Improved Efficiency in Acute Myocardial Infarction Care Through Commitment to Emergency Department Initiated Primary PCI
March 2004
This article is reprinted with permission from the Journal of Invasive Cardiology:
J INVAS CARDIOL 2003;15:693-698
From the Primary Angioplasty versus Thrombolysis in a Community Hospital (PATCH) Study, Acute Myocardial Infarction Care Improvement Initiative (AMICII) and Registry, The Park Nicollet Heart and Vascular Center at Methodist Hospital (JLT); Quality Resources, Methodist Hospital (TAG), and The Park Nicollet Institute (SA); St. Louis Park, Minnesota.
ABSTRACT: Study objective. Demonstrate improved efficiency of initial and subsequent in-hospital care following emergency department (ED) physician-initiated primary angioplasty (1ºPCI).
Methods. An observational study was undertaken in ST-elevation myocardial infarction patients presenting to a community hospital emergency department. Outcomes of patients who received ED physician-directed 1ºPCI were compared with patients previously treated by a mix of ED physician and cardiologist co-determined thrombolysis or 1ºPCI. A process improvement initiative supported the change to ED-directed 1ºPCI.
Results. The study included 287 eligible acute reperfusion patients. Median door-to-balloon time (MDBT) improved from 88 minutes (95% CI, 80-96) to 61 minutes (95% CI, 57-70; p p p = 0.0039), angiotensin-converting-enzyme inhibitor (p p = 0.0039), with favorable trends in survival to discharge, and 30-day major adverse cardiac events (MACE).
Conclusions. Conversion to ED physician-initiated 1ºPCI for ST-elevation myocardial infarction significantly improved efficiency of care as measured by MDBT, NSI and LOS. Effectiveness measures, including survival to discharge, discharge medications and 30-day MACE, demonstrated improvement or favorable trends.
This article is reprinted with permission from the Journal of Invasive Cardiology:
J INVAS CARDIOL 2003;15:693-698
Evidence continues to mount supporting primary percutaneous coronary intervention (1ºPCI) as the strategy of choice for opening acutely occluded coronary arteries.1-6 Many institutions, however, struggle to meet the American College of Cardiology and American Heart Association (ACC/AHA) 90-minute guideline specification and 60-minute benchmark for Median Door-to-Balloon Time (MDBT) in acute myocardial infarction (AMI).7 Common patient factors for delay of door-to-balloon time beyond 120 minutes include: age greater than 65 years, female gender, contraindication to thrombolytics, and absence of chest pain at presentation.8 Achieving timely reperfusion through 1ºPCI is particularly challenging on weekends, holidays and after business hours (off-hours), when MDBT is significantly longer.9 Available data suggest that, in the current model, most PCI-capable institutions limit access to 1ºPCI, with only a portion of intervention-eligible patients receiving this form of revascularization and with limited success in efforts to improve door-to-balloon time.10
Barriers to efficient 1ºPCI include delays introduced by the lack of a standard, institution-specific strategy for selection of thrombolysis versus 1ºPCI.11-13 Lacking this standardized approach, the reperfusion method is commonly determined at infarct presentation with patients presenting on weekdays most likely to receive 1ºPCI, while those arriving off-hours more commonly receive thrombolytics. Considerable time is lost during the interval from electrocardiogram (ECG) to decision, while ED physicians treating patients with confirmed ST-elevation myocardial infarction (STEMI) track down the physicians authorized to determine the specific reperfusion strategy. Similar delay was recognized in the early thrombolytic era when ED physicians were required to obtain attending or consulting physician permission prior to administration of thrombolytic and was corrected by authorizing ED physicians to initiate thrombolytic therapy without consultation. Converting to 1ºPCI for all acute interventions and requiring the emergency physician to initiate the 1ºPCI process could eliminate unnecessary and time-consuming decision points and their accompanying communication lag.
Establishing the clinical and business case for process change. A literature review, cost and feasibility analysis (Appendix 1) supported the process change to 1ºPCI for all AMI reperfusion. Prior to this 1999 analysis, the institution had achieved a modicum of success with 1ºPCI demonstrating MDBT of 88 minutes, just meeting the 90-minute guideline specification. During the feasibility analysis, it was recognized that conversion to mechanical reperfusion for all patients would result in greater utilization of interventional services off-hours, prolonging the time to intervention unless additional timesavings were realized. The requirement to contact other on-call physicians prior to contacting the Interventional Cardiologist and bringing in the interventional care team was identified as a non-value-added step in a committed 1ºPCI program. Although it was recognized, the ED physician might err in the diagnosis of STEMI experience had demonstrated ED competency to identify intervention-appropriate infarcts. Process-map analysis of primary angioplasty indicated that elimination of superfluous communication steps would reduce MDBT by 20 minutes. Further improvements in the paging process and transfer to intervention suggested the ACC/AHA 60-minute benchmark could be achieved.
The interval between decision and intervention would provide an opportunity for the interventional cardiologist to confirm the appropriateness to proceed, address any patient safety issues, and secure consent for the procedure. As in the prior mixed thrombolytic and 1ºPCI experience, the ED physician could fax the ECG and consult the interventional cardiologist via phone in questionable cases. It was predicted that the additional staffing cost (then about $400) of unnecessarily calling in an interventional team off-hours would rarely occur and have negligible effect on overall cost. Moreover, with 80% of (then) thrombolytic patients undergoing subsequent angiography and 65% receiving further intervention prior to discharge, the accompanying cost analysis suggested that for each patient converted to 1ºPCI, the cost of hospitalization would be reduced by $1,900 to $4,000. In a previous report, Stone and colleagues14 compared cost effectiveness of angioplasty versus thrombolysis in a randomized trial and failed to find significant cost advantage. That study, however, proved to be a charge comparison and was undertaken in the pre-stent era, with only 35% of patients randomized to thrombolytic receiving subsequent PCI. It was determined the institution’s analysis of its then (1999) in-house practice and use of actual institutional cost data more accurately depicted the existing situation and improvement opportunity. Favorable cost analysis with predicted improved efficiency supported the clinical decision to convert to 1ºPCI for all AMI reperfusion commencing in May 2000.
Methods
An observational study of immediate and 30-day post-hospitalization outcomes was performed on STEMI patients treated before and after implementation of process changes intended to reduce treatment time by converting to emergency physician initiated 1ºPCI. Baseline therapy consisted of a near 50:50 mix of thrombolysis or 1ºPCI, codetermined by the ED physician and cardiologist at the time of presentation (cohort A). This baseline practice had introduced delay in the therapy of thrombolytic and 1ºPCI patients alike, as the ED physician worked through noninterventional and interventional cardiologists until the method of treatment was determined. The improved process converted interventional therapy to ED-physician-directed 1ºPCI for STEMI patients and eliminated routine thrombolytic therapy. Thrombolysis would only be administered in the event that access to intervention would be delayed. A delay of greater than 90 minutes would be considered potentially excessive. The ED physician was authorized to initiate the group page established to simultaneously notify the interventional cardiologist and team of the treatment opportunity. The ED process was further modified to expedite transfer to the interventional laboratory. Standard ED therapy included aspirin, unfractionated heparin, beta-blocker, and the glycoprotein IIb/IIIa receptor inhibitor, tirofiban. Clopidogrel, 300 mg, was administered following the procedure and continued at a dosage of 75 mg per day for one month. Angiotensin-converting enzyme inhibitor, anti-lipid and smoking cessation therapies were encouraged in the first 24 hours.
Cohort A included all eligible patients treated with thrombolytic or 1ºPCI from April 1998 through June 2000. Cohort B included all eligible patients admitted from January through December 2001 treated by the improved process, following the period of monitored transition to the new process. All nontransfer patients presenting with AMI symptoms and compatible electrocardiogram were considered for entry. Entry criteria included:
Ongoing symptoms of less than 24 hours (usually Patients presenting with nondiagnostic ECG treated on the basis of subsequent ST elevation, echocardiographic evidence of acute ischemia, or clinical suspicion of evolving MI in the face of ongoing symptoms without objective ECG findings, were excluded even if they went on to intervention. Data was collected by electronic and paper chart review and compiled in the institution’s electronic acute myocardial infarction registry database. The registry contains clinical data on more than 5,000 AMI patients treated since 1990. Electronic and/or telephonic follow-up was employed to establish 30-day post-hospitalization event rates. In addition to the primary efficiency measures of MDBT, necessary subsequent interventions (NSI), and length of stay (LOS), data on age, gender, infarct location, presence of diabetes, hypertension, smoking, congestive heart failure, and presentation with shock or subsequent cardiogenic shock were collected. NSI was defined as the observed use of operator-determined additional interventional therapy (usually PCI, but occasionally coronary artery bypass surgery) prior to discharge. Though not an independent variable, NSI proved helpful in understanding patient care during the remainder of hospitalization. Secondary endpoint effectiveness measures included survival to discharge, survival in patients aged 75 years or older, discharge medications, and total major adverse cardiac events (MACE) 30 days post-discharge. MACE was defined as death, recurrent myocardial infarction, or recurrent symptoms requiring additional cardiac intervention within 30 days of discharge. As noted in the introduction, other investigators have established the effectiveness of primary PCI. The effectiveness data provided here is intended to clarify the treatment and outcome of patients before and after fundamental process changes were enacted to improve 1ºPCI process efficiency. Institutional review board approval to perform patient follow-up and submit data for external peer review was obtained on June 12, 2001. Statistical analysis and relevance of comparison cohorts. The institution’s quality resources department collected data and performed initial analyses. The research and education department provided statistical oversight. A statistical analysis of MDBT for the two cohorts supported the elimination of the unstable period separating the statistically distinct and stable baseline and post-transition intervals. MDBT for patients treated by 1ºPCI during the transition interval is reported, without further analysis. The baseline interval was chosen for relevance, as it established the preimproved state of the process and demonstrated achievement of the ACC/AHA 90-minute guideline specification. Cohort A also reflected the near 50:50 mix of thrombolysis and 1ºPCI currently practiced in many PCI-capable institutions. Cohort B demonstrated the outcomes of STEMI treated by the improved process. Outcomes of patients with unsuccessful PCI in both cohorts are reported. The test of independent medians was used to determine statistical significance of the decrease in MDBT and LOS. An independent T-test was used to determine the significance of the decrease in mean-door-to-balloon time. Fisher’s exact test confirmed reduction in need for subsequent in-hospital intervention and effectiveness outcomes with significant improvement or favorable trends of clinical interest. Results The institution treated a total of 437 of 1,303 (33.5%) consecutive AMI patients admitted between April 1, 1998 and December 31, 2001 with thrombolysis or 1ºPCI for acute reperfusion. Typical of an unselected infarct population, only a portion of this group was eligible for immediate revascularization. Not all of the 437 patients undergoing acute intervention were eligible for the study by criteria including: symptoms of less than 24 hours duration; patient consent; STEMI or left bundle branch block (LBBB) on first hospital ECG; and completed treatment permitting calculation of the actual door-to-balloon or needle time. Study ineligible, treated patients were diagnosed by subsequent ECG demonstrating ST elevation, wall motion abnormality by acute echocardiography, or persistent symptoms despite therapy prompting angiographic evaluation for refractory unstable angina. Of the 437 acute intervention patients, 45 patients treated with 1ºPCI during the transition period are included in Figure 1, but not analyzed further as statistical review showed the new process had not yet stabilized. Another 287 (65%) met the eligibility requirements for full analysis. Planned 30-day follow-up was completed in 257 of 287 (89.5%) patients. Baseline cohort A comprised 206 of 437 patients (47.1%) meeting the study’s inclusion criteria. Table 1 lists demographics of thrombolytic and 1ºPCI patients. The baseline characteristics of the thrombolytic and 1ºPCI subgroups of cohort A and cohort B 1ºPCI patients appear clinically similar. Figure 1 demonstrates the cumulative percentage of arteries opened from time of arrival at baseline (April 1998 through June 2000), during transition (July to December 2000), and as the MDBT approached benchmark (2001). The 1ºPCI success rate, defined as restoration of TIMI Grade 3 (normal epicardial) coronary artery flow, was 99 of 114 (86.8%) for cohort A and 81 of 93 (87.1%) for cohort B. This reflects a good success rate for an unselected patient population as only patients presenting in transfer from an outside institution or with a nondiagnostic first electrocardiogram were excluded. Patients with diffuse vascular disease, congestive heart failure, renal failure, cardiogenic shock, or advanced age were not excluded. The few failed PCI patients faired at least as well as thrombolytic patients. Their survival and median LOS with unsuccessful 1ºPCI was 14/15 (93%) and 4 days in cohort A and 10/12 (83%) and 4 days for Cohort B. Of cohort B patients, 20 of 81 (24.7%) were over 75 years of age, reflecting the broad application of primary angioplasty following conversion to the new process. After process change, no intervention patient was treated with thrombolysis. Using the prespecified criteria, no instance of inappropriate ED request for interventional services was identified. Four experienced interventional cardiologists performed all procedures through June 2001, when a fifth joined. The 81 cohort B patients were evaluated to determine the effects of the process change to ED physician-directed 1ºPCI. Table 2 summarizes eligible patients’ outcomes with comparison to cohort A. The primary outcomes efficiency measure MDBT improved from 88 minutes (95% CI, 80“96) to 61 minutes (95% CI, 57“70; p p = 0.0015). NSI decreased from 65.4% (95% CI, 55.6“74.4%) for thrombolytic patients in cohort A and 3% (95% CI, 0.6“8.6%) for 1ºPCI patients in cohort A to 1.2% (95% CI, 0“6.7%) for 1ºPCI patients in cohort B (p p p =0.0039), ACE inhibitor (p p = 0.0039). Survival to discharge and 30-day MACE demonstrated favorable trends. Survival to discharge improved from 93.5% (95% CI, 87.0“97.3%) among thrombolytic patients and 96% (95% CI, 90.0“98.9%) for 1ºPCI patients in cohort A to 98.8% (95% CI, 93.3“100%) in cohort B. All 20 cohort B patients over age 75 survived to discharge (p = 0.0873). Thirty-day MACE was 10.3% (95% CI, 5.2%“17.6%) for patients receiving initial thrombolysis (plus frequent [65%] additional in-hospital intervention) and 11.1% (95% CI, 5.7“19.0%) for initial 1ºPCI patients in cohort A (with only 3% NSI); this subsequently fell to 4.9% (95% CI, 1.4“12.2%) in the ED-physician directed cohort B 1ºPCI group (with 1.2% NSI) (p = 0.3139). Discussion The recent fervor for 1ºPCI is tempered by the difficulty many institutions experience providing timely reperfusion. There has been little discussion to date regarding the negative effect determining the type of reperfusion (thrombolytic versus 1ºPCI) at presentation and subsequent ED care processes exert on door-to-balloon time. Most clinicians have assumed the as fast as possible tenet applies and suffices in providing optimal care. However, the evolution of 1ºPCI as an alternative to thrombolysis has complicated the acute intervention process in institutions offering both therapies. Factors, such as who is working the ED, who is on call for cardiology and whether it is a workday or a weekend, create a complicated, uncontrolled process where communication delays prevail. The complexity of determining reperfusion in this setting has prolonged door-to-treatment time for thrombolysis and 1ºPCI alike. Paradoxically, encountering this time barrier may have deterred many intervention-capable institutions from adopting 1ºPCI as the sole method of reperfusion and achieving improved door-to-balloon times. AMI management strategies that do not permit ED physicians to initiate acute intervention are based on the concept that in order to maintain patient safety and contain costs, cardiovascular specialist oversight is required prior to initiating the primary angioplasty process. Unfortunately, time lost while ED physicians await approval to contact the interventional team prolongs door to treatment time. Communication delays associated with these interactions compound the problem and undoubtedly contribute to the nationally observed longer-than-specified MDBT. Physicians, particularly cardiologists, may be alarmed to learn they are connected to a non-value-added step in the acute MI care process. In reality, it is the placement of cardiologist control at the initiation of the process that reduces process efficiency, not the control itself. In order to meet current ACC/AHA guidelines for MDBT, initiation of the reperfusion process should be separated from the intervening cardiologist’s obligation to confirm appropriateness, safety, and patient consent to proceed. This separation requires empowerment of ED physicians to initiate 1ºPCI. The intervening cardiologist remains free to exercise diligence to insure appropriate care without delaying the process in most cases. This outcomes study demonstrates improved efficiency of care on conversion to ED-physician-initiated 1ºPCI for STEMI in an experienced, intervention-capable community hospital. The process change has reduced MDBT from the minimum acceptable 90-minute guideline specification to approach the 60-minute door to treatment benchmark. Nearly 80% of arteries successfully opened met the 90-minute guideline despite a 12% increase in off-hours interventions. This improvement in MDBT was achieved by reducing the choice of reperfusion therapy to 1ºPCI only, and equally important, delegating that decision to the emergency physician. The intervening cardiologist continued to determine patient suitability. Suitability did not prove to be a care issue in this study and does not justify delaying treatment of all intervention patients in order to prevent a minority of inappropriate requests for treatment. The results reported here may be extended to other intervention-capable institutions that currently apply a mix of thrombolysis and angioplasty or require discussion with cardiology prior to initiating their primary PCI process. The institution reporting these results is a combined community hospital and clinic of nearly 500 physicians, where the medical staff and noninterventional cardiologists elected to defer the history and physical of AMI patients to the treating interventional cardiologist. However, a change to ED-physician-initiated 1ºPCI need not disrupt normal referral patterns in a fee-for-service environment. ED-physician-initiated 1ºPCI should work equally well with the primary care attending and noninterventional cardiologist participation. Their initial patient access should be limited to the interval preceding transfer to the catheterization laboratory, and the interventional team should be summoned on diagnosis. These non-interventional physicians provide added value by assuming the tasks of post-infarction care, risk management and patient education, along with supervising progress in cardiac rehabilitation. The current national MDBT of more than 100 minutes and well-documented gap in application of evidence-supported treatment indicate the national system of AMI care has been slow to respond to evidence-based therapeutic imperatives.15,16 This outcomes study provides evidence that the ACC/AHA 60-minute benchmark for MDBT may be safely approached in capable institutions when thrombolytic therapy is eliminated and ED-physician-initiated 1ºPCI is streamlined. Improvement in subsequent care efficiency is also suggested by a reduced need for subsequent in-hospital interventions. Thrombolysis is about 50% effective, but many patients (65% in this series) require additional intervention prior to discharge. In comparison, 1ºPCI is approximately 90% effective (87% in this series), with 3% or fewer demonstrating NSI. Reduced LOS further attests to the increased efficiency of ED physician-initiated 1º PCI with improvement or favorable trends in discharge medications, survival, and 30-day MACE. Resource and evidence-based arguments support thrombolysis or 1ºPCI for STEMI according to the local institution’s capabilities and expertise. This study reveals that delays incurred by using both therapies in a suboptimal manner, adversely affect 1º PCI efficiency and may reduce the effectiveness of care. Institutions and clinicians should adopt standardized AMI care processes that broadly apply evidence-based pharmacotherapies and (in PCI capable centers) more efficient primary angioplasty. ED-physician-initiated 1ºPCI significantly improves treatment efficiency and facilitates meeting the 60-minute door to treatment benchmark. This work was supported in part by an unrestricted educational grant from Merck, Inc. Manuscript received January 28, 2003, manuscript accepted October 9, 2003. Address reprint requests to: Jackson Thatcher, MD, Park Nicollet Heart and Vascular Center at Methodist Hospital, 6490 Excelsior Boulevard, Suite 200W, St. Louis Park, MN 55426. E-mail: thatcj@parknicollet.com
1. Grines CL, Browne KF, Marco J, et al. A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1993;328:673-679.
2. Stone GW, Brodie BR, Griffin JJ, et al. Prospective, multicenter study of the safety and feasibility of primary stenting in acute myocardial infarction: In-hospital and 30-day results of the PAMI stent pilot trial. J Am Coll Cardiol 1998;31:23-30.
3. O’Neill WW, de Boer MJ, Gibbons RJ, et al. Lessons from the pooled outcome of PAMI, Zwolle, and Mayo Clinic randomized trials of primary angioplasty versus thrombolytic therapy of acute myocardial infarction. J Invas Cardiol 1998;10:4A-10A.
4. Zahn R, Schiele R, Schneider S, et al. Primary angioplasty versus intravenous thrombolysis in acute myocardial infarction: Can we identify subgroups of patients benefiting most from primary angioplasty? J Am Coll Cardiol 2001;37:1827-1835.
5. Zijlstra F, Hoorntje J, de Boer MJ, et al. Long-term benefit of primary angioplasty as compared with thrombolytic therapy for acute myocardial infarction. N Eng J Med 1999; 341: 1413-1419.
6. Berger AK, Radford MJ, Wang Y, et al. Thrombolytic therapy in older patients. J Am Coll Cardiol 2000;36:366-374.
7. Ryan TJ, Anderson JL, Antman EM, et al. ACC/AHA guidelines for the management of patients with acute myocardial infarction: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction). J Am Coll Cardiol 1996;28:1328-1428.
8. Angeja BG, Gibson CM, Chin R, et al. Predictors of door-to-balloon delay in primary angioplasty. Am J Cardiol 2002;89:1156-1161.
9. Beohar N, Chandwaney R, Goodreau L, et al. In-hospital and long-term outcomes of patients with acute myocardial infarction undergoing direct angioplasty during regular and after hours. J Invas Cardiol 2001;13:669-672.
10. Mehta RH, Montoye CK, Gallogly M, et al. Improving quality of care for acute myocardial infarction - The Guidelines Applied in Practice (GAP) Initiative. JAMA 2002;287:1269-1276.
11. Doorey A, Patel S, Reese C, et al. Dangers of delay of initiation of either thrombolysis or primary angioplasty in acute myocardial infarction with increasing use of primary angioplasty. Am J Cardiol 1998:81:1173-1177.
12. Thatcher JL, Gilseth TA. Experience with process improvement and outcomes analysis in acute myocardial infarction in a community hospital, 1990-1999. (Abstr) J Invas Cardiol 2000;12:574.
13. Thatcher JL, Gilseth TA. Experience with process improvement and outcomes analysis in acute myocardial infarction in a community hospital, 1999-2000. Cath Lab Digest 2001;9:1-10.
14. Stone GW, Grines CL, Rothbaum D, et al. Analysis of the relative costs and effectiveness of primary angioplasty versus tissue-type plasminogen activator: The Primary Angioplasty in Myocardial Infarction (PAMI) Trial. J Am Coll Cardiol 1997;29:901-907.
15. Cannon CP, Cecilia Bahit M, Haugland JM, et al. Underutilization of evidence-based medications in acute ST elevation myocardial infarction: Results of the Thrombolysis In Myocardial Infarction (TIMI) 9 Registry. Crit Path Cardiol 2002;1:44-52.
16. Aversano T, Aversano LT, Passamani E, et al. Thrombolytic therapy vs. primary percutaneous coronary intervention for myocardial infarction in patients presenting to hospitals without on-site cardiac surgery - A randomized controlled trial. JAMA 2002;287:1943-1951.
