The handoff of patients from the overnight team was more hectic than usual. Most of our emergency beds were filled with admitted patients awaiting a hospital bed, and the waiting room had many patients who had not yet been seen. Several had been waiting more than four hours. Gil, the resident working with me, dutifully took down detailed information on all the patients as I noted specific tasks that I needed to complete. (Descriptions of Gil and our patients have been altered to protect their identities.) At the end of the sign-out we discussed two “problem” patients that the departing team had alerted us about.
The first was Mr. Mares, an undocumented immigrant, who worked as a roofer and had recently developed chronic renal failure requiring ongoing dialysis. Mr. Mares was relatively young for renal failure, only 42 years old, and he smiled weakly as we walked by his cubicle and introduced ourselves as his new team of doctors. He wore a corduroy jacket with black stains from roofing tar and his lips revealed the pallor of anemia associated with renal failure. I noticed that he was breathing rapidly and mentioned this to Gil and the students and residents rounding with us. “Fluid overload,” the departing resident had said, “but his X-ray looks okay.” That resident had also explained that the overnight on-call renal fellow had agreed to evaluate whether hospital dialysis was needed right away. The problem was that because of the immigration status of the patient, Medicare would not cover the costs of dialysis as it did for most U.S. citizens, and Mr. Mares had no way to pay for the expensive procedure. This was a serious issue for undocumented immigrants with renal failure and had resulted in an awkward compromise in which the hospital agreed to dialyze patients like Mr. Mares only when they represented a true medical emergency. This approach resulted in frequent delays and return visits to the emergency department until the inevitable true emergency of a dangerously elevated potassium or pulmonary edema occurred. “Crazy system,” concluded the departing resident, and we all nodded in silent recognition of the difficult dilemma of providing appropriate care for Mr. Mares due to the health system constraints.
The other patient was an elderly gentleman named Mr. Jackson, who was waiting for a consult with the cardiology team to evaluate his new onset of chest pain. We were told that all the EKGs and lab tests were negative and that the previous team had anticipated discharge and an outpatient stress test, but since Mr. Jackson was 75 years old and had hypertension and diabetes, they wanted cardiology to evaluate the need for hospitalization. Mr. Jackson was afraid to go home in spite of his negative workup, and the team anticipated a struggle with cardiology over admission because the hospital was full. Decisions about admission for patients like Mr. Jackson had become increasingly difficult as they got handed off from one team to another with a consequent loss of information. I resolved to speak with Mr. Jackson as soon as I could so that I could verify the information and be ready for the cardiology team’s questions, but first I grabbed the chart of one of the patients who had been waiting for five hours to begin her workup.
When I returned to our work station 20 minutes later I overheard Gil arguing with the nephrology fellow. The nephrology fellow said, “We’ve determined that the patient is stable for discharge. He doesn’t need emergency dialysis today. His potassium is only 5.8 and there’s no pulmonary edema.”
“But he’s short of breath, and he’ll only get worse without dialysis,” said Gil. “He doesn’t have any transport back to the hospital.”
“He can call an ambulance. We’re not a dialysis clinic. He needs to return to Mexico where he can receive his dialysis. Otherwise we can dialyze him only in an emergency. That’s the policy.”
“That’s a ridiculous policy. He could die before he gets back here,” said Gil.
“That’s the policy.”
“Asshole,” said Gil.
The nephrology fellow came up to me. “I need to report your resident for unprofessional behavior,” he said. “Are you his attending?”
“Yes,” I said.
“Did you hear what he just said?”
“I’ll talk to him.”
Before I could say anything more the cardiology team appeared waving Mr. Jackson’s EKG in my face. “Why didn’t you do anything about this arrhythmia?” asked the attending cardiologist showing me a short run of ventricular tachycardia on an EKG.
“I don’t know,” I said. “I was told that all of the tests were negative.”
“Does this look negative?” said the cardiologist. I looked over at Gil, whose face was still red from his interaction with the nephrologist, and he came over and examined the EKG. We shook our heads, since the arrhythmia was clearly shown. The cardiologist continued, “I think we better put him in the intensive care unit. He’s been sitting down here in the ER for 15 hours without any treatment. We’re lucky he’s still alive.” The cardiologist and his team walked off to see other patients, leaving us stunned and embarrassed.
As I sat down with Gil to take stock of what had happened with our problem patients, I was exasperated and confused. Both of our patients had been victims of a health care system that was inadequate to address their problems. Mr. Mares was caught in a health policy and health system nightmare where his undocumented status limited his access to life-sustaining dialysis. The hospital would agree to perform dialysis only for life-threatening emergencies, and with renal failure these emergencies were inevitable, but the timing was unpredictable. For our other patient, Mr. Jackson, the information about his arrhythmia had gotten lost in the transitions between emergency department teams, and we had not had the time to go back over the series of EKGs he had received. The pressures to evaluate other patients who had been sitting in the emergency department for hours without treatment created an overloaded system that was vulnerable to error. Fortunately, Mr. Jackson had suffered no harm, but he might have.
The health care system had failed both Mr. Mares and Mr. Jackson in different ways. While each problem patient had a serious medical issue, the problems were not about the usual areas we concentrate on in medical education—an understanding of pathophysiology, diagnosis of an illness, development of a treatment plan, or competent performance of a complex procedure. The problems were all about systems breaking down: systems to care for undocumented residents with kidney failure; systems to transfer information between doctors over time; and an overloaded hospital with inadequate space, personnel, or equipment to meet the needs of the patients.
This would have been a wonderful time for me to sit with Gil and use these cases to illustrate systems-based care, which is one of the core competencies that he was supposed to master before graduation. Unfortunately, if I had given Gil a lecture about the history of end-stage renal disease policy and the current systems of care for undocumented immigrants, the information would not have helped him find a solution for Mr. Mares’s problem or help other patients like Mr. Mares. If I had discussed how to improve handoffs of patients like Mr. Jackson, it probably would not have prevented our missing the information about the arrhythmia, because the EKG that contained the information had been misfiled by one of the EKG techs whose shift had ended hours ago. The lost information was only recognized during a careful and concentrated review that the cardiology team conducted and was not available during the rushed patient handoff. Understanding systems like handoffs when they are working well under ideal conditions is not the same thing as understanding them when they fail.
As Gil and I discussed our two problem cases and shared our frustrations with our health system, I pondered how we might do a better job of using our education programs to prepare our trainees to understand and manage health systems effectively. Systems-based care is a topic that I teach with lectures on quality improvement and patient safety (QI/PS) and health policy. I often leave such sessions uncertain about whether the trainees will be able to apply what I have taught to actual situations. This concern has caused me to ask how we might better address systems-based care through education and how we could evaluate whether the students have developed competence in this area.
Our first problem with education about systems-based care is the lack of a common definition or understanding of what we mean by systems-based care. While we may be familiar with a system like the trauma system that involves surgeons, hospitals, ambulances, and emergency departments working together to provide the best care possible to those who are injured, we may not know the evolution of the rules for distribution of injured patients, the financing of trauma care, or the overall leadership of the system. Is the trauma system part of the public safety system with public funding? Or is it a part of the business of hospital medicine with expectations for profits from the care? If it is part of both systems, how does a future trauma surgeon or emergency physician gain competence to function effectively in the system and address problems? Many systems are so complex that any one individual can have experience and personal understanding of only a part of the system. The overall successful functioning of a system like the trauma system requires agreements, policies, funding, governance, and constant analysis of performance and improvement that have been adjusted and improved over time.
In the overall health system the participating people and organizations are even more numerous and diverse and the overall purpose less clear. Ziegelstein and Fiebach1 have suggested use of the metaphor of a village to help trainees understand “the importance of the medical community and the larger context and system of health care in providing the best care to our patients.” Like a village with its governance and various shared resources like water, roads, and police, a health system depends upon cooperation and sharing of infrastructure resources but also the sharing of knowledge and expertise. Ziegelstein and Fiebach suggest that for students to effectively use the health system, they first need to understand the component parts and how they work together.
The World Health Organization (WHO) defines a health system as consisting of “all the organizations, people, and actions whose primary interest is to promote, restore, or maintain health.”2 The WHO includes building blocks in this definition: health services, workforce, information systems, essential medical products, financing, and leadership and governance. How might a resident or student demonstrate that he or she has achieved a level of competence in systems-based practice when it is clear that as an individual he or she might have only a small role to play in the overall system performance? In 2008, Johnson et al3 described the challenge of system-based education, noting that
current efforts in medical education focus on mastering knowledge of disease, diagnostic skills, and treatment at the level of physician–patient interaction, resulting in preoccupation with system elements, while the system as a whole and its effects on patients remain invisible. The context is what has been minimized as educators try to standardize the experience for trainees.
They went on to identify gaps in our systems-based practice curricula that include the lack of an agreed-upon definition, lack of assessment methods, lack of understanding the relationship of systems-based practice to patient outcomes and safety, and lack of integration into daily practice.
Since that article was published some progress has been made in defining systems curricula and experiences for students, particularly emphasizing the opportunities to use QI/PS as an entry into systems-based practice.4 In this issue of Academic Medicine we provide several articles that address systems-based practice. I will attempt to integrate some of their messages into suggestions for improving our overall approach to the education and evaluation of systems-based practice.
In this issue, Bowe and Armstrong5 provide an overview of systems principles in medical education and introduce the concept of systems thinking that could be taught to students and faculty to sharpen their abilities to adjust and adapt to changing contexts. They also note that there is often resistance to systems approaches by administrative hierarchies, which are more comfortable dealing with problems using a reductionist approach that breaks down problems into separate discrete pieces with clear individual authority and responsibility rather than seeing the problems as part of larger systems requiring a holistic approach. I have three suggestions that I believe could help address the concerns of Bowe and Armstrong and improve our health systems in the process.
First, a reasonable starting point is to have a didactic program that introduces the vocabulary of health systems, including the history of health systems and examples of subsystems like the trauma system, public health system, payment and financing systems, education system, licensing system, and medical–legal system. During this introduction of health systems the focus would be on an understanding of a health system during optimal functioning. There should also be an introduction into health process engineering and design,6 health finance, leadership and governance, QI/PS, and health policy to provide a resident the tools to be able to later address problems in the functioning of the system. Assessment of a resident’s knowledge base could be through the same tools used to assess other knowledge-based topics in medicine: oral and written tests.
Second, we should take advantage of current pedagogies for adult learning that emphasize experience and social learning. O’Brien et al7 in this issue introduce an innovative approach to systems-based learning that attempts to do that. They describe a program that begins with a week of didactic experience followed by seven weeks of workplace learning in which the students, with guidance from faculty who are experts in systems issues, are immersed in the work of a health facility and apply the concepts introduced in the didactic presentations to a gap or problem in the health system. In this way the students are able to reflect upon their experiences and use them to deepen their understanding of concepts that were introduced at the outset of the course.
Similarly, Gonzalo et al8 in this issue describe the use of change management strategy and curriculum development principles to introduce students to authentic systems-based practice learning experiences. These experiences can change the current construct of the learning environment from a place where learning experiences can occur to a place where students also interact with and change the learning environment and in this way learn about it. Gonzalo et al describe placements of students in systems roles as health coaches for patients and as patient navigators who understand the numerous barriers for patients and can help the patients to overcome them.
As trainees begin to gain content knowledge through working with patients, there should be an opportunity to use that knowledge to address real systems problems. Trainees should be able to use data systems and other tools to detect and improve systems problems. Residents and students encounter problems in the health systems on a daily basis, such as those I described earlier. A resident like Gil who observed the failed handoff of information about Mr. Johnson would have the motivation to engage in effective systems redesign and quality improvement, particularly if he had a faculty mentor who could help. There are now increasing numbers of examples of residents providing leadership in identifying and improving quality problems. Lee et al9 have described how residents can provide stewardship of antibiotics to reduce unnecessary antibiotic use and thereby improve quality of care and reduce cost. There are also examples of residents reducing unnecessary laboratory testing10 and improving procedural performance of central line placement in the intensive care unit.11 Engagement in these activities during training will prepare residents for the payment incentives based on achieving quality measures when they enter independent practice. Assessment of trainee engagement in systems design and improvement can be through a QI/PS portfolio that describes the problem and analyzes its causes, solutions, and the trainee’s participation in the process and outcomes of the initiative.
Finally, students and residents need to make the connection between improving systems at their institutions and participating in larger systems change through health policy engagement. Residents like Gil who become frustrated with institutional policies, such as the policy about dialysis of undocumented immigrants, need to understand how to participate in changes in state and national health policy. Were Gil to attend the meetings where the health needs of undocumented immigrants were discussed, he would be able to contribute his story to influence changes in legislation or local policy.
Also, it is only though participation in the discussions about policy topics that affect health systems that trainees will be able to make the connection between the various elements of the health system. The range of important topics for engagement by residents and students is broad and could include public health issues like bicycle and motorcycle helmet laws, vaccination requirements, regulations about payment to hospitals and physicians related to quality and safety, and regulations about publicly financed programs like Medicaid and Medicare. Assessment of policy activities could also involve a portfolio that describes the issue; its history; and current challenges, policy options, and political considerations. The portfolio would include a description of the trainee’s involvement and the learning issues that were encountered.
I wonder how different medical education would be if health systems were the central organizing principle of the curriculum from which everything else flowed. There would still be the need to understand anatomy, biochemistry, pathology, genetics, and much else; students would still need to carry out many activities such as diagnosing infections and operating on broken bones. They would still need to listen to patients tell their stories and understand their concerns and values. But they would be thinking differently, using systems thinking, about how to best connect the various health resources to the patients to meet their needs, in the way that the people in Ziegelstein and Fiebach’s1 village thought about how to create safe schools and drinking water and neighborhoods. They would care when the phones in the clinics did not get answered, and would think about transportation to the clinic and the cost of care—all factors that contribute to changes in the culture of health and the health system that Gupta and Moriates,12 in this issue, feel must change and become value based.
When I reflect on the unpleasant interactions with colleagues about our problem patients, Mr. Mares and Mr. Jackson, and the accusations about unprofessional and poor-quality care, I can’t help but wonder whether these interactions reflect the cultural chasm that we in academic medicine need to cross. We stand at the precipice doing our best to solve problems with individual solutions for individual patients that often fail because of gaps in the system. Can we reach the other side of the chasm? I don’t think we will make it if we retreat into an isolationist self-protective shell. We have to embrace systems thinking and the culture that comes with it to build bridges and connections for our system. If we lead, I am sure our students and residents will be right there beside us.
David P. Sklar, MD