03111
2013
AOPXXX10.1177/1060028013503111Beik et alThe Annals of Pharmacotherapy
Rsearch Report-Emergency Medicine
Evaluation of an Institution-Wide Guideline for Hyperglycemic Emergencies at a Tertiary Academic Medical Center
Annals of Pharmacotherapy 47(10) 1260–1265 © The Author(s) 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1060028013503111 aop.sagepub.com
Nahal Beik, PharmD, BCPS1, Kevin E. Anger, PharmD, BCPS1, Allison A. Forni, PharmD, BCPS2, Komal Bawa, PharmD, BCPS3, and Paul M. Szumita, PharmD, BCPS1
Abstract Background: No previous studies exist examining implementation of an institution-wide guideline and order set for hyperglycemic emergencies (diabetic ketoacidosis [DKA] and hyperosmolar hyperglycemic state [HHS]). Objective: Evaluate the impact of an institutional guideline and order set for hyperglycemic emergencies. Methods: This retrospective descriptive study evaluated patients with a diagnosis of DKA or HHS. Two time periods were evaluated: phase 1 (PRE) assessed practice preguideline implementation, and phase 2 (POST) assessed practice postguideline and order set introduction. Results: A total of 172 patients (91 PRE and 81 POST) were included in the analysis. There was no difference in the mean hospital length of stay (LOS) in the PRE versus POST groups (5.2 ± 4 vs 5.9 ± 8.6 days, P = .49). The mean intensive care unit (ICU) LOS was shorter in the POST group (64.8 ± 19 vs 37.1 ± 74.8 hours, P < .01). The POST group had an increase in frequency of assessments for clearance of urinary ketones (18 vs 33.3%, P = .03) and β-hydroxybutyrate (16 vs 37%, P < .01). Frequency of point-of-care glucose testing (12.5 ± 4.6 vs 15.1 ± 4.7, P < .01) and time to anion gap closure (13 ± 9 vs 9.3 ± 7.4 hours, P < .01) improved in the POST group. There was no difference in the number of patients experiencing hypoglycemia or hypokalemia between both groups. Conclusions: Implementation of an institutional guideline and order set for hyperglycemic emergencies decreased ICU LOS and time to anion gap closure, with no difference in rates of hypoglycemia. Keywords hyperglycemic emergency, diabetic ketoacidosis, hyperosmolar hyperglycemic state Received 6 August 2013
Introduction Diabetic ketoacidosis (DKA) and hyperosmolar hyperglycemic state (HHS) are life-threatening acute metabolic complications of diabetes, which can occur in patients with type 1 or type 2 diabetes.1 Despite both the decline in the rate of hospital admissions for HHS and the age-adjusted mortality for DKA, the incidence of DKA continues to rise in the United States, with DKA accounting for about 136 510 hospitalizations in 2006.1-7 Implementation of a comprehensive protocol and order set for treating DKA has been found to shorten both intensive care unit (ICU) length of stay (LOS) and hospital LOS by creating a standardized level of care for all aspects of DKA.8 Additionally, protocolized care for DKA decreased the time to clearance of ketones and anion gap as well as the rate of hypoglycemia, thus, safely and effectively shortening the time to resolution of DKA.8
The American Diabetes Association (ADA) has outlined best practices to treat these hyperglycemic crises and recommends fluid therapy, insulin therapy, electrolyte repletion, and frequent laboratory monitoring to ensure successful resolution of DKA and HHS.3 In accordance with the ADA, we created a local guideline and order set, which outlines the key therapeutic modalities for the appropriate management of DKA and HHS to be implemented in the emergency department, inpatient step-down units, and ICU which differs from the previously discussed study. 1
Brigham and Women’s Hospital, Boston, MA, USA UMass Memorial Medical Center, Worcester, MA, USA 3 UCSF Medical Center, San Francisco, CA, USA 2
Corresponding Author: Nahal Beik, Department of Pharmacy, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115, USA. Email: [email protected]
1261
Beik et al Table 1. Classification of Patients. Criteria
Mild DKA (Plasma Glucose >250 mg/dL)
Moderate DKA (Plasma Glucose >250 mg/dL)
Severe DKA (Plasma Glucose >250 mg/dL)
Venous pH Serum bicarbonate (mEq/L) Urine ketone β-Hydroxybutyrate (mmol/L) Serum osmolality (mOsm/kg) Anion gap (mEq/L)
1 Normal to elevated BWH range (>5)
7.00-7.24 10-14 1-3+ >1 Normal to elevated BWH range (5-17)
17)
HHS (Plasma Glucose >600 mg/dL) >7.30 >18 Trace Trace >320 Variable
Abbreviations: DKA, diabetic ketoacidosis; HHS, hyperosmolar hyperglycemic state; BWH, Brigham and Women’s Hospital.
The objective of this study is to evaluate the impact of an institutional guideline and order set for hyperglycemic emergencies (DKA and HHS).
Research Design and Methods Setting This was a single-center retrospective descriptive study of clinical practice of adult patients admitted to our 793-bed academic medical center with a diagnosis of DKA and HHS before and after implementation of an institutional guideline and order set.
Hyperglycemic Emergencies Guideline The guideline consists of 6 components: monitoring, fluid repletion, insulin administration, electrolyte repletion, followup care, and diabetes education (Appendix A, supplementary material online). The inpatient and emergency department order sets follow the guideline (Appendix B, supplementary material online). The institutional review board reviewed and approved the study protocol prior to data collection.
Guideline Implementation Data collection occurred between 2 time periods: phase 1 (PRE) assessed practice preguideline implementation from September 2007 to September 2009 and phase 2 (POST) assessed practice postguideline and order set implementation from February 2011 to April 2012. The stepwise approach to developing and instituting the hyperglycemic emergencies guideline and order set consisted of 3 steps: development, implementation, and continuous quality improvement (CQI). The development step entailed guideline development; obtaining approval from various multidisciplinary committees, which included the Diabetes Subcommittee and Pharmacy and Therapeutics Committee; and order set development for both the inpatient and emergency department physicians. The implementation step encompassed education of medical intensivists, hospitalists, emergency department physicians, pharmacists, and
nurse educators on the guideline and order set. Another component of the implementation step involved making the guideline and order set easily accessible and available on the hospital intranet and institution-specific computerized prescriber order entry system for inpatient and emergency department prescribers. The CQI step consisted of ongoing education to the hospital staff, obtaining feedback from bedside clinicians, and updating the guideline and computerbased resources based on the feedback.
Patient Enrollment PRE patients were screened using the institutional billing database (Transition Systems Inc), where a report was generated for all patients admitted with a primary or secondary diagnosis of DKA or HHS (ICD-9 code). POST patients were screened by initiation of an order set (emergency department or inpatient) within 8 hours of presentation. POST patients were identified prospectively via an electronic alert that was sent to the investigator when an inpatient order set was activated or through review of the emergency department order entry when a patient was identified as having a hyperglycemic emergency. To be included in the analysis, patients had to fit the criteria of a hyperglycemic emergency and were classified as having DKA or HHS according to the institutional hyperglycemic emergencies management guideline (Table 1). Patient charts and electronic laboratory parameters were individually reviewed to assess guideline adherence. We excluded patients who were less than 18 years of age, who received intensive insulin therapy for critical illness that was not related to a hyperglycemic emergency, who had hyperglycemia that was attributed to metabolic abnormalities other than DKA or HHS, who did not have an order set used in the POST group, and who had incomplete medical records in the PRE group.
Data Collection Methods A single investigator collected data retrospectively, and demographic data and admission variables collected included gender, age, admitting diagnosis, history of diabetes, home
1262
Annals of Pharmacotherapy
Patients evaluated for inclusion (N = 350)
64 25 17 15
Patients Excluded N = 121 Incomplete medical record Hyperglycemia not DKA/HHS Intensive insulin therapy Incorrect coding
Patients Excluded N = 57 44 Order set not used 11 Hyperglycemia not DKA/HHS 1 HHS resolved prior to admit 1 Inappropriate use of order set
Patients included PRE N = 91
ICU N = 52
Non-ICU N = 39
Patients included POST N = 81
ICU N = 44
Non-ICU N = 37
Figure 1. Patient enrollment.
Abbreviations: DKA, diabetic ketoacidosis; HHS, hyperosmolar hyperglycemic state; PRE, practice preguideline implementation; POST, practice postguideline implementation; ICU, intensive care unit.
insulin use, precipitating factor, and ICU admission. The primary outcome was hospital LOS. Secondary outcomes included ICU LOS, time to anion gap closure, and pharmacotherapy assessments, which included laboratory and glucose assessment metrics, insulin administration metrics, and dextrose repletion metrics. Safety outcomes included the number of patients who experienced hypoglycemic (
2013
AOPXXX10.1177/1060028013503111Beik et alThe Annals of Pharmacotherapy
Rsearch Report-Emergency Medicine
Evaluation of an Institution-Wide Guideline for Hyperglycemic Emergencies at a Tertiary Academic Medical Center
Annals of Pharmacotherapy 47(10) 1260–1265 © The Author(s) 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1060028013503111 aop.sagepub.com
Nahal Beik, PharmD, BCPS1, Kevin E. Anger, PharmD, BCPS1, Allison A. Forni, PharmD, BCPS2, Komal Bawa, PharmD, BCPS3, and Paul M. Szumita, PharmD, BCPS1
Abstract Background: No previous studies exist examining implementation of an institution-wide guideline and order set for hyperglycemic emergencies (diabetic ketoacidosis [DKA] and hyperosmolar hyperglycemic state [HHS]). Objective: Evaluate the impact of an institutional guideline and order set for hyperglycemic emergencies. Methods: This retrospective descriptive study evaluated patients with a diagnosis of DKA or HHS. Two time periods were evaluated: phase 1 (PRE) assessed practice preguideline implementation, and phase 2 (POST) assessed practice postguideline and order set introduction. Results: A total of 172 patients (91 PRE and 81 POST) were included in the analysis. There was no difference in the mean hospital length of stay (LOS) in the PRE versus POST groups (5.2 ± 4 vs 5.9 ± 8.6 days, P = .49). The mean intensive care unit (ICU) LOS was shorter in the POST group (64.8 ± 19 vs 37.1 ± 74.8 hours, P < .01). The POST group had an increase in frequency of assessments for clearance of urinary ketones (18 vs 33.3%, P = .03) and β-hydroxybutyrate (16 vs 37%, P < .01). Frequency of point-of-care glucose testing (12.5 ± 4.6 vs 15.1 ± 4.7, P < .01) and time to anion gap closure (13 ± 9 vs 9.3 ± 7.4 hours, P < .01) improved in the POST group. There was no difference in the number of patients experiencing hypoglycemia or hypokalemia between both groups. Conclusions: Implementation of an institutional guideline and order set for hyperglycemic emergencies decreased ICU LOS and time to anion gap closure, with no difference in rates of hypoglycemia. Keywords hyperglycemic emergency, diabetic ketoacidosis, hyperosmolar hyperglycemic state Received 6 August 2013
Introduction Diabetic ketoacidosis (DKA) and hyperosmolar hyperglycemic state (HHS) are life-threatening acute metabolic complications of diabetes, which can occur in patients with type 1 or type 2 diabetes.1 Despite both the decline in the rate of hospital admissions for HHS and the age-adjusted mortality for DKA, the incidence of DKA continues to rise in the United States, with DKA accounting for about 136 510 hospitalizations in 2006.1-7 Implementation of a comprehensive protocol and order set for treating DKA has been found to shorten both intensive care unit (ICU) length of stay (LOS) and hospital LOS by creating a standardized level of care for all aspects of DKA.8 Additionally, protocolized care for DKA decreased the time to clearance of ketones and anion gap as well as the rate of hypoglycemia, thus, safely and effectively shortening the time to resolution of DKA.8
The American Diabetes Association (ADA) has outlined best practices to treat these hyperglycemic crises and recommends fluid therapy, insulin therapy, electrolyte repletion, and frequent laboratory monitoring to ensure successful resolution of DKA and HHS.3 In accordance with the ADA, we created a local guideline and order set, which outlines the key therapeutic modalities for the appropriate management of DKA and HHS to be implemented in the emergency department, inpatient step-down units, and ICU which differs from the previously discussed study. 1
Brigham and Women’s Hospital, Boston, MA, USA UMass Memorial Medical Center, Worcester, MA, USA 3 UCSF Medical Center, San Francisco, CA, USA 2
Corresponding Author: Nahal Beik, Department of Pharmacy, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115, USA. Email: [email protected]
1261
Beik et al Table 1. Classification of Patients. Criteria
Mild DKA (Plasma Glucose >250 mg/dL)
Moderate DKA (Plasma Glucose >250 mg/dL)
Severe DKA (Plasma Glucose >250 mg/dL)
Venous pH Serum bicarbonate (mEq/L) Urine ketone β-Hydroxybutyrate (mmol/L) Serum osmolality (mOsm/kg) Anion gap (mEq/L)
1 Normal to elevated BWH range (>5)
7.00-7.24 10-14 1-3+ >1 Normal to elevated BWH range (5-17)
17)
HHS (Plasma Glucose >600 mg/dL) >7.30 >18 Trace Trace >320 Variable
Abbreviations: DKA, diabetic ketoacidosis; HHS, hyperosmolar hyperglycemic state; BWH, Brigham and Women’s Hospital.
The objective of this study is to evaluate the impact of an institutional guideline and order set for hyperglycemic emergencies (DKA and HHS).
Research Design and Methods Setting This was a single-center retrospective descriptive study of clinical practice of adult patients admitted to our 793-bed academic medical center with a diagnosis of DKA and HHS before and after implementation of an institutional guideline and order set.
Hyperglycemic Emergencies Guideline The guideline consists of 6 components: monitoring, fluid repletion, insulin administration, electrolyte repletion, followup care, and diabetes education (Appendix A, supplementary material online). The inpatient and emergency department order sets follow the guideline (Appendix B, supplementary material online). The institutional review board reviewed and approved the study protocol prior to data collection.
Guideline Implementation Data collection occurred between 2 time periods: phase 1 (PRE) assessed practice preguideline implementation from September 2007 to September 2009 and phase 2 (POST) assessed practice postguideline and order set implementation from February 2011 to April 2012. The stepwise approach to developing and instituting the hyperglycemic emergencies guideline and order set consisted of 3 steps: development, implementation, and continuous quality improvement (CQI). The development step entailed guideline development; obtaining approval from various multidisciplinary committees, which included the Diabetes Subcommittee and Pharmacy and Therapeutics Committee; and order set development for both the inpatient and emergency department physicians. The implementation step encompassed education of medical intensivists, hospitalists, emergency department physicians, pharmacists, and
nurse educators on the guideline and order set. Another component of the implementation step involved making the guideline and order set easily accessible and available on the hospital intranet and institution-specific computerized prescriber order entry system for inpatient and emergency department prescribers. The CQI step consisted of ongoing education to the hospital staff, obtaining feedback from bedside clinicians, and updating the guideline and computerbased resources based on the feedback.
Patient Enrollment PRE patients were screened using the institutional billing database (Transition Systems Inc), where a report was generated for all patients admitted with a primary or secondary diagnosis of DKA or HHS (ICD-9 code). POST patients were screened by initiation of an order set (emergency department or inpatient) within 8 hours of presentation. POST patients were identified prospectively via an electronic alert that was sent to the investigator when an inpatient order set was activated or through review of the emergency department order entry when a patient was identified as having a hyperglycemic emergency. To be included in the analysis, patients had to fit the criteria of a hyperglycemic emergency and were classified as having DKA or HHS according to the institutional hyperglycemic emergencies management guideline (Table 1). Patient charts and electronic laboratory parameters were individually reviewed to assess guideline adherence. We excluded patients who were less than 18 years of age, who received intensive insulin therapy for critical illness that was not related to a hyperglycemic emergency, who had hyperglycemia that was attributed to metabolic abnormalities other than DKA or HHS, who did not have an order set used in the POST group, and who had incomplete medical records in the PRE group.
Data Collection Methods A single investigator collected data retrospectively, and demographic data and admission variables collected included gender, age, admitting diagnosis, history of diabetes, home
1262
Annals of Pharmacotherapy
Patients evaluated for inclusion (N = 350)
64 25 17 15
Patients Excluded N = 121 Incomplete medical record Hyperglycemia not DKA/HHS Intensive insulin therapy Incorrect coding
Patients Excluded N = 57 44 Order set not used 11 Hyperglycemia not DKA/HHS 1 HHS resolved prior to admit 1 Inappropriate use of order set
Patients included PRE N = 91
ICU N = 52
Non-ICU N = 39
Patients included POST N = 81
ICU N = 44
Non-ICU N = 37
Figure 1. Patient enrollment.
Abbreviations: DKA, diabetic ketoacidosis; HHS, hyperosmolar hyperglycemic state; PRE, practice preguideline implementation; POST, practice postguideline implementation; ICU, intensive care unit.
insulin use, precipitating factor, and ICU admission. The primary outcome was hospital LOS. Secondary outcomes included ICU LOS, time to anion gap closure, and pharmacotherapy assessments, which included laboratory and glucose assessment metrics, insulin administration metrics, and dextrose repletion metrics. Safety outcomes included the number of patients who experienced hypoglycemic (
