Greg Maynard, MD, MSc

Patient Population/Setting

This was a prospective study of patients admitted to either the medical/cardiac intensive care unit (MICU/CCU) or surgical intensive care unit (SICU) at an academic medical facility and placed on infusion insulin for >24 hours. The Institutional Review Board (IRB) approved the study for prospective chart review and anonymous results reporting without individual consent.

Patients in the SICU were initiated on infusion insulin after 2 blood glucose readings were above 150 mg/dL, whereas initiation was left to the discretion of the attending physician in the MICU/CCU. A computerized system created in‐house recommends insulin infusion rates based on point‐of‐care (POC) glucose measurements with a target range of 91 mg/dL to 150 mg/dL.

Inclusion/Exclusion Criteria

All patients on continuous insulin infusion admitted to the SICU or the MICU/CCU between May 2008 and September 2008 were evaluated for the study (Figure 1). Patients were excluded from analysis if they were on the infusion for less than 24 hours, had a liver transplant, were discharged within 48 hours of transition, were made comfort care or transitioned to an insulin pump. All other patients were included in the final analysis.

Figure 1

Transition Protocol

Data Collection/Monitoring

A multidisciplinary team consisting of hospitalists, ICU pharmacists, critical care physicians and nursing representatives was assembled during the study period. This team was responsible for protocol implementation, data collection, and surveillance of patient response to the protocol. Educational sessions with house staff and nurses in each unit were held prior to the beginning of the study as well as continued monthly educational efforts during the study. In addition, biweekly huddles to review ongoing patient transitions as well as more formal monthly reviews were held.

The primary objective was to improve glycemic control, defined as the mean daily glucose, during the first 48 hours post transition without a significant increase in the percentage of patients with hypoglycemia (41‐70 mg/dL) or severe hypoglycemia (40 mg/dL). Secondary endpoints included the percent of patients with severe hyperglycemia (300 mg/dL), length of stay (LOS) calculated from the day of transition, number of restarts back onto infusion insulin within 72 hours of transition, and day‐weighted glucose mean up to 12 days following transition for patients with diabetes.

Glucose values were collected and averaged over 6‐hour periods for 48 hours post transition. For patients with diabetes, POC glucose values were collected up to 12 days of hospitalization. Day‐weighted means were obtained by calculating the mean glucose for each hospital day, averaged across all hospital days.12

Analysis

Subjects were divided by the presence or absence of diabetes. Those with diabetes were recommended to receive basal SC insulin during the transition period. Within each group, subjects were further divided by adherence to the protocol. Failure to transition per protocol was defined as: not receiving at least 80% of the recommended basal insulin dose, receiving the initial dose of insulin after the drip was discontinued, or receiving basal insulin when none was recommended.

Descriptive statistics within subgroups comparing age, gender, LOS by analysis of variance for continuous data and by chi‐square for nominal data, were compared. Twenty‐four and 48‐hour post transition mean glucose values and the 12 day weighted‐mean glucose were compared using analysis of variance (Stata ver. 10). All data are expressed as mean standard deviation with a significance value established at P < 0.05.

Patients With Diabetes

(Table 1: Patient Demographics; Table 2: Insulin Use and Glycemic Control; Figure 2: Transition Graph).

Figure 2

A total of 62 individual patients accounted for 72 separate transitions in patients with diabetes based on past medical history or an A1c 6% (n = 14). Of these 72 transitions, 33 (46%) adhered to the protocol while the remaining 39 (54%) transitions varied from the protocol at the treatment team's discretion. Despite similar insulin infusion rates and mean glucose values pretransition, patients with diabetes following the protocol had better glycemic control at both 24 hours and 48 hours after transition than those patients transitioned without the protocol. Day 1 mean blood glucose was 168 mg/dL vs. 211 mg/dL (P = <0.001) and day 2 mean blood glucose was 176 mg/dL vs. 218 mg/dL (P = <0.001) in protocol vs. nonprotocol patients with diabetes respectively (Figure 2).

There was a severe hypoglycemic event (40 mg/dL) in 1 patient with diabetes following the protocol and 1 patient not following the protocol within 48 hours of transition. Both events were secondary to nutritional‐insulin mismatch with emesis after insulin in one case and tube feeds being held in the second case. These findings were consistent with our prior examination of hypoglycemia cases.13 Severe hyperglycemia (glucose 300mg/dL) occurred in 5 (15 %) patients following the protocol vs. 19 (49%) patients not following protocol (P = 0.002.) Patients with diabetes following the protocol received significantly more insulin in the first 24 hours (mean of 65 units vs. 27 units, P 0.001) and 24 to 48 hours after transition (mean of 61 units vs. 28 units, p0.001) than those not following protocol.

An alternate method used at our institution and others14, 15 to calculate TDD is based on the patient's weight and body habitus. When we compared the projected TDD based on weight with the TDD using the transition protocol, we found that the weight based method was much less aggressive. For patients following the protocol, the weight based method projected a mean TDD of 46.3 16.9 units whereas the protocol projected a mean TDD of 65 33.2 units (P = 0.001).

Patients with diabetes following protocol received basal insulin an average of 1.13 hours prior to discontinuing the insulin infusion versus 11.6 hours after for those not following protocol.

Three patients with diabetes following the protocol and 3 patients with diabetes not following the protocol were restarted on infusion insulin within 72 hours of transition.

LOS from final transition to discharge was similar between protocol vs. nonprotocol patients (14.6 vs. 14 days, P = 0.836).

Figure 3 demonstrates that when used correctly, the protocol provides an extended period of glycemic control up to 12 days post transition. Patients transitioned per protocol had a day‐weighted mean glucose of 155 mg/dL vs. 184 mg/dL (P = 0.043) in patients not following protocol. There was only 1 glucose value less than 40 mg/dL between days 2 to 12 in the protocol group.

Figure 3

Patients Without Diabetes

Of the 39 individual patients without diabetes there were 42 transition events, 33 transitions (78.6%) were per protocol and placed on correctional insulin only. The remaining 9 transitions failed to follow protocol in that basal insulin was prescribed, but these patients maintained comparable glycemic control without an increase in hypoglycemic events. Following transition, patients without diabetes on protocol maintained a mean glucose of 150 mg/dL in the first 24 hours and 153 mg/dL in 24 to 48 hours post transition. They required a mean daily correctional insulin dose of 3.2 units on Day 1 and 2.8 units on Day 2 despite having an average drip rate of 2.3 units/hour at the time of transition (Table 2). There were no severe hypoglycemic events and 80% of blood sugars were within the goal range of 80 mg/dL to 180 mg/dL. Only 1 patient had a single blood glucose of >300mg/dL. No patient was restarted on infusion insulin once transitioned.

Patients without diabetes had a longer LOS after transition off of infusion insulin when compared to their diabetic counterparts (22 vs. 14 days).

Patient Population/Setting

This was a prospective study of patients admitted to either the medical/cardiac intensive care unit (MICU/CCU) or surgical intensive care unit (SICU) at an academic medical facility and placed on infusion insulin for >24 hours. The Institutional Review Board (IRB) approved the study for prospective chart review and anonymous results reporting without individual consent.

Patients in the SICU were initiated on infusion insulin after 2 blood glucose readings were above 150 mg/dL, whereas initiation was left to the discretion of the attending physician in the MICU/CCU. A computerized system created in‐house recommends insulin infusion rates based on point‐of‐care (POC) glucose measurements with a target range of 91 mg/dL to 150 mg/dL.

Inclusion/Exclusion Criteria

All patients on continuous insulin infusion admitted to the SICU or the MICU/CCU between May 2008 and September 2008 were evaluated for the study (Figure 1). Patients were excluded from analysis if they were on the infusion for less than 24 hours, had a liver transplant, were discharged within 48 hours of transition, were made comfort care or transitioned to an insulin pump. All other patients were included in the final analysis.

Figure 1

Transition Protocol

Data Collection/Monitoring

A multidisciplinary team consisting of hospitalists, ICU pharmacists, critical care physicians and nursing representatives was assembled during the study period. This team was responsible for protocol implementation, data collection, and surveillance of patient response to the protocol. Educational sessions with house staff and nurses in each unit were held prior to the beginning of the study as well as continued monthly educational efforts during the study. In addition, biweekly huddles to review ongoing patient transitions as well as more formal monthly reviews were held.

The primary objective was to improve glycemic control, defined as the mean daily glucose, during the first 48 hours post transition without a significant increase in the percentage of patients with hypoglycemia (41‐70 mg/dL) or severe hypoglycemia (40 mg/dL). Secondary endpoints included the percent of patients with severe hyperglycemia (300 mg/dL), length of stay (LOS) calculated from the day of transition, number of restarts back onto infusion insulin within 72 hours of transition, and day‐weighted glucose mean up to 12 days following transition for patients with diabetes.

Glucose values were collected and averaged over 6‐hour periods for 48 hours post transition. For patients with diabetes, POC glucose values were collected up to 12 days of hospitalization. Day‐weighted means were obtained by calculating the mean glucose for each hospital day, averaged across all hospital days.12

Analysis

Subjects were divided by the presence or absence of diabetes. Those with diabetes were recommended to receive basal SC insulin during the transition period. Within each group, subjects were further divided by adherence to the protocol. Failure to transition per protocol was defined as: not receiving at least 80% of the recommended basal insulin dose, receiving the initial dose of insulin after the drip was discontinued, or receiving basal insulin when none was recommended.

Descriptive statistics within subgroups comparing age, gender, LOS by analysis of variance for continuous data and by chi‐square for nominal data, were compared. Twenty‐four and 48‐hour post transition mean glucose values and the 12 day weighted‐mean glucose were compared using analysis of variance (Stata ver. 10). All data are expressed as mean standard deviation with a significance value established at P < 0.05.

Patients With Diabetes

(Table 1: Patient Demographics; Table 2: Insulin Use and Glycemic Control; Figure 2: Transition Graph).

Figure 2

A total of 62 individual patients accounted for 72 separate transitions in patients with diabetes based on past medical history or an A1c 6% (n = 14). Of these 72 transitions, 33 (46%) adhered to the protocol while the remaining 39 (54%) transitions varied from the protocol at the treatment team's discretion. Despite similar insulin infusion rates and mean glucose values pretransition, patients with diabetes following the protocol had better glycemic control at both 24 hours and 48 hours after transition than those patients transitioned without the protocol. Day 1 mean blood glucose was 168 mg/dL vs. 211 mg/dL (P = <0.001) and day 2 mean blood glucose was 176 mg/dL vs. 218 mg/dL (P = <0.001) in protocol vs. nonprotocol patients with diabetes respectively (Figure 2).

There was a severe hypoglycemic event (40 mg/dL) in 1 patient with diabetes following the protocol and 1 patient not following the protocol within 48 hours of transition. Both events were secondary to nutritional‐insulin mismatch with emesis after insulin in one case and tube feeds being held in the second case. These findings were consistent with our prior examination of hypoglycemia cases.13 Severe hyperglycemia (glucose 300mg/dL) occurred in 5 (15 %) patients following the protocol vs. 19 (49%) patients not following protocol (P = 0.002.) Patients with diabetes following the protocol received significantly more insulin in the first 24 hours (mean of 65 units vs. 27 units, P 0.001) and 24 to 48 hours after transition (mean of 61 units vs. 28 units, p0.001) than those not following protocol.

An alternate method used at our institution and others14, 15 to calculate TDD is based on the patient's weight and body habitus. When we compared the projected TDD based on weight with the TDD using the transition protocol, we found that the weight based method was much less aggressive. For patients following the protocol, the weight based method projected a mean TDD of 46.3 16.9 units whereas the protocol projected a mean TDD of 65 33.2 units (P = 0.001).

Patients with diabetes following protocol received basal insulin an average of 1.13 hours prior to discontinuing the insulin infusion versus 11.6 hours after for those not following protocol.

Three patients with diabetes following the protocol and 3 patients with diabetes not following the protocol were restarted on infusion insulin within 72 hours of transition.

LOS from final transition to discharge was similar between protocol vs. nonprotocol patients (14.6 vs. 14 days, P = 0.836).

Figure 3 demonstrates that when used correctly, the protocol provides an extended period of glycemic control up to 12 days post transition. Patients transitioned per protocol had a day‐weighted mean glucose of 155 mg/dL vs. 184 mg/dL (P = 0.043) in patients not following protocol. There was only 1 glucose value less than 40 mg/dL between days 2 to 12 in the protocol group.

Figure 3

Patients Without Diabetes

Of the 39 individual patients without diabetes there were 42 transition events, 33 transitions (78.6%) were per protocol and placed on correctional insulin only. The remaining 9 transitions failed to follow protocol in that basal insulin was prescribed, but these patients maintained comparable glycemic control without an increase in hypoglycemic events. Following transition, patients without diabetes on protocol maintained a mean glucose of 150 mg/dL in the first 24 hours and 153 mg/dL in 24 to 48 hours post transition. They required a mean daily correctional insulin dose of 3.2 units on Day 1 and 2.8 units on Day 2 despite having an average drip rate of 2.3 units/hour at the time of transition (Table 2). There were no severe hypoglycemic events and 80% of blood sugars were within the goal range of 80 mg/dL to 180 mg/dL. Only 1 patient had a single blood glucose of >300mg/dL. No patient was restarted on infusion insulin once transitioned.

Patients without diabetes had a longer LOS after transition off of infusion insulin when compared to their diabetic counterparts (22 vs. 14 days).

Patient Population/Setting

This was a prospective study of patients admitted to either the medical/cardiac intensive care unit (MICU/CCU) or surgical intensive care unit (SICU) at an academic medical facility and placed on infusion insulin for >24 hours. The Institutional Review Board (IRB) approved the study for prospective chart review and anonymous results reporting without individual consent.

Patients in the SICU were initiated on infusion insulin after 2 blood glucose readings were above 150 mg/dL, whereas initiation was left to the discretion of the attending physician in the MICU/CCU. A computerized system created in‐house recommends insulin infusion rates based on point‐of‐care (POC) glucose measurements with a target range of 91 mg/dL to 150 mg/dL.

Inclusion/Exclusion Criteria

All patients on continuous insulin infusion admitted to the SICU or the MICU/CCU between May 2008 and September 2008 were evaluated for the study (Figure 1). Patients were excluded from analysis if they were on the infusion for less than 24 hours, had a liver transplant, were discharged within 48 hours of transition, were made comfort care or transitioned to an insulin pump. All other patients were included in the final analysis.

Figure 1

Transition Protocol

Data Collection/Monitoring

A multidisciplinary team consisting of hospitalists, ICU pharmacists, critical care physicians and nursing representatives was assembled during the study period. This team was responsible for protocol implementation, data collection, and surveillance of patient response to the protocol. Educational sessions with house staff and nurses in each unit were held prior to the beginning of the study as well as continued monthly educational efforts during the study. In addition, biweekly huddles to review ongoing patient transitions as well as more formal monthly reviews were held.

The primary objective was to improve glycemic control, defined as the mean daily glucose, during the first 48 hours post transition without a significant increase in the percentage of patients with hypoglycemia (41‐70 mg/dL) or severe hypoglycemia (40 mg/dL). Secondary endpoints included the percent of patients with severe hyperglycemia (300 mg/dL), length of stay (LOS) calculated from the day of transition, number of restarts back onto infusion insulin within 72 hours of transition, and day‐weighted glucose mean up to 12 days following transition for patients with diabetes.

Glucose values were collected and averaged over 6‐hour periods for 48 hours post transition. For patients with diabetes, POC glucose values were collected up to 12 days of hospitalization. Day‐weighted means were obtained by calculating the mean glucose for each hospital day, averaged across all hospital days.12

Analysis

Subjects were divided by the presence or absence of diabetes. Those with diabetes were recommended to receive basal SC insulin during the transition period. Within each group, subjects were further divided by adherence to the protocol. Failure to transition per protocol was defined as: not receiving at least 80% of the recommended basal insulin dose, receiving the initial dose of insulin after the drip was discontinued, or receiving basal insulin when none was recommended.

Descriptive statistics within subgroups comparing age, gender, LOS by analysis of variance for continuous data and by chi‐square for nominal data, were compared. Twenty‐four and 48‐hour post transition mean glucose values and the 12 day weighted‐mean glucose were compared using analysis of variance (Stata ver. 10). All data are expressed as mean standard deviation with a significance value established at P < 0.05.

Patients With Diabetes

(Table 1: Patient Demographics; Table 2: Insulin Use and Glycemic Control; Figure 2: Transition Graph).

Figure 2

A total of 62 individual patients accounted for 72 separate transitions in patients with diabetes based on past medical history or an A1c 6% (n = 14). Of these 72 transitions, 33 (46%) adhered to the protocol while the remaining 39 (54%) transitions varied from the protocol at the treatment team's discretion. Despite similar insulin infusion rates and mean glucose values pretransition, patients with diabetes following the protocol had better glycemic control at both 24 hours and 48 hours after transition than those patients transitioned without the protocol. Day 1 mean blood glucose was 168 mg/dL vs. 211 mg/dL (P = <0.001) and day 2 mean blood glucose was 176 mg/dL vs. 218 mg/dL (P = <0.001) in protocol vs. nonprotocol patients with diabetes respectively (Figure 2).

There was a severe hypoglycemic event (40 mg/dL) in 1 patient with diabetes following the protocol and 1 patient not following the protocol within 48 hours of transition. Both events were secondary to nutritional‐insulin mismatch with emesis after insulin in one case and tube feeds being held in the second case. These findings were consistent with our prior examination of hypoglycemia cases.13 Severe hyperglycemia (glucose 300mg/dL) occurred in 5 (15 %) patients following the protocol vs. 19 (49%) patients not following protocol (P = 0.002.) Patients with diabetes following the protocol received significantly more insulin in the first 24 hours (mean of 65 units vs. 27 units, P 0.001) and 24 to 48 hours after transition (mean of 61 units vs. 28 units, p0.001) than those not following protocol.

An alternate method used at our institution and others14, 15 to calculate TDD is based on the patient's weight and body habitus. When we compared the projected TDD based on weight with the TDD using the transition protocol, we found that the weight based method was much less aggressive. For patients following the protocol, the weight based method projected a mean TDD of 46.3 16.9 units whereas the protocol projected a mean TDD of 65 33.2 units (P = 0.001).

Patients with diabetes following protocol received basal insulin an average of 1.13 hours prior to discontinuing the insulin infusion versus 11.6 hours after for those not following protocol.

Three patients with diabetes following the protocol and 3 patients with diabetes not following the protocol were restarted on infusion insulin within 72 hours of transition.

LOS from final transition to discharge was similar between protocol vs. nonprotocol patients (14.6 vs. 14 days, P = 0.836).

Figure 3 demonstrates that when used correctly, the protocol provides an extended period of glycemic control up to 12 days post transition. Patients transitioned per protocol had a day‐weighted mean glucose of 155 mg/dL vs. 184 mg/dL (P = 0.043) in patients not following protocol. There was only 1 glucose value less than 40 mg/dL between days 2 to 12 in the protocol group.

Figure 3

Patients Without Diabetes

Of the 39 individual patients without diabetes there were 42 transition events, 33 transitions (78.6%) were per protocol and placed on correctional insulin only. The remaining 9 transitions failed to follow protocol in that basal insulin was prescribed, but these patients maintained comparable glycemic control without an increase in hypoglycemic events. Following transition, patients without diabetes on protocol maintained a mean glucose of 150 mg/dL in the first 24 hours and 153 mg/dL in 24 to 48 hours post transition. They required a mean daily correctional insulin dose of 3.2 units on Day 1 and 2.8 units on Day 2 despite having an average drip rate of 2.3 units/hour at the time of transition (Table 2). There were no severe hypoglycemic events and 80% of blood sugars were within the goal range of 80 mg/dL to 180 mg/dL. Only 1 patient had a single blood glucose of >300mg/dL. No patient was restarted on infusion insulin once transitioned.

Patients without diabetes had a longer LOS after transition off of infusion insulin when compared to their diabetic counterparts (22 vs. 14 days).

Interventions

We introduced several interventions and educational efforts throughout the course of our improvement. The 2 key interventions were as follows:

• Structured subcutaneous insulin order sets (November, 2003).

• An insulin management algorithm, described below (May 2005).

Key Intervention #1: Structured Subcutaneous Insulin Order Set Implementation

In November 2003, we introduced a paper‐based structured subcutaneous insulin order set. This order set encouraged the use of scheduled basal and nutritional insulin, provided guidance for monitoring glucose levels, and for insulin dosing. A hypoglycemia protocol and a standardized correction insulin table were embedded in the order set. This set was similar to examples of structured insulin ordering subsequently presented in the literature.9 In a parallel effort, the University of California, San Diego Medical Center (UCSDMC) was developing a computer physician order entry (CPOE) module for our comprehensive clinical information system, Invision (Siemens Medical Systems, Malvern, PA), that heretofore had primarily focused on result review, patient schedule management, and nursing documentation. In anticipation of CPOE and for the purpose of standardization, we removed outdated sliding scale insulin regimens from a variety preexisting order sets and inserted references to the standardized subcutaneous insulin order set in their stead. The medication administration record (MAR) was changed to reflect the basal/nutritional/correction insulin terminology. It became more difficult to order a stand‐alone insulin sliding scale even before CPOE versions became available. The standardized order set was the only preprinted correction scale insulin order available, and ordering physicians have to specifically opt out of basal and nutritional insulin choices to order sliding scale only regimens. Verbal orders for correction dose scales were deemed unacceptable by medical staff committees. Correctional insulin doses could be ordered as a 1‐time order, but the pharmacy rejected ongoing insulin orders that were not entered on the structured form.

We introduced our first standardized CPOE subcutaneous insulin order set in January 2004 at the smaller of our 2 campuses, and subsequently completed full deployment across both campuses in all adult medical‐surgical care areas by September 2004.

The CPOE version, like the paper version that immediately preceded it, encouraged the use of basal/bolus insulin regimens, promoted the terms basal, nutritional or premeal, and adjustment dose insulin in the order sets and the medication administration record, and was mandatory for providers wishing to order anything but a 1‐time order of insulin. Figure 1 depicts a screen shot of the CPOE version. Similar to the paper version, the ordering physician had to specifically opt out of ordering scheduled premeal and basal insulin to order a sliding scale only regimen. The first screen also ensured that appropriate POC glucose monitoring was ordered and endorsed a standing hypoglycemia protocol order. The CPOE version had only a few additional features not possible on paper. Obvious benefits included elimination of unapproved abbreviations and handwriting errors. Nutritional and correction insulin types were forced to be identical. Fundamentally, however, both the paper and online structured ordering experiences had the same degree of control over provider ordering patterns, and there was no increment in guidance for choosing insulin regimens, hence their combined analysis as structured orders.

Figure 1

Key Intervention #2: Insulin Management Algorithm

The structured insulin order set had many advantages, but also had many limitations. Guidance for preferred insulin regimens for patients in different nutritional situations was not inherent in the order set, and all basal and nutritional insulin options were offered as equally acceptable choices. The order set gave very general guidance for insulin dosing, but did not calculate insulin doses or assist in the apportionment of insulin between basal and nutritional components, and guidance for setting a glycemic target or adjusting insulin was lacking.

Recognizing these limitations, we devised an insulin management algorithm to provide guidance incremental to that offered in the order set. In April 2005, 3 hospitalists piloted a paper‐based insulin management algorithm (Figure 2, front; Figure 3, reverse) on their teaching services. This 1‐page algorithm provided guidance on insulin dosing and monitoring, and provided institutionally preferred insulin regimens for patients in different nutritional situations. As an example, of the several acceptable subcutaneous insulin regimens that an eating patient might use in the inpatient setting, we advocated the use of 1 preferred regimen (a relatively peakless, long‐acting basal insulin once a day, along with a rapid acting analog nutritional insulin with each meal). We introduced the concept of a ward glycemic target, provided prompts for diabetes education, and generally recommended discontinuation of oral hypoglycemic agents in the inpatient setting. The hospitalists were introduced to the concepts and the algorithm via 1 of the authors (G.M.) in a 1‐hour session. The algorithm was introduced on each teaching team during routine teaching rounds with a slide set (approximately 15 slides) that outlined the basic principles of insulin dosing, and gave example cases which modeled the proper use of the algorithm. The principles were reinforced on daily patient work rounds as they were applied on inpatients with hyperglycemia. The pilot results on 25 patients, compared to 250 historical control patients, were very promising, with markedly improved glycemic control and no increase in hypoglycemia. We therefore sought to spread the use of the algorithm. In May 2005 the insulin management algorithm and teaching slide set were promoted on all 7 hospitalist‐run services, and the results of the pilot and concepts of the algorithm were shared with a variety of house staff and service leaders in approximately a dozen sessions: educational grand rounds, assorted noon lectures, and subsequently, at new intern orientations. Easy access to the algorithm was assured by providing a link to the file within the CPOE insulin order set.

Figure 2

Figure 3

Measures and Analyses

Insulin Use Patterns

Figure 4 demonstrates the dramatic improvement that took place with the introduction of the structured order set. In the 6 months preceding the introduction of the structured insulin order set (May‐October 2003) 72% of 477 sampled patients with insulin orders were on sliding scale‐only insulin regimens (with no basal insulin), compared to just 26% of 499 patients sampled in the March to August 2004 time period subsequent to order set implementation (P < .0001, chi square statistic). Intermittent monthly checks on insulin use patterns reveal this change has been sustained.

Figure 4

Glycemic Control

A total of 9314 patients with 44,232 monitored patient‐days and over 120,000 POC glucose values were analyzed to assess glycemic control, which was improved with structured insulin orders and improved incrementally with the introduction of the insulin management algorithm.

The percent of patient‐days that were uncontrolled, defined as a monitored day with a mean glucose of 180 mg/dL, was reduced over the 3 time periods (37.8% versus 33.9% versus 30.1%, P < 0.005, Pearson chi square statistic), representing a 21% RR reduction of uncontrolled patient‐days from TP1 versus TP3. Table 2 shows the summary results for glycemic control, including the RR and CIs between the 3 time periods.

In a similar fashion, the percent of patients with uncontrolled patient‐stays (day‐weighted mean glucose 180 mg/dL) was also reduced over the 3 time periods (41.5% versus 37.6% versus 34.2%, P < 0.05, Pearson chi square statistic, with an RR reduction of 16% for TP3:TP1). Figure 5 depicts a statistical process control chart of the percent of patients experiencing uncontrolled patient‐stays over time, and is more effective in displaying the temporal relationship of the interventions with the improved results.

Figure 5

Uncontrolled hyperglycemic days and stays were reduced incrementally from TP3 versus TP2, reflecting the added benefit of the insulin management algorithm, compared to the benefit enjoyed with the structured order set alone.

When the analyses were repeated after excluding patients with fewer than 8 POC glucose readings (Table 3), the findings were similar, but as predicted, the effect was slightly more pronounced, with a 23% relative reduction in uncontrolled patient‐days and a 27% reduction in uncontrolled patient‐stays of TP3 versus TP1.

Hypoglycemia

Table 4 summarizes the results for hypoglycemia and severe hypoglycemia in the study population, and Table 5 summarizes the secondary analyses of hypoglycemia in the subset with at least 8 POC glucose readings.

Strengths and Limitations

Our study has several strengths. The study is large and incorporates all patients with diabetes or hyperglycemia captured by POC glucose testing, and the observation period is long enough that bias from merely being observed is not a factor. We used metrics for glycemic control, hypoglycemia, and insulin use patterns that are of high quality and are generally in line with the Society of Hospital Medicine (SHM) Glycemic Control Task force recommendations,12, 13 and examined data by both patient‐stay and patient‐day.

The increased use of anticipatory physiologic subcutaneous insulin regimens, and the subsequent decline in the use of sliding scale insulin, is the most likely mechanism for improvement. The improvements seen are fairly dramatic for an institution in absolute terms, because inpatient hyperglycemia and hypoglycemia are so common. For example, on an annualized basis for our 400‐bed medical center, these interventions prevent 124 patients from experiencing 208 hypoglycemic days.

Other institutions should be able to replicate our results. We received administrative support to create a multidisciplinary steering committee, but we did not have incremental resources to create a dedicated team for insulin management, mandated endocrinology comanagement or consultations, or manual data collection. In fact, we had only 1 diabetes educator for 400 adult beds at 2 sites, and were relatively underresourced in this area by community standards. There was some time and expense in creating the glycemic control reports, but all of the glucose data collected were part of normal care, and the data retrieval became automated.

The main limitation of this study lies in the observational study design. There were multiple interventions in addition to structured insulin orders and the insulin management algorithm, and these educational and organizational changes undoubtedly also contributed to the overall success of our program. Since we did not perform a randomized controlled trial, the reader might reasonably question if the structured order sets and insulin management algorithm were actually the cause of the improvement seen, as opposed to these ancillary efforts or secular change. However, there are several factors that make this unlikely. First, the study population was well‐defined, having diabetes or documented hyperglycemia in all 3 time periods. Second, the demographics remained constant or actually worked against improvement trends, since the markers of patient acuity suggest increased patient acuity over the observation period. Third, the temporal relationship of the improvement to the introduction of our key interventions, as viewed on statistical process control charts shown in Figure 5, strongly suggest a causal relationship. This temporal relationship was consistently observed no matter how we chose to define uncontrolled hyperglycemia, and was also seen on hypoglycemia control charts. We view the ancillary interventions (such as educational efforts) as necessary, but not sufficient, in and of themselves, to effect major improvement.

We did not analyze the impact of the improved glycemic control on patient outcomes. In the absence of a randomized controlled trial design, controlling for the various confounders is a challenging task. Also, it is likely that not all hypoglycemic events were attributable to inpatient glycemic control regimens, though the secondary analysis probably eliminated many hypoglycemia admissions.

Lessons Learned: Implications from our study

We agree with the American Association of Clinical Endocrinologists (AACE)/American Diabetes Association (ADA)2 and the SHM Glycemic Control Task Force12 about the essential elements needed for successful implementation of inpatient glycemic control programs:

• An appropriate level of administrative support.

• Formation of a multidisciplinary steering committee to drive the development of initiatives, empowered to enact changes.

• Assessment of current processes, quality of care, and barriers to practice change.

• Development and implementation of interventions, including standardized order sets, protocols, policies, and algorithms with associated educational programs.

• Metrics for evaluation of glycemic control, hypoglycemia, insulin use patterns, and other aspects of care.

Metrics to follow hypoglycemia are extremely important. The voluntary reporting on insulin‐induced hypoglycemia fluctuated widely over the course of our project. These fluctuations did not correlate well with the more objective and accurate measures we followed, and this objective data was very helpful in reducing the fear of hypoglycemia, and spreading the wider use of basal bolus insulin regimens. We strongly recommend that improvement teams formulate and follow measures of glycemic control, hypoglycemia, and insulin use, similar to those outlined in the SHM Glycemic Control Improvement Guide12 and the SHM Glycemic Control Task Force summary on glucometrics.13

Although we introduced our structured insulin order set first, with a long lag time until we introduced the insulin management algorithm, we advocate a different approach for institutions grappling with these issues. This approach is well‐described by the SHM Glycemic Control Task Force.14 An insulin management algorithm should be crafted first, integrating guidance for insulin dosing, preferred insulin regimens for different nutritional situations, a glycemic target, insulin dosing adjustment, glucose monitoring, and prompts for ordering a glycosylated hemoglobin (A1c) level. Next, the order set and the supporting educational programs should integrate this guidance as much as possible, making the key guidance available at the point of patient care.

This guidance was available in our algorithm but was not inherent in the structured insulin orders described in this report, and all basal and nutritional insulin options were offered as equally acceptable choices. This version did not calculate insulin doses or assist in the apportionment of insulin between basal and nutritional components. Only a single adjustment dose scale was offered, leaving appropriate modifications up to the end user, and from a usability standpoint, our CPOE insulin orders lacked dynamic flexibility (revising a single insulin required discontinuing all prior orders and reentering all orders). These limitations have subsequently been addressed with Version 2 of our CPOE insulin orders, and the details will soon be available in the literature.15

We are now exploring further improvement with concurrent identification and intervention of hyperglycemic patients that are not on physiologic insulin regimens or not meeting glycemic targets, and implementing protocols addressing the transition from infusion insulin.

Interventions

We introduced several interventions and educational efforts throughout the course of our improvement. The 2 key interventions were as follows:

• Structured subcutaneous insulin order sets (November, 2003).

• An insulin management algorithm, described below (May 2005).

Key Intervention #1: Structured Subcutaneous Insulin Order Set Implementation

In November 2003, we introduced a paper‐based structured subcutaneous insulin order set. This order set encouraged the use of scheduled basal and nutritional insulin, provided guidance for monitoring glucose levels, and for insulin dosing. A hypoglycemia protocol and a standardized correction insulin table were embedded in the order set. This set was similar to examples of structured insulin ordering subsequently presented in the literature.9 In a parallel effort, the University of California, San Diego Medical Center (UCSDMC) was developing a computer physician order entry (CPOE) module for our comprehensive clinical information system, Invision (Siemens Medical Systems, Malvern, PA), that heretofore had primarily focused on result review, patient schedule management, and nursing documentation. In anticipation of CPOE and for the purpose of standardization, we removed outdated sliding scale insulin regimens from a variety preexisting order sets and inserted references to the standardized subcutaneous insulin order set in their stead. The medication administration record (MAR) was changed to reflect the basal/nutritional/correction insulin terminology. It became more difficult to order a stand‐alone insulin sliding scale even before CPOE versions became available. The standardized order set was the only preprinted correction scale insulin order available, and ordering physicians have to specifically opt out of basal and nutritional insulin choices to order sliding scale only regimens. Verbal orders for correction dose scales were deemed unacceptable by medical staff committees. Correctional insulin doses could be ordered as a 1‐time order, but the pharmacy rejected ongoing insulin orders that were not entered on the structured form.

We introduced our first standardized CPOE subcutaneous insulin order set in January 2004 at the smaller of our 2 campuses, and subsequently completed full deployment across both campuses in all adult medical‐surgical care areas by September 2004.

The CPOE version, like the paper version that immediately preceded it, encouraged the use of basal/bolus insulin regimens, promoted the terms basal, nutritional or premeal, and adjustment dose insulin in the order sets and the medication administration record, and was mandatory for providers wishing to order anything but a 1‐time order of insulin. Figure 1 depicts a screen shot of the CPOE version. Similar to the paper version, the ordering physician had to specifically opt out of ordering scheduled premeal and basal insulin to order a sliding scale only regimen. The first screen also ensured that appropriate POC glucose monitoring was ordered and endorsed a standing hypoglycemia protocol order. The CPOE version had only a few additional features not possible on paper. Obvious benefits included elimination of unapproved abbreviations and handwriting errors. Nutritional and correction insulin types were forced to be identical. Fundamentally, however, both the paper and online structured ordering experiences had the same degree of control over provider ordering patterns, and there was no increment in guidance for choosing insulin regimens, hence their combined analysis as structured orders.

Figure 1

Key Intervention #2: Insulin Management Algorithm

The structured insulin order set had many advantages, but also had many limitations. Guidance for preferred insulin regimens for patients in different nutritional situations was not inherent in the order set, and all basal and nutritional insulin options were offered as equally acceptable choices. The order set gave very general guidance for insulin dosing, but did not calculate insulin doses or assist in the apportionment of insulin between basal and nutritional components, and guidance for setting a glycemic target or adjusting insulin was lacking.

Recognizing these limitations, we devised an insulin management algorithm to provide guidance incremental to that offered in the order set. In April 2005, 3 hospitalists piloted a paper‐based insulin management algorithm (Figure 2, front; Figure 3, reverse) on their teaching services. This 1‐page algorithm provided guidance on insulin dosing and monitoring, and provided institutionally preferred insulin regimens for patients in different nutritional situations. As an example, of the several acceptable subcutaneous insulin regimens that an eating patient might use in the inpatient setting, we advocated the use of 1 preferred regimen (a relatively peakless, long‐acting basal insulin once a day, along with a rapid acting analog nutritional insulin with each meal). We introduced the concept of a ward glycemic target, provided prompts for diabetes education, and generally recommended discontinuation of oral hypoglycemic agents in the inpatient setting. The hospitalists were introduced to the concepts and the algorithm via 1 of the authors (G.M.) in a 1‐hour session. The algorithm was introduced on each teaching team during routine teaching rounds with a slide set (approximately 15 slides) that outlined the basic principles of insulin dosing, and gave example cases which modeled the proper use of the algorithm. The principles were reinforced on daily patient work rounds as they were applied on inpatients with hyperglycemia. The pilot results on 25 patients, compared to 250 historical control patients, were very promising, with markedly improved glycemic control and no increase in hypoglycemia. We therefore sought to spread the use of the algorithm. In May 2005 the insulin management algorithm and teaching slide set were promoted on all 7 hospitalist‐run services, and the results of the pilot and concepts of the algorithm were shared with a variety of house staff and service leaders in approximately a dozen sessions: educational grand rounds, assorted noon lectures, and subsequently, at new intern orientations. Easy access to the algorithm was assured by providing a link to the file within the CPOE insulin order set.

Figure 2

Figure 3

Measures and Analyses

Insulin Use Patterns

Figure 4 demonstrates the dramatic improvement that took place with the introduction of the structured order set. In the 6 months preceding the introduction of the structured insulin order set (May‐October 2003) 72% of 477 sampled patients with insulin orders were on sliding scale‐only insulin regimens (with no basal insulin), compared to just 26% of 499 patients sampled in the March to August 2004 time period subsequent to order set implementation (P < .0001, chi square statistic). Intermittent monthly checks on insulin use patterns reveal this change has been sustained.

Figure 4

Glycemic Control

A total of 9314 patients with 44,232 monitored patient‐days and over 120,000 POC glucose values were analyzed to assess glycemic control, which was improved with structured insulin orders and improved incrementally with the introduction of the insulin management algorithm.

The percent of patient‐days that were uncontrolled, defined as a monitored day with a mean glucose of 180 mg/dL, was reduced over the 3 time periods (37.8% versus 33.9% versus 30.1%, P < 0.005, Pearson chi square statistic), representing a 21% RR reduction of uncontrolled patient‐days from TP1 versus TP3. Table 2 shows the summary results for glycemic control, including the RR and CIs between the 3 time periods.

In a similar fashion, the percent of patients with uncontrolled patient‐stays (day‐weighted mean glucose 180 mg/dL) was also reduced over the 3 time periods (41.5% versus 37.6% versus 34.2%, P < 0.05, Pearson chi square statistic, with an RR reduction of 16% for TP3:TP1). Figure 5 depicts a statistical process control chart of the percent of patients experiencing uncontrolled patient‐stays over time, and is more effective in displaying the temporal relationship of the interventions with the improved results.

Figure 5

Uncontrolled hyperglycemic days and stays were reduced incrementally from TP3 versus TP2, reflecting the added benefit of the insulin management algorithm, compared to the benefit enjoyed with the structured order set alone.

When the analyses were repeated after excluding patients with fewer than 8 POC glucose readings (Table 3), the findings were similar, but as predicted, the effect was slightly more pronounced, with a 23% relative reduction in uncontrolled patient‐days and a 27% reduction in uncontrolled patient‐stays of TP3 versus TP1.

Hypoglycemia

Table 4 summarizes the results for hypoglycemia and severe hypoglycemia in the study population, and Table 5 summarizes the secondary analyses of hypoglycemia in the subset with at least 8 POC glucose readings.

Strengths and Limitations

Our study has several strengths. The study is large and incorporates all patients with diabetes or hyperglycemia captured by POC glucose testing, and the observation period is long enough that bias from merely being observed is not a factor. We used metrics for glycemic control, hypoglycemia, and insulin use patterns that are of high quality and are generally in line with the Society of Hospital Medicine (SHM) Glycemic Control Task force recommendations,12, 13 and examined data by both patient‐stay and patient‐day.

The increased use of anticipatory physiologic subcutaneous insulin regimens, and the subsequent decline in the use of sliding scale insulin, is the most likely mechanism for improvement. The improvements seen are fairly dramatic for an institution in absolute terms, because inpatient hyperglycemia and hypoglycemia are so common. For example, on an annualized basis for our 400‐bed medical center, these interventions prevent 124 patients from experiencing 208 hypoglycemic days.

Other institutions should be able to replicate our results. We received administrative support to create a multidisciplinary steering committee, but we did not have incremental resources to create a dedicated team for insulin management, mandated endocrinology comanagement or consultations, or manual data collection. In fact, we had only 1 diabetes educator for 400 adult beds at 2 sites, and were relatively underresourced in this area by community standards. There was some time and expense in creating the glycemic control reports, but all of the glucose data collected were part of normal care, and the data retrieval became automated.

The main limitation of this study lies in the observational study design. There were multiple interventions in addition to structured insulin orders and the insulin management algorithm, and these educational and organizational changes undoubtedly also contributed to the overall success of our program. Since we did not perform a randomized controlled trial, the reader might reasonably question if the structured order sets and insulin management algorithm were actually the cause of the improvement seen, as opposed to these ancillary efforts or secular change. However, there are several factors that make this unlikely. First, the study population was well‐defined, having diabetes or documented hyperglycemia in all 3 time periods. Second, the demographics remained constant or actually worked against improvement trends, since the markers of patient acuity suggest increased patient acuity over the observation period. Third, the temporal relationship of the improvement to the introduction of our key interventions, as viewed on statistical process control charts shown in Figure 5, strongly suggest a causal relationship. This temporal relationship was consistently observed no matter how we chose to define uncontrolled hyperglycemia, and was also seen on hypoglycemia control charts. We view the ancillary interventions (such as educational efforts) as necessary, but not sufficient, in and of themselves, to effect major improvement.

We did not analyze the impact of the improved glycemic control on patient outcomes. In the absence of a randomized controlled trial design, controlling for the various confounders is a challenging task. Also, it is likely that not all hypoglycemic events were attributable to inpatient glycemic control regimens, though the secondary analysis probably eliminated many hypoglycemia admissions.

Lessons Learned: Implications from our study

We agree with the American Association of Clinical Endocrinologists (AACE)/American Diabetes Association (ADA)2 and the SHM Glycemic Control Task Force12 about the essential elements needed for successful implementation of inpatient glycemic control programs:

• An appropriate level of administrative support.

• Formation of a multidisciplinary steering committee to drive the development of initiatives, empowered to enact changes.

• Assessment of current processes, quality of care, and barriers to practice change.

• Development and implementation of interventions, including standardized order sets, protocols, policies, and algorithms with associated educational programs.

• Metrics for evaluation of glycemic control, hypoglycemia, insulin use patterns, and other aspects of care.

Metrics to follow hypoglycemia are extremely important. The voluntary reporting on insulin‐induced hypoglycemia fluctuated widely over the course of our project. These fluctuations did not correlate well with the more objective and accurate measures we followed, and this objective data was very helpful in reducing the fear of hypoglycemia, and spreading the wider use of basal bolus insulin regimens. We strongly recommend that improvement teams formulate and follow measures of glycemic control, hypoglycemia, and insulin use, similar to those outlined in the SHM Glycemic Control Improvement Guide12 and the SHM Glycemic Control Task Force summary on glucometrics.13

Although we introduced our structured insulin order set first, with a long lag time until we introduced the insulin management algorithm, we advocate a different approach for institutions grappling with these issues. This approach is well‐described by the SHM Glycemic Control Task Force.14 An insulin management algorithm should be crafted first, integrating guidance for insulin dosing, preferred insulin regimens for different nutritional situations, a glycemic target, insulin dosing adjustment, glucose monitoring, and prompts for ordering a glycosylated hemoglobin (A1c) level. Next, the order set and the supporting educational programs should integrate this guidance as much as possible, making the key guidance available at the point of patient care.

This guidance was available in our algorithm but was not inherent in the structured insulin orders described in this report, and all basal and nutritional insulin options were offered as equally acceptable choices. This version did not calculate insulin doses or assist in the apportionment of insulin between basal and nutritional components. Only a single adjustment dose scale was offered, leaving appropriate modifications up to the end user, and from a usability standpoint, our CPOE insulin orders lacked dynamic flexibility (revising a single insulin required discontinuing all prior orders and reentering all orders). These limitations have subsequently been addressed with Version 2 of our CPOE insulin orders, and the details will soon be available in the literature.15

We are now exploring further improvement with concurrent identification and intervention of hyperglycemic patients that are not on physiologic insulin regimens or not meeting glycemic targets, and implementing protocols addressing the transition from infusion insulin.

Interventions

We introduced several interventions and educational efforts throughout the course of our improvement. The 2 key interventions were as follows:

• Structured subcutaneous insulin order sets (November, 2003).

• An insulin management algorithm, described below (May 2005).

Key Intervention #1: Structured Subcutaneous Insulin Order Set Implementation

In November 2003, we introduced a paper‐based structured subcutaneous insulin order set. This order set encouraged the use of scheduled basal and nutritional insulin, provided guidance for monitoring glucose levels, and for insulin dosing. A hypoglycemia protocol and a standardized correction insulin table were embedded in the order set. This set was similar to examples of structured insulin ordering subsequently presented in the literature.9 In a parallel effort, the University of California, San Diego Medical Center (UCSDMC) was developing a computer physician order entry (CPOE) module for our comprehensive clinical information system, Invision (Siemens Medical Systems, Malvern, PA), that heretofore had primarily focused on result review, patient schedule management, and nursing documentation. In anticipation of CPOE and for the purpose of standardization, we removed outdated sliding scale insulin regimens from a variety preexisting order sets and inserted references to the standardized subcutaneous insulin order set in their stead. The medication administration record (MAR) was changed to reflect the basal/nutritional/correction insulin terminology. It became more difficult to order a stand‐alone insulin sliding scale even before CPOE versions became available. The standardized order set was the only preprinted correction scale insulin order available, and ordering physicians have to specifically opt out of basal and nutritional insulin choices to order sliding scale only regimens. Verbal orders for correction dose scales were deemed unacceptable by medical staff committees. Correctional insulin doses could be ordered as a 1‐time order, but the pharmacy rejected ongoing insulin orders that were not entered on the structured form.

We introduced our first standardized CPOE subcutaneous insulin order set in January 2004 at the smaller of our 2 campuses, and subsequently completed full deployment across both campuses in all adult medical‐surgical care areas by September 2004.

The CPOE version, like the paper version that immediately preceded it, encouraged the use of basal/bolus insulin regimens, promoted the terms basal, nutritional or premeal, and adjustment dose insulin in the order sets and the medication administration record, and was mandatory for providers wishing to order anything but a 1‐time order of insulin. Figure 1 depicts a screen shot of the CPOE version. Similar to the paper version, the ordering physician had to specifically opt out of ordering scheduled premeal and basal insulin to order a sliding scale only regimen. The first screen also ensured that appropriate POC glucose monitoring was ordered and endorsed a standing hypoglycemia protocol order. The CPOE version had only a few additional features not possible on paper. Obvious benefits included elimination of unapproved abbreviations and handwriting errors. Nutritional and correction insulin types were forced to be identical. Fundamentally, however, both the paper and online structured ordering experiences had the same degree of control over provider ordering patterns, and there was no increment in guidance for choosing insulin regimens, hence their combined analysis as structured orders.

Figure 1

Key Intervention #2: Insulin Management Algorithm

The structured insulin order set had many advantages, but also had many limitations. Guidance for preferred insulin regimens for patients in different nutritional situations was not inherent in the order set, and all basal and nutritional insulin options were offered as equally acceptable choices. The order set gave very general guidance for insulin dosing, but did not calculate insulin doses or assist in the apportionment of insulin between basal and nutritional components, and guidance for setting a glycemic target or adjusting insulin was lacking.

Recognizing these limitations, we devised an insulin management algorithm to provide guidance incremental to that offered in the order set. In April 2005, 3 hospitalists piloted a paper‐based insulin management algorithm (Figure 2, front; Figure 3, reverse) on their teaching services. This 1‐page algorithm provided guidance on insulin dosing and monitoring, and provided institutionally preferred insulin regimens for patients in different nutritional situations. As an example, of the several acceptable subcutaneous insulin regimens that an eating patient might use in the inpatient setting, we advocated the use of 1 preferred regimen (a relatively peakless, long‐acting basal insulin once a day, along with a rapid acting analog nutritional insulin with each meal). We introduced the concept of a ward glycemic target, provided prompts for diabetes education, and generally recommended discontinuation of oral hypoglycemic agents in the inpatient setting. The hospitalists were introduced to the concepts and the algorithm via 1 of the authors (G.M.) in a 1‐hour session. The algorithm was introduced on each teaching team during routine teaching rounds with a slide set (approximately 15 slides) that outlined the basic principles of insulin dosing, and gave example cases which modeled the proper use of the algorithm. The principles were reinforced on daily patient work rounds as they were applied on inpatients with hyperglycemia. The pilot results on 25 patients, compared to 250 historical control patients, were very promising, with markedly improved glycemic control and no increase in hypoglycemia. We therefore sought to spread the use of the algorithm. In May 2005 the insulin management algorithm and teaching slide set were promoted on all 7 hospitalist‐run services, and the results of the pilot and concepts of the algorithm were shared with a variety of house staff and service leaders in approximately a dozen sessions: educational grand rounds, assorted noon lectures, and subsequently, at new intern orientations. Easy access to the algorithm was assured by providing a link to the file within the CPOE insulin order set.

Figure 2

Figure 3

Measures and Analyses

Insulin Use Patterns

Figure 4 demonstrates the dramatic improvement that took place with the introduction of the structured order set. In the 6 months preceding the introduction of the structured insulin order set (May‐October 2003) 72% of 477 sampled patients with insulin orders were on sliding scale‐only insulin regimens (with no basal insulin), compared to just 26% of 499 patients sampled in the March to August 2004 time period subsequent to order set implementation (P < .0001, chi square statistic). Intermittent monthly checks on insulin use patterns reveal this change has been sustained.

Figure 4

Glycemic Control

A total of 9314 patients with 44,232 monitored patient‐days and over 120,000 POC glucose values were analyzed to assess glycemic control, which was improved with structured insulin orders and improved incrementally with the introduction of the insulin management algorithm.

The percent of patient‐days that were uncontrolled, defined as a monitored day with a mean glucose of 180 mg/dL, was reduced over the 3 time periods (37.8% versus 33.9% versus 30.1%, P < 0.005, Pearson chi square statistic), representing a 21% RR reduction of uncontrolled patient‐days from TP1 versus TP3. Table 2 shows the summary results for glycemic control, including the RR and CIs between the 3 time periods.

In a similar fashion, the percent of patients with uncontrolled patient‐stays (day‐weighted mean glucose 180 mg/dL) was also reduced over the 3 time periods (41.5% versus 37.6% versus 34.2%, P < 0.05, Pearson chi square statistic, with an RR reduction of 16% for TP3:TP1). Figure 5 depicts a statistical process control chart of the percent of patients experiencing uncontrolled patient‐stays over time, and is more effective in displaying the temporal relationship of the interventions with the improved results.

Figure 5

Uncontrolled hyperglycemic days and stays were reduced incrementally from TP3 versus TP2, reflecting the added benefit of the insulin management algorithm, compared to the benefit enjoyed with the structured order set alone.

When the analyses were repeated after excluding patients with fewer than 8 POC glucose readings (Table 3), the findings were similar, but as predicted, the effect was slightly more pronounced, with a 23% relative reduction in uncontrolled patient‐days and a 27% reduction in uncontrolled patient‐stays of TP3 versus TP1.

Hypoglycemia

Table 4 summarizes the results for hypoglycemia and severe hypoglycemia in the study population, and Table 5 summarizes the secondary analyses of hypoglycemia in the subset with at least 8 POC glucose readings.

Strengths and Limitations

Our study has several strengths. The study is large and incorporates all patients with diabetes or hyperglycemia captured by POC glucose testing, and the observation period is long enough that bias from merely being observed is not a factor. We used metrics for glycemic control, hypoglycemia, and insulin use patterns that are of high quality and are generally in line with the Society of Hospital Medicine (SHM) Glycemic Control Task force recommendations,12, 13 and examined data by both patient‐stay and patient‐day.

The increased use of anticipatory physiologic subcutaneous insulin regimens, and the subsequent decline in the use of sliding scale insulin, is the most likely mechanism for improvement. The improvements seen are fairly dramatic for an institution in absolute terms, because inpatient hyperglycemia and hypoglycemia are so common. For example, on an annualized basis for our 400‐bed medical center, these interventions prevent 124 patients from experiencing 208 hypoglycemic days.

Other institutions should be able to replicate our results. We received administrative support to create a multidisciplinary steering committee, but we did not have incremental resources to create a dedicated team for insulin management, mandated endocrinology comanagement or consultations, or manual data collection. In fact, we had only 1 diabetes educator for 400 adult beds at 2 sites, and were relatively underresourced in this area by community standards. There was some time and expense in creating the glycemic control reports, but all of the glucose data collected were part of normal care, and the data retrieval became automated.

The main limitation of this study lies in the observational study design. There were multiple interventions in addition to structured insulin orders and the insulin management algorithm, and these educational and organizational changes undoubtedly also contributed to the overall success of our program. Since we did not perform a randomized controlled trial, the reader might reasonably question if the structured order sets and insulin management algorithm were actually the cause of the improvement seen, as opposed to these ancillary efforts or secular change. However, there are several factors that make this unlikely. First, the study population was well‐defined, having diabetes or documented hyperglycemia in all 3 time periods. Second, the demographics remained constant or actually worked against improvement trends, since the markers of patient acuity suggest increased patient acuity over the observation period. Third, the temporal relationship of the improvement to the introduction of our key interventions, as viewed on statistical process control charts shown in Figure 5, strongly suggest a causal relationship. This temporal relationship was consistently observed no matter how we chose to define uncontrolled hyperglycemia, and was also seen on hypoglycemia control charts. We view the ancillary interventions (such as educational efforts) as necessary, but not sufficient, in and of themselves, to effect major improvement.

We did not analyze the impact of the improved glycemic control on patient outcomes. In the absence of a randomized controlled trial design, controlling for the various confounders is a challenging task. Also, it is likely that not all hypoglycemic events were attributable to inpatient glycemic control regimens, though the secondary analysis probably eliminated many hypoglycemia admissions.

Lessons Learned: Implications from our study

We agree with the American Association of Clinical Endocrinologists (AACE)/American Diabetes Association (ADA)2 and the SHM Glycemic Control Task Force12 about the essential elements needed for successful implementation of inpatient glycemic control programs:

• An appropriate level of administrative support.

• Formation of a multidisciplinary steering committee to drive the development of initiatives, empowered to enact changes.

• Assessment of current processes, quality of care, and barriers to practice change.

• Development and implementation of interventions, including standardized order sets, protocols, policies, and algorithms with associated educational programs.

• Metrics for evaluation of glycemic control, hypoglycemia, insulin use patterns, and other aspects of care.

Metrics to follow hypoglycemia are extremely important. The voluntary reporting on insulin‐induced hypoglycemia fluctuated widely over the course of our project. These fluctuations did not correlate well with the more objective and accurate measures we followed, and this objective data was very helpful in reducing the fear of hypoglycemia, and spreading the wider use of basal bolus insulin regimens. We strongly recommend that improvement teams formulate and follow measures of glycemic control, hypoglycemia, and insulin use, similar to those outlined in the SHM Glycemic Control Improvement Guide12 and the SHM Glycemic Control Task Force summary on glucometrics.13

Although we introduced our structured insulin order set first, with a long lag time until we introduced the insulin management algorithm, we advocate a different approach for institutions grappling with these issues. This approach is well‐described by the SHM Glycemic Control Task Force.14 An insulin management algorithm should be crafted first, integrating guidance for insulin dosing, preferred insulin regimens for different nutritional situations, a glycemic target, insulin dosing adjustment, glucose monitoring, and prompts for ordering a glycosylated hemoglobin (A1c) level. Next, the order set and the supporting educational programs should integrate this guidance as much as possible, making the key guidance available at the point of patient care.

This guidance was available in our algorithm but was not inherent in the structured insulin orders described in this report, and all basal and nutritional insulin options were offered as equally acceptable choices. This version did not calculate insulin doses or assist in the apportionment of insulin between basal and nutritional components. Only a single adjustment dose scale was offered, leaving appropriate modifications up to the end user, and from a usability standpoint, our CPOE insulin orders lacked dynamic flexibility (revising a single insulin required discontinuing all prior orders and reentering all orders). These limitations have subsequently been addressed with Version 2 of our CPOE insulin orders, and the details will soon be available in the literature.15

We are now exploring further improvement with concurrent identification and intervention of hyperglycemic patients that are not on physiologic insulin regimens or not meeting glycemic targets, and implementing protocols addressing the transition from infusion insulin.

Hospitalists

Endocrinologists

Education

• Marcia D. Draheim, RN, CDE. Program Supervisor, Diabetes Center, St Luke's Hospital, Cedar Rapids, Iowa.

• Sharon Mahowald, RN, CDE. Inpatient Diabetes Coordinator, Hennepin County Medical Center, Minneapolis, Minnesota.

Financial

• Adam Beck, MHS, FABC. MedStar Research Institute, Washington, DC.

Pharmacists

• Stuart T. Haines, PharmD, FASHP, FCCP, BCPS. Associate Professor/Vice Chair for Education, University of Maryland School of Pharmacy Baltimore, Maryland.

Case Management

Dietetics

• Carrie Swift, MS, RD, BC‐ADM. Dietetics Coordinator, Veterans Affairs Medical Center, Walla Walla, Washington.

SHM Staff Members

• Geri Barnes and Joy Wittnebert.

Glycemic Control Resource Room Project Team

• Greg Maynard, Jason Stein, David Wesorick, Mary Ann Emanuele, Kevin Larsen, Geri Barnes, Joy Wittnebert, and Bruce Hansen.

Hospitalists

Endocrinologists

Education

• Marcia D. Draheim, RN, CDE. Program Supervisor, Diabetes Center, St Luke's Hospital, Cedar Rapids, Iowa.

• Sharon Mahowald, RN, CDE. Inpatient Diabetes Coordinator, Hennepin County Medical Center, Minneapolis, Minnesota.

Financial

• Adam Beck, MHS, FABC. MedStar Research Institute, Washington, DC.

Pharmacists

• Stuart T. Haines, PharmD, FASHP, FCCP, BCPS. Associate Professor/Vice Chair for Education, University of Maryland School of Pharmacy Baltimore, Maryland.

Case Management

Dietetics

• Carrie Swift, MS, RD, BC‐ADM. Dietetics Coordinator, Veterans Affairs Medical Center, Walla Walla, Washington.

SHM Staff Members

• Geri Barnes and Joy Wittnebert.

Glycemic Control Resource Room Project Team

• Greg Maynard, Jason Stein, David Wesorick, Mary Ann Emanuele, Kevin Larsen, Geri Barnes, Joy Wittnebert, and Bruce Hansen.

Hospitalists

Endocrinologists

Education

• Marcia D. Draheim, RN, CDE. Program Supervisor, Diabetes Center, St Luke's Hospital, Cedar Rapids, Iowa.

• Sharon Mahowald, RN, CDE. Inpatient Diabetes Coordinator, Hennepin County Medical Center, Minneapolis, Minnesota.

Financial

• Adam Beck, MHS, FABC. MedStar Research Institute, Washington, DC.

Pharmacists

• Stuart T. Haines, PharmD, FASHP, FCCP, BCPS. Associate Professor/Vice Chair for Education, University of Maryland School of Pharmacy Baltimore, Maryland.

Case Management

Dietetics

• Carrie Swift, MS, RD, BC‐ADM. Dietetics Coordinator, Veterans Affairs Medical Center, Walla Walla, Washington.

SHM Staff Members

• Geri Barnes and Joy Wittnebert.

Glycemic Control Resource Room Project Team

• Greg Maynard, Jason Stein, David Wesorick, Mary Ann Emanuele, Kevin Larsen, Geri Barnes, Joy Wittnebert, and Bruce Hansen.

Building The Team

Implementing a medical centerwide standardized insulin infusion order set with supporting policies, protocols, monitoring standards, and the requisite educational programs is a major task for any hospital. This is not a simple maneuver involving only 1 or 2 interested individuals and requires much more than selecting a published protocol and disseminating it to various patient care units. Instead, to manage the full spectrum of diabetes programs and protocols, an institution must convene a multidisciplinary steering committee or task force. This should include representation from nursing, nursing administration, pharmacy, nutrition services, and the quality improvement department. Physicians should include hospitalists, intensivists, and endocrinologists but may also involve anesthesiologists and surgeons, as applicable. At times, additional members may need to be recruited according to project needs.

Identifying the Stakeholders and Current Practices

In developing or improving currently utilized IIPs, the multidisciplinary committee would benefit from careful background work before moving forward. First, administrative and institutional support must be secured to endorse uniform standards for insulin infusions, and to provide the important infrastructure needed to facilitate the work involved. Clinical and administrative stakeholders from the key departments then need to be identified.

All insulin infusion orders and policies/procedures presently used in the institution should be identified and examined. The developers and/or users of these order sets should be engaged in a dialogue and encouraged to share their experiences regarding their current practice and the attendant work flow, glucose monitoring, and data collection and reporting. Immediate concerns should be clearly addressed. Measurement systems for glycemic control, hypoglycemia, and insulin use patterns should assess current practice and the impact of subsequent modifications of the protocol or initiation of new protocols. We recommend using glucometrics consistent with those endorsed by the Society of Hospital Medicine (SHM) Glycemic Control Task Force elsewhere in this supplement.18

Addressing the Burden of Change

Through this process, the committee will uncover barriers, dysfunctional and inconsistent practices, and individuals who will pose challenges. Identifying these issues should not discourage the team, but rather it should guide the interventional strategies, and help build consensus that change may be required. There must be caution not to exclude significant individuals simply because they resist changes. Indeed, if they are included and have the opportunity to contribute to the process, success is much more likely.

It is important for process leaders to understand the implications of what is proposed, particularly for nursing services.19 For example, it has been shown that IIPs require about 5 minutes per patient per hour for glucose monitoring and dose adjustments.20 Acknowledging and attempting to address this burden proactively (often well over 2 hours per day) can gain staff acceptance more effectively than a laissez faire approach. In this regard, some effort should be invested in nursing education of the benefits of tight glycemic management on critical care outcomes. The difficult‐to‐quantify work involved when patients' blood glucose is not well controlled (eg, paging physicians for stat insulin orders) should also be part of this discussion.

Identifying and Addressing Barriers to IIP implementation

There are numerous potential barriers to implementation of IIPs. Table 1 identifies some of the most frequent ones along with potential strategies or solutions. Very common barriers include skepticism surrounding the benefits of tight glycemic control, fear of hypoglycemia, and difficulty agreeing on glycemic targets.

Whereas the national guidelines call for a glycemic target in critical care areas with an upper limit of 110 mg/dL, there is room for debate, and tailoring of the glycemic target to fit individual patient circumstances is often advisable. Starting with a less aggressive glycemic target can be good politics (and perhaps good medicine as well). Once the higher glycemic targets are achieved safely, it can pave the way for a more aggressive approach.

Fear of hypoglycemia is one of the most potent barriers to intensive insulin infusion implementation. Because hyperglycemia is such a common condition in critical care units, nursing and physician staff may have developed a skewed view of the definition of hypoglycemia, at times fearing for their patient when the glucose values reach a level of 100 mg/dL or so. Polling the nurses on what they think their own fasting glucose levels are, and then actually measuring them can be an effective strategy (the nurses may be surprised that the patient's scary 90 mg/dL reading is higher than their own). It should also be emphasized that properly designed and implemented protocols may actually decrease the incidence of hypoglycemia when compared to standard care which may involve individually and sometimes improperly adjusted intravenous (IV) insulin (discussed further below).

Structural Differences in Protocols

Protocols that vary by level of insulin sensitivity generally use column methods with the individual columns representing different categories of insulin sensitivity, placing the most sensitive category on the left with the highest level of insulin resistance on the far right. These methods use a multiplier to adjust for sensitivity. They are constructed according to the rule of 1500, 1700, or 1800, thereby adjusting for changes in insulin sensitivity that follow surgery or other changes in physiologic stress in the acute setting. Rate of change is addressed by shifting to the right if correction of hyperglycemia is too slow and to the left if the glucose is dropping too rapidly.

Most institutions using the column/emnsulin sensitivity method, implement paper orders as first published by Markovitz.28 The same concepts have been used to develop computer‐assisted insulin infusion protocols. One published method is the Glucommander,27 but a number of institutions are using similar computer‐assisted methods developed locally.

Other methods use the present glucose and change from last glucose to constantly adjust to any situation.7, 2125 They usually involve 2 or 3 steps and often require more calculations by the nurse. These methods are purported to be more agile or flexible but there have been no direct comparisons with the column methods looking at effectiveness, nursing errors, or hypoglycemic risk. The Yale Protocol23, 24 and the Portland Protocol22 are 2 prominent examples of this type of protocol. Adjustments are defined as units, percent change, or a combination of both.

Limitations of the IIP Literature

There had been few published insulin protocols aimed at reaching specific glucose goals when the Leuven surgical ICU experience7 was published in 2001. These early publications featured algorithms that adjusted insulin infusions solely on the basis of glucose level, and did not take the velocity of change, direction of change, proximity to glycemic target, or different insulin sensitivities into account. Also, the targeted glucose goals in those reports were not consistent with the present standards.2931 Other published protocols featured glucose‐insulin‐potassium infusion (GIK) protocols that focused on the amount of insulin administered and failed to attain appropriately defined glucose targets.32, 33 These publications offer no real guidance in crafting a modern glycemic targetoriented protocol.

Whereas more than 20 modern IIPs directed at glycemic targets have been published,7, 1517, 2127, 3450 many of the published protocols represent local modifications of ones previously published elsewhere, and no published, prospective, head‐to‐head comparisons of the best‐known IIPs are available.

Several reviews of previous published reports include comparisons of IIPs with varied areas of emphasis.5154 The reliability of such comparisons is limited by the inconsistency in methodology between studies and the different populations studied. For example, medical populations are generally harder to control and are more prone to hypoglycemia than surgical populations,7, 21, 23, 24, 50 and some studies include only patients with diabetes.4, 5, 38 In addition, definitions of hypoglycemia and methods of analyzing glycemic control are highly variable, making comparisons of IIPs challenging.39 As a result, attempts to compare published IIP results without consideration of the population studied could lead to erroneous conclusions.

In spite of the aforementioned limitations, there are several lessons we can learn from the IIP literature and accumulated clinical experience at a variety of centers.

Glycemic Targets and Other IIP Features Can Evolve over Time

Revisions of protocol details, including glycemic targets, often evolve over time. This deliberate approach can facilitate staff acceptance by demonstrating safe achievement of higher glycemic targets. The Yale team initially selected a conservative glycemic target of 100‐139 mg/dL, as depicted in Table 1.23, 24 They subsequently lowered the target range to 90‐119 mg/dL and increased the initial insulin bolus amount by 40%.50 The modified protocol displayed improved performance and yet retained safety, with only 1 glucose reading < 40 mg/dL in 101 patients over 117 runs of insulin infusion.

The Portland Protocol, used primarily in cardiac surgery patients, has also evolved over time. This group has altered the IIP in a number of ways at least 4 times. These protocols, including the most recent protocol targeting glucose levels of 70‐110 mg/dL, is found at www.portlandprotocol.org.4, 5, 55, 56

Medical and Surgical Patients Are Different

The most convincing evidence for stringent glycemic control evolved from studies of surgical patients.4, 5, 7 Acknowledging this fact, along with the greater degree of difficulty in achieving glycemic targets safely in critically ill medical patients, have led many to endorse higher glycemic targets for certain populations than others.57 Although we generally favor a uniform glycemic target for a unit serving a particular patient population, adopting a less stringent glycemic target in medical ICU settings compared to surgical ICU settings is reasonable and prudent in many institutional settings. However, the accompanying challenges regarding boarding patients and floating nurses between units would also need to be considered.

Local Factors and Implementation Methods Matter

The success or failure of a protocol likely depends on local factors and implementation methods as much as it does on the structure of the protocol itself. IIP development and implementation is a process that must be approached systematically and with attention to detail.19 Errors in approach can delay or abort the implementation, or potentially lead to an ineffective or unsafe protocol for the institution.

The Yale experience again provides us with a salient example.58 Initial efforts to implement an IIP failed due to a number of factors: a complicated protocol, insufficient nursing involvement, and inadequate training and education led to incomplete buy‐in, and nursing concerns over hypoglycemia that actually was within the goal range of the protocol. Successful implementation was not achieved until the leaders learned from their mistakes. Nurses and other clinical allies were involved and educated. Important stakeholders, who were not included earlier, were now involved, and front‐line nursing staff were engaged in proactive troubleshooting.

In another example, multisite studies like VISEP59 and Glucontrol9 have tried to adapt the Leuven protocol, only to struggle with excessive hypoglycemia and an inability to replicate the tight glycemic control enjoyed by van den Berghe and colleagues.7, 21 The Leuven care team augmented the performance of the IIP by their experience with the protocol, physician involvement, and adjustments made by the nursing staff.

Some IIPs Have Lower Hypoglycemia Rates than Others

Recent articles have highlighted the association of intensive insulin treatment, hypoglycemia, and mortality in the medical ICU population.60 Concerns that excessive hypoglycemia could reduce or reverse the overall benefits of intensive insulin therapy has led some to call for moderation of critical care glycemic targets.61 Comparisons of IIPs5254, 61 that include information about the propensity of the IIP to produce hypoglycemia are therefore important.

The Leuven IIP has been reported to yield high rates of hypoglycemia with almost 20% of patients suffering from severe hypoglycemia (glucose < 40 mg/dL) in some studies.9, 21, 59 By way of contrast, several different IIP regimens have frequently achieved identical or very similar glycemic control with less than 5% of the patients experiencing severe hypoglycemia.27, 50, 62 Although implementation factors, severity of illness, and other population factors play some role, some protocols have an inherent structural propensity to produce more hypoglycemia than others. In the case of the Leuven protocol, there is more of a role for clinician adjustment, and the written protocol itself does not call for much active adjustment as the patient enters into the hypoglycemic range.52 In contrast, more sophisticated and complicated regimens adjust more aggressively in this range and consistently induce less hypoglycemia.

Successful Methods to Manage the Complexity of an IIP

Many modern IIPs use bolus insulin to expedite control and adjust the infusion rate based on the velocity and direction of glucose change, not just the glucose value.52 Insulin resistance is taken into account in some models, and multiple calculation steps are required in several reported protocols.52 The improved automation and control refinements come at a cost of increased complexity.

Intensive implementation efforts and strong leadership can overcome some issues, as demonstrated by the Yale experience, but 2 other strategies have commonly demonstrated success. First, focused expert glycemic management teams that directly oversee many aspects of the IIP, or even assume direct management roles, can be quite effective,25 especially during early implementation.

Automation of calculations and computerization is another method to consider, and many recent reports involve applications of web‐based or other computerized models.27, 40, 42, 43, 46, 48, 63 Comparisons of computerized versus manual methods, computerized versus column methods, and a computerized nomogram‐versus‐chart method are now available.42, 45, 46, 48, 62 Computerized protocols show significant promise and would be expected to reduce dosing errors. These instruments generally present the nurse with a specific infusion rate after each monitored glucose and then recommend an interval for the next glucose determination. In direct comparisons, they tend to perform as well or often better than conventional methods. For example, in the most recent randomized trial comparing the Glucommander to a paper‐based column method in an ICU, Newton and colleagues found the computerized method reached the goal more rapidly and reached a lower mean glucose without increasing the rate of severe hypoglycemia (< 40 mg/dL).62 Computerized methods also facilitate data collection for analytical purposes.

Computerized systems (both commercial and home‐grown versions) are becoming more common and appear to hold significant benefits as long as they are backed by a validated algorithm. Whereas many institutions are not yet in a position to integrate such protocols into their standard or electronic record systems, we expect the trend for increased implementation to continue.

Standardizing Insulin Infusion Preparations and Priming New Tubing

Varied concentrations or types of insulin for insulin infusions can lead to serious errors. Insulin infusions should generally be centrally prepared with a standard concentration of regular insulin in the pharmacy (usually 1 unit/cc), and the infusion concentration should be included in your infusion order set. Insulin binding to IV tubing can lead to false elevation of insulin requirements, potentially followed by serious hypoglycemia. When nurses change IV tubing or initially set up an insulin drip, education/emnstructions on priming new tubing with a small amount of insulin infusion to saturate the binding to the polyvinyl chloride tubing should be incorporated into their routine. Although 50 mL has often been recommended for priming, a recent study64 found that 20 mL of insulin infusion is enough to reach the saturation point.

Avoiding Over‐Reliance on the Insulin Protocol

Nurse‐driven insulin infusion protocols automate frequent insulin adjustment and reduce unnecessary calls to the physician. Although this is generally a decided advantage, the care team can be lulled into a sense of false security by the presence of orders that allow for such adjustment. Increasing the rate of an insulin infusion without thoughtful attention to factors that may be playing a role in this increased requirement (such as developing sepsis, other medical decompensation, steroid boluses, or an increase in carbohydrate intake) can have serious consequences. By the same token, an unanticipated rapid decrease in insulin requirement should lead to a reassessment of the infusion, and an inquiry about cessation of glucocorticoid therapy or nutrition. Rarely, a pharmacy or nursing error may induce a pseudo‐change in insulin requirements. The protocol should lead the nurse to seek advice and alert the physician to review potential causes of dramatic changes in insulin requirements, rather than simply adjusting insulin or nutrition to correct the present abnormal value.

Interruption of the Insulin Infusion

Interruption of insulin infusions may occur for many reasons, either intended or unintended. At times, the doctors or nurses may temporarily stop the protocol to allow for delivery of blood products or medications when IV access is limited. Infusions may mistakenly not be restarted, or deliberate discontinuation may not be adequately communicated, potentially leading to worsening hyperglycemia or even the development of ketoacidosis, and other adverse clinical outcomes. Therefore, the algorithm should have clear orders for the nurses to contact the ordering physician if the infusion is stopped for any reason, other than protocol‐driven cessation due to falling blood glucose concentrations.

Interruption of Nutrition, Field Trips, and Communication

Insulin infusion commonly provides both basal and nutritional insulin requirements. Interruptions in nutritional intake are extremely common in the inpatient setting, with a potential to cause serious hypoglycemia. Feeding tubes are often pulled out without warning; enteral nutrition may also need be halted if high gastric residuals are noted or during certain diagnostic tests. At times, IV carbohydrate sources (dextrose, partial parenteral nutrition, total parenteral nutrition) may be interrupted as well. In some cases, field trips out of the critical care units to the operating room, imaging studies, or other hospital locations add another layer of challenges to managing the IIP. Staff in these various areas may not be familiar with the IIP or monitoring standards and techniques, and potentially may not even be aware that the patient is on an insulin infusion. It is therefore crucial to anticipate these pitfalls and develop effective institutional procedures for addressing them. For example, many institutions use D10 solution to replace the carbohydrate calories that are lost when tube feedings have to be interrupted in a gram‐per‐gram fashion. Patients should be clearly identified as being on an insulin infusion. The requirement for consistent glucose monitoring, hypoglycemia recognition and treatment, and insulin infusion adjustment requires either critical care nurse care of the patient on the field trip, or training in the same skills in areas such as endoscopy, interventional radiology, and operating rooms including the preoperative and postoperative care units. In any case, all services should be involved in crafting solutions that will ensure a consistent approach to glycemic control as the patient travels off‐unit. Monitoring and treatment equipment needs to be readily available in all sites, and hypoglycemia protocols need to be distributed and supported in all areas.

Preventing and Treating Hypoglycemia

Some hypoglycemia will occur with infusion protocols, no matter how carefully a protocol is crafted and how well it is administered. Hypoglycemia protocols should therefore be incorporated directly into an infusion order set. Treatment of hypoglycemic events with a full 50 mL of D50 solution is equivalent to 25 g of carbohydrate, which will raise glucose levels in the average patient by 125 mg/dL. Many institutions discourage the overcorrection of hypoglycemic events by encouraging giving lesser aliquots of D50 based on the degree of hypoglycemia.

Preventing hypoglycemia by recognizing hypoglycemia risk factors, proper monitoring, and anticipating reductions in insulin requirements from decreasing severity of illness, nutritional intake, or steroid dosing can also reduce the frequency of hypoglycemic events.

Steroids

Steroid boluses are commonly an integral part of regimens targeting a variety of conditions, such as transplant rejection, reactive airways disease, certain infections, cancer, and a variety of autoimmune disorders. This can lead to glycemic excursions and rapidly varying insulin requirements. Educational efforts and treatment regimens should address the disproportionate impact that steroids have on postprandial glycemic excursions. To minimize the glycemic impact of glucocorticoid therapy, a team should investigate promoting the use of steroid infusions in situations when a bolus is not absolutely necessary.

Dealing with the Eating Patient and Other Sources of Carbohydrate‐Induced Glycemic Excursions

Glucose levels can be difficult to control in patients who are eating while on insulin infusion, because the infusion chases the glycemic excursions through frequent adjustments, often with a late overshoot and inappropriate reduction in dose. We instead recommend providing bolus nutritional insulin to cover the expected glycemic excursion caused by carbohydrate ingestion. Carbohydrate counting and using a unit of insulin for each 10‐15 g of carbohydrate consumed can smooth out the rapid fluctuations in glucose. Guidance for this should be incorporated into the order set.

Transition Off of Insulin Infusion

Rational strategies for dealing with this transition are covered in detail elsewhere in this supplement.65 Guidance for managing this transition should be integrated into your insulin infusion and SC order sets. The transition to SC insulin may represent a separate order set but is sometimes best integrated into the IV insulin infusion order set itself.

IIPs Outside of the Critical Care Setting

IIPs are most commonly used in the critical care setting. In some institutions, IV insulin protocols are safely and effectively employed outside the ICU. Obviously, the number of nurses and other personnel who must be familiar with such protocols is much higher outside the ICU, and protocol errors are therefore likely to be somewhat higher. In addition the nurse‐per‐patient ratio is usually lower outside of the critical care setting. As a result, suggestions for safe implementation of insulin infusion regimens outside of the critical care setting include:

• Choose an infusion protocol with a higher glycemic target.

• Limit the medical and surgical units where this expertise will be developed.

• Consider simplified infusion protocols but stay consistent with format.

• Automated or computerized assistance of calculations may reduce human error and nursing burden.

Monitoring, Recording, and Analyzing Glycemic Control Data

Once the IIP is implemented, it is critical that the impact on glycemic control, hypoglycemia, insulin use, and other factors be analyzed and used for improving the IIP and care delivery. Frequent monitoring of glucose levels is necessary for the safe infusion of insulin. Guidance for how often the monitoring is required must be explicit and included in the infusion order set. Intermittent auditing for compliance with the frequency of glucose testing and appropriate dose selection is good practice. Attention should be paid to how the glucose level is obtained, recorded, and made available to the health care team in your institution. All glucose readings should be recorded electronically for ongoing analyses and retrieval, and ideally, this could be done in an automated or single‐step method. Try to eliminate duplication of effort, such as asking the nurse to record the glucose level and their reaction to it on paper and again in an electronic format. Your team should also provide guidance about the potential problems of using point‐of‐care glucose testing in settings with hypotension, sepsis, pressor use, and other conditions that may impair the accuracy of capillary glucose readings.

Reports on the time to reach the glycemic target, glycemic control while on infusion, and the incidence of hypoglycemia should be reviewed by the multidisciplinary steering committee. The Society of Hospital Medicine Glycemic Control Task Force recommends analysis by patient day and by patient stay (or insulin infusion run) as preferred methodologies for analysis of glycemic control and hypoglycemia rates over the method of using each individual glucose reading as the unit of analysis. (The latter tends to under‐value the frequency of hypoglycemia.) Detailed practical recommendations for analyzing and summarizing glycemic control data are available elsewhere in this supplement.18 These data should drive decisions on modification of glycemic targets and the protocol structure. Patients meeting prespecified criteria should be referred to the improvement team for review. For example, patients who experience any glucose readings of 40 mg/dL, or who take more than 12 hours to reach the upper limit glycemic target should be referred to the team for a case review.

Assessing Adherence to the Protocol and Ease‐of‐Use Issues

Focused audits in the pilot and early implementation phases should look for nonadherence to the protocol. Deviations should be evaluated according to the patterns identified. For example, variation in application in some cases is specific for an individual and in others is characteristic of a specific group or the whole. Accordingly, this may point to gaps in education or attitudes about the importance of this endeavor. Front‐line staff may deviate from the protocol because they find it ineffective, unsafe, or impractical for certain situations or specific patients. Many IIPs are the subject of nursing errors related to the knowledge and acceptance of the nurse but also the complexity of the protocol. Appropriate modifications to the protocol based on these cases can frequently improve the ease of use and effectiveness of the protocol. The ongoing review process should identify issues that must be addressed with permanent solutions rather than accepting frequent individual alterations to meet goals. Revisions require supplementary education and rapid and wide dissemination. Although educational efforts and monitoring are often most intense in the early implementation phase, periodic retraining should continue to achieve optimal results and safety. Educational tools must consider nursing time commitments and will often include an interactive web‐based module that gives more flexibility for trainers and clinical nurses alike.

Building The Team

Implementing a medical centerwide standardized insulin infusion order set with supporting policies, protocols, monitoring standards, and the requisite educational programs is a major task for any hospital. This is not a simple maneuver involving only 1 or 2 interested individuals and requires much more than selecting a published protocol and disseminating it to various patient care units. Instead, to manage the full spectrum of diabetes programs and protocols, an institution must convene a multidisciplinary steering committee or task force. This should include representation from nursing, nursing administration, pharmacy, nutrition services, and the quality improvement department. Physicians should include hospitalists, intensivists, and endocrinologists but may also involve anesthesiologists and surgeons, as applicable. At times, additional members may need to be recruited according to project needs.

Identifying the Stakeholders and Current Practices

In developing or improving currently utilized IIPs, the multidisciplinary committee would benefit from careful background work before moving forward. First, administrative and institutional support must be secured to endorse uniform standards for insulin infusions, and to provide the important infrastructure needed to facilitate the work involved. Clinical and administrative stakeholders from the key departments then need to be identified.

All insulin infusion orders and policies/procedures presently used in the institution should be identified and examined. The developers and/or users of these order sets should be engaged in a dialogue and encouraged to share their experiences regarding their current practice and the attendant work flow, glucose monitoring, and data collection and reporting. Immediate concerns should be clearly addressed. Measurement systems for glycemic control, hypoglycemia, and insulin use patterns should assess current practice and the impact of subsequent modifications of the protocol or initiation of new protocols. We recommend using glucometrics consistent with those endorsed by the Society of Hospital Medicine (SHM) Glycemic Control Task Force elsewhere in this supplement.18

Addressing the Burden of Change

Through this process, the committee will uncover barriers, dysfunctional and inconsistent practices, and individuals who will pose challenges. Identifying these issues should not discourage the team, but rather it should guide the interventional strategies, and help build consensus that change may be required. There must be caution not to exclude significant individuals simply because they resist changes. Indeed, if they are included and have the opportunity to contribute to the process, success is much more likely.

It is important for process leaders to understand the implications of what is proposed, particularly for nursing services.19 For example, it has been shown that IIPs require about 5 minutes per patient per hour for glucose monitoring and dose adjustments.20 Acknowledging and attempting to address this burden proactively (often well over 2 hours per day) can gain staff acceptance more effectively than a laissez faire approach. In this regard, some effort should be invested in nursing education of the benefits of tight glycemic management on critical care outcomes. The difficult‐to‐quantify work involved when patients' blood glucose is not well controlled (eg, paging physicians for stat insulin orders) should also be part of this discussion.

Identifying and Addressing Barriers to IIP implementation

There are numerous potential barriers to implementation of IIPs. Table 1 identifies some of the most frequent ones along with potential strategies or solutions. Very common barriers include skepticism surrounding the benefits of tight glycemic control, fear of hypoglycemia, and difficulty agreeing on glycemic targets.

Whereas the national guidelines call for a glycemic target in critical care areas with an upper limit of 110 mg/dL, there is room for debate, and tailoring of the glycemic target to fit individual patient circumstances is often advisable. Starting with a less aggressive glycemic target can be good politics (and perhaps good medicine as well). Once the higher glycemic targets are achieved safely, it can pave the way for a more aggressive approach.

Fear of hypoglycemia is one of the most potent barriers to intensive insulin infusion implementation. Because hyperglycemia is such a common condition in critical care units, nursing and physician staff may have developed a skewed view of the definition of hypoglycemia, at times fearing for their patient when the glucose values reach a level of 100 mg/dL or so. Polling the nurses on what they think their own fasting glucose levels are, and then actually measuring them can be an effective strategy (the nurses may be surprised that the patient's scary 90 mg/dL reading is higher than their own). It should also be emphasized that properly designed and implemented protocols may actually decrease the incidence of hypoglycemia when compared to standard care which may involve individually and sometimes improperly adjusted intravenous (IV) insulin (discussed further below).

Structural Differences in Protocols

Protocols that vary by level of insulin sensitivity generally use column methods with the individual columns representing different categories of insulin sensitivity, placing the most sensitive category on the left with the highest level of insulin resistance on the far right. These methods use a multiplier to adjust for sensitivity. They are constructed according to the rule of 1500, 1700, or 1800, thereby adjusting for changes in insulin sensitivity that follow surgery or other changes in physiologic stress in the acute setting. Rate of change is addressed by shifting to the right if correction of hyperglycemia is too slow and to the left if the glucose is dropping too rapidly.

Most institutions using the column/emnsulin sensitivity method, implement paper orders as first published by Markovitz.28 The same concepts have been used to develop computer‐assisted insulin infusion protocols. One published method is the Glucommander,27 but a number of institutions are using similar computer‐assisted methods developed locally.

Other methods use the present glucose and change from last glucose to constantly adjust to any situation.7, 2125 They usually involve 2 or 3 steps and often require more calculations by the nurse. These methods are purported to be more agile or flexible but there have been no direct comparisons with the column methods looking at effectiveness, nursing errors, or hypoglycemic risk. The Yale Protocol23, 24 and the Portland Protocol22 are 2 prominent examples of this type of protocol. Adjustments are defined as units, percent change, or a combination of both.

Limitations of the IIP Literature

There had been few published insulin protocols aimed at reaching specific glucose goals when the Leuven surgical ICU experience7 was published in 2001. These early publications featured algorithms that adjusted insulin infusions solely on the basis of glucose level, and did not take the velocity of change, direction of change, proximity to glycemic target, or different insulin sensitivities into account. Also, the targeted glucose goals in those reports were not consistent with the present standards.2931 Other published protocols featured glucose‐insulin‐potassium infusion (GIK) protocols that focused on the amount of insulin administered and failed to attain appropriately defined glucose targets.32, 33 These publications offer no real guidance in crafting a modern glycemic targetoriented protocol.

Whereas more than 20 modern IIPs directed at glycemic targets have been published,7, 1517, 2127, 3450 many of the published protocols represent local modifications of ones previously published elsewhere, and no published, prospective, head‐to‐head comparisons of the best‐known IIPs are available.

Several reviews of previous published reports include comparisons of IIPs with varied areas of emphasis.5154 The reliability of such comparisons is limited by the inconsistency in methodology between studies and the different populations studied. For example, medical populations are generally harder to control and are more prone to hypoglycemia than surgical populations,7, 21, 23, 24, 50 and some studies include only patients with diabetes.4, 5, 38 In addition, definitions of hypoglycemia and methods of analyzing glycemic control are highly variable, making comparisons of IIPs challenging.39 As a result, attempts to compare published IIP results without consideration of the population studied could lead to erroneous conclusions.

In spite of the aforementioned limitations, there are several lessons we can learn from the IIP literature and accumulated clinical experience at a variety of centers.

Glycemic Targets and Other IIP Features Can Evolve over Time

Revisions of protocol details, including glycemic targets, often evolve over time. This deliberate approach can facilitate staff acceptance by demonstrating safe achievement of higher glycemic targets. The Yale team initially selected a conservative glycemic target of 100‐139 mg/dL, as depicted in Table 1.23, 24 They subsequently lowered the target range to 90‐119 mg/dL and increased the initial insulin bolus amount by 40%.50 The modified protocol displayed improved performance and yet retained safety, with only 1 glucose reading < 40 mg/dL in 101 patients over 117 runs of insulin infusion.

The Portland Protocol, used primarily in cardiac surgery patients, has also evolved over time. This group has altered the IIP in a number of ways at least 4 times. These protocols, including the most recent protocol targeting glucose levels of 70‐110 mg/dL, is found at www.portlandprotocol.org.4, 5, 55, 56

Medical and Surgical Patients Are Different

The most convincing evidence for stringent glycemic control evolved from studies of surgical patients.4, 5, 7 Acknowledging this fact, along with the greater degree of difficulty in achieving glycemic targets safely in critically ill medical patients, have led many to endorse higher glycemic targets for certain populations than others.57 Although we generally favor a uniform glycemic target for a unit serving a particular patient population, adopting a less stringent glycemic target in medical ICU settings compared to surgical ICU settings is reasonable and prudent in many institutional settings. However, the accompanying challenges regarding boarding patients and floating nurses between units would also need to be considered.

Local Factors and Implementation Methods Matter

The success or failure of a protocol likely depends on local factors and implementation methods as much as it does on the structure of the protocol itself. IIP development and implementation is a process that must be approached systematically and with attention to detail.19 Errors in approach can delay or abort the implementation, or potentially lead to an ineffective or unsafe protocol for the institution.

The Yale experience again provides us with a salient example.58 Initial efforts to implement an IIP failed due to a number of factors: a complicated protocol, insufficient nursing involvement, and inadequate training and education led to incomplete buy‐in, and nursing concerns over hypoglycemia that actually was within the goal range of the protocol. Successful implementation was not achieved until the leaders learned from their mistakes. Nurses and other clinical allies were involved and educated. Important stakeholders, who were not included earlier, were now involved, and front‐line nursing staff were engaged in proactive troubleshooting.

In another example, multisite studies like VISEP59 and Glucontrol9 have tried to adapt the Leuven protocol, only to struggle with excessive hypoglycemia and an inability to replicate the tight glycemic control enjoyed by van den Berghe and colleagues.7, 21 The Leuven care team augmented the performance of the IIP by their experience with the protocol, physician involvement, and adjustments made by the nursing staff.

Some IIPs Have Lower Hypoglycemia Rates than Others

Recent articles have highlighted the association of intensive insulin treatment, hypoglycemia, and mortality in the medical ICU population.60 Concerns that excessive hypoglycemia could reduce or reverse the overall benefits of intensive insulin therapy has led some to call for moderation of critical care glycemic targets.61 Comparisons of IIPs5254, 61 that include information about the propensity of the IIP to produce hypoglycemia are therefore important.

The Leuven IIP has been reported to yield high rates of hypoglycemia with almost 20% of patients suffering from severe hypoglycemia (glucose < 40 mg/dL) in some studies.9, 21, 59 By way of contrast, several different IIP regimens have frequently achieved identical or very similar glycemic control with less than 5% of the patients experiencing severe hypoglycemia.27, 50, 62 Although implementation factors, severity of illness, and other population factors play some role, some protocols have an inherent structural propensity to produce more hypoglycemia than others. In the case of the Leuven protocol, there is more of a role for clinician adjustment, and the written protocol itself does not call for much active adjustment as the patient enters into the hypoglycemic range.52 In contrast, more sophisticated and complicated regimens adjust more aggressively in this range and consistently induce less hypoglycemia.

Successful Methods to Manage the Complexity of an IIP

Many modern IIPs use bolus insulin to expedite control and adjust the infusion rate based on the velocity and direction of glucose change, not just the glucose value.52 Insulin resistance is taken into account in some models, and multiple calculation steps are required in several reported protocols.52 The improved automation and control refinements come at a cost of increased complexity.

Intensive implementation efforts and strong leadership can overcome some issues, as demonstrated by the Yale experience, but 2 other strategies have commonly demonstrated success. First, focused expert glycemic management teams that directly oversee many aspects of the IIP, or even assume direct management roles, can be quite effective,25 especially during early implementation.

Automation of calculations and computerization is another method to consider, and many recent reports involve applications of web‐based or other computerized models.27, 40, 42, 43, 46, 48, 63 Comparisons of computerized versus manual methods, computerized versus column methods, and a computerized nomogram‐versus‐chart method are now available.42, 45, 46, 48, 62 Computerized protocols show significant promise and would be expected to reduce dosing errors. These instruments generally present the nurse with a specific infusion rate after each monitored glucose and then recommend an interval for the next glucose determination. In direct comparisons, they tend to perform as well or often better than conventional methods. For example, in the most recent randomized trial comparing the Glucommander to a paper‐based column method in an ICU, Newton and colleagues found the computerized method reached the goal more rapidly and reached a lower mean glucose without increasing the rate of severe hypoglycemia (< 40 mg/dL).62 Computerized methods also facilitate data collection for analytical purposes.

Computerized systems (both commercial and home‐grown versions) are becoming more common and appear to hold significant benefits as long as they are backed by a validated algorithm. Whereas many institutions are not yet in a position to integrate such protocols into their standard or electronic record systems, we expect the trend for increased implementation to continue.

Standardizing Insulin Infusion Preparations and Priming New Tubing

Varied concentrations or types of insulin for insulin infusions can lead to serious errors. Insulin infusions should generally be centrally prepared with a standard concentration of regular insulin in the pharmacy (usually 1 unit/cc), and the infusion concentration should be included in your infusion order set. Insulin binding to IV tubing can lead to false elevation of insulin requirements, potentially followed by serious hypoglycemia. When nurses change IV tubing or initially set up an insulin drip, education/emnstructions on priming new tubing with a small amount of insulin infusion to saturate the binding to the polyvinyl chloride tubing should be incorporated into their routine. Although 50 mL has often been recommended for priming, a recent study64 found that 20 mL of insulin infusion is enough to reach the saturation point.

Avoiding Over‐Reliance on the Insulin Protocol

Nurse‐driven insulin infusion protocols automate frequent insulin adjustment and reduce unnecessary calls to the physician. Although this is generally a decided advantage, the care team can be lulled into a sense of false security by the presence of orders that allow for such adjustment. Increasing the rate of an insulin infusion without thoughtful attention to factors that may be playing a role in this increased requirement (such as developing sepsis, other medical decompensation, steroid boluses, or an increase in carbohydrate intake) can have serious consequences. By the same token, an unanticipated rapid decrease in insulin requirement should lead to a reassessment of the infusion, and an inquiry about cessation of glucocorticoid therapy or nutrition. Rarely, a pharmacy or nursing error may induce a pseudo‐change in insulin requirements. The protocol should lead the nurse to seek advice and alert the physician to review potential causes of dramatic changes in insulin requirements, rather than simply adjusting insulin or nutrition to correct the present abnormal value.

Interruption of the Insulin Infusion

Interruption of insulin infusions may occur for many reasons, either intended or unintended. At times, the doctors or nurses may temporarily stop the protocol to allow for delivery of blood products or medications when IV access is limited. Infusions may mistakenly not be restarted, or deliberate discontinuation may not be adequately communicated, potentially leading to worsening hyperglycemia or even the development of ketoacidosis, and other adverse clinical outcomes. Therefore, the algorithm should have clear orders for the nurses to contact the ordering physician if the infusion is stopped for any reason, other than protocol‐driven cessation due to falling blood glucose concentrations.

Interruption of Nutrition, Field Trips, and Communication

Insulin infusion commonly provides both basal and nutritional insulin requirements. Interruptions in nutritional intake are extremely common in the inpatient setting, with a potential to cause serious hypoglycemia. Feeding tubes are often pulled out without warning; enteral nutrition may also need be halted if high gastric residuals are noted or during certain diagnostic tests. At times, IV carbohydrate sources (dextrose, partial parenteral nutrition, total parenteral nutrition) may be interrupted as well. In some cases, field trips out of the critical care units to the operating room, imaging studies, or other hospital locations add another layer of challenges to managing the IIP. Staff in these various areas may not be familiar with the IIP or monitoring standards and techniques, and potentially may not even be aware that the patient is on an insulin infusion. It is therefore crucial to anticipate these pitfalls and develop effective institutional procedures for addressing them. For example, many institutions use D10 solution to replace the carbohydrate calories that are lost when tube feedings have to be interrupted in a gram‐per‐gram fashion. Patients should be clearly identified as being on an insulin infusion. The requirement for consistent glucose monitoring, hypoglycemia recognition and treatment, and insulin infusion adjustment requires either critical care nurse care of the patient on the field trip, or training in the same skills in areas such as endoscopy, interventional radiology, and operating rooms including the preoperative and postoperative care units. In any case, all services should be involved in crafting solutions that will ensure a consistent approach to glycemic control as the patient travels off‐unit. Monitoring and treatment equipment needs to be readily available in all sites, and hypoglycemia protocols need to be distributed and supported in all areas.

Preventing and Treating Hypoglycemia

Some hypoglycemia will occur with infusion protocols, no matter how carefully a protocol is crafted and how well it is administered. Hypoglycemia protocols should therefore be incorporated directly into an infusion order set. Treatment of hypoglycemic events with a full 50 mL of D50 solution is equivalent to 25 g of carbohydrate, which will raise glucose levels in the average patient by 125 mg/dL. Many institutions discourage the overcorrection of hypoglycemic events by encouraging giving lesser aliquots of D50 based on the degree of hypoglycemia.

Preventing hypoglycemia by recognizing hypoglycemia risk factors, proper monitoring, and anticipating reductions in insulin requirements from decreasing severity of illness, nutritional intake, or steroid dosing can also reduce the frequency of hypoglycemic events.

Steroids

Steroid boluses are commonly an integral part of regimens targeting a variety of conditions, such as transplant rejection, reactive airways disease, certain infections, cancer, and a variety of autoimmune disorders. This can lead to glycemic excursions and rapidly varying insulin requirements. Educational efforts and treatment regimens should address the disproportionate impact that steroids have on postprandial glycemic excursions. To minimize the glycemic impact of glucocorticoid therapy, a team should investigate promoting the use of steroid infusions in situations when a bolus is not absolutely necessary.

Dealing with the Eating Patient and Other Sources of Carbohydrate‐Induced Glycemic Excursions

Glucose levels can be difficult to control in patients who are eating while on insulin infusion, because the infusion chases the glycemic excursions through frequent adjustments, often with a late overshoot and inappropriate reduction in dose. We instead recommend providing bolus nutritional insulin to cover the expected glycemic excursion caused by carbohydrate ingestion. Carbohydrate counting and using a unit of insulin for each 10‐15 g of carbohydrate consumed can smooth out the rapid fluctuations in glucose. Guidance for this should be incorporated into the order set.

Transition Off of Insulin Infusion

Rational strategies for dealing with this transition are covered in detail elsewhere in this supplement.65 Guidance for managing this transition should be integrated into your insulin infusion and SC order sets. The transition to SC insulin may represent a separate order set but is sometimes best integrated into the IV insulin infusion order set itself.

IIPs Outside of the Critical Care Setting

IIPs are most commonly used in the critical care setting. In some institutions, IV insulin protocols are safely and effectively employed outside the ICU. Obviously, the number of nurses and other personnel who must be familiar with such protocols is much higher outside the ICU, and protocol errors are therefore likely to be somewhat higher. In addition the nurse‐per‐patient ratio is usually lower outside of the critical care setting. As a result, suggestions for safe implementation of insulin infusion regimens outside of the critical care setting include:

• Choose an infusion protocol with a higher glycemic target.

• Limit the medical and surgical units where this expertise will be developed.

• Consider simplified infusion protocols but stay consistent with format.

• Automated or computerized assistance of calculations may reduce human error and nursing burden.

Monitoring, Recording, and Analyzing Glycemic Control Data

Once the IIP is implemented, it is critical that the impact on glycemic control, hypoglycemia, insulin use, and other factors be analyzed and used for improving the IIP and care delivery. Frequent monitoring of glucose levels is necessary for the safe infusion of insulin. Guidance for how often the monitoring is required must be explicit and included in the infusion order set. Intermittent auditing for compliance with the frequency of glucose testing and appropriate dose selection is good practice. Attention should be paid to how the glucose level is obtained, recorded, and made available to the health care team in your institution. All glucose readings should be recorded electronically for ongoing analyses and retrieval, and ideally, this could be done in an automated or single‐step method. Try to eliminate duplication of effort, such as asking the nurse to record the glucose level and their reaction to it on paper and again in an electronic format. Your team should also provide guidance about the potential problems of using point‐of‐care glucose testing in settings with hypotension, sepsis, pressor use, and other conditions that may impair the accuracy of capillary glucose readings.

Reports on the time to reach the glycemic target, glycemic control while on infusion, and the incidence of hypoglycemia should be reviewed by the multidisciplinary steering committee. The Society of Hospital Medicine Glycemic Control Task Force recommends analysis by patient day and by patient stay (or insulin infusion run) as preferred methodologies for analysis of glycemic control and hypoglycemia rates over the method of using each individual glucose reading as the unit of analysis. (The latter tends to under‐value the frequency of hypoglycemia.) Detailed practical recommendations for analyzing and summarizing glycemic control data are available elsewhere in this supplement.18 These data should drive decisions on modification of glycemic targets and the protocol structure. Patients meeting prespecified criteria should be referred to the improvement team for review. For example, patients who experience any glucose readings of 40 mg/dL, or who take more than 12 hours to reach the upper limit glycemic target should be referred to the team for a case review.

Assessing Adherence to the Protocol and Ease‐of‐Use Issues

Focused audits in the pilot and early implementation phases should look for nonadherence to the protocol. Deviations should be evaluated according to the patterns identified. For example, variation in application in some cases is specific for an individual and in others is characteristic of a specific group or the whole. Accordingly, this may point to gaps in education or attitudes about the importance of this endeavor. Front‐line staff may deviate from the protocol because they find it ineffective, unsafe, or impractical for certain situations or specific patients. Many IIPs are the subject of nursing errors related to the knowledge and acceptance of the nurse but also the complexity of the protocol. Appropriate modifications to the protocol based on these cases can frequently improve the ease of use and effectiveness of the protocol. The ongoing review process should identify issues that must be addressed with permanent solutions rather than accepting frequent individual alterations to meet goals. Revisions require supplementary education and rapid and wide dissemination. Although educational efforts and monitoring are often most intense in the early implementation phase, periodic retraining should continue to achieve optimal results and safety. Educational tools must consider nursing time commitments and will often include an interactive web‐based module that gives more flexibility for trainers and clinical nurses alike.

Building The Team

Implementing a medical centerwide standardized insulin infusion order set with supporting policies, protocols, monitoring standards, and the requisite educational programs is a major task for any hospital. This is not a simple maneuver involving only 1 or 2 interested individuals and requires much more than selecting a published protocol and disseminating it to various patient care units. Instead, to manage the full spectrum of diabetes programs and protocols, an institution must convene a multidisciplinary steering committee or task force. This should include representation from nursing, nursing administration, pharmacy, nutrition services, and the quality improvement department. Physicians should include hospitalists, intensivists, and endocrinologists but may also involve anesthesiologists and surgeons, as applicable. At times, additional members may need to be recruited according to project needs.

Identifying the Stakeholders and Current Practices

In developing or improving currently utilized IIPs, the multidisciplinary committee would benefit from careful background work before moving forward. First, administrative and institutional support must be secured to endorse uniform standards for insulin infusions, and to provide the important infrastructure needed to facilitate the work involved. Clinical and administrative stakeholders from the key departments then need to be identified.

All insulin infusion orders and policies/procedures presently used in the institution should be identified and examined. The developers and/or users of these order sets should be engaged in a dialogue and encouraged to share their experiences regarding their current practice and the attendant work flow, glucose monitoring, and data collection and reporting. Immediate concerns should be clearly addressed. Measurement systems for glycemic control, hypoglycemia, and insulin use patterns should assess current practice and the impact of subsequent modifications of the protocol or initiation of new protocols. We recommend using glucometrics consistent with those endorsed by the Society of Hospital Medicine (SHM) Glycemic Control Task Force elsewhere in this supplement.18

Addressing the Burden of Change

Through this process, the committee will uncover barriers, dysfunctional and inconsistent practices, and individuals who will pose challenges. Identifying these issues should not discourage the team, but rather it should guide the interventional strategies, and help build consensus that change may be required. There must be caution not to exclude significant individuals simply because they resist changes. Indeed, if they are included and have the opportunity to contribute to the process, success is much more likely.

It is important for process leaders to understand the implications of what is proposed, particularly for nursing services.19 For example, it has been shown that IIPs require about 5 minutes per patient per hour for glucose monitoring and dose adjustments.20 Acknowledging and attempting to address this burden proactively (often well over 2 hours per day) can gain staff acceptance more effectively than a laissez faire approach. In this regard, some effort should be invested in nursing education of the benefits of tight glycemic management on critical care outcomes. The difficult‐to‐quantify work involved when patients' blood glucose is not well controlled (eg, paging physicians for stat insulin orders) should also be part of this discussion.

Identifying and Addressing Barriers to IIP implementation

There are numerous potential barriers to implementation of IIPs. Table 1 identifies some of the most frequent ones along with potential strategies or solutions. Very common barriers include skepticism surrounding the benefits of tight glycemic control, fear of hypoglycemia, and difficulty agreeing on glycemic targets.

Whereas the national guidelines call for a glycemic target in critical care areas with an upper limit of 110 mg/dL, there is room for debate, and tailoring of the glycemic target to fit individual patient circumstances is often advisable. Starting with a less aggressive glycemic target can be good politics (and perhaps good medicine as well). Once the higher glycemic targets are achieved safely, it can pave the way for a more aggressive approach.

Fear of hypoglycemia is one of the most potent barriers to intensive insulin infusion implementation. Because hyperglycemia is such a common condition in critical care units, nursing and physician staff may have developed a skewed view of the definition of hypoglycemia, at times fearing for their patient when the glucose values reach a level of 100 mg/dL or so. Polling the nurses on what they think their own fasting glucose levels are, and then actually measuring them can be an effective strategy (the nurses may be surprised that the patient's scary 90 mg/dL reading is higher than their own). It should also be emphasized that properly designed and implemented protocols may actually decrease the incidence of hypoglycemia when compared to standard care which may involve individually and sometimes improperly adjusted intravenous (IV) insulin (discussed further below).