The use of percutaneous coronary intervention (PCI) for coronary artery revascularization has grown considerably since the first balloon angioplasty in 1977. Improvements in equipment and drug therapies plus the development of coronary artery stents have all contributed to the growth of PCI as an alternative to surgical revascularization. Compared with the early days of balloon angioplasty, when 1–2.5% of patients died and 1.9–5.8% required urgent coronary artery bypass graft (CABG) surgery, recent data from high-volume centers show an in-laboratory mortality rate of 0.2% or less and a 0.3–0.6% incidence of urgent CABG.1–4 As the safety and outcomes of PCI have steadily improved, the indications for the procedure have also expanded. PCI is now the preferred therapy for patients with ST-segment elevation myocardial infarction (STEMI), and in this setting is called primary PCI.5 However, the superior outcomes of primary PCI are adversely affected by time delays that may be encountered before the patient arrives in the cardiac catheterization (cath) laboratory.6–8 To avoid the inherent delays that occur when transferring patients and to extend this therapy to a larger number of patients, some facilities began to perform primary PCI in the absence of on-site cardiac surgery and reported excellent outcomes.9,10 As the number of patients with STEMI at any location is relatively small, some facilities also started performing elective PCIs in an effort to maintain proficiency among the support staff and to have adequate procedure volumes to justify the operation of such programs.11–13 Although accepted in many countries abroad, the performance of PCI without on-site cardiac surgery remains controversial in the US.14
Percutaneous Coronary Intervention without On-site Surgery in the US
Data are not easily found on the performance of PCI at facilities without on-site surgical back-up in the US. In 2007, the Society for Cardiovascular Angiography and Interventions (SCAI) published an Expert Consensus Document on the topic of PCI without on-site surgical back-up that contained survey data on the performance of PCI without on-site surgery in the US.15 This is a dynamic situation as some states have changed their regulations since that initial survey.
An updated US map now shows only seven states where neither primary nor elective PCI is allowed without on-site surgery, nine states where only primary PCI is allowed, and 34 states where both primary and elective PCI without on-site surgery are performed, although in some states it is allowed only as part of a research study or demonstration project (see Figure 1). The number of patients receiving PCI at facilities without on-site surgery in the US is unknown, but recent data from the CathPCI™ Registry of the National Cardiovascular Data Registry (NCDR™) suggest the number is increasing.16
Guideline Statements About Percutaneous Coronary Intervention without On-site Surgery
Guidelines documents for PCI procedures have been published by professional organizations within the US and abroad (see Table 1).17–25 Current guidelines from professional organizations in the US do not endorse elective PCI in facilities without on-site surgical back-up, but acknowledge that primary PCI for STEMI is acceptable under carefully controlled and monitored circumstances. Additional recommendations from the US guidelines state that primary PCI may be performed by higher-volume operators experienced in both elective PCI and primary PCI with ongoing activity levels of more than 75 elective PCI procedures per year and, ideally, annual primary PCI activity levels of at least 11 per year.17 In contrast, the 2005 European Society of Cardiology (ESC) guidelines do not comment on PCI without on-site cardiac surgery,18 and guidelines from other countries generally consider this form of PCI delivery more favorably.
Published Data on Percutaneous Coronary Intervention without On-site Surgery
The first reports of PCI performed without on-site surgery appeared in the literature in the early 1990s and there are now about 40 published reports on this topic.15 Simple aggregation or meta-analysis of these studies would be problematic for several reasons. First, some studies focus on either primary or elective PCI performed without on-site surgical back-up, while others include all PCI patients. Second, some studies apply strict screening criteria to identify only low-risk patients for PCI without on-site surgery, whereas others include several high-risk subgroups. Finally, these studies span a time period from 1990 to 2008 and thus incorporate many changes in the procedure, such as glycoprotein IIb/IIIa inhibitors and coronary artery stents, which alter both the short- and long-term outcome. Even the total patient number within some reports is not easily derived because some studies are expanding experiences within the same registry and thus may duplicate early patient experiences.10,11 As coronary artery stents resulted in a substantial decrease in the need for emergency CABG for a failed PCI it is a logical point of separation, but the use of stents is not consistently reported among these studies.
Most recent reports show that PCI without on-site surgical back-up is performed with a high success rate, a low in-hospital mortality rate, and a low rate of urgent cardiac surgery.11–13,26 However, these data are subject to a number of limitations. All of the published studies rely on retrospective reviews or registries and thus are subject to unintentional bias and other methodological concerns. There is also the possibility of publication bias as there is no requirement for public reporting of programs that have not succeeded at PCI without on-site surgical back-up.
Recent Reports of Percutaneous Coronary Intervention without On-site Cardiac Surgery
A contemporary view of the status of PCI without on-site surgery can be acquired from examining recent registry data and publications from the US and abroad.
Large Registry Reports
Three large registries have reported data on the performance of PCI without on-site cardiac surgery in a total of almost 420,000 patients (see Table 2). The British Cardiovascular Interventional Society (BCIS), the Swedish Coronary Angiography and Angioplasty Registry (SCAAR), and the NCDR all show low mortality rates at facilities without on-site surgical back-up.27–29 Furthermore, the need for urgent bypass surgery at such facilities is uniformly lower than at facilities with surgery on-site. This is likely a reflection of an effective screening process to exclude high-risk cases from facilities without on-site surgery.
Reports from the US
The Mayo Clinic has supported a satellite program at a community hospital without on-site surgery for several years (see Table 3). Ting et al. reported their experience in 1,007 PCIs performed at a facility approximately 85 miles from their main facility, where cardiac surgery was available.11 PCIs at the facility without on-site surgery were matched one-to-one with those performed at the main facility with surgery on-site. Clinical outcomes (in-hospital procedural success, death, any myocardial infarction [MI], Q-wave MI, emergency CABG surgery, and follow-up survival) were compared. Among 722 elective PCIs, procedural success was higher (97%) at the facility without on-site surgery compared with the facility that had surgery (95%; p=0.046). Despite the case-matching process, this was likely related to the preferential selection of lower-risk cases to be performed at the facility without on-site surgery. Among 285 primary PCIs, procedural success was not different among the two facilities (93% at the facility without and 96% at the facility with on-site surgery; p=0.085). No patients at the facility without on-site surgery required emergent transfer for cardiac surgery, and survival after two years of follow-up was similar for both elective PCI and primary PCI patients at the two facilities.
A series of 1,090 elective PCIs was recently reported by Frutkin and colleagues at a community hospital 15 miles from the supporting facility (Mid-America Heart Institute) with cardiac surgery.26 In this series, only elective PCIs were performed at this facility and patients with STEMI were preferentially transferred to the facility with on-site surgery. The elective PCI program used experienced interventionalists, technicians, and nurses, a tested helicopter transport protocol, a well-equipped cath laboratory, and a quality assurance process. Patients treated at the facility without on-site surgery had unadjusted frequencies of procedural success (93 versus 94%; p=NS), Q-wave MI (0.3 versus 0.3%; p=NS), emergency CABG surgery (0.2 versus 0.03%; p=0.09), vascular complications (0.6 versus 0.6%; p=NS), and in-hospital death (0.1 versus 0.8%; p=0.002) compared with the facility that had on-site surgery. Two patients required transfer for emergency CABG surgery without adverse effects. To adjust for the lower-risk cases performed at the hospital without on-site surgery, observed frequencies of in-hospital death, a combined end-point of Q-wave MI/emergency CABG surgery, and vascular complications were compared using prediction models developed by the Northern New England Cardiovascular Disease Study Group.30 Fewer in-hospital deaths and vascular complications were observed at the facility without on-site surgery than predicted by the model, thus confirming the safety of this approach when strict program requirements are used. A study within the Veterans Affairs system that excluded patients with STEMI but included patients with other acute coronary syndromes also reported excellent outcomes.31
Reports from Abroad
Many of the early studies related to PCI without on-site surgical back-up were from Europe, where this practice has been adopted more widely than in the US (see Table 3). A contemporary study from Norway showed excellent safety and in-hospital outcomes at a facility without on-site cardiac surgery compared with a regional facility that has surgery onsite.32 Likewise, excellent safety and outcomes were reported in a series of 1,348 patients at a remote facility without on-site surgery in Tasmania.33 In this series, no patients were excluded and 18% of the patients presented with a STEMI (mortality 3.7%). Only one patient required transfer for urgent CABG because of an ascending aortic dissection caused by the guiding catheter.
Unanswered Questions Related to Percutaneous Coronary Intervention without On-site Surgical Back-up
Although the use of PCI without on-site surgery in the US is expanding, it remains a controversial matter both within and external to the interventional community. Data from the Global Registry of Acute Coronary Events (GRACE) shows an increase in the use of reperfusion therapy among patients with STEMI, but only 44% of patients receive primary PCI.34 Despite the benefits of primary PCI, timely access to this service in the US remains a challenge. Only 25% of the acute care hospitals in the US are capable of providing PCI.35 One interpretation of these facts would be that more PCI centers are needed. However, the actual impact of opening more PCI centers at facilities without on-site surgery is questionable. Using census data from 2000, it was estimated that nearly 80% of the adult population live within 60 minutes of a PCI-capable hospital and among those living closer to non-PCI hospitals, almost three-quarters would experience <30 minutes of additional delay by direct referral to a PCI hospital.36 Furthermore, a recent study examining data from Michigan estimated that providing PCI without onsite surgery improved access to <5% of the population.37
Currently, there are three models for the delivery of PCI care in patients with STEMI. One model is to develop PCI programs at community hospitals without on-site surgery in an attempt to provide rapid primary PCI to patients in their local community.9,14 Although several reports document that this can be performed safely, it requires a high level of physician and facility support and a commitment to maintain high standards of quality. Most would argue that it is inappropriate to open more low-volume PCI centers if they are not firmly based on the health needs of the community. Opening a low-volume PCI program in close proximity to a high-volume program, thereby degrading the high-volume program, is not necessarily in the best interests of patients or the community. However, many factors besides distance can define a geographical area, including the level and availability of emergency transport services, response times of emergency medical transport, immediate availability of qualified cath lab personnel, and coverage by interventional cardiologists.
An alternative model is the ‘hub-and-spoke’ model where a referral network is established to transfer all STEMI patients to a central high-volume facility.38–40 In this model, the central hospital works closely with outlying hospitals to develop treatment and transfer protocols designed to standardize care and minimize transfer times. This model has been promoted successfully at a state-wide level.41 A variant of the hub-and-spoke model is the so-called ‘bypass model.’ In this model, there is an enhanced effort to recognize patients with STEMI in the field and then transport patients directly to PCI-capable facilities, thereby bypassing facilities without PCI capability.42
The need to develop a national strategy for the timely treatment of STEMI has been highlighted along with the potential barriers to this goal.43,44 Although debate has focused on whether facilities that offer PCI without on-site surgery should exist, a more meaningful approach would focus on the goal of providing the best possible care to patients who require PCI, regardless of the setting. In some areas, the appropriate solution may be the development of a hub-and-spoke or bypass system for the efficient transfer of patients to a PCI facility. However, in other areas, developing a PCI program at a hospital without on-site surgery may be preferable. It is clear that there are multiple opportunities to improve the care of patients requiring PCI, and the best solution for one area may not be the best solution for another area.
Desires for personal or institutional financial gain, prestige, market share, or other similar motives should not be part of the decision-making process in determining the need for a PCI program. These considerations apply equally to those wishing to start a new PCI program without on-site back-up and those wishing to protect existing programs with on-site back-up. Recent changes in reimbursement policies have substantially reduced payments for elective PCI procedures, making the incentive to start a PCI program at a community hospital less attractive. In the final analysis, every PCI procedure, regardless of where it is performed, should be of the highest possible quality. This means the PCI is performed for appropriate clinical indications and by a skilled operator with documented satisfactory outcomes in a laboratory with appropriate equipment and personnel that has careful tracking of patient outcomes and corrective mechanisms in place to manage individual operator or laboratory outcome data that fall below national standards. Ensuring that all PCI programs meet appropriate performance metrics is likely to save more lives than requiring all PCI programs to have on-site surgery.