Understanding Factors Associated with
Ethical & Compliance Lapses

Supported by grants:

BF Foundation seed grant (DuBois, PI)

1R21RR026313-01 R21 NIH (DuBois, PI)

1R01AG043527-01 (DuBois, PI)

We study lapses in professionalism with the aim of improving training programs and policies. Solid research data are needed to develop programs and policies that are no more burdensome than necessary and associated with positive outcomes for researchers, institutions, patients, participants, and other stakeholders.

In studying lapses in professionalism, we aim to identify factors that contribute to real world breaches. Prospective, experimental designs are often viewed as most rigorous, but they have limitations: They typically manipulate only one or two variables, and they occur in highly controlled conditions that frequently differ from real world situations. For these reasons, we have turned to “historiometric” methods: Statistical analysis of large sets of coded, historical, narrative data (e.g., from court cases, investigative journalism reports, or board investigations).

A first step in understanding professional wrongdoing was identifying the different kinds of ethical lapses that exist to enable heterogeneous sampling and tracking of how behaviors cluster. Through a systematic literature review, examination of 300 cases, interviews with experts, and a test of inter-rater reliability, we developed a taxonomy of wrongdoing in medical research and practice.

DuBois, J.M., Kraus, E., Vasher, M. (2012) “The Development of a taxonomy of wrongdoing in medical practice and research,” American Journal of Preventive Medicine, 42 (1): 89-98. PMC3244684

We identified 14 – 15 primary kinds of wrongdoing in each domain (research and practice).

We next examined the literature to determine which environmental variables might contribute to lapses in professionalism. We identified 10 factors that appeared in cases and were supported by experimental studies, e.g., playing conflicting roles, vulnerable victims, and oversight failures.

DuBois, J.M., et al., (2012) “Environmental factors contributing to wrongdoing in medicine: A criterion-based review of studies and cases,” Ethics & Behavior, 22 (3): 163-188.

We then researched 120 cases of wrongdoing in research and medical practice, wrote a synopsis of each case, had a team of five experts rate the severity of the wrongdoing, and tried to identify factors that predict severe wrongdoing.

DuBois, J.M. et al. (2013) “Understanding the severity of wrongdoing in healthcare delivery and research: Lessons learned from a historiometric study of 100 cases,” American Journal of Bioethics, Primary Research, 4(3): 39-48.

While we found that some factors were present across all kinds of cases (e.g., a preponderance of males, repeated misbehavior, and a duration of 4 years), we also found that different variables are associated with different kinds of wrongdoing. That is, professional wrongdoing cannot be approached as one coherent construct such as “organizational deviance.”

This finding was reinforced when we contrasted cases of research misconduct (data falsification, fabrication, and plagiarism) to other kinds of wrongdoing.

DuBois, J.M. (2013) “Understanding research misconduct: A comparative analysis of 120 cases of professional wrongdoing,” Accountability in Research, 20(5-6):320-338.

In this paper we found that the primary factors associated with research misconduct are individual rather than environmental: Self-centered thinking, unwarranted certainty, and recklessness with data or oversight of personnel.

We are currently examining 100 cases of abuse of opioid prescribing privileges, 100 cases of sexual abuse of patients, and 100 cases of performing unnecessary invasive procedures. In these studies, we track approximately 70 variables that describe the case setting, dates, duration, and victims, as well as the wrongdoers and environmental factors that might explain the misbehavior (such as conflicts of interest, oversight failures, and substance abuse). We then develop a theory of the case, and identify the variables that appear most salient in understanding the motives, means, and opportunity for the misbehavior found in the case.

Next steps will include applying our methodology to investigate 100 cases of human subjects research scandals. Preliminary data analysis suggests that the factors associated with such scandals are quite different from those found in research misconduct cases.