| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
• Objectives To compare intensive glucose control with modified conventional control in the medical intensive care unit and to assess compliance with glucose targets, incidence of hypoglycemia, and staffing adequacy.
• Methods Medical intensive care patients who had been receiving mechanical ventilation for less than 24 hours were randomized to intensive or modified conventional protocols for glucose control. Nurses were trained before participating in the study and were interviewed after its completion.
• Results Five subjects were randomized to each protocol. Mean blood glucose levels were 5.8 (SD 1.5) mmol/L (105.3 [SD 26.3] mg/dL) for the intensive group and 9.8 (SD 2.5) mmol/L (177.4 [SD 45.5] mg/dL) for the modified conventional group (P < .001). Fifty percent of glucose levels met target values in the intensive group, and 72% of glucose levels met target values in the modified conventional group (P < .001). Severe hypoglycemia (glucose <2.2 mmol/L [<40 mg/dL]) occurred rarely and without complication. Nurses suggested protocols might be improved by using smaller steps in adjusting insulin dosage and reported that simultaneously caring for more than 1 study subject was taxing.
• Conclusions Target levels for blood glucose were achieved with both protocols. Severe hypoglycemia was rare and uncomplicated regardless of type of glucose control. Additional staffing may be needed for intensive glucose control.
In 2001, Van den Berghe et al4 reported that inhospital mortality was reduced 33% (7.2% vs 10.7%) when critically ill surgical patients were randomized to an intensive insulin therapy group with target blood glucose levels of 4.4 to 6.1 mmol/L (80–110 mg/dL). They furthermore reported reduction of bloodstream infections by 46%, acute renal failure by 41%, red blood cell transfusions by 50%, and critical illness polyneuropathy by 44% when intensive insulin therapy was used. Control of blood glucose level, not the amount of exogenous insulin administered, accounted for the mortality benefit.5,6
| Use of target blood glucose levels of 4.4 to 6.1 mmol/L (80–110 mg/dL) reduced in-hospital mortality by 33% in critically ill surgical patients.
|
Less information is available about whether the benefits of intensive control of blood glucose levels seen in postsurgical patients would apply to critically ill medical patients. Krinsley7 reported that after a protocol for intensive glucose control was implemented to keep blood glucose levels at less than 7.8 mmol/L (140 mg/dL) in a medical-surgical ICU, in-hospital mortality was reduced 29% (14.8% vs 20.9%) compared with historical controls. It is also unclear whether the usual patient-to-nurse ratio in a North American medical ICU (MICU) can support the implementation of intensive control of glucose levels. Van den Berghe et al4 used a team of nurses and a physician not participating in clinical decisions to manage blood glucose levels in their study subjects. Krinsley7 reported no increase in nurse staffing after implementing intensive control to keep blood glucose levels at less than 7.8 mmol/L (140 mg/dL).
We performed a collaborative pilot study in which critically ill patients in the MICU were randomized to receive intensive control of glucose levels or a modification of conventional blood glucose control. The goals of the pilot study were several:
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The committee decided, as had Van den Berghe et al,4 that a double-blind study of blood glucose control was not feasible with the resources available. A double-blind study would require continuous availability of a blood glucose management team separate from the treating staff. Also it would be difficult to conceal blood glucose values from the treating staff because blood glucose measurements are obtained from the clinical laboratory at least once daily for most patients. An alternative study design in which blood glucose control and outcomes would be compared before and after unit-wide implementation of intensive control of blood glucose levels was rejected as being less scientifically powerful than a randomized design that does not rely on historical controls. We chose to use a study design in which patients were randomized to intensive or modified conventional control of blood glucose levels. In order to maintain a group of nurses trained to manage blood glucose levels according to the study protocol, the study was limited geographically to the MICU at the sponsoring institution.
The pilot study consisted of 10 randomly assigned patients, 5 to the intensive control group and 5 to the modified conventional control group. The study had 2 phases: intervention and monitoring. The intervention phase involved nurses’ management of blood glucose levels for up to 28 days according to a protocol while the patient was in the MICU. The intervention phase was discontinued when or shortly before the patient was discharged from the MICU. All patients were monitored through study day 28. The intervention phase was discontinued at study day 28 if the patient was still in the MICU.
An enrollment sheet was set up with 10 consecutive enrollment positions. Five of those positions were randomly chosen (by using a random numbers table) to be in the intensive control group, and the other positions were assigned to the modified conventional control group. From this enrollment sheet, the statistician provided a stack of 10 numbered sealed envelopes, each containing instructions for assignment to a study group. Patients entering the MICU were screened for inclusion and exclusion criteria. If a patient was eligible and informed consent had been obtained, he or she was randomized into the intensive control group or the modified conventional control group according to the instructions in the top envelope (next lowest number) in the stack. Potential participants in the study did not know their group assignment until after informed consent had been obtained. With one exception, the study investigator who obtained informed consent also had no knowledge of the group assignment until after informed consent had been obtained. The exception occurred for the 10th patient; the investigator knew that the patient would be assigned to the modified control group because the intensive control group already had 5 patients.
Sample
The MICU at Loma Linda University Medical Center is a 20-bed unit providing care for critically ill medical patients with a wide variety of predominantly noncardiac disorders. The patient-to-nurse ratio is typically 2:1 or 1:1. One full-time nurse practitioner works in the unit. Physician staffing is provided by a team of internal medicine house staff supervised by full-time teaching attending physicians. All patients are transferred to the care of the MICU physician service when admitted to the unit.
All patients admitted to the MICU who had been receiving mechanical ventilation for less than 24 hours were eligible for inclusion in the study. The rationale for excluding patients not receiving mechanical ventilation is as follows. Because the ultimate goal of the study was to determine if intensive control of glucose levels changed the mortality of MICU patients, the study committee decided to exclude patients for whom such control was unlikely to alter survival. Patients who were not receiving mechanical ventilation were thought to be likely to survive regardless of blood glucose control. Moribund patients and those with a do-not-resuscitate order were excluded in order to exclude high-mortality patients not likely to benefit from intensive glucose control. Patients were also excluded if they were participating in another clinical trial, had a primary diagnosis of diabetic ketoacidosis or a diagnosis of diabetic hyperosmolar nonketotic syndrome. Patients were also excluded if they had medical insurance that required transfer to another hospital when feasible, or if informed consent was not provided by the patient or the patient’s surrogate.
Instrumentation
Blood samples were obtained by finger lancet or from an arterial catheter. Blood glucose level was measured at the bedside with Accu-Chek comfort curve strips and an Accu-Chek Advantage Model 777 glucometer (Roche Diagnostics Corp, Indianapolis, Ind). B Braun Outlook 200 Safety Infusion System pumps (B. Braun Medical Inc, Bethlehem, Pa) were used to deliver insulin infusions consisting of 100 units of regular human insulin in 100 mL of isotonic sodium chloride solution.
Procedures
The committee developed protocols for nurses’ management of blood glucose levels for the 2 study groups (Table 1
). The study was performed under the supervision of the institutional review board of the sponsoring university and was carried out in accordance with the ethical standards set forth in the Helsinki Declaration of 1975. Before the first patient was enrolled, all MICU bedside nurses and charge nurses participated in a 30-minute orientation session in which they learned about the rationale for the study, became familiar with the protocols for management of blood glucose levels, and had their questions answered. After the 10th patient had completed the blood glucose management part of the study, the MICU nurse manager interviewed all the nurses who had cared for study patients and asked the nurses for suggestions to improve the study. The nurse manager recorded and grouped these suggestions.
|
. Patients not receiving tube feedings or total parenteral nutrition received an intravenous infusion of 5% dextrose in water by pump at a minimum rate of 50 mL/h. Patients transported out of the MICU had total parenteral nutrition and tube feedings continued, unless ordered to have no enteral intake, in which case tube feedings were stopped and an infusion of 5% dextrose in water was started, if such an infusion was not already being given. Insulin infusions were stopped at the time of transportation. Blood glucose level was checked 1 hour later. Nutrition and insulin infusions were resumed at previous settings as soon as possible after the patients returned to the MICU. The dose of insulin was decreased or temporarily stopped when tube feedings were stopped for procedures such as endoscopy or extubation. The glucose control protocol was discontinued when patients were transferred from the MICU to an intermediate care unit. The accepting physician in the new unit was informed of the most recent blood glucose level and the insulin infusion rate before the infusion was stopped. Before transfer, a patient could receive, at the discretion of the treating physicians, a dose of subcutaneous NPH insulin and/or regular insulin. The patient could then be monitored for several hours in the MICU to ensure stability of the blood glucose level before transfer. A log was kept of patients screened for the study, with reasons for those excluded. All blood glucose values measured at the bedside were recorded while each patient was in the glucose management phase of the study. Survival status 28 days after enrollment was obtained by a review of the medical record or by a telephone call.
Analysis
Statistical analysis of data was accomplished by using SPSS software (version 10, SPSS Inc, Chicago, Ill). Results were considered significantly different at the P < .05 level. The difference in blood glucose levels between the 2 groups was assessed by comparing means via a t test. The difference between the percentages of blood glucose values within the target range for the 2 groups was assessed by using a z test for proportions. The difference in the incidence of moderate hypoglycemia between the 2 groups was also assessed by using a z test for proportions.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
. The patients enrolled in this study had a mean age of 56.7 (SD 15.2) years. Of the 10 patients, 7 were men and 3 were women. The mean length of stay was 32.1 (SD 23.8) days. Median length of stay was 28 days. Principal diagnoses obtained from the hospital discharge summaries were as follows: 7 patients had acute respiratory failure, 1 had severe sepsis, 1 had acute respiratory distress syndrome, and 1 had out-of-control seizures. A mean of 4.2 concurrent illnesses per subject were also listed on the discharge summaries (Table 3).
|
|
For the 8 patients in whom glucose control was analyzed, the mean blood glucose level was 5.8 (SD 1.5) mmol/L (105.3 [SD 26.3] mg/dL) for the intensive group and 9.8 (SD 2.5) mmol/L (177.4 [SD 45.5] mg/dL) for the modified conventional group (P < .001). In the intensive control group, 50% of 1444 blood glucose measurements were within the target range, compared with 72% of 1348 measurements for the modified conventional group (P < .001). For the intensive control group, severe hypoglycemia (blood glucose <2.2 mmol/L [<40 mg/dL]) occurred only twice—both times in the same subject. Similarly, for the modified conventional group, 2 blood glucose measurements were less than 2.2 mmol/L (40 mg/dL), again both in the same subject. Severe hypoglycemia occurred in 0.1% of blood glucose measurements for both groups and caused no demonstrable clinical complications.
Moderate hypoglycemia (blood glucose 2.2–3.3 mmol/L [40–60 mg/dL]) occurred in an additional 2.1% of the blood glucose measurements in the intensive control group (in 4 of 5 patients) and 0.2% of the blood glucose measurements in the modified conventional group (in 3 of 5 patients; P = .15). Moderate hypoglycemia was also not associated with medical complications. On the basis of intent to treat, 4 of the 5 patients in the intensive control group were alive after 28 days compared with 3 of 5 in the other group. This study did not have sufficient power to test for survival differences between the 2 groups.
| Blood glucose levels in the intensive therapy group were significantly lower than levels in the conventional therapy group.
|
Subjective Results
On the basis of informal interviews with participating bedside nurses, the MICU nurse manager reported that the intensive control and modified conventional protocols had been adequate for "coarse tuning" of blood glucose control. It was the consensus of the nurses, however, that a protocol permitting smaller steps when adjusting insulin dosage (0.1 units/h vs the 0.5 units/h steps in our protocol) might have enabled them to achieve even better compliance with target levels for blood glucose. Surveyed nurses deemed it desirable for patients in the study to have arterial catheters because the patients were otherwise subjected to finger sticks to obtain blood samples no less often than every 2 hours. After the target level for blood glucose was reached, nurses found that some patients had stable blood glucose values and could be safely managed with blood glucose checks every 4 hours rather than every 2 hours as specified in the protocol. Nurses also thought that mild hypoglycemia (blood glucose 3.3–3.9 mmol/L [60–70 mg/dL]) could be successfully managed by temporarily shutting off the insulin infusion rather than by giving intravenous dextrose or using an enteral high-carbohydrate liquid as called for in the study protocol. The few nurses who were assigned to provide care simultaneously for 2 newly enrolled patients in the study reported that it was difficult to meet all the requirements of the study protocol in addition to the usual patient care activities.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Nurses participating in our study thought that even better compliance with blood glucose targets could be achieved by using smaller steps when adjusting insulin dose and by taking less drastic action for mild hypoglycemia. Because the protocol called for frequent measurements of blood glucose levels, arterial catheters were deemed desirable to reduce patients’ discomfort from frequent finger sticks. In some patients, however, blood glucose levels became so stable that the 2-hour maximum interval between blood glucose measurements could have been extended to 4 hours without loss of control of blood glucose levels.
The protocol of Van den Berghe et al4 permitted up to 4 hours between blood glucose checks for patients with stable blood glucose levels who were receiving insulin infusions; however, the protocol used by Krinsley7 required hourly measurements of blood glucose levels for patients receiving insulin infusions. At least in the early hours after enrollment of patients in our study, the nursing effort associated with intensive glucose control was marked. The few nurses who were assigned 2 patients to care for simultaneously encountered difficulty meeting all the requirements of the study protocol in addition to the usual patient care activities.
| Severe hypoglycemia was rare in both groups. The incidence of moderate hypoglycemia did not differ between the intensive glucose control group and the conventional group, and no hypoglycemic complications were noted.
|
In this pilot study of intensive versus modified conventional control of glucose levels in acutely ill medical patients receiving mechanical ventilation, a nurse-managed protocol for control of blood glucose levels resulted in acceptable compliance with target levels for blood glucose. Severe hypoglycemia was rare in both groups and was not associated with medical complications. Moderate hypoglycemia was uncommon, and its incidence was similar in the 2 groups. Protocol refinements such as smaller steps (0.1 units/h) for adjusting insulin dosage and less drastic responses to mild hypoglycemia may further improve stability of blood glucose levels and reduce the incidence of mild hypoglycemia. The additional nursing effort associated with intensive control of glucose levels is not trivial. This extra effort may be reduced in some patients receiving insulin infusions who have stable blood glucose levels by increasing to 4 hours the interval between measurements of blood glucose levels.
|
ACKNOWLEDGMENTS |
|---|
We received financial support for this study from the Jang Foundation and a Loma Linda University School of Medicine Research Support Grant.
To purchase reprints, contact The InnoVision Group, 101 Columbia, Aliso Viejo, CA 92656. Phone, (800) 809-2273 or (949) 362-2050 (ext 532); fax, (949) 362-2049; e-mail, reprints{at}aacn.org.
|
REFERENCES |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. S. Wiener, D. C. Wiener, and R. J. Larson Benefits and Risks of Tight Glucose Control in Critically Ill Adults: A Meta-analysis JAMA, August 27, 2008; 300(8): 933 - 944. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Aragon Evaluation of nursing work effort and perceptions about blood glucose testing in tight glycemic control. Am. J. Crit. Care., July 1, 2006; 15(4): 370 - 377. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
