Strict versus liberal insulin therapy in the cardiac surgery patient: An evidence-based practice development, implementation and evaluation project

Accepted Manuscript Strict versus liberal insulin therapy in the cardiac surgery patient: An evidence-based practice development, implementation and evaluation project

Jacqueline M. Gordon, Lori S. Lauver, Harleah G. Buck PII: DOI: Reference:

S0897-1897(16)30255-5 doi:10.1016/j.apnr.2017.11.013 YAPNR 50996

To appear in:

Applied Nursing Research

Received date: Revised date: Accepted date:

6 October 2016 20 October 2017 2 November 2017

Please cite this article as: Jacqueline M. Gordon, Lori S. Lauver, Harleah G. Buck , Strict versus liberal insulin therapy in the cardiac surgery patient: An evidence-based practice development, implementation and evaluation project. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Yapnr(2017), doi:10.1016/j.apnr.2017.11.013

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Title: Strict versus Liberal Insulin Therapy in the Cardiac Surgery Patient: An Evidence-Based Practice Development, Implementation and Evaluation Project Author names and affiliations:

IP

T

Jacqueline M. Gordon, DNP, RN, CCNS, CCRNa [email protected] (corresponding author)

CR

Lori S. Lauver, PhD, RN, CNP, CNEa [email protected]

US

Harleah G. Buck, PhD, RN, FPCN, FAANa1 [email protected] a

AN

The Pennsylvania State University College of Nursing 315D Nursing Sciences Building University Park, PA, 16802, USA

AC

CE

PT

ED

Present address: University of South Florida 12901 Bruce B. Downs Blvd. MDC22 Tampa, FL 33612-4766, USA

M

1

ACCEPTED MANUSCRIPT 2 Abstract Background Hyperglycemia post-cardiac surgery is associated with poor clinical outcomes. Recent studies suggest maintaining liberal glycemic control (<180mg/dL) using a continuous insulin infusion

T

(CII) versus strict control achieves optimal outcomes and prevents hypoglycemia.

IP

Purpose

CR

To develop, implement and evaluate a nurse managed liberal CII protocol. Methods

US

Retrospective review of 144 strict CII patient records and 147 liberal CII patient records.

AN

Results

Mean blood glucose was 159.8 mg/dL (liberal CII) compared to 143.3 mg/dL (strict CII) (p =

M

<.001). No surgical site infections occurred in either group. Mean ICU length of stay was 4.5

ED

days (liberal) versus 4.4 days (strict) (p = 0.74). Two 30-day mortalities occurred for the liberal cohort compared to no deaths in the strict group (p = 0.49). Hypoglycemia incidence within 24

PT

hours after surgery was 0.1% (liberal) compared to 0.3% (strict) compared to (p = 0.16).

CE

Conclusion

Use of a nurse managed liberal CII resulted in similar outcomes with fewer incidents of

AC

hypoglycemia. Keywords:

Glycemic control, cardiac surgery, insulin, evidence-based practice

ACCEPTED MANUSCRIPT 3 Introduction Nearly 500,000 cardiac surgeries are performed in the United States each year (Centers for Disease Control and Prevention [CDC], 2010). Postoperative hyperglycemia can potentially occur in any patient undergoing cardiac surgery whether there is a preoperative diagnosis of

T

diabetes or not (Knapik et al., 2009). Factors that contribute to this situational hyperglycemia in

IP

cardiac surgery patients include the use of pharmacological agents, disturbances in blood glucose

CR

homeostasis due to insulin resistance, and failure of beta cell function caused by the systemic inflammatory response syndrome after cardiopulmonary bypass (CPB) (Ascione, Rogers,

US

Rajakaruna, & Angelini, 2008).

AN

Postoperative hyperglycemia poses an increased risk for poor outcomes, such as surgical site infections, increased length of stay, and mortality following cardiac surgery (Furnary & Wu,

M

2006). Surgical site infections and increased lengths of stay result in escalated healthcare costs

ED

and potential readmissions (Greco et al., 2015). Postoperative hypoglycemia also poses an increased risk for poor outcomes in this population. Hypoglycemia (<70mg/dL) occurred in 21%

PT

of patients receiving a strict insulin infusion to maintain blood glucose 80-126mg/dL and was

CE

significantly associated with mortality (D’Ancona et al., 2011). In addition, hypoglycemia is independently associated with other complications such as increased risk for respiratory

2011).

AC

complications and prolonged hospital and intensive care unit lengths of stay (Stamou et al.,

The use of continuous insulin infusions (CIIs) is standard of care following cardiac surgery to prevent hyperglycemia without inducing hypoglycemia (Reddy, Duggar, & Butterworth, 2014). Ideal target blood glucose levels using CIIs have been debated in the literature; however, recent studies have shown similar patient outcomes using both strict and

ACCEPTED MANUSCRIPT 4 liberal blood glucose targets with the advantage of liberal targets decreasing the incidence of hypoglycemia (Giakoumidakis et al., 2013; McDonnell et al., 2013; Lazar et al., 2011). The Society of Thoracic Surgeons (STS) published guidelines in 2009 recommending that cardiac surgery patients with and without the diagnosis of diabetes and elevated serum glucose receive a

T

CII to maintain liberal serum glucose less than 180 mg/dL for the duration of their intensive care

IP

(Lazar et al., 2009).

CR

The Surgical Care Improvement Project (SCIP) is a national quality partnership of organizations interested in improving surgical care by significantly reducing surgical

US

complications (The Joint Commission [TJC], 2014). SCIP recognized cardiac glycemic control

AN

as a priority quality measure beginning in 2008; however, in January 2015, the measure was indefinitely suspended and then officially removed in October 2015 because it did “not reflect

M

current [STS] clinical guidelines, and there [were] concerns that it may adversely affect the way

ED

clinicians and hospitals provide care” (The Joint Commission [TJC], 2015, p. 1). The conflicting evidence regarding optimal target glucose range for cardiac surgery patients with and

PT

without diabetes coupled with inconsistent national and organizational directives poses a

CE

challenge for the interprofessional team charged with identifying the target glucose to achieve using a CII for all cardiac surgery patients. The presence of a strict CII protocol in the clinical

AC

setting in the face of guidelines recommending liberal CII protocols was the impetus for this project. Therefore, the purpose of this paper is to describe the development, implementation and evaluation of a liberal CII protocol and to report outcomes of surgical site infection (SSI), intensive care unit (ICU) length of stay (LOS), 30-day mortality, and incidence of hypoglycemia pre and post CII protocol change. Protocol Development and Implementation

ACCEPTED MANUSCRIPT 5 The Stevens ACE Star Model of Knowledge Transformation© provided the framework which guided the liberal CII protocol development and the overarching evidence-based practice project (Stevens, 2012). The model consists of five stages: discovery research, evidence summary, translation to guidelines, practice integration, and process, outcome evaluation.

T

Primary resources retrieved during the literature review on the topic of glycemic control and the

IP

postoperative cardiac surgery patient comprised the “discovery stage” concept of the model. The

CR

“evidence summary involved using retrieved primary sources to synthesize literature regarding glycemic control in the cardiac surgery patient. The concepts uncovered during the evidence

US

summary were integrated into an evidence based continuous insulin infusion protocol with a goal

AN

of avoiding hypoglycemia during the “translation to guidelines” stage of the model. The “practice integration” stage involved collaborating with institutional committees and

M

stakeholders to implement the continuous insulin infusion protocols in the intensive care unit.

ED

Finally, planned evaluation of outcomes following implementation of the evidence-based liberal CII encompasses the final stage of the model, “process and outcome evaluation.”

PT

First, an evaluation of current practice in the project setting was conducted. Current

CE

practice was a strict CII protocol targeted at maintaining blood glucose values 100-140 mg/dL immediately postoperatively. However, a sampling of medical records (n=10) resulted in

AC

identification of patterns of non-adherence to the strict CII protocol. When nurses in the practice setting were informally queried the authors received information indicating that 1) the frequency of titrations required 2) complexity of the strict CII protocol and 3) perceived potential for hypoglycemia incidence led to non-adherence. For example, bedside calculations and titrations were required potentially every 15 minutes for the strict CII protocol. The strict CII protocol was also in paper format, taped to the computer located in the patient’s room.

ACCEPTED MANUSCRIPT 6 Next, revising existing clinical practice necessitates collaboration among the interprofessional team. Content experts from nursing, endocrinology, pharmacy, cardiac surgery, and critical care developed the new liberal CII protocol under the leadership of the critical care clinical nurse specialist (CCNS). The CCNS and critical care pharmacist navigated

T

the protocol approval process with institutional committees to assure safety and efficacy. An

CR

liberal CII protocol to be available in the electronic health record.

IP

electronic protocol was created in collaboration with pharmacy informatics that enabled the

The limitations of the previous strict protocol were taken into consideration during the

US

development phase of the liberal CII protocol. The newly implemented liberal CII protocol was

AN

incorporated into the electronic medication administration record to improve accessibility to the nurses. The titration frequency was decreased from a potential of every 15 minutes to hourly.

M

The liberal CII protocol required point of care (POC) blood glucose monitoring hourly until the

ED

target range was reached for three subsequent hours, then monitoring decreased to every two hours. These changes increased usability of the protocol to nurses, streamlined administration of

PT

the protocol, and decreased nursing workload. Finally, the liberal CII protocol included

CE

guidance for physicians and advanced practice clinicians regarding glycemic control when patients resumed oral intake. This enhancement assisted with transitions of care when patients

AC

began recovering following surgery. Educational modules were developed by the CCNS and pharmacist to train clinicians on the new protocol and the importance of the practice change to a liberal CII protocol. All clinicians (medical, nursing, pharmacy) completed the training. The liberal CII protocol was implemented in March 2015. A pharmacist collaborated with physicians and advanced practice clinicians to transition existing patients from the strict CII protocol to the liberal CII protocol

ACCEPTED MANUSCRIPT 7 safely on the day of implementation. The CCNS facilitated implementation of the new practice with nurses, physicians and advanced practice clinicians. Evaluation Methods Design and sample participants

T

Following protocol implementation, a retrospective electronic medical record analysis

IP

was performed to evaluate patients receiving the strict and liberal CII protocols. The project

CR

setting was a 551- bed academic medical center in central Pennsylvania that provides comprehensive cardiovascular services including cardiac surgery, advanced heart failure

US

management, ventricular assist device implantation, and extracorporeal membrane oxygenation.

AN

All patients were admitted to a 30-bed heart and vascular critical care unit that provides care to patients immediately following cardiac surgery. On average, 300 cardiac surgeries are

M

performed each year. Patient records were included if the patient was more than 18 years old,

ED

with or without a diagnosis of diabetes and had undergone cardiac surgery including coronary artery bypass grafting (CABG), valve procedures, or a combination of CABG and valve

PT

procedures. Patients undergoing emergency cardiac surgery, cardiac transplantation, ventricular

CE

assist device implantation, and those who died in the operating room were excluded as this

2014).

AC

patient population has historically been excluded in glycemic reporting requirements (TJC,

Data collection

Following institutional review board approval, Business Objects InfoView for Systems, Applications, and Products in Data Processing (SAP)® was used to retrieve records of cardiac surgery patients who received the strict CII from June 1, 2014 to December 31, 2014 and patients who received the liberal CII from March 1, 2015 to September 30, 2015. Queries of the

ACCEPTED MANUSCRIPT 8 electronic medical record were conducted for patients’ baseline demographics including age, gender, diagnosis of diabetes, type of surgery performed, and body mass index (BMI). Outcomes of interest to this project included ICU LOS, SSI, 30-day mortality, and hypoglycemia. ICU LOS was operationalized as the difference between the date and time of

T

ICU admission and the date and time of ICU discharge, this data was retrieved via the InfoView

IP

report. Surgical site infection outcomes were obtained via the Centers for Disease Control

CR

National Healthcare Safety Network (NHSN) database. Following identification of patients on the two separate protocols retrieved from the InfoView report, the NHSN database was queried

US

by the Heart and Vascular Institute (HVI) Infection Prevention coordinator and unique identifiers

AN

were linked. Thirty-day mortality rates were obtained via the Society for Thoracic Surgery (STS) database. Following identification of patients from the InfoView report, the STS database

M

was queried and unique identifiers were linked. Incidence of hypoglycemia (operationalized as

ED

blood glucose values <70 mg/dL documented during ICU stay) and blood glucose values within

PT

the first 24 hours were retrieved via the InfoView report.

CE

Data analysis

Data from the InfoView report and queries were exported to excel spreadsheets,

AC

tabulated, and then statistical tests performed using the program R (R: A language and environment for statistical computing. Version, 2013). Demographic data were statistically analyzed to determine differences between groups. Continuous variables (age, BMI, blood glucose, and ICU LOS) were compared using t-tests. Categorical variables (gender, diagnosis of diabetes, surgery type hypoglycemia, SSI, 30-day mortality) were compared using Fisher’s exact test.

ACCEPTED MANUSCRIPT 9 Results A total of 148 patients underwent cardiac surgery and received the strict CII between June 1, 2014- December 31, 2014. Four patients were excluded for surgery type leaving 144 patients in the strict CII group. A total of 149 patients underwent cardiac surgery and received

T

the liberal CII between March 1, 2015- September 30, 2015. Two patients were excluded

IP

because of length of stay, leaving 147 patients in the liberal CII group. The baseline

CR

characteristics of the groups were similar as shown in Table 1. No statistically significant

US

differences were found between the strict and liberal cohorts for age, BMI, and diabetes history. During the first 24 hours after surgery, the strict CII cohort’s mean blood glucose

AN

measurement was significantly higher at p <0.001 (143.3 mg/dL (SD = 18)) when compared to

M

the liberal CII cohort’s mean blood glucose measurement (159.8 mg/dL (SD = 26.6)). None of the other glucose measures were significant. 48.5% of blood glucose results within the first 24

ED

hours following surgery were within target range for the strict CII cohort (100-140 mg/dL)

PT

versus 24.9% of blood glucoses for the liberal CII cohort (140-160 mg/dL). The proportion of all blood glucose measurements for the duration of ICU stay within target range was 45.7% for

CE

the strict CII cohort and 35.4% for the liberal CII cohort.

AC

Clinical outcomes are reported in Table 2. There were no SSIs in either group. There was no significant difference in ICU LOS between groups. The strict CII group mean ICU LOS was 4.4 days (SD = 3.5) versus 4.5 days in the liberal CII group (SD = 3.8); (p = 0.74). There were zero deaths in the strict CII cohort compared to two 30-day mortalities in the liberal CII cohort (p = 0.49). The incidence of hypoglycemia within 24 hours after surgery was low in both groups. There were 10 hypoglycemic events in the strict group (0.3%) compared to 3 hypoglycemic events in the liberal group (0.1%) (p = 0.16). The 10 hypoglycemic events

ACCEPTED MANUSCRIPT 10 occurred in 7 patients (4.9%) receiving the strict CII protocol while the 3 events in the liberal CII protocol occurred in 3 patients. Discussion

T

This nurse led protocol development, implementation, and evaluation project compared

IP

mean blood glucose levels and clinical outcomes (SSI, ICU LOS, and 30- day mortality) in

CR

cardiac surgery patients receiving a liberal CII protocol versus a previous strict CII protocol. The evidence-based, nurse managed liberal CII protocol implemented STS recommendations to

US

maintain blood glucose < 180 mg/dL in the first 24 hours following cardiac surgery. The results of this project provide evidence that equivalent clinical outcomes of SSI, ICU LOS, and 30-day

AN

mortality are achieved when administering the newer liberal CII protocol in the first 24 hours

M

following cardiac surgery.

ED

The first landmark randomized controlled trial evaluating strict and liberal blood glucose targets reported that a strict target blood glucose of 80-110 mg/dL reduced mortality during

PT

intensive care from 8.0% with conventional treatment to 4.6% with intensive insulin treatment

CE

(van den Berghe et al., 2001). Subsequently, the NICE-SUGAR trial showed that strict glycemic control increased mortality when 27.5% of patients in the intensive control group died (Finfer, et

AC

al., 2009). More recent trials have found that maintaining a blood glucose level between 121 and 180 mg/dL is as effective in promoting positive patient outcomes and has the added benefit of decreasing the incidence of postoperative hypoglycemia and its associated risks (Desai et al., 2012). Despite this lack of clarity on specific glucose target ranges for CII protocols in the cardiac surgery patient, this current project’s liberal CII protocol provided similar clinical outcomes when compared to the previous strict CII protocol.

ACCEPTED MANUSCRIPT 11 The average blood glucose achieved in the first 24 hours following surgery in this project was significantly different, 143.3 mg/dL (strict group) versus 159.8 mg/dL (liberal group). This demonstrates that the protocols impacted glucose values. However, only 48.5% of blood glucoses were in target range (100-140 mg/dL) for the strict group and 24.9% for the liberal

T

group (140-160 mg/dL) in the first 24 hours postoperatively. These results suggest that the

IP

protocol may need to be amended to achieve and maintain blood glucose values within target

CR

range. Similar results have been reported previously in the literature with administration of nurse managed CIIs. Passarelli and colleagues (2016) found that only 43.9% of blood glucose

US

values were within range when using a nurse managed insulin protocol.

AN

Hypoglycemia occurred infrequently in both the strict and liberal groups. As expected,

M

the incidence of hypoglycemia was higher in the strict CII group compared to the liberal group; however, this was not statistically significant. It is uncertain why there was an overall low

ED

incidence of hypoglycemia in the strict CII group. The strict CII had a higher target range (100-

PT

140 mg/dL) compared to the literature, most starting at 80 mg/dL. This higher target range is a possible explanation of the low incidence of hypoglycemia in the strict CII group. It is also

CE

possible that nurses, using sound clinical judgment, perceived that the strict CII could induce

AC

hypoglycemia, leading to concerns with fully implementing the previous protocol. This strict CII protocol non-adherence may have contributed to our lower incidence of hypoglycemia. Nurses in our academic medical center are highly educated and engaged in evidence based practices. It is possible that there was an informal use of a higher blood glucose threshold to manage their patients. Protocol non-adherence has been previously reported in nurse managed CIIs (Passarelli et al., 2016). However, in that case, protocol non-adherence resulted in

ACCEPTED MANUSCRIPT 12 increased hypoglycemia in the cardiac surgery patient population in Passarelli et al. (2016) unlike our findings of decreased hypoglycemia. Limitations

T

The primary limitation of this project is the use of retrospective electronic medical record

IP

data. Medical record analysis relies on user input of data into the electronic health record. It is

CR

possible that incorrect blood glucoses were entered into the electronic health record and included in this project. Clinical outcomes of SSI and 30-day mortality were obtained from the NHSN

US

database and STS database, respectively. The possibility exists that all 30-day mortalities were not reported. The STS database information is derived from secondary sources such as

AN

obituaries, there is no direct measure of mortality for this database. Furthermore, the sample size

M

was small and involves a single site.

ED

A further limitation is that this project only evaluated blood glucose values for the first 24 hours following cardiac surgery. The STS guidelines suggest that optimal patient outcomes are

PT

achieved if blood glucose is controlled throughout the entire postoperative hospitalization, on

CE

average 4.5 days. Inferences from this project’s results can only be drawn to blood glucose values obtained within 24 hours postoperatively. Finally, protocol violations and non-adherence

AC

were not captured in this retrospective review, information that could be useful in explaining the low incidence of hypoglycemia and blood glucose values achieved. Implications Implementing evidence based, nurse-managed liberal CIIs in the postoperative period following cardiac surgery requires concerted efforts from an interprofessional team. Evaluation of glycemic control and patient outcomes is necessary following any CII protocol

ACCEPTED MANUSCRIPT 13 implementation or change. Future work is recommended related to glycemic control following cardiac surgery. More rigorous research is needed using prospectively collected data evaluating multiple protocols to determine the optimal target blood glucose range following cardiac surgery. The literature surrounding hyperglycemia during any inpatient hospitalization and its

T

deleterious effects abounds (Moghissi et al., 2009). Institutions are no longer required to track

IP

and report glycemic control for cardiac surgery patients following CMS’s final rule to remove

CR

the SCIP measure. Voluntary programs exist to highlight institutions that provide optimal

US

glycemic control including The Joint Commission disease specific certification for inpatient diabetes (TJC, 2016). However, there are no regulations outlining mandatory reporting of

AN

inpatient glycemic control as part of the payment structure. CMS rules and policy changes must

M

be developed to ensure healthcare organizations are being held accountable for glycemic control.

ED

Conclusion

A nurse managed liberal CII protocol with the goal of maintaining blood glucose 140-160

PT

mg/dL provides similar outcomes in SSI, ICU LOS, and 30-day mortality when compared to a

CE

nurse managed strict CII protocol w in the first 24 hours following cardiac surgery. Findings from this project support current STS recommendations to implement liberal CII protocols in the

AC

postoperative period following cardiac surgery. The target blood glucose was not achieved the majority of the time in this retrospective review warranting further investigation into these findings in the project setting.

ACCEPTED MANUSCRIPT 14

Acknowledgement

T

We would like to thank the nurses, advanced practice nurses, pharmacists, and physicians who

IP

supported this project. We would also like to thank Victoria Schirm, PhD, RN for her guidance

AC

CE

PT

ED

M

AN

US

CR

throughout this project.

ACCEPTED MANUSCRIPT 15

References

T

Ascione, R., Rogers, C. A., Rajakaruna, C., & Angelini, G. D. (2008). Inadequate blood glucose

IP

control is associated within-hospital mortality and morbidity in diabetic andnondiabetic

CR

patients undergoing cardiac surgery. Circulation, 118, 113-123. http://dx.doi.org/10.1161/CIRCULATIONAHA.107.706416

US

Centers for Disease Control. (2010). Inpatient surgery. Retrieved from

AN

http://www.cdc.gov/nchs/fastats/inpatient-surgery.htm

Centers for Medicare and Medicaid Services. (2015). Removal of SCIP-Inf-4 cardiac surgery

M

patients with controlled postoperative blood glucose. Retrieved from

ED

https://www.federalregister.gov/articles/2015/08/17/2015-19049/medicare-programhospital-inpatient-prospective-payment-systems-for-acute-care-hospitals-and-the#h-375

PT

D’Ancona, G., Bertuzzi, F., Sacchi, L., Pirone, F., Stringi, V., Arcadipane, A., ... Pilato, M.

CE

(2011). Iatrogenic hypoglycemia secondary to tight glucose control is an independent determinant for mortality and cardiac morbidity. European Journal of Cardio-thoracic

AC

Surgery, 40(2), 360-366. http://dx.doi.org/10.1016/j.ejcts.2010.11065 Desai, S. P., Henry, L. L., Holmes, S. D., Hunt, S. L., Martin, C. T., Hebsur, S., & Ad, N. (2012, February 1). Strict versus liberal target range for perioperative glucose in patients undergoing coronary artery bypass grafting: A prospective randomized controlled trial. Journal of Thoracic and Cardiovascular Surgery, 143(2), 318-325. http://dx.doi.org/10.1016/j.jtcvs.2011.10.070.

ACCEPTED MANUSCRIPT 16 Dickerson, R. N., Johnson, J. L., Maish III, G. O., Minard, G., & Brown, R. O. (2012). Evaluation of nursing adherence to a paper-based graduated continuous intravenous regular human infusion algorithm. Nutrition, 28(10), 1008-1011. http://dx.doi.org/10.1016/j.nut2012.01.010

T

Finfer, S., Chittock, D. R., Su, S. Y., Blair, D., Foster, D., Bellomo, R., ... Ronco, J. (2009).

IP

Intensive versus conventional glucose control in critically ill patients. The New England

CR

Journal of Medicine, 360(13), 1283-1297. http://dx.doi.org/10.1056/NEJMoa0810625 Furnary, A. P., Wu, Y. (2006). Clinical effects of hyperglycemia in the cardiac surgery

AN

http://dx.doi.org/10.4158/EP.12.S3.22

US

population: The portland diabetic project. Endocrine Practice, 12(Suppl3), 22-26.

Giakoumidakis, K., Eltheni, R., Patelarou, E., Theologou, S., Patris, V., Michopanou, N., ...

M

Brokalaki, H. (2013, March 1). Effects of intensive glycemic control and outcomes of

ED

cardiac surgery. Heart & Lung: The Journal of Acute and Critical Care, 42(2), 146-151. http://dx.doi.org/10.1016/j.hrtlng.2012.12.007

PT

Greco, G., Michler, R. E., Meltzer, D. O., Ailawadi, G., Hohmann, S. F., Thourani, V. H., ...

CE

Gardner, T. J. (2015). Costs associated with health care-associated infections in cardiac surgery. Journal of the American College of Cardiology, 65(1), 15-23.

AC

http://dx.doi.org/10.1016/j.jacc.2014.09.079 Greco, G., Ferket, B. S., D’Aessandro, D. A., Shi, W., Horvath, K. A., Rosen, A…Moskowitz, A. J. (2016). Diabetes and the association of postoperative hyperglycemia with clinica1 and economic outcomes in cardiac surgery. Diabetes Care, 39(3), 408-417. http://dx.doi.org/10.2337/dc15-1817

ACCEPTED MANUSCRIPT 17 Haga, K. K., McClymont, K. L., Clarke, S., Grounds, R. S., Ying B Ng, K., Clyde, D. W., ... Alston, R. P. (2011). The effect of tight glycemic control, during and after cardiac surgery, on patient mortality and morbidity: A systematic review and meta-analysis. Journal of Cardiothoracic Surgery, 6(3). http://dx.doi.org/10.1186/1749-8090-6-3

T

Hargraves, J. D. (2014). Glycemic control in cardiac surgery: Implementing an evidence-based

IP

insulin infusion protocol. American Journal of Critical Care, 23(3), 250-258.

CR

http://dx.doi.org/10.4037/ajcc2014236

Ichai, C., Preiser, J., Annane, D., Bougle, A., Chiolero, R., Damoisel, C., ... Wion-Barbot, N.

US

(2010). International recommendations for glucose control in adult non diabetic critically

AN

ill patients. Critical Care, 14. http://dx.doi.org/10.1186/cc9258 Knapik, P., Nadziakiewicz, P., Urbanska, E., Saucha, W., Herdynska, M., & Zembala, M.

M

(2009). Cardiopulmonary bypass increases postoperative glycaemia and insulin

ED

consumption after coronary surgery. The Annals of Thoracic Surgery, 87(6), 1859-1865. http://dx.doi.org/10.1016/j.athoracsur.2009.02.066

PT

Lazar, H. L., McDonnell, M., Chipkin, S. R., Furnary, A. P., Engelman, R. M., Sadhu, A. R., ...

CE

Shemin, R. J. (2009). The Society of Thoracic surgeons practice guideline series: Blood glucose management during adult cardiac surgery. Annals of Thoracic Surgery, 87, 663-

AC

669. http://dx.doi.org/10.1016/j.athoracsur.2008.11.011 McDonnell, M. E., Alexanian, S. M., Junqueira, A., Cabral, H., & Lazar, H. L. (2013). Relevance of the surgical care improvement project on glycemic control in patients undergoing cardiac surgery who receive continuous infusions. The Journal of Thoracic and Cardiovascular Surgery, 145(2), 590-597. http://dx.doi.org/10.1016/j.jtcvs.2012.09.031

ACCEPTED MANUSCRIPT 18 Melnyk, B. M. & Fineout-Overholt, E. (2011). Evidence-based practice in nursing & healthcare: A guide to best practice. (Ed.). Philadelphia, PA: Lippincott Williams & Wilkins.. Moghissi, E. S., Korytkowski, M. T., DiNardo, M., Einhorn, D., Hellman, R., Hirsch, I. B., ... Umpierrez, G. E. (2009). American Association of Clinical Endocrinologists and

T

American Diabetes Association consensus statement on inpatient glycemic control.

IP

Endocrine Practice, 15(4), 1-17. Retrieved from

CR

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact =8&ved=0ahUKEwjWzt3GxazLAhUCOj4KHau1BgAQFggiMAA&url=https%3A%2F

US

%2Fwww.aace.com%2Ffiles%2Finpatientglycemiccontrolconsensusstatement.pdf&usg=

AN

AFQjCNHJOHv39GGVS3bFOUiZXoyhPLYZRg&sig2=pKDFhomQYEpg7GfRhlb3dQ Reddy, P., Duggar, B., & Butterworth, J. (2014). Blood glucose management in the patient

M

undergoing cardiac surgery: A review. World Journal of Cardiology, 6(11), 1209–1217.

ED

http://doi.org/10.4330/wjc.v6.i11.1209

Stamou, S. C., Nussbaum, M., Carew, J. D., Dunn, K., Skipper, E., Robicsek, F., & Lobdell, K.

PT

W. (2011). Hypoglycemia with intensive insulin therapy after cardiac surgery:

CE

Predisposing factors and association with mortality. The Journal of Thoracic and

AC

Cardiovascular Surgery, 142(1), 166-173. http://dx.doi.org/10.1016/j.jtcvs.2010.09.064 Stevens, K. R. (2012). ACE Star Model of EBP: Knowledge Transformation. Retrieved from http://www.acestar.uthscsa.edu The Joint Commission. (2014). Surgical care improvement project. Retrieved from http://www.jointcommission.org/surgical_care_improvement_project/

ACCEPTED MANUSCRIPT 19 The Joint Commission. (2015). Effective immediately: Suspension of data collection for performance measure SCIP-Inf-4. Retrieved from http://www.jointcommission.org/assets/1/23/jconline_January_28_15.pdf The Joint Commission. (2016). Certification in Inpatient Diabetes. Retrieved from

T

http://www.jointcommission.org/certification/inpatient_diabetes.aspx

IP

Van den Berghe, G., Wouters, P., Weekers, F., Verwaest, C., Bruyninckx, F., Schetz, M., ...

AC

CE

PT

ED

M

AN

US

England Journal of Medicine, 345(19), 1359-1367.

CR

Bouillon, R. (2001, November 8). Intensive insulin therapy in critically ill patients. New

ACCEPTED MANUSCRIPT 20

Table 1

Strict (n=144)

Liberal (n=147)

P value

39 (27.1)

41 (27.9)

66.8 ± 10.2

BMI

31 ± 7.7

History of diabetes—No. (%)

53 (36.8)

AC

CABG + Valve

CE

Valve

M

0.896 0.197

30. 3± 6.5

0.413

45 (30.6)

0.268 0.744

92 (63.8)

90 (61.2)

28 (19.4)

34 (23.1)

24 (16.7)

23 (15.6)

PT

CABG

ED

Procedures—No. (%)

65.2 ± 11.4

AN

Age

US

Female sex—No. (%)

CR

IP

Variable

T

Baseline Characteristics of the Patients

Note. Plus-minus values are means ± SD, BMI = body mass index, CABG = coronary artery bypass grafting

ACCEPTED MANUSCRIPT

US

CR

IP

T

21

AN

Table 2

M

Outcome variables

Strict (n=144)

Liberal (n=147)

P value

SSI—No. (%)

0 (0)

0 (0)

—

105.1 (4.4)

108.5 (4.5)

0.74

0 (0)

2 (1.4)

0.49

10 (.3)

3 (.1)

0.16

CE

ICU LOS—hours (days)

PT

ED

Variable

AC

30-day mortality—No. (%) Hypoglycemia—No. (%)

Note. SSI = surgical site infection, ICU LOS = intensive care unit length of stay, hypoglycemia = # of blood glucoses < 70 mg/dL within 24 hours for all patients

ACCEPTED MANUSCRIPT 22 Highlights

T IP CR US AN M ED PT



CE



Liberal continuous insulin infusions provide similar outcomes of surgical site infection, intensive care length of stay, and mortality compared to strict continuous insulin infusions in the postoperative cardiac surgery patient population. Interprofessional collaboration is required to implement liberal continuous insulin infusion protocols following cardiac surgery in the practice environment. Adherence to nurse-managed continuous insulin infusion protocols is warranted to evaluate incidence of hypoglycemia.

AC

