Improved hospital safety incident reporting begins with medical residents

A research interest in understanding how to encourage employees to report unsafe work conditions has led to important findings aimed at ultimately improving safety incident reporting procedures in hospitals.

Rangaraj Ramanujam, Associate Professor of Management at Vanderbilt's Owen Graduate School of Management, co-authored a recent study focusing on medical residents -- doctors fresh out of medical school and beginning their training. The findings showed that while medical residents report errors at a fairly low rate, even enacting a series of simple steps may help increase their rates of incident reporting.

The key to involving more residents in the process of improving patient safety is additional academic training to emphasize and encourage engagement in incident reporting, Ramanujam said. The study was conducted at a major medical center in the Midwest.

"When I started asking questions about medical error reporting among doctors, I discovered that the best way to understand the cultural norms is to go back to their early training," Ramanujam said.

"Physicians are highly professionally socialized. Their early training tends to have lasting effects on their attitudes and behaviors later in their careers," he said. So it made sense to look for early patterns.   

Although Ramanujam's prior research focused on error reporting in many industries, including financial services, there are few places where error reporting is more critical than hospitals -- from a cost-saving, patient safety and patient outcome standpoint.

"Incident reporting is typically a voluntary activity, which is not required in the formal job description. Questions such as when and why employees choose to go beyond the call of duty and report incidents are conceptually interesting. But even more, these questions have great practical significance in settings such as hospitals where incidents can have life-or-death consequences. Therefore, in this research I am especially interested in developing interventions or changes to get residents, and ultimately physicians, more involved with incident reporting," he said. "It's not a formal part of your job, yet if you want the organization to be safe, you want to encourage employees to do this."

As Ramanujam and his co-author, Dr. Lia Logio of Indiana University School of Medicine, prepared to launch their study, they found little previous data on error reporting among residents and sparse institutional training, other than a few safety classes.

Most medical residents did not know the basic procedures for submitting incident reports in the hospitals where they undergo training, the research showed. Moreover, even the residents who knew about reporting procedures reported incidents only infrequently. These findings were surprising, Ramanujam said, because a number of previous studies by other researchers discovered that mistakes and unsafe conditions occur regularly in hospital settings. So, anyone with even minimal experience, including a resident, would almost certainly encounter situations that warrant incident reporting.

However, he noted that it is important to keep in mind that not all safety incidents are medical errors. Some are just deviations from normal procedure but still worthy of correction.

The research also showed that residents' reporting behaviors seem to be shaped by the culture of the different hospitals within the same academic center that the residents worked in during their rotational training. The researchers examined the incident reporting behaviors of 305 residents who had each worked in all five hospitals in the academic center. They learned that residents were more likely to have reported incidents in some hospitals than in others.

Each of the training hospitals played an important role in shaping residents' safety communication behaviors. Therefore, individual hospitals – not just the major academic center -- should encourage residents to report incidents and emphasize their role in improving the whole system, the researchers concluded.

On the positive side, the study reported that residents frequently discussed safety incidents with peers and some faculty on an informal basis, demonstrating awareness that even small incidents merit attention.

The study involved two online surveys of more than 900 medical residents and fellows as they rotated among five IUSM-affiliated hospitals, including a large community hospital; a university referral hospital with expertise in tertiary care; a well-known children's hospital; a VA hospital and a public county facility.

The study, the largest of its kind, is also one of the first to explore whether and how residents' reporting behaviors change as they move among hospitals. Ramanujam applauded IUSM's participation and desire to better understand and improve incident reporting among medical residents.

"The underlying goal for the study is to get at what can we do to train physicians to become more engaged from the get-go in the process of continuous improvement in patient safety," he said.

The findings were reported in an article, "Medical Trainees' Formal and Informal Incident Reporting Across a Five-Hospital Academic Medical Center," which appeared in the January 2010 issue of the Joint Commission Journal on Quality and Patient Safety.

"The findings are important in an era of health care reform. While the main impact of better incident reporting by residents will be seen once they move along in their careers and have more responsibility for safe patient care, it will also mean fewer mistakes that can be costly for patient safety and the bottom line," Ramanujam said.

"Some of the reasons residents don't report more incidents are mundane. So the proposed solutions are simple but their long-term effects are potentially profound."

Based on their findings, Ramanujam and Logio recommended a number of steps to improve incident reporting:
  • Adopt an institution-wide, comprehensive patient safety curriculum that emphasizes the importance of communicating safety incidents, no matter how minor.
  • Encourage faculty physicians to discuss incidents that might involve or be witnessed by residents. 
  • Provide formal and informal opportunities for residents to discuss incidents.
  • Provide anecdotal information about lessons learned or improvements made from previous incident reporting. 
  • Work with hospitals where residents are trained to ensure there are procedures in place to make residents aware of reporting processes.
  • Use orientation to inform residents of hospital protocol.
  • Instill a sense of responsibility for improving the system.
  • Empower residents with greater "ownership" of patient care.

Ramanujam's broader research focuses on the organizational causes and consequences of operational errors in a variety of institutional settings. The basic premise of this work is that the causes of many major errors, and hence their remedies, are organizational.

"So I am especially interested in understanding the structures and processes that enable organizations to effectively manage errors. That is, why are some organizations consistently more effective in minimizing errors, detecting errors, containing the fallout from errors and learning from errors?" he said. "Although there is no simple answer to these questions, I'd say that one common feature of successful organizations is that they seem to be good at encouraging their employees to discuss errors openly. If employees can talk about errors freely, that makes the early detection and correction of errors a lot easier."

His earlier work with Dr. Subra Tangirala of University of Maryland focused on a different kind of communication behavior—silence. Specifically, he explored the reasons that nurses who are highly committed to patient care hold back information when they think there is potential harm to patients. His results focused on the issue of perceived fairness.

"We found that if the managers in a unit are not seen as being fair procedurally, then our data showed pretty clearly that nurses are more likely to remain silent. A safe culture is also a fair culture," Ramanujam said.

In general, safety communication tends to be more effective in organizations where employees at all levels, particularly at the lower levels, are willing to go beyond the formal requirements of their jobs and voluntarily share information about unsafe conditions that they encounter in their work, feel comfortable about sharing such potentially sensitive information, and believe that the information they provide will lead to meaningful change in work processes and outcomes, he said.

Ramanujam cited recent examples when lower-level employees had concerns that were dismissed by top management. Most notably, in the case of the Challenger and Columbia explosions, NASA employees had concerns early on about the products that contributed to the disasters. Employees also had misgivings about Vioxx, an anti-inflammatory drug prescribed for arthritis patients that was shown to have harmful side effects. It ultimately was withdrawn from the market and class action suits have followed.

"It's really important for organizations in any number of settings to create conditions where employees feel comfortable reporting errors, because errors are valuable as a basis for organizational learning and improvement," Ramanujam said.

MBA students should be trained to talk about errors as well, he said. "You have to learn to share information."

Ramanujam's next project focuses on how incident reporting behaviors among medical residents change over time. " For example, we'd like to verify whether the safety communication behaviors of first-year residents and third-year residents are different," he said. "We are also trying to see if there are systematic differences between whether residents report differences across different specialties."

In addition, a third piece of work in progress concerns organizational efforts to learn from medical error reporting. "Part of my interest is in learning how organizations purposefully go about analyzing data and making changes," he said.


Published May 3, 2010 in Vanderbilt Business Intelligence
Copyright 2010 Vanderbilt Owen Graduate School of Management