Evidence-Based Practice for Nurses

You are asked to join an interdisciplinary committee formed to examine glucose

management in intensive care unit (ICU) patients. You note that the current ICU practice

is to administer a sliding scale subcutaneous insulin dose based on the bedside glucose

obtained at 6-hour time intervals. A retrospective chart review of 50 ICU patients receiving

the sliding scale insulin dosage for 24 hours reveals that glycemic control in ICU patients

is not present. Consequently, the committee develops an evidence-based intravenous (IV)

insulin protocol based on hourly glucose measurements. Glycemic control in ICU patients is

the outcome selected, and you are going to use a blood sugar result of less than 180 mg/dL

within 8 hours of instituting the IV protocol as a measurement indicator. Staff nurses collect

data on 50 patients over a 3-month time period and compare those to the baseline data

shown here:

Mean Blood Glucose

Level Before Change

Mean Blood Glucose

Level After Change

On admission 250 248

8 hours after admit 235 194

12 hours after admit 230 180

24 hours after admit 212 181

Although these reductions were statistically significant (p = .02), the outcome was not met.

What other variables could be accounting for the fact that the outcome was not met? Would

you continue to use the IV insulin protocol?

CRITICAL THINKING EXERCISE 18-3

absence of bleeding defined in a measurable way? Because bleeding might be
interpreted in several different ways, a precise definition of bleeding should
have been provided to ensure consistency in reporting. Second, when should
patients be assessed for absence of bleeding? Is the absence of bleeding to
be assessed when the patient first ambulates or at a later time? Input from
the staff prior to changing the nursing protocol could have clarified these
questions, resulting in more reliable results.

Another consideration in outcome evaluation is to obtain data relative to
current practice for comparison purposes. To document the need for a practice
change and to support a new protocol, baseline data might need to be collected
to demonstrate limitations of the current standard of care. Consideration must
be given to all extraneous variables that may be influencing the outcome, such
as time, equipment, safety, and costs. It could be helpful to involve a statistician
to perform complex analyses.

18.3 Evaluating the Outcomes 481