By Christopher Guadagnino, Ph.D.
Medical researchers are not in agreement that the benefits to patients outweigh the increased risks associated with PCI without open heart backup on-site, and current guidelines by the American College of Cardiology and the American Heart Association (ACC/AHA) warn that the practice is still contraindicated based on the absence of compelling evidence from randomized trials of a favorable risk/benefit ratio. Current Pa. hospital regulations require that elective PCIs be performed only at health care facilities that have on-site open heart surgery, as a backup to treat unexpected complications of the intervention.
A little over four years ago, the DOH had initially denied requests by several hospitals for exceptions to that regulation, prompting hospitals to file administrative appeals, which were resolved by settlement agreements between the DOH and each of the hospitals, according to a Legislative Budget and Finance Committee (LBFC) report released last April. The settlement agreements created two-year waivers during which time elective PCIs could be performed without on-site open heart surgery.
During the course of the waiver program, a nationwide clinical trial was being launched to study the safety of the procedure. Led by Johns Hopkins University School of Medicine cardiologist and researcher Thomas Aversano, M.D., and conducted by the Atlantic Cardiovascular Patient Outcomes Research Team (CPORT), the 18,000-patient study is comparing elective PCI outcomes at hospitals with and without surgical backup and hopes to have its conclusions by 2008.
When the DOH threatened not to renew the waivers unless they agreed to enroll in the CPORT trial, hospitals balked – some because they did not want to lose 25 percent of their elective PCI patients to tertiary institutions, as required by the randomized trial protocol, and others because they could not meet the trial’s minimum annual PCI volume requirement of 200. Some hospitals sued the department, while Pa. legislators also began to move a bill that would permit the hospitals to continue to perform elective PCI without CPORT trial participation.
Pressured by litigation and looming legislation, the DOH has now agreed to resolve the impasse by offering waiver extensions to hospitals that do one of three things: join the CPORT trial, join an in-state registry that collects and studies quality outcomes, or agree to a supervisory and cross-training arrangement with a tertiary hospital to oversee quality.
One of the original PCI waiver sites – Jeanes Hospital in Philadelphia – has since launched on open heart surgery program, obviating the need for a waiver, and illustrating the behavior of the free market in the absence of Certificate of Need.
Critics question the safety of allowing the waivers to continue without the strict oversight and protocols of the national trial. An institution’s procedure volume remains a widely accepted surrogate measure of quality, and only the CPORT and registry options have minimum annual procedure volume requirements.
LBFC’s report, Quality Assurance for Specialized Clinical Services – which was mandated by a House Resolution which Pa. Rep. Phyllis Mundy (D-Luzerne) sponsored and which had a unanimous vote – criticized the manner in which the DOH monitors cardiac procedures, particularly the PCI demonstration project, and recommended several reforms, including that the DOH should require all PCI waiver sites to join the CPORT trial and comply with all of its supervisory requirements and protocols. ACC’s Pa. Chapter also said that requiring existing PCI waiver sites to participate in the CPORT trial would be an effective way to monitor the safety of the programs and identify outliers early on.
Solid evidence is lacking that elective PCI without on-site surgery is as safe as with on-site surgery, and only a prospective, randomized trial can produce meaningful comparisons among sites, says Aversano. The CPORT trial errs on the conservative side of requiring some minimum elective PCI volume from each institution: at least 200 per year, he maintains. ACC/AHA recommends that institutions (with cardiac surgery back-up) perform 400 PCIs per year, or 200 per year for geographically isolated institutions; once volume falls below 200, there is good data that outcomes will become worse, says Aversano. “That doesn’t mean that volume will become worse at a single institution, but outcomes at many sites below 200 are clearly worse,” he adds. It is also difficult to assess quality in real-time at low-volume institutions because of the rarity of PCI complications which require cardiac surgery. It would take ten years for a center that performs 100 PCIs per year to experience a coronary perforation, which Aversano says happens once in 1,000 cases, on average.
To test the hypothesis that outcomes of elective PCI performed at hospitals without on-site cardiac surgery are not inferior to outcomes at hospitals with on-site surgery, Aversano’s trial is using mortality six weeks after PCI as its primary endpoint, and is also collecting and analyzing additional outcomes data for a nine-month interval, including the comparative incidence of death, heart attack, stroke, bleeding, heart failure, bypass surgery, vascular surgery, and target vessel revascularization; comparative incidence and classification of heart failure and angina; comparative angiographic and clinical success rates; and the comparative cost of care.
Excluded from trial participation are patients who have unprotected critical left main coronary artery disease, patients who have poor heart function, and patients who are thought to require the use of niche devices that are generally unavailable at non-surgical cath labs – known in the industry as “rippers and scrapers,” Aversano says.
Interventional cardiologists who perform the elective PCIs must also meet ACC/AHA competency guidelines, including the minimum performance of 75 PCIs per year.
The CPORT trial is being monitored at three layers, Aversano explains: each participating hospital reports any adverse events to its own Institutional Review Board, to John’s Hopkins’ Institutional Review Board, and to an Independent Data Safety and Monitoring Board. Any of those three bodies can terminate the participation of a physician or hospital at any time for protocol violations, for example, if a site recruits a patient without obtaining informed consent, says Aversano.
The entire trial itself has “stopping rules” if the mortality difference between surgical and non-surgical sites exceeds a statistically significant threshold, Aversano notes.
The only PCI waiver site proceeding with the CPORT option is Chambersburg Hospital, according to Richard Lee, DOH’s deputy secretary for quality assurance.
Pressure to Abandon CPORT Requirement
Other Pa. waiver hospitals objected to CPORT’s requirements – either because they did not want to lose 25 percent of their elective PCI patients to tertiary institutions, or because they could not meet the trial’s minimum annual PCI volume requirement of 200 – and approached legislators to permit the waivers to continue without the CPORT requirement.
After negotiations failed to persuade the DOH to remove the CPORT requirement, Pa. Rep. Tom H. Killion (R-Chester/Delaware) sponsored a bill that would do that – HB 2443 – by mandating automatic renewal of waivers to hospitals that perform 200 PCIs annually, and mandating renewal for lower-volume hospitals that agree to close supervisory relationship with a hospital that has a cardiac surgery program.
According to Mundy, the waiver hospitals convinced their legislators that in order to compete financially with larger hospitals they needed to be permitted to perform elective PCIs, while they also asserted that since they had been granted a waiver and had invested in the equipment and staff they should be allowed to continue. The hospitals made a case that access to this procedure should be available close to home, which proved to be compelling arguments on the House Floor, says Mundy. “The issue of compromised quality was a difficult sell. The hospitals assured their representatives that quality and patient safety were fine,” she adds.
The House passed Killion’s bill, and the Senate amended it and sent it back to the House for a final vote. “My goal wasn’t to pass that bill,” says Killion. “My legislation was a hammer to get the DOH to budge,” he notes.
It worked, and the DOH agreed in late June to work out alternative waiver extension models with hospitals, says Killion.
Lee says the DOH saw movement of Killion’s bill as a statement of policy by the General Assembly to seek a way to extend the PCI waivers while satisfying safety and quality concerns.
The DOH regards the 200 annual PCI volume (combined elective and emergency) as a proxy for safety, says Lee, and has approved a waiver extension model that preserves that requirement, while adding an outcome reporting and analysis infrastructure involving one of two entities: Quality Insights of Pennsylvania or researchers at Temple University Hospital.
The option requires waiver hospitals to continue using ACC’s elective PCI patient selection criteria, as well as meet its proficiency volume standard for individual physicians (75 procedures annually), says Lee. Four PCI waiver hospitals – Jameson, Uniontown, Holy Redeemer and Montgomery – are planning to adopt this model, he notes.
Holy Redeemer Hospital has performed about 250 elective PCIs over each of the past two years with no emergency transfers because of complications, and it believes its approach to the procedure is safe and well-established – offering patients ready access to a life-saving procedure without the hospital needing to consider adding a costly cardiac surgery program that would further dilute bypass graft surgery volume at other hospitals, according to its Chief Medical Officer Charles Wagner, M.D. Although it met the PCI procedure volume requirements of CPORT, Holy Redeemer felt that the trial’s case randomization protocol would unnecessarily confuse patients, who would be told by their doctors that they might have to go elsewhere for the procedure, says Wagner.
When Holy Redeemer’s initial PCI waiver ended this August, the DOH agreed to let the hospital continue to perform the procedures while finalizing arrangements with Temple University Hospital’s Chief of Cardiology Alfred Bove, M.D. to participate in a PCI patient registry study, says Wagner. While Holy Redeemer continues to submit elective PCI outcome data to ACC’s National Cardiovascular Data Registry (NCDR) database, those data end upon patient discharge. Under its new arrangement, Temple will send quarterly follow-up surveys to Holy Redeemer’s elective PCI patients for one year to track their status, Wagner adds.
The survey will use many of the NCDR quality benchmarking parameters – including mortality, comorbidities, incidence of heart attack, hospital readmission, and additional PCIs performed – then compare the survey data to nationwide post-procedure outcomes, says Bove. Montgomery Hospital also plans to enroll in Temple’s PCI patient study, he notes.
The DOH is granting a two-year PCI waiver extension to hospitals using this model, and is requiring Bove to report back his study’s results, although Bove says he and the hospitals will ask for an additional year to track 12 months of outcome data for patients receiving elective PCI during the second year of the waiver extension.
Similar registry contracts are expected to be worked out soon between Quality Insights of Pennsylvania – the state’s Quality Improvement Organization – and Uniontown and Jameson Hospitals, according to Daniel Jones, Jr., Quality Insights’ chief operating officer. Hospitals would agree to allow Quality Insights to monitor their quarterly NCDR data and look for outcome and complication outliers relative to nationwide benchmarks, says John Brehm, M.D., chief medical officer of the West Virginia Medical Institute – the parent company of Quality Insights. If it detected an unusual frequency of a complication, Quality Insights would notify the DOH and broker a root cause analysis by an ACC consultant, notes Brehm, who says the ACC is pleased with this model and is promoting it in other states.
Lee acknowledges that the registry option does not achieve the research rigor of CPORT, but he says its volume and data monitoring features do fulfill the department’s primary goal of safety. “It was a compromise. We got quality protections, but gave up research rigor. Three or four hospitals had sued us to continue the original waiver conditions, objecting to the CPORT requirement. We could have spent the entire fall litigating, but they were interested in this other approach and we decided that it offered satisfactory quality,” says Lee.
Some hospitals approached the DOH with a third waiver extension option which involves tight coordination between a community hospital and a tertiary hospital with on-site cardiac surgery, while requiring no minimum annual procedure volume, says Lee. The model, as presented in a study published online March 10, 2006 in the Journal of the American College of Cardiology, was proven successful by the Mayo Clinic, which cross-trained and cross-credentialed cardiologists performing elective PCI procedures at one of its community hospitals without on-site surgical backup. The community hospital adhered to the Mayo Clinic’s strict protocols for case selection and PCI program requirements, while Mayo Clinic physicians also observed the community hospital cardiologists’ elective PCI procedures real-time through telemedicine links.
The DOH’s initial PCI waivers had no procedure volume requirements and two waiver hospitals – Somerset and Monongahela Valley Hospital – each had performed under 100 per year, with no opportunity to meet the 200-per-year requirement of the CPORT or registry options, according to Lee.
In late August the DOH reach a PCI waiver settlement agreement under which Somerset Hospital would be supervised by Conemaugh Valley Memorial Hospital, and the department anticipates a similar agreement between Mon Valley and another tertiary hospital, says Lee.
The two-year settlement agreement requires both Somerset and Conemaugh to submit their research design to provide elective PCI at Somerset to an Institutional Review Board that contains no members who are directly responsible for managing the financial operations of either hospital. Program requirements under the model include the following:
· Each cardiologist must perform a minimum of 75 PCIs per year, including 11 emergency PCIs.
· All Somerset cardiologists and allied health staff providing elective PCI shall be trained and credentialed at Conemaugh, and periodically rotate through its cardiac cath lab.
· Somerset and Conemaugh shall follow the same periprocedural and postprocedural protocols for patient care.
· For the first 12 months of the agreement, an interventional cardiologist credentialed by Conemaugh, who is not performing the PCI, shall observe and provide real-time consultation for every elective PCI performed at Somerset, and review Somerset PCI clinical data immediately following the procedure. These requirements are applicable unless the Conemaugh cardiologist determines and documents in writing that such observation and review is not necessary for each particular procedure.
· Somerset and Conemaugh shall develop and regularly test a protocol for rapid transport of patients with procedural complications requiring emergency cardiac surgery, capable of transport within 60 minutes.
· Both Conemaugh and Somerset shall collect postprocedural PCI patient outcome data at 6, 12 and 24 months – to be analyzed by an independent entity, and reported to both hospitals and to the DOH. The analysis shall include comparisons of the number and types of complications that occurred and comparisons to national benchmarks.
· The DOH can terminate the waiver and exclude elective PCI procedures from Somerset if the department determines that there is any threat to patient health and safety.
While the Mayo Clinic study involved telemedicine links between the tertiary and community hospitals, Somerset Hospital CEO Michael Farrell says that requirement would be impractical and cost prohibitive – costing over $200,000 for the equipment plus $5,000 to $6,000 per month to maintain the connection lines. Under its settlement agreement, Somerset gets clearance from a Conemaugh interventional cardiologist who reviews an elective PCI candidate before Somerset proceeds, and the two hospitals ensure that a 60 minutes emergency transfer capability is in place at that time, Farrell notes. Somerset has chosen Quality Insights of Pennsylvania to perform the comparison and benchmarking study, he adds.
Farrell says Somerset has performed more than 200 elective PCIs in nearly four years, with no reportable complications or emergency transfers. The hospital has one full-time interventional cardiologist on staff, who has been doing PCIs for six years. “That’s what drove us to request the waiver in the first place – we had been sending all the elective angioplasty cases up to Conemaugh, when we had all the facilities here – including the cardiologist.” Farrell says Somerset’s interventional cardiologist is “very conservative” and will send elective PCI patients to the tertiary institution if he believes there are any opportunities for complications.
Three hospitals will see their original waivers expire next July and will have to request another exception to the DOH regulations in order to continue performing PCI: Mount Nittany Medical Center, Grandview Hospital and Riddle Memorial Hospital.
Riddle performed about 60 elective PCIs during the first 10 months of its two-year waiver, and their goal is to be performing over 200 PCIs annually, meeting the requirement of the research registry waiver extension option, according to Janet Webb, RN, Riddle’s vice president of Clinical Services. Another interventional cardiology group has just started performing PCI at the hospital which she says should significantly increase the number performed.
Riddle had one emergency transfer of a PCI patient for bypass surgery. The patient got to the tertiary hospital in less than one hour, was stabilized and had open heart surgery later that afternoon with a good outcome. “That’s the myth of the on-site surgery requirement,” says Webb. “Even if the patient were at the hospital where the surgery was performed, it would not have been performed any sooner than it was. There are no surgical teams standing by in an OR during angioplasties.”
Riddle has the support of both Crozer Chester Medical Center and Lankenau Hospital to perform PCIs and has agreements with both hospitals to receive an emergency transfer within one hour 24 hours/7 days a week.
While Riddle plans to request another waiver by choosing the research registry option, Webb indicates that if necessary Riddle can participate in the Mayo Clinic option as a back up plan. “Riddle follows the quality standards that are part of that alternative now anyway, so it would be relatively easy to comply,” she says.
“CPORT is not an option as the study would require randomizing and transferring 25 percent of the hospital’s elective PCIs. This would decrease our numbers, I believe, even more than 25 percent due to patient choice of not entering the program which contradicts the argument for volume,” says Webb.
Initiating an open heart program at Riddle’s is “still on the table and if necessary, our CEO would support it,” Webb indicates, but Riddle has been trying to avoid doing open heart primarily because it makes less sense with the number of existing programs in the region, and would result in reducing the nearby hospitals’ volume, which is already declining with current treatments including drug therapy and angioplasty, adds Webb.
Criticism of Registry and Supervisory Approaches
Some critics continue to object to the waivers under any circumstance. Proliferation of health care services erodes quality when procedure and treatment volume is diluted at each individual facility, while duplication of equipment and staff at more facilities also adds costs to the system, maintains Mundy. “In my view the waivers should never have been granted in the first place. I believe the public should be able to rely on the Department of Health to assure the public’s interest in high quality health care is protected,” she says. “Some of the waiver hospitals in southeastern Pa. have a hard time making a case for greater access when there are so many heart programs nearby. If this is an example of free market medicine, I suggest that model is not serving citizens well,” she adds.
In addition to specific requirements of each of the three PCI waiver extension models, all waiver hospitals continue to report performance data to the American College of Cardiology National Cardiovascular Data Registry (NCDR), which feeds back confidential institutional outcome reports on a quarterly and annual basis that compare performance to national and volume-based peer benchmarks, according to Lee. Waiver hospitals must share their institutional reports with the DOH, which monitors outcomes to prevent or minimize harm to patients, says Lee.
Another layer of oversight, he adds, is the Pennsylvania Patient Safety Authority’s incident reporting system, which alerts the DOH to specific cases of concern. “We’ve checked a few incidents already, some involving deaths. No protocols were violated and everyone involved in those incidents did exactly what they were supposed to do,” says Lee.
There is a crucial need for reliable data on the safety of elective PCI at community hospitals, and only a randomized trial with a control group can produce them, says Aversano, who believes the waiver program options Pa. has permitted are not methodologically capable of producing those data. “It is unethical to have patients involved in a research study that is not designed properly. What morbidity or mortality statistic has to be reached before the Pa. studies should be stopped? I have no idea – the number of cases being compared is too small to be able to know,” Aversano declares. “This is a research experiment. You need clear questions. How will you know when the questions have been answered?”
Aversano says the Mayo clinic study is “a totally different ball of wax: the procedures were all performed in Mayo institutions, the standards and practitioners were the same in both the tertiary and community hospital, real-time review by Mayo Clinic physicians took place, and all outcomes were scrutinized by Mayo.” Replicating the model elsewhere raises concerns about who is supervising the tertiary-community hospital cooperation, who is reviewing and validating the data, and what statistical analysis will be used to judge the comparisons, Aversano adds.
Lee says the DOH talked with Aversano several times, and heard his arguments that a randomized trial like CPORT is the only safe and reliable way to answer the question whether elective PCI is as safe at community hospitals without surgical back-up as it is at hospitals with back-up.
“We agree with him – that’s why we originally required CPORT participation as a condition of waiver renewal,” says Lee. But the department found itself bogged down in litigation with hospitals alleging that the state cannot change the conditions of its waiver program, and faced passage of a Pa. law that not only would bypass CPORT, but would jettison volume requirements for any waiver participants, he adds.
Lee says the state succeeded in preserving the 200-procedure volume requirement for several of the waiver hospitals, along with some research and oversight infrastructure for all of them, which he says secures some level of patient safety.
In the meantime, the DOH continues its moratorium on PCI waivers to additional hospitals and awaits results from CPORT – or a change in ACC guidelines – to determine the next steps it will take on the matter.