Keeping track of mistakes that don't happen

Near-miss tracking project for residents aims to improve hospital care.

Visualize an attending physician inserting a central line into a patient when the senior resident notices that something doesn't seem right. “The senior resident says, ‘No, that actually looks arterial to me. Stop!’” described Spencer G. Nabors, ACP Associate Member. The attending stops, realizes the resident is right, and avoids potential complications for the patient.


If this hypothetical encounter occurred in New York State before 2005, the resident could do little more than pat himself on the back and move on to the next patient. But with the recent expansion of an ACP-led near-miss tracking program, residents in any of New York's 64 teaching hospitals can go online anonymously to report near-miss events and possibly prevent similar mistakes from occurring in the future.

The researchers behind the project are looking for reports on everything from incorrect drug orders to sending the wrong patient for a scan—any kind of error that is caught before it harms the patient, said Dr. Nabors, a member of the project's advisory committee. The New York Department of Health already has a mandatory reporting system for adverse events. The hope is that an analysis of near-miss events will provide insight into ways to reduce the number of events that actually do cause harm.

“It's widely believed that there are far more near-misses than fully committed errors,” said project leader Ethan D. Fried, FACP, director of graduate medical education at St. Luke's-Roosevelt Hospital Center. “We can gather a lot of data about medical errors in a relatively short period of time because we're collecting near-misses.”

Anonymity ensures reporting safety

Data have already begun flowing into the program's Web site, which is a cooperative effort of the New York chapter of ACP, the Association of Program Directors in Internal Medicine New York Special Interest Group, the Committee of Interns and Residents, and the New York State health department.

The pilot phase, which was conducted at five hospitals starting in 2005, collected almost 50 reports in three months. “The majority of near-misses in the pilot phase were communication issues—cases where a covering team was unaware of an allergy, or unaware that a patient had already been started on a medication that was not effective,” said Dr. Fried.

Momentum has been building slowly for the larger-scale project since its launch last year. Dr. Fried is traveling from hospital to hospital in New York, explaining the program to residents during noon conferences or grand rounds. He distributes log-in information, so that the residents are ready to report anonymously on any near-miss events they are involved in.

That anonymity was a key point for designers of the project, noted Dr. Fried. “The more anonymous it is, the safer it will be to report events, and the more likely people will report things.”

The Web site reporting form does not even collect the name of the hospital where the near-miss occurred, although it does ask enough detailed questions to gain useful data about the demographics of the hospital, such as its size, scheduling systems and level of technology. The project originally grew out of program directors' concerns about the impact that work-hour reductions might be having on patient care, but it has since broadened in scope.

Preventing mistakes from becoming adverse events

Once a baseline quantity of data has been collected, the program will put out a quarterly newsletter for participating hospitals to update them on the findings, and present the research that is relevant to specific institutions. “We'll be able to say, based on these characteristics, here's what you're at risk for,” said Dr. Fried. “These are the errors we see in large urban centers with computerized order entry and night float whereas these are the type of errors we're seeing in smaller rural hospitals that don't have computer order entry.”

In addition to looking at what causes the mistakes, the survey also attempts to uncover the factors that prevented them from becoming adverse events, whether it was intervention by another hospital staff member or some kind of equipment safeguard. “We can see what the most effective barriers are that protect the patients. Strengthening those barriers, we think, is going to be the ultimate outcome,” said Dr. Fried.

Dr. Nabors also has other, more personal hopes for what the project will accomplish. “In medicine, we breed this culture of ‘We need to be right all the time.’ To me, that also creates an environment where those problems and near-misses are going to occur more, because people are going to be less inclined to question or admit their concerns about their skill level,” he said. By focusing attention on near-misses and encouraging residents to address them, the tracking program could encourage a culture shift in which the possibility of error is more openly acknowledged, Dr. Nabors said.

His and other questions about the impact of near-misses should be answered, at least in part, in a little over two years, when Department of Health funding for the three-year project runs out, and when Dr. Fried and colleagues will analyze and publish their data.

However, that may not be the end of the investigation into near-misses, said Linda Lambert, executive director of ACP's New York Chapter. “We're hoping at some point to find a stable source of funding that will allow the project to continue and even grow. Our ultimate goal is to make it not just internal medicine resident reports, but reports for everyone in the health care system—meaning nurses, attendings, lab techs and cleaning staff who discover something that prevented an accident from happening.”

Program organizers would also eventually like to expand the tracking system to all hospitals in New York, not just teaching facilities. But whether or not that happens, the findings from this study should be useful to hospitals around the country, Dr. Fried predicted.

“I think New York is so diverse that the results will be generalizable. There are rural hospitals. There are urban centers, larger urban centers, small urban centers, VA hospitals, all kinds of environments, all kinds of venues for patients,” he said.

The determining factor will be whether hospitals decide—as those in New York have done by signing onto the tracking project—that looking at near-misses is necessary to improve their patient care. “It's really been very important that the hospitals buy into the fact that a risk-free, anonymous reporting system might ultimately help them design or redesign their safety systems,” said Dr. Fried.