High-quality cardiac surgery through teamwork

Abstract

The high-quality cardiothoracic surgery program is primed for mindful effective surgery. The challenge lies in attaining mindful skills and efficiency. Herein is one journey toward high departmental quality over two decades.

The unexamined life is not worth living. —Plato¹

A RUDE AWAKENING

On a dreary Tuesday in winter 2005, HealthGrades.com publicly released a report of their proprietary comparative ratings of Dallas–Fort Worth area heart surgery programs. Out of possible three stars, we at Baylor University Medical Center (BUMC) were assigned one star. Though these same reports were ultimately found to only loosely relate to the reality of a program and were largely fishing trips for “consulting” gigs, the rating still stung. We surgeons were intentionally oblivious to the unsolicited outside-looking-in reports consistent with Halstedian tradition from time immemorial. Administrators, however, were not.

The early 2000s brought great national attention to medical quality via the shocking reports that thousands of patients were harmed or killed by shoddy work in hospitals. The Institute of Medicine’s To Err Is Human in 2000,² followed by Crossing the Quality Chasm in 2001,³ referenced thousands of hospital errors contributing to the carnage. Though salacious, parts of the report were correct. There were numerous issues to be addressed, not just in surgery but systemically in hospitals across the country.

The BUMC cardiothoracic surgery chairman was sternly queried regarding the poor rating; alas, no cogent explanation was borne. In turn, a kind invitation was proffered to a junior department member about his personal departmental governance aspirations. Innocently answering in the affirmative, he was “field promoted,” becoming the department’s ninth chairman (Table 1), and duly forwarded the invitation to the executive suites to explain the rating.

Table 1.

Chairs of cardiothoracic surgery, Baylor University Medical Center at Dallas

Surgeon (final training)	Dates	Years (n)
Robert R. Shaw (Michigan)	1945–1961	16
Donald L. Paulson (Mayo Clinic)	1962–1976	14
Ben F. Mitchell (BUMC)	1977–1986	9
Hal C. Urschel Jr. (Mass General)	1986–1987	1
Maurice Adams (Tulane)	1987–1991	3
Thomas P. Myers (Wisconsin)	1991–1995	4
A. Carl Henry III (Texas Heart Institute)	1996–1998	2
Richard E. Wood (Kentucky)	1999–2006	7
Baron L. Hamman (Alabama)	2007–2014	7
G. V. Gonzalez-Stawinski (Cleveland Clinic)	2014–2017	3
Robert F. Hebeler Jr. (Emory)	2017–2018	1
Charles S. Roberts (Royal Brompton)	2018–present	2+

Warmly decorated rich walnut-walled offices belied the coarse setting, and congenial introductions were quickly over. The business of our business was front and center. “How could this be?” was asked more than once, and aggressively. The meeting focus was clear: Management expected an immediate total review of the current performance with recommendations to follow—and soon. With guidance from the Baylor Quality Institute led by Dr. David Ballard, the new chairman proceeded to investigate the source and rationale of our real quality.⁴

REVIEW PROCESSES

We surgeons convened an old-fashioned all-hands-on-deck mortality and morbidity (M&M) review of the most recent performance. Attendance was compulsory. Reviewers were fair and balanced. Panel members included three active surgeons, a nurse educator, a data manager, and the program coordinator. The chairman of the review panel was a disinterested practicing surgeon with the added qualifications of past department chairman and experience in combat in the US Navy Medical Corps. The M&M panel convened every Tuesday promptly at 3 pm and lasted to finis. Data were replete. Words were not minced.

There was plenty to review. Every case had a complete paper chart, original films, and the operating surgeon’s contemporaneous description and case context. Culpability and timing were assigned fairly and openly. This laborious, critical, top-down, anecdotal, and punitive review labored painstakingly for 8 months. Years later, this technique was cited by the Society of Thoracic Surgeons (STS) president as a phase-of-care mortality analysis.⁵

The findings were described for our colleagues. We learned that “tipping-point” events occurred in an even distribution between the three phases of care: preoperative preparation, accounting for 43%; intraoperative conduct, 20%; and postoperative management, 37% (Table 2). The panel also illuminated three discernible points. Each department member had sometimes applied “too optimistic” an indication for surgery. Operationally, we had allowed use of the computed tomography (CT) scanner during evenings, which had created an operational problem for the care team of sending the case off unit with relatively inexperienced nurses during night hours. In the hemodialysis units, we were similarly asking our nursing staff to manage patients in a way that we could improve in the intensive care unit (ICU) with a portable dialysis unit.

Table 2.

Factors in mortality in 37 cardiovascular surgery cases, 2005

Process phase of care	Factor	n	% within process group
Preoperative workup (n = 30, 43% of all factors)	Judgment on indication and methods	14	47%
	Cardiac risk factors	7	23%
	Noncardiac risk factors	5	17%
	Preparation for surgery	3	10%
	Evaluation completeness/method	1	3%
Operative conduct (n = 14, 20% of all factors)	Surgeon activity	10	71%
	Cardiopulmonary bypass	3	21%
	Anesthesia	1	8%
Postoperative care (n = 26, 37% of all factors)	Intensive care unit: Surveillance/recognition/treatment	2	8%
	Intensive care unit: Catastrophic event, unforeseen	7	44%
	Intensive care unit: Hemodynamic management	6	23%
	Postoperative floor	10	38%
	Discharge	1	4%

Operational changes were made. We mandated CT scanning for ICU patients only during daylight hours. Concomitantly, a policy of portable dialysis on the surgical unit was started. We embraced preoperative surgical checklists.⁶ Departmentally, we strongly “recommended” the use of a second opinion for all high-risk cases. These easy changes also highlighted the problem we were all addressing, and mortality was cut in half. That was not enough.

We presumed for good reason that all surgeons were quite good at their craft and surmised that scrutiny of their own data would result in organic self-improvement. Though only one stroke away, par golf is far harder to attain than bogie golf. With less than complete buy-in to the paradigm of mutual cooperation without malice, we redemonstrated that trust was critical to success. With the “low-hanging fruit” harvested, the sour grapes of hard programmatic quality improvement work loomed. Using data from the M&M review and the knowledge developed and taught at the Baylor Best Care Institute, we could see issues with our data and practice.⁴ Through standardization of terms, formats, and outcomes using the STS data set nomenclature (Figure 1), we vastly improved abstraction and reporting. BUMC surgeons were made aware of a standard preoperative consultation template and STS risk calculator that addressed the employed elements of risk. Further, we all augmented STS dataset clarity by encouraging all surgeons to actively participate in data acquisition and outcomes reporting.

Figure 1.

A sample 2012 report from the Society for Thoracic Surgeons with associated star ratings.

Systematizing and socializing our own data, knowledge, and expertise remained ahead. Several members of the team went to the Northern New England data group meeting in Maine for additional help. This entity was the first data collaborative in the nation that broke trade barriers, consequently allowing shared improvements. Using their model of cooperation,⁷ we formalized preoperative data-gathering tools and templates and shared these across practices among our four programs. We emulated the Northern New England Cooperative Group and formalized data and outcomes review with a group that would become the Cardiovascular Outcomes Department. The preoperative “testing” template was formally structured so that all elective cardiovascular surgery cases filtered through the same preoperative flow chart to ensure completeness of standard preoperative data availability.

Recognizing that our teams could be better organized and run than the traditional operating room model, we leaned heavily on the aviation quality paradigm: cockpit resource management.⁸ Surgeons had been groomed for individual academic achievement and success in brutal repeated selection processes in college, medical school, internship, residency, and even fellowship. Suddenly, this same person was assigned leadership of a real team involved in high-risk activity, although teamwork was often not his or her purview. Through cockpit resource management techniques, we sought to improve clear communication and to facilitate speaking up.

DATA SHARING

Though improved, our risk-adjusted isolated coronary artery bypass graft (CABG) surgery mortality remained stubbornly slightly above the national average—at 2.5%. At the local Best Care Meeting, the national champion of community data cooperatives presented his model of cooperation in Michigan.⁹ In his personal communication about risk-adjusted isolated CABG mortality, we concluded that some of the high-acuity cases seen in BUMC and other tertiary centers contributed uncommon elements to the database. Numerous cases involved problems such as advanced cirrhosis and/or transplantation,¹⁰ which while contributing much risk to mortality and morbidity did not populate the national database copiously and consequently did not impact the preoperative risk models.

Baylor physicians and administrators helped create the regional outcomes sharing initiative housed in the Dallas–Fort Worth Hospital Council Foundation “safe house” offices. The Texas Quality Initiative (TQI) is a unique forum for clinical outcomes sharing among institutions solely for quality purposes.¹¹ TQI was one of five large regional collaboratives that allowed sharing of outcomes data expressly for learning and exploration of successful methods among trusted neighboring institutions.

Leadership of the TQI includes the academics and the dynamos. The first activity of the greater group was to inspect the quality of data input from all institutions. Both completeness and veracity were deficient in some data points that were not minor.¹² With 2 years of system methodology comparisons, we were able to share best practices regarding data input for information creation. With the candid comparison of cooperating and competing programs’ methods, we were uniquely positioned to highlight systemic and analytic errors. Our self-description allowed the national STS to institutionalize a requirement for hard-stop 30-day mortality data vs assumptions. We were still far from knowledge and even further from wisdom.

The cardiologists and surgeons at Baylor lifted a page from the Baylor heart transplant “playbook” to form a comprehensive clinic for complex structural heart problems.¹³ Cardiologists Stoler and Grayburn and the surgeons and staff together made the “Valve Clinic”—a multidisciplinary, one-stop patient experience that truly facilitated a contemporaneous review of all the patient’s data with acknowledged experts. This program incorporated the not new but instrumental “heart team” methodology,¹⁴ allowing wise collaborative decisions to be made in real time.

OTHER EDUCATIONAL AND STAFFING CHANGES

Our academic colleagues embraced surgeons’ participation in the noon cardiology conferences and pathological seminars. These conferences were “show and tell” but were deliberately and emphatically transformed into truly academic sessions under Drs. Grayburn, Wells, and McCullough. Monday’s Journal Club was formally improved under Dr. McCullough and expanded to include invited professors’ lectures. Further, Dr. Wheelan and others improved the existing combined cardiology and cardiac surgery bimonthly forum into a truly operational surgical and medicine educational forum.

Surgeons embraced and mandated routine use of online STS outcomes prediction tools. The surgeons embraced the STS star rating for self-improvement. Together we created the Friday morning high-risk cardiovascular surgery multidisciplinary case conference where we openly discussed high-risk cases (those with a predicted mortality >5%) before the planned procedure. In this conference, we presented any and all controversial or unusual cases.

In the hospital setting, registered nurse practitioners were allowed to practice at the top of their licensure to closely attend the ICU while surgeons operated. After a trial use of nurse practitioners on the telemetry unit, we expanded intensivists’ activity to include advance practice nurses in the ICU. This model ultimately expanded to 24/7 coverage for the cardiothoracic ICU. Our communications education module was honed into a real case simulation module that is now part of the onboarding training for all ICU nurses. Consequently, we folded the expertise of other ICUs’ silos to improve all units through interchangeable staff, knowledge, and management.

Following the same valve clinic paradigm, we formally launched an integrated vascular and cardiothoracic surgery department aortic conference and clinic on Thursday mornings. Finally, we assimilated into the newly developed Texas A&M Medical School as formal teachers of residents and students. Table 3 shows a timeline of changes made to improve the program.

Table 3.

Timeline of program element incorporation, volume, and quality ratings

Program element	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
Formal M&M and periodic program review	******	******	******	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *
ICU/HD remain home policy			******	******	******	******	******	******	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *
Quarterly review: STS statistics	* * * *	* * * *	* * * *	******	******	******	******	******	******	******	******	******	******	******	******	******
Recommended second opinions PROM >5%			* * * *	* * * *	* * * *	******	******	******	******	******	******	******	******	******	******	******
Mandated second opinions						******	******	******	******	******	******	******	******	******	******	******
BUMC departmental POCMA	*	* *	******	******	******	******	******	******	******	******	******	******	******	******	******	******
Systemwide POCMA				******	******	******
BC meetings general data		* *	* * * *	******	******	******	******	******	******	******	******	******	******	******	******	******
BC unblinded program data report				******	******	******	******	******	******	******	******	******	******	******	******	******
BC unblinded surgeon data report					******	******	******	******	******	******	******	******	******	******	******	******
TQI citywide STS data sharing and learning				* * * *	******	******	******	******	******	******	******	******	******	******	* *	* *
Catheterization conference	* *	* *	* *	* *	* *	* *	* *	* *	* *	* *	******	******	******	******	******	******
Echocardiography conference	* *	* *	* *	* *	* *	* *	* *	* *	* *	* *	******	******	******	******	******	******
Cardiology journal club and rounds	* *	* *	* *	* *	* *	* *	* *	* *	* *	* *	******	******	******	******	******	******
Pathology cardiac case review	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	* * * *	******	******	******	******
Valve clinic staffing						*	**	**	***	******	******	******	******	******	******	******
Aortic conference and center															* *	* * * *
Transcatheter aortic valve replacement									21	30	54	58	99	164	184	139
High-risk cardiac surgery review											64	120	121	92	73	82
Isolated CABG	653	608	554	421	414	351	351	324	243	218	213	250	261	286	303	405
Open heart	1110	1057	947	821	786	712	727	370	326	350	295	372	365	411	394	524

BC indicates Best Care; BUMC, Baylor University Medical Center; CABG, coronary artery bypass grafting; HD, hemodialysis; ICU, intensive care unit; M&M, morbidity and mortality; POCMA, phase of care mortality analysis; PROM, predicted risk of mortality; STS, Society of Thoracic Surgeons; TQI, Texas Quality Initiative.

^*, pilot project; **, start with some staff; ****, relaxed policy with exceptions;******, policy.

The results are clear. In the last 6 years we have scored either at the cusp of or at 3 stars (out of the 3-star STS rating system) in major quality metrics: isolated CABG, isolated aortic valve replacement, and CABG plus aortic valve replacement (Figure 2). In isolated CABG, which accounts for the lion’s share of cases, we have scored 3 stars for each of the last 3 years. Simultaneously, we have enjoyed an increased volume of CABG cases in the setting of most program declines. More importantly, we have all developed a real pride in working in an expert, well-functioning program steeped with mutual respect and camaraderie (Figure 3).

Figure 2.

Baylor University Medical Center (BUMC) program score with confidence intervals compared with Society for Thoracic Surgeons (STS) mean score. For “Overall” star rating, BUMC improved to exceed the mean score for STS from 2010 on.

Figure 3.

BUMC cardiovascular surgery department members, 2020. Left to right: A. Carl Henry, MD, Rick Hebeler, MD, Baron L. Hamman, MD, Aldo Rafael, MD, Charles Roberts, MD, and Dan Meyer, MD. Not pictured: Daniel Enter, MD.

DISCUSSION: OUTCOMES AND APPLICATION OF TEAMWORK PRINCIPLES

Our department journey can be described as the development of genuine mutually respectful teamwork. Teamwork itself is not new, but it was new to the department. Most anthropologists relate teamwork specifically to the development of the spoken language.¹⁵ Consequently, it has allowed humans to dominate the world but necessarily through cooperation, respect, and altruism.

All of us have a credo to which we adhere. Whether from church, sports, or a social club, adherence to one’s credo is necessary for personal and departmental growth. Most vary little. All successful credos subscribe to listening and respectful consideration of all team members.

We at Baylor showed that embracing courtesy, bravery, trust, and reverence for the outcomes facilitates mutual respect. Ted Koppel of CBS News reported on his review of many acclaimed leaders and summarized their common principles in a televised interview.¹⁶ Three leaders from widely different walks of life illuminated their “elements” of leadership. Common elements were humble reflection, clear communication, and deep care. In business terms, our progression is that of improved relational coordination. Jody Hoffer-Gitell of Brandeis School of Business explained that through active listening, conscious recognition of all team members’ contributions, validated opinions, and a license to speak up, the degree of mutual respect increases, causing the team to perform beyond the sum of its parts.¹⁷ When commenting on problematic academic departments, Professor Crookston of Brigham Young University opined that there are six steps required for lasting faculty change: clarify values, make policy, trust, reflect, listen, and act.¹⁸ Each step represents making the agency of change tangible and genuine.

New to the BUMC cardiothoracic surgery department over two decades is the recognition of subspecialized expertise and development of keen mutual respect. We have a high regard for the patient, a high regard for the individual doctor, and a high regard for every team member. We have actively listened to “foreign” ideas and incorporated many into a marque practice. Consequently, we have all developed high regard for one another, and relational coordination has soared. In the last 10 years, we have stripped away the facade of individual intellectual superiority and organically grown teamwork by embracing outcomes measurement as our standard. With leadership development, decorum, respect, and cooperation have finally become the norm (Table 4).

Table 4.

Our steps to a highly reliable program in cardiac surgery

1.	Become self-aware.
2.	Develop an interest in quality.
3.	License a team for programmatic change management.
4.	Agree on measures—validity and accuracy.
5.	Set goals that are laudable and attainable and work selflessly toward them.
6.	Make measurements well.
7.	Honor the measure through praise, comments, and criticism.
8.	Celebrate advances and exchange feedback.
9.	Nurture mutual respect for expertise and success through decorum.
10.	Trust in others, trust in the team.

^*Adapted from several sources, including Moon,¹⁹ Crookston,¹⁸ and others.

Examination renders data, information, and then knowledge. Knowledge informs the wise. The wise lead the uninitiated. The good leader is a humble, empathetic listener with the power and alacrity to act wisely. Steadfast servant leadership fosters a high-quality cardiac surgery program.