Original Article
Lessons in Flying: Crew Resource Management as a Quality Improvement Method for Acute Coronary Syndromes Care Phillip D. Levy, MD, MPH,* JaNeen N. Dancy, PharmD,† Stephanie A. Stowell, MPhil,† James W. Hoekstra, MD,‡ Crystal L. Arthur, MD,§ Charles H. Wilson, MD,║ John M. Bednar, MD,║ Todd Dorman, MD, FCCM,** and Brian Hiestand, MD, MPH, FACEP‡ Abstract: Providing timely, high-quality, guideline-based care to patients with acute coronary syndromes (ACS) who present to the emergency department is critically dependent on cooperation, coordination, and communication between emergency medicine physicians and cardiologists. However, to achieve sustained improvement at the individual institution level, consistent implementation of quality improvement (QI) activities is needed. We describe a QI initiative for ACS care in the emergency setting that combined clinical education with a curriculum based on crew resource management (CRM) principles—a set of tools and techniques for communication, teamwork, and error avoidance used in the aviation industry and with proven applicability in the healthcare setting. Educational training sessions were open to multidisciplinary healthcare teams at 3 hospital sites, and participants were provided practical tools and resources to enhance communication, teamwork, and patientcentered care. Through patient chart reviews, participant surveys, and clinician interviews, baseline assessments of clinical performance measures and team communication-, logistics-, and skills-based efficiencies were performed and reported before the educational training was delivered at each QI site. Reviews of pre- and postinitiative participant surveys demonstrated improvement in knowledge and confidence in the delivery of appropriate and effective ACS care; however, reviews of pre- and postinitiative patient charts revealed limited process improvements. Altogether, this multicenter study of a continuing medical education program based on CRM principles was associated with improvements in provider knowledge and confidence regarding the delivery of appropriate ACS care, but had limited impact on clinical performance measures. Key Words: acute coronary syndromes, quality improvement, emergency medicine (Crit Pathways in Cardiol 2014;13: 36–42)
C
ooperation, coordination, and communication between emergency medicine physicians and cardiologists are critical to
Received for publication October 16, 2013; accepted November 18, 2013. From the *Department of Emergency Medicine and Cardiovascular Research Institute, Wayne State University School of Medicine, Detroit, MI; †Med-IQ, Baltimore, MD; ‡Department of Emergency Medicine, Wake Forest University Health Sciences, Winston-Salem, NC; §Department of Emergency Medicine, Wayne State University School of Medicine, Detroit, MI; ║Department of Emergency Medicine, Cone Health and Heart Vascular Center, Greensboro, NC; and **Department of Critical Care, Johns Hopkins University School of Medicine, Baltimore, MD. This initiative was supported by independent, unrestricted educational grants from the Bristol-Myers Squibb/Sanofi Partnership and the Daiichi Sankyo, Inc./Eli Lilly and Company Alliance. The funding sources had no role in the execution, analysis, or development of the resulting article associated with this initiative. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.critpathcardio.com). Reprints: Phillip D. Levy, MD, MPH, Department of Emergency Medicine, Wayne State University School of Medicine, UHC-6G, 4201 St. Antoine, Detroit, MI 48201. E-mail: [email protected]. Copyright © 2014 by Lippincott Williams & Wilkins ISSN: 1003-0117/14/1301-0036 DOI: 10.1097/HPC.0000000000000002
36 | www.critpathcardio.com
providing timely, high-quality, guideline-based care to patients with acute coronary syndromes (ACS) presenting to the emergency department (ED). Patients must be rapidly triaged, diagnosed, and treated, particularly patients with ST-segment elevation myocardial infarction (STEMI), who suffer an increased risk of mortality for every 30-minute delay in door-to-balloon time.1 Mortality rates for patients with ACS have declined; however, wide variation exists between high- and low-performing hospitals.2,3 A national study of approximately 4600 hospitals found a nearly 2-fold difference in 30-day acute myocardial infarction (AMI) mortality rates—10.9% at high-performing hospitals compared with 24.9% at low-performing hospitals.3 Because outcomes are improved with the application of evidence-based measures for ACS, the formalization of standards for care quality has become a national focus.4,5 Guidelines put forth by the American Heart Association (AHA) and the American College of Cardiology (ACC) have been used by a variety of government and quality agencies to develop performance and quality measures to improve ACS management.6–11 Undoubtedly, such guidelines are important to ACS patient care, yet the mere presence of protocols and processes does not directly translate into quality care.12,13 Studies have shown that AMI mortality rates are lower when organizational cultures promote creative problem solving, empowerment of staff at both the physician and nurse levels, and strong communication and coordination across disciplines, departments, and management levels. Crew resource management (CRM) is one type of procedural system that has shown improved communication, problem solving, decision making, and teamwork. The system was established in the 1980s after the aviation industry determined that more than 70% of air crashes involved human factors—primarily deficiencies in leadership, communication, and team coordination.14,15 After adopting this technique, which cultivates a culture where authority can be respectfully questioned, accidents declined by 50%, and air travel is now one of the safest modes of transportation.16 Several similarities exist between medicine and the aviation industry—involvement of teams consisting of highly trained professionals, the use of advanced technology, the high costs associated with errors, and high-risk, high-stress situations—suggesting applicability of CRM to the healthcare setting. Studies examining the use of CRM in various treatment settings have noted improved perception of patient safety, methods for reporting errors, and adherence to evidence-based guidelines, but its use in the management of patients with ACS is unknown.17–19 The purpose of this study was to examine the application of a communication and team coordination framework developed by the aviation industry combined with medical education to improve the execution of guideline-recommended protocols for the initial management of patients with ACS. To this end, a quality improvement (QI) initiative was designed to promote situational awareness and foster cognitive and interpersonal skills necessary to manage the vital healthcare resources required by patients with ACS presenting to the ED.
Critical Pathways in Cardiology • Volume 13, Number 1, March 2014
Critical Pathways in Cardiology • Volume 13, Number 1, March 2014
METHODS Design of the Initiative Hospitals within the United States were assessed for participation interest in a QI activity using CRM techniques to improve ACS patient care. This assessment was performed in agreement with the Emergency Medicine Cardiac Research and Education Group– International. After discussions with hospital leadership, 3 hospitals agreed to participate in the QI initiative: Moses Cone Hospital (MCH) in Greensboro, NC; Detroit Receiving Hospital (DRH) in Detroit, MI; and Harper University Hospital (HUH), a facility connected to DRH by an underground tunnel. MCH is a community-based, urban, level II trauma center with on-site 24-hour-a-day percutaneous coronary intervention (PCI) capability for STEMI care. Both DRH (a level I trauma center) and HUH (a tertiary-care facility) are academic teaching hospitals affiliated with the Wayne State University School of Medicine and share a common cardiac catheterization suite, which is staffed 24 hours a day by an in-house coronary intervention team. At each of the 3 hospitals, CRM faculty conducted an initial strategic assessment. The assessment consisted of a series of surveys, interviews, focus groups, and observations of ED operations. Findings were presented to the institution’s leadership for the purpose of 2 goals: (1) outline current practices and identify areas that would potentially benefit from application of CRM principles, and (2) ensure that hospital leadership could be champions of change by understanding the needs for and benefits of the initiative and then by sharing this information with staff members. In addition, the QI initiative and its associated educational sessions were promoted to hospital cardiology and ED staff using a series of posters, e-mails, and faxes. Hospital leadership also contacted community cardiologists to inform them of the QI initiative and encourage participation in the educational sessions. At each hospital, three 5-hour live educational sessions were conducted over a period of 2 days to allow for maximum participation. Sessions took place in March, August, and September 2011 at MCH, DRH, and HUH, respectively. The educational intervention was open to emergency medical service professionals, emergency physicians, cardiologists, nurses, pharmacists, support staff, and any other member of the healthcare team involved in the acute care of patients with STEMI and unstable angina/non-STEMI. All sessions were certified for continuing medical education credits for physicians, nurses, and pharmacists. By using interactive educational techniques, including role-playing, group discussions, and audio and video segments, the sessions presented CRM principles (3.5 h) and traditional clinical evidence regarding initial ACS management (1.5 h). The role of human factors in complex, high-risk team situations was discussed. Participants were then coached on using strategic countermeasures, such as team briefings, standardized safety checklists, read-backs and verifications, and the use of a common language. During the final portion of the session, participants applied both CRM techniques and ACS management guidelines to promote appropriate and timely interventions to presented patient case scenarios. After the educational sessions, participants and other hospital staff members received a print resource highlighting the need for adhering to ACS management protocols and practical strategies for reducing preventable medical errors and enhancing ACS patient– centered care in the ED. To reinforce CRM techniques, brief monthly e-mail communications and a newsletter were sent to all program participants.
Data Collection and Definitions Retrospective patient-level data were collected by research personnel specifically trained in study data abstraction methods. At each site, 120 charts were randomly selected from a pool of all charts © 2014 Lippincott Williams & Wilkins
ACS Care Quality Improvement
assigned International Classification of Diseases Ninth Revision code 410.XX (AMI) at hospital discharge, with inclusion if the patient was 18 years or older and had symptoms indicating potential ACS during their initial ED evaluation. Charts were excluded if the patient was younger than 18 years or if ACS was not listed as a primary or secondary diagnosis by the treating physician during the initial ED evaluation. No identifying information on the patients or healthcare professionals was collected during retrospective review of the charts. This study was determined by the Institution Review Board at Wayne State University to be a QI activity and was, thus, considered exempt from full board review. A set of performance measures derived from the then-current ACC/AHA Task Force guidelines for ACS management was used to assess the appropriateness of ACS management before and after the educational intervention.7,20 Specific outcomes of interest included the following guideline-based process measures: use of standardized riskassessment methods, timing of ACS evaluation and assessment [eg, door-to-electrocardiogram (ECG) times], timing of ACS intervention (eg, reperfusion times), differential use and timing of administration for recommended ACS protocol medication (anticoagulation, thienopyridines, and aspirin), and rate of patient transfer from the ED to inpatient units. Door-to-ECG times were listed as 0 minutes if the ECG was performed before formal triage or ED arrival (prehospital ECG). In accordance with guidelines, a goal of 10 minutes or less was used for door-to-ECG times, and a goal of 90 minutes or less was used for door-to-PCI for STEMI cases. Patients were characterized as receiving anticoagulation if any of the following were administered: unfractionated heparin, enoxaparin, dalteparin, bivalirudin, or fondaparinux. Patients who received either clopidogrel or prasugrel were classified as having received a thienopyridine (ticagrelor had not received approval by the Food and Drug Administration at the time of the intervention), and patients who received aspirin either before arrival (emergency medical service or home dose) or within the first 24 hours of hospitalization were considered to have met recommendations for aspirin therapy. We required specific documentation of a validated risk-stratification score (Thrombolysis In Myocardial Infarction,21 Global Registry of Acute Coronary Events,22 or Platelet glycoprotein IIb/IIIa in Unstable angina: Receptor Suppression Using Integrilin)23 in the chart to meet this recommendation. The timeframe for patient chart review varied at each site depending on the date of the educational intervention. Accordingly, charts at MCH were examined between March and December 2010 for the preintervention period and between April 2011 and November 2011 for the postintervention period. At both DRH and HUH, the preintervention time period was January to June 2011; the postintervention period was September 2011 to May 2012 at DRH and October 2011 to March 2012 at HUH. In addition to patient-level process data, the confidence, knowledge, and competency of participating hospital staff were assessed through a written survey administered immediately before (presurvey) and after (postsurvey) the educational intervention. All participants were asked to complete the survey again 30 days after the intervention (30-day postintervention) to assess retention. The survey (Table S1, Supplemental Digital Content, http://links.lww. com/HPC/A197) consisted of 15 multiple-choice knowledge and 3 confidence questions with a 4- or 5-point scale (not confident at all, somewhat confident, moderately confident, extremely confident, or unsure). Knowledge responses were evaluated individually and in aggregate; the latter was used to derive a summary score based on percent correct. To minimize nonresponse bias, only complete surveys were included in aggregate score analyses; partial surveys were excluded. To maintain confidentiality and encourage transparency, all surveys were anonymous, and no participant identifiers were collected. www.critpathcardio.com | 37
Critical Pathways in Cardiology • Volume 13, Number 1, March 2014
Levy et al
Statistical Analysis The medical encounter was used as the unit of analysis in the performance review. We used the Shapiro–Wilk test24 to evaluate the parametricity of continuous variables; because the results indicated a departure from a normal distribution, we used a nonparametric Wilcoxon rank-sum analysis to compare medians between pre- and postintervention performance. The Kruskal–Wallis test (also nonparametric) was used to compare continuous variables across sites within the same analytic time period (pre- or postintervention). For comparisons involving categorical data, Fisher exact test was used. QI measures for ACS management were classified as meeting or not meeting ACC/AHA guideline recommendations, and changes in guideline adherence from pre- to postintervention were also compared using Fisher exact test. Of note, individual results were not compared across charts, that is, the patients whose charts were reviewed initially were not the same patients whose charts were subsequently reviewed, and no attempts were made to link charts. Because no individual participant identifiers were collected, the preintervention, postintervention, and 30-day results of the confidence, knowledge, and competency surveys were compared at the group level using Fisher exact test. Statistical significance was set at a P value less than 0.05. For this QI intervention, the formal sample size was not analyzed a priori, and no adjustments were made to the alpha value for multiple comparisons. Statistics were calculated using Stata IC 11.2 (StataCorp, College Station, TX).
TABLE 1. Educational Session Participants Detroit Receiving Hospital n = 113 Provider type Physician EMT NP/PA Nurse Other Not specified Specialty/department Cardiology Emergency medicine Critical care Other Not specified
Harper University Hospital n = 95
Moses Cone Hospital n = 144
51% 2% 2% 30% 11% 4%
3% 3% 3% 68% 12% 11%
18% 3% 13% 48% 7% 10%
4% 73%
3% 54%
10% 64%
5%
0%
0%
10%
22%
15%
8%
21%
10%
EMT, emergency medical technician; NP, nurse practitioner; PA, physician assistant.
RESULTS The educational interventions were collectively attended by 352 healthcare professionals; participant characteristics are summarized in Table 1. The majority of participants worked in the ED (64%). Nurses were most widely represented at the HUH and MCH sessions, whereas mostly physicians participated in the DRH sessions.
Survey Results Of the 352 attendees, 76% participants completed both the preand postintervention surveys, and a limited sample (22%) completed the 30-day follow-up. With the exception of a single pretest knowledge question, no statistically significant difference was noted in survey responses between the group who completed the first 2 surveys (ie, preand postintervention) and those who completed all 3 (data not shown). Participant confidence levels improved significantly after the educational intervention (Table 2). When asked about confidence in identifying process-related factors that may lead to medical errors in the ED setting, the participants reporting extreme confidence more than doubled (13% vs. 32%; P < 0.001). Regarding confidence in the ability to use CRM leadership techniques to enhance patient-centered care, a significant and notable improvement was observed. Whereas 51% of the participants were unfamiliar with CRM techniques at presurvey, all participants were familiar at postsurvey, and the majority of respondents (89%) indicated high confidence (extremely or moderately confident) with CRM techniques after the intervention. Significant improvements were also detected in the participants’ confidence in their ability to successfully implement current guidelines for the treatment of patients with ACS who present to the ED. In addition to increased confidence, knowledge levels also improved (Table 2). A statistically significant improvement was demonstrated in 10 of the 15 knowledge questions among participants who completed both the pre- and postintervention surveys (Table S1). The largest improvements were seen in questions surrounding the use of checklists, the qualities of effective advocacy, and the ability to identify clinical factors that mark an increased risk of adverse outcomes in patients with ACS. Overall, the percentage of correct responses across all knowledge questions improved significantly from pre- to postsurvey (61% vs. 73%; P = 0.003). 38 | www.critpathcardio.com
For those who completed the 30-day survey, significant and consistent longer-term gains were observed for all 3 confidence questions (Table 3). Specifically, at the time of the preintervention survey, 18% of these participants were extremely confident in their ability to successfully implement current guideline-recommended treatment strategies for the care of patients with ACS who present to the ED compared with 38% (P = 0.002) at the time of the postintervention survey. From postintervention to 30-day follow-up, extreme confidence decreased marginally (38% vs. 35%; P = 0.40) but was still notably higher than at baseline. Longer-term increases were also observed in participant confidence in the ability to use CRM leadership techniques to enhance patient-centered care and in the ability to identify process-related factors that can lead to medical errors in the ED. The limited sample of participants also exhibited knowledge gains immediately after the intervention (61% vs. 72%; P
Lessons in Flying: Crew Resource Management as a Quality Improvement Method for Acute Coronary Syndromes Care Phillip D. Levy, MD, MPH,* JaNeen N. Dancy, PharmD,† Stephanie A. Stowell, MPhil,† James W. Hoekstra, MD,‡ Crystal L. Arthur, MD,§ Charles H. Wilson, MD,║ John M. Bednar, MD,║ Todd Dorman, MD, FCCM,** and Brian Hiestand, MD, MPH, FACEP‡ Abstract: Providing timely, high-quality, guideline-based care to patients with acute coronary syndromes (ACS) who present to the emergency department is critically dependent on cooperation, coordination, and communication between emergency medicine physicians and cardiologists. However, to achieve sustained improvement at the individual institution level, consistent implementation of quality improvement (QI) activities is needed. We describe a QI initiative for ACS care in the emergency setting that combined clinical education with a curriculum based on crew resource management (CRM) principles—a set of tools and techniques for communication, teamwork, and error avoidance used in the aviation industry and with proven applicability in the healthcare setting. Educational training sessions were open to multidisciplinary healthcare teams at 3 hospital sites, and participants were provided practical tools and resources to enhance communication, teamwork, and patientcentered care. Through patient chart reviews, participant surveys, and clinician interviews, baseline assessments of clinical performance measures and team communication-, logistics-, and skills-based efficiencies were performed and reported before the educational training was delivered at each QI site. Reviews of pre- and postinitiative participant surveys demonstrated improvement in knowledge and confidence in the delivery of appropriate and effective ACS care; however, reviews of pre- and postinitiative patient charts revealed limited process improvements. Altogether, this multicenter study of a continuing medical education program based on CRM principles was associated with improvements in provider knowledge and confidence regarding the delivery of appropriate ACS care, but had limited impact on clinical performance measures. Key Words: acute coronary syndromes, quality improvement, emergency medicine (Crit Pathways in Cardiol 2014;13: 36–42)
C
ooperation, coordination, and communication between emergency medicine physicians and cardiologists are critical to
Received for publication October 16, 2013; accepted November 18, 2013. From the *Department of Emergency Medicine and Cardiovascular Research Institute, Wayne State University School of Medicine, Detroit, MI; †Med-IQ, Baltimore, MD; ‡Department of Emergency Medicine, Wake Forest University Health Sciences, Winston-Salem, NC; §Department of Emergency Medicine, Wayne State University School of Medicine, Detroit, MI; ║Department of Emergency Medicine, Cone Health and Heart Vascular Center, Greensboro, NC; and **Department of Critical Care, Johns Hopkins University School of Medicine, Baltimore, MD. This initiative was supported by independent, unrestricted educational grants from the Bristol-Myers Squibb/Sanofi Partnership and the Daiichi Sankyo, Inc./Eli Lilly and Company Alliance. The funding sources had no role in the execution, analysis, or development of the resulting article associated with this initiative. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.critpathcardio.com). Reprints: Phillip D. Levy, MD, MPH, Department of Emergency Medicine, Wayne State University School of Medicine, UHC-6G, 4201 St. Antoine, Detroit, MI 48201. E-mail: [email protected]. Copyright © 2014 by Lippincott Williams & Wilkins ISSN: 1003-0117/14/1301-0036 DOI: 10.1097/HPC.0000000000000002
36 | www.critpathcardio.com
providing timely, high-quality, guideline-based care to patients with acute coronary syndromes (ACS) presenting to the emergency department (ED). Patients must be rapidly triaged, diagnosed, and treated, particularly patients with ST-segment elevation myocardial infarction (STEMI), who suffer an increased risk of mortality for every 30-minute delay in door-to-balloon time.1 Mortality rates for patients with ACS have declined; however, wide variation exists between high- and low-performing hospitals.2,3 A national study of approximately 4600 hospitals found a nearly 2-fold difference in 30-day acute myocardial infarction (AMI) mortality rates—10.9% at high-performing hospitals compared with 24.9% at low-performing hospitals.3 Because outcomes are improved with the application of evidence-based measures for ACS, the formalization of standards for care quality has become a national focus.4,5 Guidelines put forth by the American Heart Association (AHA) and the American College of Cardiology (ACC) have been used by a variety of government and quality agencies to develop performance and quality measures to improve ACS management.6–11 Undoubtedly, such guidelines are important to ACS patient care, yet the mere presence of protocols and processes does not directly translate into quality care.12,13 Studies have shown that AMI mortality rates are lower when organizational cultures promote creative problem solving, empowerment of staff at both the physician and nurse levels, and strong communication and coordination across disciplines, departments, and management levels. Crew resource management (CRM) is one type of procedural system that has shown improved communication, problem solving, decision making, and teamwork. The system was established in the 1980s after the aviation industry determined that more than 70% of air crashes involved human factors—primarily deficiencies in leadership, communication, and team coordination.14,15 After adopting this technique, which cultivates a culture where authority can be respectfully questioned, accidents declined by 50%, and air travel is now one of the safest modes of transportation.16 Several similarities exist between medicine and the aviation industry—involvement of teams consisting of highly trained professionals, the use of advanced technology, the high costs associated with errors, and high-risk, high-stress situations—suggesting applicability of CRM to the healthcare setting. Studies examining the use of CRM in various treatment settings have noted improved perception of patient safety, methods for reporting errors, and adherence to evidence-based guidelines, but its use in the management of patients with ACS is unknown.17–19 The purpose of this study was to examine the application of a communication and team coordination framework developed by the aviation industry combined with medical education to improve the execution of guideline-recommended protocols for the initial management of patients with ACS. To this end, a quality improvement (QI) initiative was designed to promote situational awareness and foster cognitive and interpersonal skills necessary to manage the vital healthcare resources required by patients with ACS presenting to the ED.
Critical Pathways in Cardiology • Volume 13, Number 1, March 2014
Critical Pathways in Cardiology • Volume 13, Number 1, March 2014
METHODS Design of the Initiative Hospitals within the United States were assessed for participation interest in a QI activity using CRM techniques to improve ACS patient care. This assessment was performed in agreement with the Emergency Medicine Cardiac Research and Education Group– International. After discussions with hospital leadership, 3 hospitals agreed to participate in the QI initiative: Moses Cone Hospital (MCH) in Greensboro, NC; Detroit Receiving Hospital (DRH) in Detroit, MI; and Harper University Hospital (HUH), a facility connected to DRH by an underground tunnel. MCH is a community-based, urban, level II trauma center with on-site 24-hour-a-day percutaneous coronary intervention (PCI) capability for STEMI care. Both DRH (a level I trauma center) and HUH (a tertiary-care facility) are academic teaching hospitals affiliated with the Wayne State University School of Medicine and share a common cardiac catheterization suite, which is staffed 24 hours a day by an in-house coronary intervention team. At each of the 3 hospitals, CRM faculty conducted an initial strategic assessment. The assessment consisted of a series of surveys, interviews, focus groups, and observations of ED operations. Findings were presented to the institution’s leadership for the purpose of 2 goals: (1) outline current practices and identify areas that would potentially benefit from application of CRM principles, and (2) ensure that hospital leadership could be champions of change by understanding the needs for and benefits of the initiative and then by sharing this information with staff members. In addition, the QI initiative and its associated educational sessions were promoted to hospital cardiology and ED staff using a series of posters, e-mails, and faxes. Hospital leadership also contacted community cardiologists to inform them of the QI initiative and encourage participation in the educational sessions. At each hospital, three 5-hour live educational sessions were conducted over a period of 2 days to allow for maximum participation. Sessions took place in March, August, and September 2011 at MCH, DRH, and HUH, respectively. The educational intervention was open to emergency medical service professionals, emergency physicians, cardiologists, nurses, pharmacists, support staff, and any other member of the healthcare team involved in the acute care of patients with STEMI and unstable angina/non-STEMI. All sessions were certified for continuing medical education credits for physicians, nurses, and pharmacists. By using interactive educational techniques, including role-playing, group discussions, and audio and video segments, the sessions presented CRM principles (3.5 h) and traditional clinical evidence regarding initial ACS management (1.5 h). The role of human factors in complex, high-risk team situations was discussed. Participants were then coached on using strategic countermeasures, such as team briefings, standardized safety checklists, read-backs and verifications, and the use of a common language. During the final portion of the session, participants applied both CRM techniques and ACS management guidelines to promote appropriate and timely interventions to presented patient case scenarios. After the educational sessions, participants and other hospital staff members received a print resource highlighting the need for adhering to ACS management protocols and practical strategies for reducing preventable medical errors and enhancing ACS patient– centered care in the ED. To reinforce CRM techniques, brief monthly e-mail communications and a newsletter were sent to all program participants.
Data Collection and Definitions Retrospective patient-level data were collected by research personnel specifically trained in study data abstraction methods. At each site, 120 charts were randomly selected from a pool of all charts © 2014 Lippincott Williams & Wilkins
ACS Care Quality Improvement
assigned International Classification of Diseases Ninth Revision code 410.XX (AMI) at hospital discharge, with inclusion if the patient was 18 years or older and had symptoms indicating potential ACS during their initial ED evaluation. Charts were excluded if the patient was younger than 18 years or if ACS was not listed as a primary or secondary diagnosis by the treating physician during the initial ED evaluation. No identifying information on the patients or healthcare professionals was collected during retrospective review of the charts. This study was determined by the Institution Review Board at Wayne State University to be a QI activity and was, thus, considered exempt from full board review. A set of performance measures derived from the then-current ACC/AHA Task Force guidelines for ACS management was used to assess the appropriateness of ACS management before and after the educational intervention.7,20 Specific outcomes of interest included the following guideline-based process measures: use of standardized riskassessment methods, timing of ACS evaluation and assessment [eg, door-to-electrocardiogram (ECG) times], timing of ACS intervention (eg, reperfusion times), differential use and timing of administration for recommended ACS protocol medication (anticoagulation, thienopyridines, and aspirin), and rate of patient transfer from the ED to inpatient units. Door-to-ECG times were listed as 0 minutes if the ECG was performed before formal triage or ED arrival (prehospital ECG). In accordance with guidelines, a goal of 10 minutes or less was used for door-to-ECG times, and a goal of 90 minutes or less was used for door-to-PCI for STEMI cases. Patients were characterized as receiving anticoagulation if any of the following were administered: unfractionated heparin, enoxaparin, dalteparin, bivalirudin, or fondaparinux. Patients who received either clopidogrel or prasugrel were classified as having received a thienopyridine (ticagrelor had not received approval by the Food and Drug Administration at the time of the intervention), and patients who received aspirin either before arrival (emergency medical service or home dose) or within the first 24 hours of hospitalization were considered to have met recommendations for aspirin therapy. We required specific documentation of a validated risk-stratification score (Thrombolysis In Myocardial Infarction,21 Global Registry of Acute Coronary Events,22 or Platelet glycoprotein IIb/IIIa in Unstable angina: Receptor Suppression Using Integrilin)23 in the chart to meet this recommendation. The timeframe for patient chart review varied at each site depending on the date of the educational intervention. Accordingly, charts at MCH were examined between March and December 2010 for the preintervention period and between April 2011 and November 2011 for the postintervention period. At both DRH and HUH, the preintervention time period was January to June 2011; the postintervention period was September 2011 to May 2012 at DRH and October 2011 to March 2012 at HUH. In addition to patient-level process data, the confidence, knowledge, and competency of participating hospital staff were assessed through a written survey administered immediately before (presurvey) and after (postsurvey) the educational intervention. All participants were asked to complete the survey again 30 days after the intervention (30-day postintervention) to assess retention. The survey (Table S1, Supplemental Digital Content, http://links.lww. com/HPC/A197) consisted of 15 multiple-choice knowledge and 3 confidence questions with a 4- or 5-point scale (not confident at all, somewhat confident, moderately confident, extremely confident, or unsure). Knowledge responses were evaluated individually and in aggregate; the latter was used to derive a summary score based on percent correct. To minimize nonresponse bias, only complete surveys were included in aggregate score analyses; partial surveys were excluded. To maintain confidentiality and encourage transparency, all surveys were anonymous, and no participant identifiers were collected. www.critpathcardio.com | 37
Critical Pathways in Cardiology • Volume 13, Number 1, March 2014
Levy et al
Statistical Analysis The medical encounter was used as the unit of analysis in the performance review. We used the Shapiro–Wilk test24 to evaluate the parametricity of continuous variables; because the results indicated a departure from a normal distribution, we used a nonparametric Wilcoxon rank-sum analysis to compare medians between pre- and postintervention performance. The Kruskal–Wallis test (also nonparametric) was used to compare continuous variables across sites within the same analytic time period (pre- or postintervention). For comparisons involving categorical data, Fisher exact test was used. QI measures for ACS management were classified as meeting or not meeting ACC/AHA guideline recommendations, and changes in guideline adherence from pre- to postintervention were also compared using Fisher exact test. Of note, individual results were not compared across charts, that is, the patients whose charts were reviewed initially were not the same patients whose charts were subsequently reviewed, and no attempts were made to link charts. Because no individual participant identifiers were collected, the preintervention, postintervention, and 30-day results of the confidence, knowledge, and competency surveys were compared at the group level using Fisher exact test. Statistical significance was set at a P value less than 0.05. For this QI intervention, the formal sample size was not analyzed a priori, and no adjustments were made to the alpha value for multiple comparisons. Statistics were calculated using Stata IC 11.2 (StataCorp, College Station, TX).
TABLE 1. Educational Session Participants Detroit Receiving Hospital n = 113 Provider type Physician EMT NP/PA Nurse Other Not specified Specialty/department Cardiology Emergency medicine Critical care Other Not specified
Harper University Hospital n = 95
Moses Cone Hospital n = 144
51% 2% 2% 30% 11% 4%
3% 3% 3% 68% 12% 11%
18% 3% 13% 48% 7% 10%
4% 73%
3% 54%
10% 64%
5%
0%
0%
10%
22%
15%
8%
21%
10%
EMT, emergency medical technician; NP, nurse practitioner; PA, physician assistant.
RESULTS The educational interventions were collectively attended by 352 healthcare professionals; participant characteristics are summarized in Table 1. The majority of participants worked in the ED (64%). Nurses were most widely represented at the HUH and MCH sessions, whereas mostly physicians participated in the DRH sessions.
Survey Results Of the 352 attendees, 76% participants completed both the preand postintervention surveys, and a limited sample (22%) completed the 30-day follow-up. With the exception of a single pretest knowledge question, no statistically significant difference was noted in survey responses between the group who completed the first 2 surveys (ie, preand postintervention) and those who completed all 3 (data not shown). Participant confidence levels improved significantly after the educational intervention (Table 2). When asked about confidence in identifying process-related factors that may lead to medical errors in the ED setting, the participants reporting extreme confidence more than doubled (13% vs. 32%; P < 0.001). Regarding confidence in the ability to use CRM leadership techniques to enhance patient-centered care, a significant and notable improvement was observed. Whereas 51% of the participants were unfamiliar with CRM techniques at presurvey, all participants were familiar at postsurvey, and the majority of respondents (89%) indicated high confidence (extremely or moderately confident) with CRM techniques after the intervention. Significant improvements were also detected in the participants’ confidence in their ability to successfully implement current guidelines for the treatment of patients with ACS who present to the ED. In addition to increased confidence, knowledge levels also improved (Table 2). A statistically significant improvement was demonstrated in 10 of the 15 knowledge questions among participants who completed both the pre- and postintervention surveys (Table S1). The largest improvements were seen in questions surrounding the use of checklists, the qualities of effective advocacy, and the ability to identify clinical factors that mark an increased risk of adverse outcomes in patients with ACS. Overall, the percentage of correct responses across all knowledge questions improved significantly from pre- to postsurvey (61% vs. 73%; P = 0.003). 38 | www.critpathcardio.com
For those who completed the 30-day survey, significant and consistent longer-term gains were observed for all 3 confidence questions (Table 3). Specifically, at the time of the preintervention survey, 18% of these participants were extremely confident in their ability to successfully implement current guideline-recommended treatment strategies for the care of patients with ACS who present to the ED compared with 38% (P = 0.002) at the time of the postintervention survey. From postintervention to 30-day follow-up, extreme confidence decreased marginally (38% vs. 35%; P = 0.40) but was still notably higher than at baseline. Longer-term increases were also observed in participant confidence in the ability to use CRM leadership techniques to enhance patient-centered care and in the ability to identify process-related factors that can lead to medical errors in the ED. The limited sample of participants also exhibited knowledge gains immediately after the intervention (61% vs. 72%; P
