some physicians continued to prefer to not treat until blood glucose exceeded 250 mg/dl. Based upon this, the following ranges were used: less than 70 mg/dl, 70-150 mg/dl, 151-250 mg/dl, greater than 250 mg/dl.

Data Analysis Results

Descriptive statistics were calculated using Microsoft Excel. Comparisons of standard deviations and means were done using Microsoft Excel Analysis ToolPak[R]. Chi square analysis was done manually with the assistance of the hospital statistician. All data were analyzed and tested at p = [less than or equal to] 0.05 level of significance. Statistically significant data are highlighted within the respective tables.

Sample Characteristics

The study included 60 patients (30 males and 30 females). Ten patients were included each from units A, B, and C. Subjects' age range was 34-92, and length of stay was 11S days (see Tables 1 and 2). While the mean length of stay did not result in a statically significant difference, there was a significant difference in the standard deviation of the bolus group.

Patients were categorized according to risk of difficulty controlling blood glucose as previously described. Of the 60 patients, two were in the low-risk category, four were in the moderate-risk category, and 54 (90% of the sample) were in the high-risk category. All patients in the low-risk and moderate-risk categories were in the bolus treatment group. All patients in the basal-bolus group were in the high-risk category.

Bolus Group Results

Tables 3 and 4 show the results of blood glucose for patients in the bolus group. The bolus treatment group had 357 glucose results. Results were compared for patients with and without diabetes, as well as males and females. There was no statistically significant variation within the bolus group.

Basal-Bolus Group Results

Tables 5 and 6 show 351 glucose results for patients in the basal-bolus group. Results were compared for patients with and without diabetes, as well as males and females. No statistically significant variation occurred within the basal-bolus group.

Comparative Results

For the 60 patients included in the study, 708 individual blood glucose tests were performed. In a comparison of the two groups, 357 blood glucose measures (50.4%) were completed for patients in the bolus group and 351 blood glucose measures (49.6%) were completed for patients in the basal-bolus group.

The average blood glucose for patients in the basal-bolus group was significantly lower than blood glucose of bolus-only patients (see Table 7). The variation among glucose values is also significantly smaller for the basal-bolus group. Figure 1 depicts the difference in standard deviations for individual patients.


The target range for blood glucose determined for medical patients at the study hospital was 70-150 mg/dl. Of patients within the bolus group, 23.8% of blood glucose results were within the target range. In comparison, the basal-bolus group resulted in 62.7% of blood glucose results within the target range. Further analysis was done to examine glucose results in smaller increments (see Figure 2). Results of blood glucose measurement were reviewed by increments of 40 mg/dl. In the 71-110 mg/dl range, 5.94% of blood glucose values from the bolus group fell within this range compared to 26.28% of blood glucose values from the basal-bolus group. Significantly fewer glucose results were in each of the ranges above 191 mg/dl for patients in the basal-bolus treatment group. Glucose results above 310 mg/dl only occurred in the bolus group of patients. Review of individual glucose values in this range consisted of results from 9 different patients of the 30 in the bolus only group.

Additional Analysis

For the 60 patients included in the study, 708 individual glucose results were recorded. Further analysis in the bolus group yielded an average of 11.83 glucose values per patient. An average of 11.72 glucose values was completed for patients in the basal-bolus group. For both groups, this is consistent with routine orders for glucose measurement to be checked four times per day during the hospital stay.



The mean length of hospital stay was 0.7 days shorter for the basal-bolus group than for the bolus groups (see Table 8). Although this was not significant statistically, the standard deviation between the two groups was significantly different with the length of stay for the basal-bolus patients consistently lower. The results also showed a statistically significant, shorter length of stay for patients with diabetes treated by the basal-bolus method than for patients treated by the bolus method.

Blood glucose values were grouped into four categories as determined by the study hospital: less than 70 mg/dl, 70-150 mg/dl, 151-250 mg/dl, and greater than 250 mg/dl. Results in each category were compared using chi square analysis. This analysis resulted in significantly more basal-bolus patients in the acceptable range of 70-150 mg/dl, and fewer results in the two highest categories (see Figure 3). Significant differences existed in results between units and sexes for the bolus group, but those differences were not present within the basal-bolus group.


The data support management of blood glucose by use of basal-bolus subcutaneous insulin. This is evidenced by the statistically significant, larger number of blood glucose values in the target range of 70-150 mg/dL within the basal-bolus group compared to blood glucose values in the bolus group. Previous research has shown improved patient outcomes in the areas of decreased mortality, reduction in infections, decreased risk of renal complications, and decreased length of stay in critically ill patients by predominantly utilizing insulin infusion for glucose control within a target range of 70-110 mg/dL. Since the initiation of this project, a randomized study comparing bolus sliding scale insulin to basal-bolus insulin on 130 patients with type 2 diabetes (RABBIT 2) was published (Umpierrez et al., 2007). The RABBIT 2 Trial resulted in significantly improved blood glucose values in medical-surgical patients with use of basal-bolus insulin (receiving insulin glargine and aspart) when compared to patients receiving bolus insulin (regular insulin) alone. The findings of this study were consistent with RABBIT 2 in showing improved blood glucose control with basal-bolus dosing compared to bolus-only dosing. Both studies demonstrated improved control without an increased incidence of hypoglycemia.

The need was identified for a system for patients on medical units to have consistency and improved glucose control. Order sets were developed by a multidisciplinary team for use outside critical care settings. The team included staff from various services: pharmacy, endocrinology, nursing, hospital education, dietary, laboratory, diabetes education, and nursing management. The proposed orders provided an option to dose utilizing basal and bolus coverage. After the trial and staff education, a revised plan was developed based on preliminary data to facilitate glucose management for patients outside the critical care setting. The revised plan was implemented on all medical units within the organization and was the location of data collection for this study.

Comparison between Groups

The average blood glucose for patients in the basal-bolus group was significantly lower than the bolus group, while achieving less variation in glucose values. Additionally, the percentage of glucose values in preferred therapeutic ranges was greater for the basal-bolus group. The large variability and glucose values above 310 mg/dl were noted only in the bolus group. These findings support not only improved glucose values collectively during the hospital stay, but also more consistent glucose control for each patient receiving basal-bolus therapy.

The noted variation between the treatment groups also supports the need for further standardization of treatment methods. The bolus group exhibited significant differences between sexes and units while the basal-bolus group had no significant differences. This indicates the basal-bolus protocol is neither unit nor sex specific and the protocol is being followed consistently regardless of location in the hospital.

Due to use of standardized orders within the basal-bolus group, there was no variation in treatment regimens present in the bolus group. Because collected data did not include transcription of specific physician orders, detailed differences in treatments were not captured. Physicians wrote their own orders for insulin dosage based upon blood glucose values and indicated when treatment should be initiated. This lack of consistency within the bolus group further reinforces need for use of standardized treatment methods.


The current sample was small at 60 patients, with $4 having diabetes. This represents approximately S% of patients with a diagnosis of diabetes on these units during the time this study was conducted. Even with a small sample, statistically significant variation was noted between treatment groups. The sample also was not stratified to include an equal number of patients from each risk category; therefore correlation to confounding factors affecting blood glucose values was not possible. The basal-bolus group experienced statistically improved glucose control despite all patients in this group being in the high-risk category. In addition, use of a single facility prevents results from being easily generalized to other facilities. However, there is support for further evaluation of glucose management within the study hospital.

The method of treatment between the groups was determined by the ordering physician as opposed to being randomized. Randomization was not an option due to the nature of this retrospective review. Admission blood glucose values were included in this study for both treatment groups even though they were not a result of interventions.


The need for continued quality improvement is at the forefront of care within hospitals across the nation. The knowledge gained from research has countless applications and benefits to the patients as well as health care organizations. Regulatory agencies have begun to examine not only quality indicators that require mandatory reporting, but also the processes in place and the utilization of evidence-based practice.

This study resulted in a higher percentage of blood glucose values within the target range with treatment by basal-bolus subcutaneous insulin when compared with bolus-only dosing. Patients in the basal-bolus group had higher numbers of blood glucose values near normal range without a significant increase in hypoglycemia. If the demonstrated outcome benefits among critically ill patients were truly a result of improved glycemic control, medical patients also might demonstrate improved benefits related to mortality, infections, renal complications, and hospital length of stay. Further research is needed to compare the influence of basal-bolus insulin to bolus insulin dosing and determine the impact on patient outcomes.


Centers for Disease Control and Prevention. (2011). National diabetes fact sheet 2011. Retrieved from

Childs, B.P (2004). Hyperglycemia in the hospital: Changing the way we think. Diabetes Spectrum, 17(2), 69-114.

Clement, S., Braithwaite, S.S., Magee, M.E, Ahmann, A., Smith, E.P, Schafer, R.G., & Hirsch, I.B. (2004). Management of diabetes and hyperglycemia in hospitals. Diabetes Care, 27(2), 553-591.

Hirsch, I.B. (2004). Inpatient management of hyperglycemia and diabetes: Implementing the new recommendations. Diabetes Digest, 8, 1-6.

Institute for Healthcare Improvement. (2007). Improvement methods. Retrieved from

National Diabetes Information Clearinghouse (NDIC). (2011). National diabetes statistics 2011. Retrieved from

World Health Organization. (2011). Diabetes fact sheet. Retrieved from


Buttaro, T.M., Trybulski, J., Bailey, P.P., & Sandberg-Cook, J. (2008). Primary care: A collaborative approach (3rd ed.). Philadelphia: Mosby Elsevier.

Centers for Medicare and Medicaid Services. (2010). Hospital-acquired conditions (HAC) in acute inpatient prospective payment system (IPPS) hospitals. Retrieved from

Coursin, D.B., Connery, L.E., & Ketzler, J.T. (2004). Perioperative diabetic and hyperglycemic management issues. Critical Care Medicine, 32(4), 116-125.

Garber, A.J., Moghissi, E.S., Bransome, E.D., Clark, N.G., Clement, S., Cobin, R.H., & Zamudio, V. (2004). American College of Endocrinology Position Statement on Inpatient Diabetes and Metabolic Control. Endocrine Practice, 10(2), 77-82.

Ghandi, G.Y., Nuttall, G.A., & Abel, M.D. (2005). Intraoperative hyperglycemia and perioperative outcomes in cardiac surgery patients. Mayo Clinic Procedures, 80, 862-866.

Krinsley, J.S. (2006). Perioperative glucose control. Current Opinion in Anaestheology, 19, 111-116.

Malmberg, K., Norhammar, A., Wedel, H., & Ryden, L. (1999). Glycometabolic state at admission: Important risk marker of mortality in conventionally treated patients with diabetes mellitus and acute myocardial infarction. Circulation, 99, 2626-2632.

Moghissi, E. (2004). Hospital management of diabetes: Beyond the sliding scale. Cleveland Clinic Journal of Medicine, 71(10), 801-808.

Molitch, M. (2007, January). Insulin administration in the hospital Paper presented at the American Diabetes Association conference, Lafayette, IN.

Songer, T., Ettaro, L., & Economics of Diabetes Project Panel. (2010). Studies on the cost of diabetes. Retrieved from

Umpierrez, G.E., Smiley, D., Zisman, A., Prieto, L.M., Palacio, A., Ceron, M., ... Meijia, R. (2007). Randomized study of basal-bolus insulin therapy in the inpatient management of patients with type 2 diabetes RABBIT 2. Diabetes Care, 30(9), 2181-2186.

Van den Berghe, G., Wouther, M., Weekers, F., Verwaest, C., Bruyninckx, E, Schetz, M., & Bouillon, R. (2001). Intensive insulin therapy in critically ill patients. New England Journal of Medicine, 345(19), 1359-1367.

Kristin Marie Deno, RN, NP-C, is Nurse Practitioner, Franciscan Saint Elizabeth Health, Lafayette, IN.

Judith Schaper, MS, is Statistical Consultant, Los Angeles, CA.

Note: The authors and all MEDSURG Nursing Editorial Board members reported no actual or potential conflict of interest in relation to this continuing nursing education article.

Age of All Patients

                                            F Value
                               Standard     Standard
                  Mean         Deviation    Deviation

Bolus             67.6         15.6
Basal-Bolus       67.3         12.4

                  F            T Value      T
                  Critical     Mean         Critical

                  1.84         0.08         1.69


Length of Stay

                                            F Value
                               Standard     Standard
                  Mean         Deviation    Deviation

Bolus             3.9          3.0
Basal-Bolus       3.2          1.7

                  F            T Value      T
                  Critical     Mean         Critical

                  1.84         1.11         1.69


Bolus Patients

Number of Glucose Results by Category

                               <70 mg/dl         70-150 mg/dl

Bolus        No diabetes       0                 16
             Diabetes          8                 69
All bolus treatment (n=357)    8                 85

                               151-250 mg/dl     >250

Bolus        No diabetes       27                3
             Diabetes          144               90
All bolus treatment (n=357)    171               93


Bolus Patients: Blood Glucose by Sex

Blood                                  F Value
Glucose                   Standard     Standard
by Sex       Mean         Deviation    Deviation

Male         204.7        78.9
Female       211.7        85.7

Glucose      F            T Value      T
by Sex       Critical     Mean         Critical

             2.48         0.23         2.13


Basal-Bolus Patients

Number of Glucose Results by Category

                       <70 mg/dl    170-150      151-250      >250
                                    mg/dl        mg/dl

All basal-bolus        9            220          114          8
treatment (n=351)


Basal-Bolus Patients: Blood Glucose by Sex

                                       F Value
                          Standard     Standard
             Mean         Deviation    Deviation

Male              136.7         44.5
Female            138.0         40.8

             F            T Value      T
             Critical     Mean         Critical

                   2.48        0.083         2.13

All Patients: Blood Glucose Values

                                            F Value
                               Standard     Standard
                  Mean         Deviation    Deviation

Bolus             208.4        82.2
Basal-Bolus       137.6        42.7

                  F            T Value      T
                  Critical     Mean         Critical

                  1.92         4.22         1.69


Patients with Diabetes: Length of Stay

                                            F Value
                               Standard     Standard     F
                  Mean         Deviation    Deviation    Critical

Bolus             4.3          3.2
                                            3.54         1.9
Basal-Bolus       3.2          1.7

                               T            T
                  T Value      Critical     Critical
                  Mean         0.05         0.10

                  1.5          1.69         1.3
COPYRIGHT 2011 Jannetti Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:CNE SERIES
Author:Deno, Kristin Marie; Schaper, Judith
Publication:MedSurg Nursing
Article Type:Report
Date:Sep 1, 2011
Previous Article:What is the future of nursing?
Next Article:Perspectives of patients and families about the nature of and reasons for call light use and staff call light response time.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters