CCAS 2020 Meeting Review
Session II: Delivery of Care for Patients with CHD: An International Perspective
By Katherine Zaleski, MD
The second morning session, moderated by Dr. Susan Nicholson, consisted of a series of four lectures that dealt with the concept of clinical care regionalization for patients with congenital heart disease (CHD) as a means of improving morbidity and mortality outcomes.
The Landscape of Congenital Cardiac Care and Outcomes in the United States
Sara K. Pasquali, MD, MHS
Dr. Sara Pasquali delivered the first lecture of the series in which she detailed the current landscape of congenital cardiac care in the United States and discussed potential drivers of improvement including collaborative data/analytics, public reporting, and the regionalization of care.
At present, congenital cardiac care in the United States is both highly fragmented and highly variable. As of 2019, there were more than 150 congenital heart surgery programs in 39 states, 66% of which are located within 25 miles of another center and 44% of which are considered low-volume centers (<150 cases/year).
This variability in volume holds true for index cases as well – only 5% of centers perform > 20 Norwood operations per year, while 81% perform less than 10 per year (51% perform less than 5 per year). Across centers, there exists a high degree of variability in staffing, program structure, care processes, and outcomes. Since the late 1990s, overall operative mortality has dropped significantly, especially for STAT4 and 5 categories, and as we would expect, high-risk categories are associated with increased mortality (e.g. STAT1 0.5% v. STAT5 15.8%) and morbidity (e.g. STAT1 3% v. STAT5 30%).
This improvement, however, has not been consistent across programs. When one looks at case-mix-adjusted outcomes across centers, there is a three-fold variation in operative mortality (upper decile 50% v. lower decile 1.6%) and a greater than three-fold variation in the rate of major complications (upper decile 20.3% v. lower decile 6.0%). Perhaps more surprisingly, the magnitude of variation across centers (inter-decile ratio) in terms of mortality, length of stay, and complications does not differ between STAT 1-3 and STAT 4-5 cases.
Dr. Pasquali next addressed potential mechanisms (public reporting, collaborative quality improvement, regionalization with centers of excellence, innovation and research) to drive outcomes improvement and reduce variability in care. As a means of introduction, she presented the results of a poll conducted at the AHA 2017 Sessions in which 62% of respondents chose collaborative quality improvement (over transparency/public reporting (15%), regionalization of care (15%), and innovative research (8%)) as the best way to drive national improvements in pediatric cardiac care.
Public Reporting: Public reporting has a number of purported benefits including increasing transparency, incentivizing improvement, and allowing patients and their families to steer towards the “correct” and/or “better” centers. There is currently no data to support the assertion that public reporting improves outcomes in congenital cardiac care, while in other specialties, the data is conflicting. When public reporting leads to risk aversion, there is a potential for patient harm. In the adult population, it has been shown that less high-risk percutaneous coronary artery interventions are performed in states with public reporting and that these states also have a higher mortality rate. In a poll conducted at the AHA 2019 meeting, 49% of respondents answered “yes” when asked if their center has discussed or declined surgery on a high-risk patient due to potential concerns about the impact on publicly reported data. Further work will be needed in order to translate public reporting into improved outcomes.
Collaborative Quality Improvement: The aim of cross-center collaboration is to share data and expertise in order to establish best practices, reduce practice variation, and shift the mean towards better outcomes. There are numerous organizations, consortiums, and collaboratives (STS, PC4, CCAS, PAC3, IMPACT, NPC-QIC, Pedimacs, ELSO, PHTS, CNOC, ACTION) that maintain multicenter registries of pediatric cardiovascular data. Work done within Cardiac Networks United (an integrated pediatric and congenital cardiovascular research and improvement network including PC4, PAC3, CNOC, ACTION, and NPC-QIC) has shown that collaborative projects can lead to profound improvements in outcomes as well as reduced costs. Collaborative efforts are challenged by sustainability, the participation of individual centers, and the ability of participating centers to effect change (resources, personnel, leadership).
Centers or Excellence/Regionalization:Although there have been voluntary efforts by states, regions, and the Department of Defense, there is no national mandate for regionalization of pediatric cardiac care. A regionalization simulation (Welke et al, 2020), consolidation of care from 153 hospitals with varying case volumes to 37 hospitals with case volumes of >300 cases/year increased mean travel distance by 31 miles, but lead to a decrease of 263 deaths, 124,602 hospital days, and 1504 major complications over a four-year period. Beyond the lack of a national mandate, regionalization faces a number of challenges in the US including changes to staffing, revenue generation considerations, effects on post-graduate medical education, and care coordination.
Dr. Pasquali concluded her talk re-asking the question “What is the best way to optimize national CHD care and outcomes?” and quoted Professor Martin Elliot, “Families made it clear that they did not just want data or to have the option of traveling to the best hospital. Rather, they expected that we would provide excellent care, everywhere.”
Regionalization of Congenital Cardiac Care: Lessons Learned from the UK
Alistair J. Cranston, FRCA
Dr. Alistair Cranston followed with a presentation of the past, present, and potential future disposition of congenital cardiac care in the UK. He opened his lecture with a series of slides depicting the geographic location of the UK centers providing congenital cardiac care in 1995 (12 centers) and at present (11 centers) as well as the six national inquires and reviews undertaken in the interval years, leading to the question, “So after 25 years and six inquiries/reviews are we pretty much in the same place?”
Dr. Cranston’s historical overview began with the pediatric cardiac surgery program at the Bristol Royal Infirmary in the late 1980s and early 1990s. In retrospective analysis, cardiac surgical register and hospital episode statistics data from 1991 through March 1995 demonstrate an inverse relationship between surgical volume (under one year of age) and mortality.
In 1989, Dr. Steven Bolsin became a consultant anesthetist at Bristol and began to notice that there was a high mortality rate for both Arterial Switch (20/38) and AV canal (9/15) procedures. He brought his concerns regarding significant “excess” mortality to the hospital leadership, but no action was taken. He eventually took his concerns to the media which led to a GMC hearing in 1998, a formal public inquiry, and ultimately, the Kennedy Report (2001).
Statistical analysis performed as part of the inquiry demonstrated that Bristol was an outlier (1984-1996) in terms of mortality for open procedures in children under one year of age (Lancet 2001; 358:181-87). Dr. Cranston shared several excerpts from the Kennedy Report (The Report of the Public Inquiry into Children’s Heart Surgery at the Bristol Royal Infirmary 1984-1995: Learning from Bristol) that highlighted a number of issues including a hospital “club culture” with an imbalance of power and too much control in the hands of a few individuals, general failings in the National Health Service (NHS) with no standards for evaluating performance and confusion as to who is responsible for monitoring quality of care, as well as more general issues within the program at Bristol such as lack of insight, leadership, teamwork, and communication.
The Kennedy Report put forth 198 recommendations that addressed issues specific to Bristol as well as those within the NHS in general. It suggested that there were too many cardiac surgical centers, that complex surgery should only be performed in specialist centers, and that care should be both standards-based and better governed. Dr. Cranston recommended a lecture by Prof. Martin Elliot, entitled “The Bristol Scandal and its Consequences: Politics, Rationalisation and the Use and Abuse of Information” (available at gresham.ac.uk) for a more in-depth look.
Next, Dr. Cranston presented brief synopses of the five subsequent reviews that have been conducted since the Kennedy Report, starting with the Monro Report (2003), a wide-ranging committee review that considered all aspects of congenital cardiac care for the all of the UK. The report made standards-based recommendations including that the number of centers performing procedures should be reduced (300 surgeries, three surgeons per center for an almost 50% reduction in units). The committee’s recommendations were not acted upon on the basis that “no concrete evidence of unsafe practice had emerged.”
Next came the Summit of Experts in 2006, which involved the “tzars” of cardiology and pediatrics from all existing centers. Again, the current configuration was deemed “unsustainable” and there was a general agreement on the need for change including a reduction in number of centers (11 to seven). These views were further echoed by the Royal College of Surgeon’s “Surgery for Children: Delivering a First Class Service” (2007). In 2010, the NHS published “Safe and Sustainable,” an interim review of children’s congenital heart services in England, which again noted a need for reorganization and produced a set of recommendations very similar to those of the Kennedy Report nine years prior.
Despite a total unity of professional support at the time, legal challenges from units likely to close lead the Secretary of State for Health to convene the Independent Reconfiguration Panel to examine the proposals. The Secretary concluded in 2013 that the “outcome of the review was based on flawed analysis … proposals cannot go ahead ... the review is suspended.” The NHS put forth “new” congenital heart disease standards in 2015, which although comprehensive, were similar to previous standards. Surgical centers were to have a minimum of four whole time equivalent (WTE) surgeons each undertaking 125 procedures per year and were given five years to achieve this.
Dr. Cranston remarked that perhaps two centers are currently compliant. Only one center, Oxford (single surgeon, low-volume, suspect mortality rate), has closed during this time after four deaths prompted suspension of service in March 2010 with permanent closure in September of the same year.
Dr. Cranston then introduced NICOR (National Institute for Cardiovascular Outcomes Research) which collects and analyzes data related to quality of care and outcomes. He showed several funnel plots illustrating representative data for Arterial switch (2009-2012) and Norwood Procedures (2009-2012, 2014-2017) as well as data from the NICOR National Congenital Heart Disease Audit 2019 Summary Report demonstrating a reduction of 30-day hospital mortality from 3.2 to 1.4% over the past ten years.
Other measured outcomes included acute neurological events, ECLS post-procedure, renal replacement therapy, unplanned pacemaker insertion, prolonged pleural drainage, and catheter-related complications (need for surgical intervention, device embolization). He spoke of other quality indicators including the rate of antenatal diagnosis, the data quality index score (DQI), and the PRAiS2 risk score as well as the upcoming Fetal Database (2020). The “Understanding Children’s Heart Surgery Outcomes” website (https://childrensheartsurgery.info) is available to help patients/families navigate and make sense of NICOR data.
Dr. Cranston concluded his talk with a summary of what has been learned (there have been four major reviews with more or less the same conclusions and no planned implementation, but results continue to improve and local/regional politics are important), where we are now (11 centers performing surgeries with reviews ongoing but no changes yet), and the future (including considerations about cardiac transplantation).
Regionalization of Congenital Cardiac Care: Lessons We are Still Learning in the Netherlands
Nigel Turner, MB, ChB, PhD, MMe
Dr. Nigel Turner delivered the third lecture of the series, which provided a rough overview of healthcare in the Netherlands, insight into the current discussion regarding further centralization of pediatric cardiac surgery, and a comparison between the UK and the Netherlands with regards to regionalization.
The Netherlands has approximately 17 million residents, 3.4 million of which are under 18 years of age. Of the 180,000 pediatric operations that are performed each year, 1200 are for congenital heart disease. Among the roughly 1.2 million people that are employed in the healthcare field, 1740 are anesthesiologists, 100 of which are based in pediatric centers, and only 12 of which are involved in pediatric cardiac care.
Healthcare spending as a percent of GDP is 10%, roughly in-line with that of Canada (11%) and less than that of the United States (17%). The average distance to a hospital in the Netherlands is 4.7 km or about five minutes by car as compared with 11.4-15.8 minutes by car to the nearest hospital emergency facility in the US. Within the Netherlands, general pediatric surgery and anesthesia is highly centralized – only seven centers have PICU and NICU capabilities. There were initially six pediatric cardiac centers, however, two merged and another closed, leaving four centers (1 center/~4 million inhabitants) with roughly equal mortality rates. The anesthesiologists involved in pediatric cardiac care vary in both their cross-training background (i.e. cardiac anesthesia, pediatric intensive care, and/or pediatric cardiac anesthesia) and their current practice mix.
Having described the Netherlands’ healthcare system, Dr. Turner next described the history of its pediatric cardiac centers beginning with calls for centralization in both 1993 (National Hearth Council Report, congenital heart disease) and 1998 (Dutch Heart Foundation and professional bodies joint statement, all age categories).
Sweden centralized its pediatric heart surgery into two centers in 1993 and saw a reduction in mortality. With the exception of concerns about waiting lists and PICU bed shortages, patient organizations were otherwise happy with the standard of care and closer cooperation was recommended by the Ministry of Health. In 2007, the Minister of Health and the Dutch Parliament set up a committee (Lie Commission) to explore whether concentration of care leads to quality improvement. A national QA system based on international comparisons was established and further decisions regarding centralization were deferred to 2009 with all six centers remaining open with a temporary license until then.
In 2009, the Lie Commission concluded that higher volumes can give better results (350 operations per year reduced mortality) but that single-hero centers are highly vulnerable, travel distances and costs would likely increase, university hospitals fear the loss of CHD patients, and that centers should cooperate immediately.
The disadvantages of regionalization for families could be offset by regional cooperation of care teams, early transfer back to local hospitals, efficient planning of appointments, affordable/subsidized accommodations for parents, and health insurer reimbursement of travel costs. Dr. Turner noted that there is precedent within the Netherlands for a single center – a single center has already been introduced for pediatric oncology. The question of how many pediatric cardiac centers would be ideal was not as clear-cut. If there were only to be two centers (five-six surgeons per center), then it made sense to consolidate into the two largest centers (Rotterdam and Amsterdam/Leiden); however, geographically, a third center (Groningen) was a more attractive arrangement.
The Commission ultimately settled on three centers (deadline 2011) and recommended that they cooperate with each other and other hospitals, that non-university hospitals that treat CHD must stop immediately, and that hospitals referring patients abroad (Maastricht) must comply with Dutch QA requirements. Not much has changed since the Lie Commission – the deadline was extended and all centers have retained their licenses but are expected to cooperate such that one “virtual” center would be formed. A progress report from the Health Inspectorate in 2012 demonstrated no evidence of cooperation (fear actually caused less cooperation), while reports in 2018 and 2019 described limited cooperation in clusters of centers but no clinical cooperation between clusters.
Dr. Turner concluded that although the Netherlands has highly centralized pediatric anesthesia and that patients/parents can understand and appreciate centralization, the Netherlands is not currently particularly centralized for pediatric cardiac surgery.
Dolly D. Hansen, MD Lecture-Lessons Learned from Failure of Congenital Cardiac Care Programs: Regionalization (Virtual) of Congenital Cardiac Care Program in New England
Paul R. Hickey, MD
The second morning section concluded with the Dolly D. Hansen, MD Lecture, given in honor of Dr. Dolly D. Hansen, the “Mother of Pediatric Cardiac Anesthesia” and the Chief of Cardiac Anesthesia at Boston Children’s Hospital from 1980-2002. This year’s invited lecturer was Dr. Paul Hickey. His presentation touched on a number of themes including the birth of congenital cardiac care in 1938, the arguments (and evidence) for regionalization both domestically and abroad, the evolution of regionalization of cardiac care in New England, and the potential limits of regionalization (“how big is too big?”).
Dr. Hickey began his talk sharing memories of starting out as a cardiac anesthesiologist under Dr. Hansen’s guidance. He then gave a brief history of the Boston Children’s Hospital Division of Cardiac Anesthesia which began with Dr. Hansen as the sole cardiac anesthesiologist and has now expanded to a diverse group of 18 anesthesiologists. Dr. Hickey next turned his attention to the first congenital cardiac anesthetic (mask cyclopropane) that was delivered by nurse anesthetist Betty Lank in 1938 at Boston Children’s Hospital for an open thoracotomy and PDA ligation. He showed pictures of Betty Lank and her anesthesia record, Dr. Robert Gross’ surgical diagram, and their patient, Lorraine Sweeney, both on post-op day one and then at age 88, surrounded by family at a recent Red Sox game.
Since the first congenital cardiac surgery, the field of pediatric cardiac care has undergone tremendous growth and change, but this has not been without failures and uncertainties. The second portion of the talk addressed these shortcomings and the calls for regionalization that they have generated, both domestically and abroad.
Internationally, the Bristol heart babies scandal and the centralization of pediatric cardiac heart surgery in Sweden highlighted failures within their respective systems and drew attention to the potential mortality benefit of regionalization. A study published in the 2018 (Ann Thorac Surg 2018; 105(5): 1436-1440) analyzing data from the European Congenital Heart Surgeons Association database demonstrated that the annual volume of operations was an independent risk factor for operative mortality when less than 60 neonatal cases were performed per year.
In Europe, there have been pushes for regionalization - in Sweden, regionalization (consolidation of four into two centers) resulted in a drop in mortality from 9.5 to 1.9%; in England, the NHS has proposed regionalization so that each center would have four surgeons, performing 125 operations per surgeon per year (500 cases per center); the European Congenital Heart Surgeons Association has recommended three surgeons per center with 125 operations per surgeon per year and 60 neonatal operations per year per center.
High-profile failures have also made the news in the United States. Analysis of the STS Congenital Heart Surgery Database has demonstrated that; 1. hospital-specific observed-to-expected ratios for operative mortality exhibit a high level of variance across institutions, and 2. there is an inverse relationship between the odds ratio of mortality and annual case volume with an inflection point around 300 cases per year (J Thorac Cardiovasc Surg 2009 May; 137(5): 1133-1140).
Regionalization simulation analysis using hospital discharge data from California showed that up to 83 deaths (over three years) could be potentially avoided if patients from low and medium-volume centers were redistributed to the nearest high-volume center. Failed congenital heart programs have a number of common denominators including small size (less than three surgeons, low case volumes) and lack of specialized cardiac anesthesia, cardiology, and/or cardiac intensive care teams and resources.
A number of factors have enabled congenital cardiac care in New England to be largely regionalized to Boston Children’s Hospital. The first PDA closure in 1938 gave Boston a substantial head-start in terms of establishing a congenital heart program, although Dr. Gross did reject Helen Taussig’s idea for shunting tetralogy of Fallot patients and so the first Blalock-Taussig shunt (BTS) was performed elsewhere.
In 1972, Dr. Aldo Castaneda, the “Father of Neonatal Corrective Surgery,” was appointed Cardiac Surgeon-in-Chief. In 1980, the Cardiac Anesthesia Division was formed (Dr. Hansen and Dr. Hickey) and allowed for more consistent and innovative care. Dr. James Lock became the Cardiologist-in-Chief in 1993 and pioneered catheter-based interventions for congenital heart disease, greatly expanding the capabilities and volume of the catherization laboratory. In 2010, the Boston Children’s Hospital Cardiovascular Program began clinical relationships with Maine Medical Center, Dartmouth-Hitchcock Medical Center, Hasbro Children’s Hospital (RI), Connecticut Children’s Hospital, and Vermont Medical Center. In 2012, the program agreed to provide a team (pediatric cardiac surgeons, pediatric cardiac anesthesiologists, and perfusionists) to do RACHS-1 cases at Massachusetts General Hospital after their pediatric cardiac surgery program was suspended. And finally, in 2017, the only other remaining program within Boston (Floating Hospital/Tufts) closed.
Unfortunately, the question now is “how big is too big - does the volume-outcome relationship fail at the high end?” Unlike with highly standardized and protocolized procedures such as hip replacements in the elderly where outcome should scale with volume, customized procedures in heterogeneous populations, such as the spectrum of congenital heart disease, outcome may not scale with volume and the disadvantages of increased size may lead to adverse events.
Within the cardiac OR at Boston Children’s Hospital (1000 cases per year), there are 1309 possible team combinations (anesthesiologists, surgeons, perfusionists) on any given day. In the cath lab, EP, and cardiac imaging suite (3000 cases per year) there are 1785 possible EP and 1530 Interventional team combinations, raising the question as to whether interventional care should be centralized along with surgical care. The possible team combination number balloons to 209,044 when ECMO is required in the cardiac OR.