Improving Teamwork to Reduce Surgical Mortality

Operating rooms are among the most complex political, social, and cultural structures that exist, full of ritual, drama, hierarchy, and too often conflict. For instance, in a surgical morbidity and mortality conference in which a roomful of surgeons were reviewing the delayed transport of a trauma patient from the emergency department to the operating room, the senior surgeon asked a junior surgeon what he would do differently next time. He replied, “I’m going to push harder, I’m going to transport them to the operating room faster.”¹ This response highlights what is so great but also so myopic about medicine. Physicians strongly advocate for patients, believing they alone are responsible for patient outcomes. Yet pushing harder is an ineffective intervention and may result in conflict rather than better patient outcomes. Health care, and especially surgical care, is too often viewed as primarily a solo activity rather than as a team function.

Over the last few decades, surgical and anesthesia care have dramatically improved, by virtue of advances in surgical techniques and anesthesia interventions, enabling surgeons and anesthesiologists to master ever greater technical challenges. Patients have benefitted substantially, by surgeons and anesthesiologists meticulously improving technical work and by significant increases in the availability and use of less invasive techniques, even for complex operations. Yet despite these advances, patients still experience preventable harm, much of it from less than optimal teamwork rather than from technical shortcomings or failures.

Poor teamwork contributes prominently to most adverse events, including those in the operating room. Using methods from aviation, researchers and clinicians have implemented interventions to improve teamwork in the operating room.² In aviation's past, pilots ignored the input of copilots because in their hierarchical culture, the copilot's input had been deemed relatively unimportant. Now, to be licensed, pilots must demonstrate both technical and teamwork competencies. Health care can learn from these examples in aviation. For instance, in many health care sentinel events, a member of the health care team knew something was wrong but either did not speak up or spoke up and was ignored (P.J.P., unpublished data, 2008).

Several studies have evaluated the use of checklists and briefings for reducing adverse events in surgical care. In a pilot study evaluating an operating room checklist with some technical elements (eg, use of antibiotics) and use of briefings and debriefings, Makary et al² demonstrated improvements in teamwork and staff members' perceptions of risks. However, some of the clinicians participating in the study resisted adopting these interventions, not convinced that improved teamwork would really benefit their patients. In a subsequent study, Haynes et al³ reported that use of a surgical checklist was associated with significant reductions in mortality, preventing 1 of every 2 deaths within 30 days of surgery. The study evaluated patients in 1 to 4 operating rooms at 8 sites and used a before-and-after design, but lacked concurrent controls, collected a large number of complex outcomes, and focused predominantly on technical work (completing the checklist items) and less on teamwork (encouraging interdisciplinary dialogue about risks). The authors reported a questionably large treatment effect, 1 death prevented per 143 patients, whereas other studies suggest that the preventable hospital mortality rate is approximately 1 in 400 patients and that 1 in 20 hospital deaths may be preventable.^{4 - 5}

The study by Neily et al,⁶ reported in this issue of JAMA, provides strong evidence that improving teamwork was associated with reduced surgical mortality. This large and robustly designed study demonstrated that using a multifaceted intervention of medical team training, involving teamwork training, ongoing coaching, and checklists to trigger operating room briefings and debriefings, was associated with a reduction in mortality. In this retrospective health services cohort study using a contemporaneous control group, surgical mortality declined 18% at the 74 hospitals that implemented the Medical Team Training program, compared with a 7% mortality reduction in the 34 control hospitals. Importantly, there was a dose-response relationship, in which greater use of teamwork was associated with a greater reduction in mortality. The absolute magnitude of the reductions in mortality were plausible from 17 deaths per 1000 procedures per year at baseline in the trained facilities to 14 deaths per 1000 procedures per year.

The study by Neily et al⁶ demonstrates that patient safety studies can use robust design. With 180 000 procedures from 108 hospitals, the study was appropriately powered to detect changes in mortality. Moreover, the investigators used a multiple time-series design with concurrent controls, accounted for selection bias using a propensity score, and included a dose-response analysis. They also supplemented quantitative data with qualitative data to analyze the degree to which the intervention was implemented, providing deeper insights into use of the intervention and potential reasons it worked.

Thus, the intervention in this study was far more than a checklist. It included building teamwork skills through training and coaching, and providing tools to support teamwork. The authors developed evidence-based training yet appropriately encouraged local adaption. Teams did not solely check off items on the checklist, they used the checklist to trigger a conversation.

This type of robust study was possible because of the Veterans Affairs substantial investment in health information technology and quality assurance processes. This study also provides insights into the conduct of large-scale patient safety efforts. Such programs require leadership and centralized support for collecting robust data and producing evidence-based interventions and training materials. Nevertheless, these structured programs must also encourage local modification and innovation in how interventions are implemented. Similar lessons emerged in previous studies of interventions designed to reduce central line–associated bloodstream infections.⁷ A culture of teamwork, accountability with measures, and use of checklist items are needed.^{8 - 9}

The study by Neily et al⁶ has important implications for health care. First, health systems should consider implementing team training and coaching to build teamwork skills. Although questions remain about the optimal methods and frequency of teamwork training and how to monitor competency in teamwork, hospitals should implement interventions to improve teamwork, such as those that were used in the Veterans Affairs system.¹⁰ Second, hospitals should ensure that operating room briefings and debriefings reliably occur. Locally modified checklists can be used to ensure patients receive recommended interventions and to trigger crucial conversations among all team members about risks and ways to reduce risks. Although briefings and debriefings should be standard in all procedures, the specific technical checklist items must vary; for instance, pediatric surgery risks differ from vascular surgery. Third, anesthesia and surgery boards should consider requiring teamwork competency skills, similar to that required for pilot licensing. Moreover, health systems should consider teamwork competencies in making hiring decisions and in the credentialing process for delineation of privileges. Fourth, additional research is needed to determine the most effective and efficient ways to improve teamwork. This will require investments, but patient safety research is sorely underfunded. Few academic medical centers have the psychologists, sociologists, anthropologists, behavioral economists, system and human factors engineers, or health services researchers to conduct robust patient safety research. Fifth, residency training programs should place greater emphasis on building teamwork skills, without compromising training in technical skills. Although some medical schools have incorporated patient safety and teamwork training in their curricula, few residency programs have and even fewer provide interdisciplinary training.

The study by Neily et al⁶ demonstrates that a multifaceted intervention was associated with improved teamwork and reduced surgical mortality. There is no quick fix to implementing interventions to reduce surgical-related adverse outcomes; doing so requires investment, training, coaching, and reflective practice. The health care community must make good teamwork the norm rather than the exception. In too many hospitals and too many operating rooms, clinicians may still perceive that they are battling each other, each trying to push harder rather than help others, too often forgetting they are on the same team. Health care must have as much improvement in teamwork skills as there has been in technical skills. Physicans and all other members of the health care team have an imperative to improve safety and outcomes and to reduce surgical mortality—patients deserve nothing less.