Clinicians know there is a clear ethical obligation to disclose errors to patients. However, the individual who finds a diagnostic error may not be a clinician in direct contact with the patient. Instead, it might be a researcher who is reviewing a chart long after a patient was discharged.

As a fellow at Texas Children’s Hospital in Houston, Grant J. Shafer, MD, performed retrospective evaluations of charts that included information regarding pediatric patients who had been discharged. Researchers found some diagnostic errors made in patients who had been discharged from the neonatal ICU. This raised the question: What are researchers’ ethical obligations if they find an error no one else had discovered?

“The field of diagnostic errors is relatively new to patient safety as a whole,” Shafer says.

“Surprisingly, there hasn’t been a lot of guidance in this field. The focus has been on the clinician/patient dyad,” says Shafer, currently a neonatologist at Children’s Hospital of Orange County in Fullerton, CA. When a patient/physician relationship exists, the ethical obligations are clearer. When conducting patient safety research based on retrospective chart reviews, there is no contact with patients.

“We looked at the literature, and there was really nothing to give specific guidance on people who found errors retrospectively,” Shafer says.

Under these circumstances, the researcher is looking at the case when it is past the phase of diagnosis, sometimes even years past.

“What makes it so challenging is that we end up going back and reviewing charts, and we don’t necessarily understand what the clinician was seeing at the time,” says Frank X. Placencia, MD, MS, an attending neonatologist and one of the founding members of the ethics consult service at Texas Children’s Hospital.

These situations can make it difficult to determine if someone made an error. “We may be judging the clinician’s actions in a way that isn’t totally fair. That makes it hard to create policies,” says Placencia, an assistant professor at Baylor College of Medicine.

Shafer and Placencia co-authored a paper about this topic.1 Diagnostic error researchers are going to need guidance on this ethically complex issue. “This is a discussion that’s going to continue to come up,” Shafer predicts.

One question is whether errors found during retrospective reviews demand disclosure directly to the patient. If so, it is unclear exactly who should contact the patient. “As researchers, they aren’t our patients directly, but they are patients of the health system as a whole,” Placencia notes.

Some errors, such as missed syphilis diagnosis, will continue to cause harm to the patient. Thus, this would create an urgent need to alert the patient. There also is the ethical obligation to disclose errors to treating clinicians so they can be aware and improve the care they provide.

Researchers need to respect the boundaries of the patient/physician relationship. “We did not want to recommend to a researcher to do an end-run around that line of communication,” Placencia says.

Ideally, it is the clinician who communicates with the patient directly about the error. Clinicians are not necessarily still working at the institution at the time the error is found. For research using de-identified data and large databases that specifically do not include patient identifiers, it is impossible to follow up with the individual patients.

Researchers can use existing institutional disclosure policies to find guidance on all these scenarios. Still, policies alone are not enough, according to Placencia: “It’s incumbent on hospital ethics committees to foster a culture of disclosure.”


  1. Shafer GJ, Placencia FX. The ethics of disclosing diagnostic errors: What is the researcher’s duty? JAMA Pediatr 2020;174:405-406.