The Code of Professional Conduct:

Grappling with Its Failure

 

Towards the end of second year, the situation seemed hopeless. In an email to friends seeking advice for their last rotation, I tried to warn them. I told them that the worst was coming out of everyone and that you would see and do things that you would be ashamed of. The code of professional conduct was failing and standards of professionalism were dissolving.

As I was struggling with my own failures, I found comfort in learning more about the problem both at Duke and nationally. Around that time, Dr. Danielle Ofri wrote an article called “The Darkest Year of Medical School,” in which she described the decline in empathy and professionalism during medical students’ first year on the wards. The trend was well documented around the nation. Also around that time, Dr. Chudgar gave a lecture on professionalism, during which it became clear that the spectrum of what students considered acceptable behavior had grown tremendously. The black-and-white of right-and-wrong held by most students at the beginning of the year transformed into a world of gray where ideas about professionalism depended on a range of questions: who was on your team, how many hours did you sleep, was your supervisor evaluating you, when was your SHELF exam, etc. This was further evidenced at a student retreat at the conclusion of our second year. The advisory deans revealed that nearly everyone witnessed honor code violations, but very few reported them. Many students, including myself, who failed to report violations, cited “it could have been me.” The honor code became a system of implicit collusion and dishearteningly, our class’ survey results were “typical.”

Feeling disheartened at the end of my second year, I struggled to understand the cause of these changes. Students emerging from the experience still have good values, but in hindsight we seem to agree that we “bent the rules.” In a public policy course covering game theory, it became clear to me that behavior on the wards was not perplexing, but completely rational. The forces at play can be modeled as individuals choosing between communal responsibilities to clinical duties and individual responsibilities to study for exams and achieve high marks.

For example, consider a group of 10 medical students rotating together. Each student has 10 units of energy, which they invest either in patient care or in personal activities (e.g., studying, sleeping, etc.). The units of energy dedicated towards patient care go into a common pool that upholds the code of professional conduct and fosters trust in the work place. Every unit invested in the common pool is doubled and then divided equally among the 10 students to signify the incentive to invest in the community.

On one hand, the higher the investment in the common pool, the higher the payout to all students. If every student chooses to invest 8 units of energy into upholding the code of professional conduct, the 80 units are doubled and each student is returned 16 units of energy. Adding the 2 units held onto for personal activities, every student ends up with 18 units, nearly double what they started out with. Thus, collective investment can make life better for everyone.

However, individual students are also incentivized to not invest in the common pool. Regardless of how well students perform on clinical duties, obtaining Honors on most rotations requires countless hours of studying. A unit of energy invested in the common pool is a unit of energy not spent studying. If student A contributes a unit of energy into the common pool, this increases the common pool by 2 units after the doubling. But 1.8 of the increment is distributed to the other 9 students; student A only gets back 0.2 units. From this perspective, contributing nothing into the common pool is strategic. For example, if student A keeps his 10 units of energy to study while the other 9 students each invest 8 units into the common pool, student A leaves with 24.4 units (10 plus dividends from the common pool) and every other student leaves with 16.4 units (2 plus dividends from the common pool). The same reasoning can be applied to any student who wants to spend time studying, but hopes to be a “free rider” on the efforts of others to fulfill clinical duties.

Countless situations during second year reveal this tension between collective and individual interests. A classic example is a clinical requirement, say a lecture, the day before a SHELF exam. Collectively, students want each other to attend the lecture to uphold the code of professional conduct and maintain trust in the work environment. However, students, like myself, individually feel pressure to miss the lecture and invest that time and energy into studying. The dominant strategy would be to miss the lecture and be a “free rider” on the effort of others.

Game theorists describe this set up as a multiperson prisoners’ dilemma. Players are assumed to act out of individual interest and abide by their dominant strategy, which in this case would be to defect and violate the code of professional conduct. However, if all students play their dominant strategy, the common pool is left empty. Despite how well students perform on SHELF exams, trust is eliminated from the work place. Medical students lose their sense of professionalism and supervisors lose their respect for students. One supervisor, surprised that another medical student and I were showing up to work, told us he became accustomed to students escaping to the library.

Strategies to improve the Duke University School of Medicine code of professional conduct can be found in The Art of Strategy, by Avinash Dixit and Barry Nalebuff. The authors propose ways to achieve successful cooperation in multiperson prisoners’ dilemmas. New methods of detecting cheating must be introduced, because the current system of peer-reporting is failing. If the detection is fast and accurate, repercussions can be immediate. This reduces the gain from cheating while increasing its cost, and thus increases prospects for successful cooperation. Second, repercussions must be appropriate. If students fear that honor code violations will permanently damage an individual’s record, violations will persist unreported. Third, boundaries of acceptable behavior must be revised. Expectations must not remain an unattainable ideal, but must be grounded in the reality of the medical student experience. If certain violations are truly ubiquitous, we must question the appropriateness of regulating those behaviors. If the majority of students feel that a clinical experience wastes their time, must they be required to participate when they could instead be learning information necessary to practice medicine? Fourth, student groupings must be made more stable and enduring. New teams of students who do not have a stake or a history of participation in a collective arrangement are less likely to cooperate. Similarly, if an established group expects to be disrupted, incentives increase to cheat and take extra benefit right now. Thus, the transience of team composition during second year rotations encourages honor code violations. Finally, students must have confidence that there will be repercussions for defection and that cooperation will be rewarded.

The multiperson prisoners’ dilemma provides great insight into the failure of the medical student code of professional conduct. Instead of blaming individual students for shirking clinical responsibilities or individual supervisors and clerkship directors for a lack of oversight, understanding the incentive structure helps explain the frequency of violations. But despite what feels like complacency with a failing system, the situation is not hopeless – the system can be revised to enhance cooperation and standards of professionalism. The medical school has a responsibility to refine the code of professional conduct in order to graduate students who can more skillfully balance personal and collective interests.