HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Respond to This Article View Responses to This Article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Malesker, M. A. Articles by Hilleman, D. E. Search for Related Content PubMed PubMed Citation Articles by Malesker, M. A. Articles by Hilleman, D. E.

An Efficiency Evaluation of Protocols for Tight Glycemic Control in Intensive Care Units

Corresponding author: Mark A. Malesker, PharmD, Creighton University Medical Center, 2500 California Plaza, Omaha, NE 68178 (e-mail: malesker{at}creighton.edu).

	Abstract

Top Abstract Methods Results Discussion Conclusion and Recommendations References

 
Background The efficiency of protocols for tight glycemic control is uncertain despite their adoption in hospitals.

Objectives To evaluate the efficiency of protocols for tight glycemic control used in intensive care units.

Methods Three separate studies were performed: (1) a third-party observer used a stopwatch to do a time-motion analysis of patients being treated with a protocol for tight glycemic control in 3 intensive care units, (2) charts were retrospectively reviewed to determine the frequency of deviations from the protocol, and (3) a survey assessing satisfaction with and knowledge of the protocol was administered to full-time nurses.

Results Time-motion data were collected for 454 blood glucose determinations from 38 patients cared for by 47 nurses. Mean elapsed times from blood glucose result to therapeutic action were 2.24 (SD, 1.67) minutes for hypoglycemia and 10.65 (SD, 3.24) minutes for hyperglycemia. Mean elapsed time to initiate an insulin infusion was 32.56 (SD, 12.83) minutes. Chart review revealed 734 deviations from the protocol in 75 patients; 57% (n = 418) were deviations from scheduled times for blood glucose measurements. The mean number of deviations was approximately 9 per patient. Of 60 nurses who responded to the workload survey, 42 (70%) indicated that the protocol increased their workload; frequency of blood glucose determinations was the most common reason.

Conclusions Nurses spend substantial time administering protocols for tight glycemic control, and considerable numbers of deviations occur during that process. Further educational efforts and ongoing assessment of the impact of such protocols are needed.

Despite the adoption of glycemic targets in hospitals and the increasing number of hospitals that are adopting protocols for tight glycemic control, questions about the implementation and efficiency of these protocols remain unanswered. Glycemic control in the inpatient setting is often suboptimal and is viewed as a secondary concern to other patient care issues.¹² Several different tight glycemic control protocols (TGCPs) have been used in published studies,^5–8 from relatively simple fixed-dose glucose-insulin infusions to more complicated protocols that require frequent testing of blood glucose levels. It is currently unknown if one of these protocols is superior to the others with regard to the effect on clinical outcomes, potential to require increased nursing time, greater expenditures for supplies and equipment, and higher rates of medication errors.

The purpose of our study was to evaluate the relative efficiency of TGCPs used in the intensive care units (ICUs) of hospitals in the Omaha, Nebraska, metropolitan area. Study objectives were 3-fold: (1) to conduct a time-motion analysis to determine the amount of time required by nursing staff to administer the TGCP, (2) to perform a retrospective chart review to evaluate the frequency of deviations from the protocol in patients being treated with a TGCP, and (3) to assess nursing satisfaction with the TGCP and the extent of their knowledge about the TGCP by using a survey with both open-ended and multiple-choice questions.

Even a single elevated glucose value at admission is associated with adverse patient outcomes.

 

	Methods

Top Abstract Methods Results Discussion Conclusion and Recommendations References

 
Time-Motion Study
A time-motion analysis of patients being treated with a TGCP in 3 hospital ICUs was performed by third-party observers using a stopwatch. Elapsed time was defined as the interval between the time a nurse picked up the point-of-care glucometer and the time activities related to that blood glucose result were completed. Elapsed time was divided into 3 intervals: (1) time required to obtain a blood glucose result by using a handheld point-of-care glucometer, (2) time required to take appropriate therapeutic action based on a blood glucose result, and (3) time to note the blood glucose result (manually or electronically) in the patient’s chart.

The time to take appropriate therapeutic action was further stratified into 4 intervals: (1) interval when the blood glucose result was high enough to trigger the initiation of an insulin infusion in patients not receiving an insulin infusion, (2) interval when the blood glucose result exceeded the upper threshold of normal and required an increase in the dosage of a patient’s insulin infusion, (3) interval when the blood glucose result was in the therapeutic range and no change in the insulin infusion rate was needed or no change in therapy was indicated, and (4) interval when the blood glucose result indicated hypoglycemia.

Chart Review of TGCP Deviations
Charts of consecutive patients being treated with a TGCP at an ICU in a fourth hospital (level I trauma center) that was not participating in the time-motion study were reviewed retrospectively. Protocol deviations were categorized as follows: (1) obtaining blood glucose determinations at a time different than the time indicated in the TGCP, (2) administering an insulin dose different from that indicated in the TGCP, and (3) failure to follow any other instructions in the TGCP. Deviations for improper timing were defined as blood glucose determinations obtained more than 15 minutes before or after the appropriately scheduled time. Deviations in administering an accurate insulin dose were defined as administration of any dose other than that calculated or indicated by the TGCP. Deviations in following instructions in the TGCP algorithms included any deviation from instructions in the correct algorithm (Figure 1).

View larger version (46K): [in this window] [in a new window]   Figure 1 Protocol for tight glycemic control

View larger version (28K): [in this window] [in a new window]   Figure 2 Nursing Satisfaction Survey

	Results

Top Abstract Methods Results Discussion Conclusion and Recommendations References

View larger version (27K): [in this window] [in a new window]   Figure 3 Summary of elapsed times in the time-motion study.

Elapsed time for starting an insulin infusion appeared excessive and included contacting the physician, checking the policy, and waiting for the infusion to be mixed and sent.

 

The mean time from meter pickup to noting a blood glucose result that indicated hypoglycemia in the patient’s chart or electronic medical record was 32.65 (SD, 7.68) minutes. Included in this elapsed time was stopping the insulin infusion, administering glucose, and performing other patient care activities. The greatest part of this time was 19.46 (SD, 8.14) minutes spent in the patients’ rooms providing care after stopping the insulin infusion. The elapsed time between obtaining a result that indicated hypoglycemia and adjusting or stopping the insulin infusion was only 2.24 (SD, 1.67) minutes.

The mean elapsed time from meter pickup to noting a blood glucose result that was out of the upper target range in the chart/record of a patient already receiving an insulin infusion was 29.67 minutes. Important components of this elapsed time included 10.65 minutes spent reviewing the chart and the algorithm for insulin infusion in the protocol (and in some instances calling physicians to report the change), 5.40 minutes spent performing routine activities related to patient care, and 8.42 minutes from the time of leaving the patient’s room until the result was actually noted in the chart.

About 9 protocol deviations (time, dose, algorithm) occurred per patient.

 

The elapsed time from picking up the meter to noting a result indicating hyperglycemia in the chart of a patient requiring initiation of an insulin infusion was 51.57 minutes. The largest percentage of this time was spent with the nurse contacting the physician, checking the chart and/or TGCP, and waiting for the insulin infusion to be mixed and sent from the pharmacy to the ICU (32.56 minutes).

TGCP Deviations
A total of 75 consecutive patients were treated with a TGCP at a fourth ICU that did not participate in the time-motion study. The protocol in use at this university-affiliated institution’s ICU had been adapted from the protocol published by Trence et al.¹³ This ICU has 25 beds, and the TGCP had been in place for approximately 4 months. An intensivist medical team is available 24 hours a day, but mandatory consultations are not required. Patients admitted to this ICU have a wide variety of medical, surgical, and trauma diagnoses. Characteristics of this population included a mean age of 63 years, 68% men, 57% surgical cases, and 43% medical cases. Cardiac disease was the most common admitting diagnosis; next were surgery, trauma, and neurological disorders. A history of diabetes was documented in 41% of the patients. After initiation of the TGCP, the patients were followed up for a mean of 20 hours. During this time, 981 blood glucose levels were ordered (range, 9–20 per patient).

A total of 734 deviations from the protocol were documented: 57% (n=418) were related to deviations in times for scheduled blood glucose readings, 38% (n = 279) were for incorrect insulin doses, and 5% (n = 37) were related to improper execution of the individual algorithm instructions. The mean number of deviations was slightly more than 9 per patient (range, 1–23 per patient). Most patients had deviations in more than a single category. A total of 47% had deviations related to both time and dose, and 39% had deviations in all categories (time, dose, and algorithm). The timing of the blood glucose measurement was the only deviation in 13% of patients, and 1% had a deviation in dose only. Approximately half of the 75 patients achieved their target glycemic control during the 20-hour observation period.

Approximately 75% of all measured blood glucose levels were associated with a protocol deviation. Because of the retrospective nature of this part of the study, we could not determine if any of the deviations resulted in clinically important adverse outcomes for patients. After these results were obtained, hospital administrators implemented a quality improvement initiative and an educational campaign about the use of the TGCP in the ICU at this institution.

Forty-two respondents (70%) reported that tight glycemic control increased their workload.

 

Nursing Survey
A total of 220 full-time nurses at the 4 ICUs were sent a copy of the survey. A total of 75 surveys (34%) were completed and returned (Table 2). Among the respondents, 45 (60%) had been in practice more than 5 years, and 31 (41%) had practiced more than 10 years. A total of 39 respondents (52%) had used a TGCP in more than 20 patients. A total of 55 respondents (73%) agreed that the protocol was effective in controlling hyperglycemia, and 44 (59%) agreed that the protocol was not associated with excess risk of hypoglycemia. A total of 42 respondents (56%) agreed that the protocol was easy to administer; 27 (36%) disagreed and indicated that the TGCP was difficult to administer.

When nurses were asked about the optimal blood glucose range in the critical care setting, 22 different blood glucose ranges were specified. The most common response, offered by 35% of respondents, was 70 to 150 mg/dL (to convert to millimoles per liter, multiply by 0.0555). A total of 56% of the respondents stated that the minimum blood glucose goal should be 70 mg/dL, and 74% indicated that the maximal blood glucose goal should be between 120 and 150 mg/dL.

	Discussion

Top Abstract Methods Results Discussion Conclusion and Recommendations References

 
Studies^5–10 published to date indicate that tight glycemic control improves outcomes and reduces the cost of care in critically ill patients. As a result, a growing number of hospitals are implementing TGCPs in their ICUs. Relatively little has been published on how implementing a TGCP affects nurses’ workloads. The number of patients receiving insulin infusions is substantially higher in ICUs where a TGCP has been implemented than in other ICUs. This increase in the number of patients receiving insulin infusions and the requirement of frequent (every 1–2 hours) blood glucose monitoring could increase nurses’ workload. In our study, times required to carry out various phases of a TGCP were recorded, nurses were surveyed to assess attitudes and experience with a TGCP, and charts were reviewed retrospectively to assess the frequency and types of protocol deviations that occurred during administration of a TGCP. This study is the first comprehensive evaluation designed to assess the efficiency of using TGCPs in ICUs.

Elapsed times (Figure 3) to achieve various milestones in the TGCP observed in the ICUs in our study seem quite long for what may have been assumed to be relatively simple tasks. The elapsed time from meter pickup to obtaining a blood glucose result from the glucometer was slightly more than 5 minutes. However, a number of activities occurred during this time, including meter startup, moving from the nurse’s station into a patient’s room, obtaining a blood sample from the patient, and placing the blood sample into the glucometer. In addition, once the blood sample was placed into the glucometer, nurses would often take care of the patient’s other care needs while waiting for the blood glucose result to appear on the readout of the meter. Similar observations were made during other intervals of elapsed time.

The time from obtaining a normal blood glucose value until the time that the result was actually noted in the medical record was slightly greater than 19 minutes. It obviously does not take this amount of time to perform this task, but nurses performed other activities during this time before actually entering the result in the medical record, often entering the result at the same time that other medical information was being noted in the chart. The elapsed time from obtaining a blood glucose reading that indicated hypoglycemia to the time of therapeutic intervention (ie, stopping the insulin infusion) was slightly greater than 2 minutes.

Elapsed times for certain milestones did seem excessive. For example, an elapsed time of almost 33 minutes to start an insulin infusion seems excessive. Most of the elapsed time was spent waiting for insulin infusions to be delivered from the pharmacy. Efforts to reduce the time required for the pharmacy to provide an insulin infusion seem warranted. One other elapsed time that seemed excessive was the time to increase an insulin infusion dose in response to a blood glucose reading indicating hyperglycemia. The elapsed time of almost 11 minutes after obtaining a result indicating hyperglycemia should be shorter. Most of the time spent during this time interval was used to review the medical chart and to calculate the new insulin infusion dose on the basis of the TGCP. Calculating insulin doses was listed as a cause of difficulty in administering a TGCP in the survey, so attempts to simplify dose calculations seem worthwhile.

Determining whether most of the times observed during the time-motion analysis should be considered within the accepted standard of practice is difficult. In the only other published study¹⁴ of times to achieve milestones during administration of a TGCP in an ICU, the mean time to obtain a blood glucose result and change the insulin infusion was only 4.7 minutes (range, 3.1–8.2). The clinical importance of this finding, which is a full 10 minutes shorter than our time to adjust an insulin infusion, is uncertain. In the study by Aragon,¹⁴ only 21 timed blood glucose observations were evaluated in 4 different ICUs. No other information was provided about these observations. It is not certain how many different nurses were involved in the measurements, how many patients were included in the analysis, how long a patient had been receiving an insulin infusion when samples were being collected, or at what time intervals (every 1 hour or every 4 hours) the samples were being collected. Isolated observations made in what could be construed as "staged" timing events certainly would produce results different from the elapsed times for samples being collected longitudinally during the routine course of patient care in a busy ICU, as in our study.

In our time-motion analysis, attempts were made to observe nurses during 12-hour blocks of time per patient, during which the blood samples for glucose testing were collected. This approach included the impact of shift change, nurses’ leaving the unit for breaks and meals, and the usual variations in the intensity of care that occur over a long period. Additional time-motion analyses performed in a fashion similar to the one we used must be done to determine the optimal times for achieving the various milestones in administering a TGCP in an ICU.

The survey results we collected on nurses’ attitudes about workload related to administration of a TGCP in an ICU were generally similar to the findings of Aragon.¹⁴ In our survey, which included 75 respondents, 74% of the nurses thought that their TGCP was effective in controlling hyperglycemia and 59% agreed that the TGCP was not associated with hypoglycemia, but only 56% agreed that the TGCP was easy to administer. Of the 60 nurses who responded to a question about the impact of the TGCP on nursing workload, 42 (70%) indicated that TGCP increased their workload. Of those 42, 24 indicated that the number of blood glucose determinations was the primary reason for the increased workload.

In the Aragon survey,¹⁴ in which nurses were asked questions that differed slightly from ours about the use of TGCP, 49 of 66 respondents (74%) indicated that the TGCP took too much time, was too much work, or was a waste of time. The data from the survey by Aragon also indicated that the frequency of blood glucose determinations was a major source of increased workload. In addition, the need to obtain blood via finger sticks was an obstacle. In the Aragon study, a substantial percentage of nurses indicated a desire to use arterial catheters or central venous catheters to collect blood rather than collecting blood via finger sticks. A total of 76% percent of the nurses in the Aragon study also indicated that they would be willing to dedicate an intravenous catheter to blood glucose measurements.

Much of the elapsed time in treating hyperglycemia was spent calculating new insulin infusion doses from the protocol.

 

Communication with physicians and difficulty using the TGCPs in patients receiving oral or intravenous feedings were some of the more common problems identified by our respondents. As previously mentioned, calculating insulin infusion doses also was a problem with the TGCPs. Problems with the insulin infusion rate were identified as the most common source of error with a TGCP. These obstacles could be lessened by improved education about TGCP administration and/or by altering or adjusting the TGCP.

Perhaps one of the most surprising findings of our efficiency evaluation was the high number of deviations from the protocol. Clearly, a mean of approximately 9 deviations per patient represents substandard care. Because the times for blood glucose determinations were outside the accepted range of ±15 minutes more than 75% of the time, perhaps the nurses simply did not appreciate the importance of the TGCP. Although the reasons for the excessive number of protocol deviations cannot be determined, the TGCP had been in place for only 4 months when the chart review took place. After the results of the chart review were disseminated, nursing administrators (in the unit where the deviations occurred) initiated an ongoing education and training program for nurses that focused on administration of the TGCP. The high rate of protocol deviations seems consistent with the finding that most survey respondents could not state the optimal blood glucose range used in the TGCP.

	Conclusion and Recommendations

Top Abstract Methods Results Discussion Conclusion and Recommendations References

 
Both the time-motion analysis and the responses to the nursing survey indicate that substantial nursing time is involved in the administration of a TGCP. In addition, a substantial number of protocol deviations occurred during administration of a TGCP. Additional educational efforts and refinements in TGCP administration are needed. Methods designed to simplify TGCPs and to streamline documentation of health information should be implemented. Ongoing assessment of the effect of TGCPs on nurses’ workloads also should be done.

FINANCIAL DISCLOSURES
This study was funded by a grant from LifeScan, Inc, Milpitas, California.

eLetters
Now that you’ve read the article, create or contribute to an online discussion about this topic using eLetters. Just visit www.ajcconline.org and click "Respond to This Article" in either the full-text or PDF view of the article.

To purchase electronic or print reprints, contact The InnoVision Group, 101 Columbia, Aliso Viejo, CA 92656. Phone, (800) 809-2273 or (949) 362-2050 (ext 532); fax, (949) 362-2049; e-mail, reprints{at}aacn.org.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion and Recommendations References

 

1. Wolfe RR, Herndon DN, Jahoor F, Miyoshi H, Wolfe M. Effect of severe burn injury on substrate cycling by glucose and fatty acids. N Engl J Med. 1987;317(7):403–408.[Abstract]
2. Shangraw RE, Jahoor F, Miyoshi H, et al. Differentiation between septic and postburn insulin resistance. Metabolism. 1989;38(10):983–989.[Medline]
3. Umpierrez GE, Isaacs SD, You X, Thaler LM, Kitabchi AE. Hyperglycemia: an independent marker of in-hospital mortality in patients with undiagnosed diabetes. J Clin Endocrinol Metab. 2002;87(3):978–982.[Abstract/Free Full Text]
4. Capes SE, Hunt D, Malmberg K, Gerstein HC. Stress hyper-glycemia and increased risk of death after myocardial infarction in patients with and without diabetes: a systematic overview. Lancet. 2000;355(9206):773–778.[Medline]
5. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med. 2001;345(19): 1359–1367.[Abstract/Free Full Text]
6. Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. AnnThorac Surg. 1999;67(2):352–362.[Abstract/Free Full Text]
7. Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H, Apstein CS. Tight glycemic control in diabetic coronary artery bypass graft patients improves perioperative outcomes and decreases recurrent ischemic events. Circulation. 2004;109 (12):1497–1502.[Abstract/Free Full Text]
8. Krinsley JS. Effect of an intensive glucose management protocol on the mortality of critically ill adult patients. Mayo Clin Proc. 2004;79(8):992–1000.[Abstract/Free Full Text]
9. Van den Berghe G, Wouters PJ, Kesteloot K, Hilleman DE. Analysis of healthcare resource utilization with intensive insulin therapy in critically ill patients. Crit Care Med. 2006;34(3):612–616.[Medline]
10. Krinsley JS, Jones R. Cost analysis of intensive glycemic control in critically ill patients. Chest. 2006;129(3):644–650.[Medline]
11. Garber AJ, Moghissi ES, Bransome ED Jr, et al. American College of Endocrinology position statement on inpatient diabetes and metabolic control. Endocr Pract. 2004;10(1):77–82.[Medline]
12. Levetan CS, Passaro M, Jablonski K, Kass M, Ratner RE. Unrecognized diabetes among hospitalized patients. Diabetes Care. 1998;21(2):246–249.[Abstract]
13. Trence DL, Kelly JL, Hirsch IB. The rationale and management of hyperglycemia for in-patients with cardiovascular disease: time for change. J Clin Endocrinol Metab. 2003;88(6):2430–2437.[Abstract/Free Full Text]
14. Aragon D. Evaluation of nursing work effort and perceptions about blood glucose testing in tight glycemic control. Am J Crit Care. 2006;15(4):370–377.[Abstract/Free Full Text]

This article has been cited by other articles:

				J. Blaha, P. Kopecky, M. Matias, R. Hovorka, J. Kunstyr, T. Kotulak, M. Lips, D. Rubes, M. Stritesky, J. Lindner, et al. Comparison of Three Protocols for Tight Glycemic Control in Cardiac Surgery Patients Diabetes Care, May 1, 2009; 32(5): 757 - 761. [Abstract] [Full Text] [PDF]

				R. S. Wiener, D. C. Wiener, and R. J. Larson Benefits and Risks of Tight Glucose Control in Critically Ill Adults: A Meta-analysis JAMA, August 27, 2008; 300(8): 933 - 944. [Abstract] [Full Text] [PDF]

eLetters:

This Article Abstract Full Text (PDF) Respond to This Article View Responses to This Article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Malesker, M. A. Articles by Hilleman, D. E. Search for Related Content PubMed PubMed Citation Articles by Malesker, M. A. Articles by Hilleman, D. E.