The study, published in Global Surgical Education, the journal of the Association for Surgical Education, set out to answer a deceptively simple question: can a validated instrument for tracking operative trust, originally built for general surgery, work in an environment where every case is long, complex, and uniformly life-or-death? The answer, according to the research team led by Megan L. Schultz and Gurjit Sandhu, is a resounding yes—and the data they gathered reveal just how tightly trust and hands-on opportunity are linked.

The stakes could hardly be higher. Cardiac surgery training is in trouble. Work-hour restrictions, heightened patient safety concerns, and growing scrutiny of surgical outcomes have all chipped away at the traditional model of graded responsibility, in which trainees earn progressively more independence as their skills mature. Meanwhile, the patients themselves have changed: advances in interventional cardiology mean that the straightforward teaching cases of previous generations are increasingly rare, and those who do reach the operating table carry more comorbidities than ever. Cardiac surgery, as a leader in outcomes transparency, has felt these pressures acutely.

The consequences are measurable. Many cardiothoracic trainees now graduate without routinely performing common procedures such as off-pump coronary artery bypass grafting or mitral valve repair. More than half of recent graduates report insufficient exposure to minimally invasive, robotic, and congenital cardiac operations. In 2024, more than a quarter of examinees failed the thoracic oral board certification exam on their first attempt—the highest failure rate in more than a decade. Layered on top of this is a looming workforce crisis: the Health Resources and Services Administration projects a 34 percent shortage of cardiothoracic surgeons in the United States by 2037. Against that backdrop, understanding how trust is built—or withheld—in the operating room becomes a matter of national workforce planning, not just pedagogy.

The instrument at the heart of the study is called OpTrust, a tool previously validated in general, vascular, and plastic surgery. It measures two intertwined phenomena: faculty entrustment, the willingness of the attending surgeon to delegate, and trainee entrustability, the behaviors the trainee displays that warrant that delegation. Raters score five behavioral domains—the types of questions asked, the operative plan, instruction, problem solving, and trainee leadership—on a scale from 1, indicating low entrustability, to 4, indicating full entrustability. Because prior work has shown the five domains collapse into a single construct, the scores are averaged into one number for each surgeon and trainee, then combined into a total dyad score ranging from 1 to 7, where 7 means the trainee operated autonomously.

What makes the new work distinctive is the Cardiac OpTrust supplement, a specialty-specific addition that grounds those abstract trust scores in fifteen concrete surgical benchmarks drawn from the shared sequence of nearly every cardiac operation using cardiopulmonary bypass. The benchmarks run from incision and sternotomy through internal mammary artery harvest, cannulation stitches, heparin administration, aortic cannulation, initiation of bypass, application and removal of the cross-clamp, cessation of bypass, protamine administration, decannulation, chest tube placement, and closure. For each benchmark, a trained third-party observer simply recorded who performed it: the faculty surgeon or the trainee. That binary record, stacked against the trust scores, turns the rhythm of a cardiac case into a granular map of autonomy.

Between November 2022 and June 2023, trained research assistants observed forty-nine elective cardiac operations at the University of Michigan, a large quaternary care academic medical center. The raters underwent standardized training and had to achieve at least 80 percent interrater reliability before collecting data independently. They observed from incision until the attending surgeon left the room, scoring faculty and trainees simultaneously but independently. The observed cases included twenty mitral or tricuspid valve repairs or replacements, twenty aortic valve or aortic root operations, eight coronary bypass grafting cases, and one combined bypass and mitral repair, generating scores for twenty-six unique faculty-trainee pairings drawn from twelve trainees and seven attending surgeons.

The results were strikingly consistent. Mean faculty entrustment was 2.46 on the four-point scale, mean trainee entrustability 2.47, and mean total dyad entrustment 3.93 on the seven-point scale, with trainees completing an average of 8.33 of the fifteen benchmarks per case. Crucially, entrustment scores were significantly higher for benchmarks the trainee completed than for those the attending kept. Spearman’s rank correlation confirmed strong positive associations between trainee benchmark completion and faculty entrustment, trainee entrustability, and total entrustment, with correlation coefficients of 0.66, 0.62, and 0.65 respectively.

Poisson regression quantified the relationship with even sharper precision. Each one-unit increase in faculty entrustment corresponded to a 32 percent increase in the number of benchmarks performed by the trainee, rising from an expected 5.36 benchmarks at low entrustment to 12.61 at full entrustment. Each unit increase in trainee entrustability behaviors predicted a 36 percent increase, from 5.17 to 13.00 benchmarks. And each unit increase in total dyad entrustment predicted a 16 percent increase, from 5.17 to 13.02. All associations were statistically significant at p less than 0.001. In plain terms: when trust goes up, the scalpel changes hands far more often.

Two benchmarks—protamine administration and cessation of cardiopulmonary bypass—did not reach statistical significance, though they followed the same trend. The authors note that these two tasks uniquely capture leadership and team management rather than pure technical skill, hinting at a fascinating distinction: entrusting a trainee to sew may be psychologically different from entrusting a trainee to direct the perfusionist, anesthesiologist, and nurses through the delicate transition off the heart-lung machine. Whether the principles of entrustment extend to these so-called soft skills remains an open question for future research.

The study’s limitations are candidly acknowledged. It was a single-institution pilot, so variability across cases, participants, and centers may shape how Cardiac OpTrust performs elsewhere. All faculty observed were male, reflecting the composition of the department at the time, while 51 percent of trainees were female—programs with different demographics may show different patterns. The tool also demands significant human resources, since it relies on independent raters physically present in the operating room. That investment, however, buys something self-report tools like SIMPL and Zwisch cannot: real-time, third-party observation free of recall bias. Previous work suggests the observer’s presence does not disrupt the operative experience, and the Hawthorne effect, while impossible to eliminate entirely, appears limited.

There is also an important conceptual caveat: OpTrust measures entrustment, not technical proficiency. A trainee may be handed tasks without being competent, or possess skill without being granted autonomy. The binary benchmark design records who performed each step, not how well. Still, because greater entrustment has been linked to operative performance in prior studies, optimizing trust in the cardiac operating room may yield tangible benefits for trainee competency and confidence. The benchmarks also deliberately exclude skills performed during the cross-clamp period, such as distal anastomoses in bypass grafting, since the pilot’s case volume was insufficient to draw meaningful conclusions about case-specific techniques; future iterations may add them.

The timing of this work is significant for structural reasons as well. Accredited general surgery programs have already integrated Entrustable Professional Activities into their curricula, vascular surgery programs are following, and cardiothoracic programs will likely face the same transition toward competency-based training soon. Tools like Cardiac OpTrust could provide the measurement backbone for that shift, replacing crude case counts with nuanced, behaviorally anchored assessments of when and why trainees are actually allowed to operate.

For now, the message from Ann Arbor is both simple and quietly radical: trust in the operating room is not an ineffable chemistry but an observable, countable phenomenon—and it can be studied, taught, and perhaps deliberately cultivated. As the authors argue, trainees can even use the OpTrust behavioral descriptions as a roadmap, incorporating the specific behaviors that signal readiness for greater independence. With further multi-center validation, the moment-by-moment choreography of trust in the cardiac operating room may finally become something educators can see, measure, and improve—one sternotomy, one cannulation, one cross-clamp at a time.

Subject of Research: Measuring intraoperative entrustment and trainee autonomy in cardiac surgery operating rooms

Article Title: Understanding and measuring intraoperative entrustment in the cardiac surgery operating room

Article References: Schultz, M. L., Theall, A. C. P., Ozuna-Harrison, R., Buchanan, D., Tubbs, D., Evans, J., Matusko, N., Bolling, S. F., & Sandhu, G. (2026). Understanding and measuring intraoperative entrustment in the cardiac surgery operating room. Global Surgical Education – Journal of the Association for Surgical Education, 5(1), Article 193. https://doi.org/10.1007/s44186-026-00595-3

Image Credits: AI Generated

DOI: 10.1007/s44186-026-00595-3

Keywords: cardiac surgery, surgical education, entrustment, trainee autonomy, OpTrust, operating room, cardiopulmonary bypass, competency-based training, surgical benchmarks, faculty-trainee dyad, psychometrics, workforce shortage