The medical literature pertaining to patient safety is voluminous, especially since the publication of the 1999 Institute of Medicine (IOM) report estimating that between 44,000 and 98,000 Americans die from medical errors each year.1 When focusing on hospitalists and the matter of near misses, however, where the outcome of the error has been intercepted before harm can occur, the volume of published information is far more limited.2
“There are few good data about the role that hospitalists play in preventing adverse events,” says Saul Weingart, MD, PhD, vice president for patient safety at the Dana-Farber Cancer Institute, Boston, “so most accounts are anecdotal. My own sense is that hospital medicine plays a critical role in preventing and mitigating errors in the acute care setting.”
One study, conducted by two hospitalists in a 200-bed academic hospital and published in 2003 in the Journal of General Internal Medicine, distinguished near misses from adverse events.3 The investigators set out to determine the frequency, types, and consequences of errors that can be detected by hospitalists during routine clinical care, and to compare the types of errors first discovered by hospitalists to those discovered by other providers.
Sarwat Chaudhry, MD, a former hospitalist who is now a health services researcher and assistant professor of medicine at the Yale School of Medicine in New Haven, Conn., was the lead investigator in that study. “Oftentimes the errors that are ‘sexier’ and more exciting for people to talk about are the adverse events,” she says. “But I think what are perhaps more common and more informative are the near misses.”
For purposes of Chaudhry, et al.’s study, “adverse events” were defined as adverse outcomes that resulted from errors.3 Errors that did not result in patient harm, but could have, were categorized as near misses. The umbrella definition of error fit that which was used in the IOM report; that is, “the failure of a planned action to be completed as intended or the use of a wrong plan to achieve an aim.”1 Cases in which a bad outcome occurred (such as death or disability) without a preventable cause were not considered errors and, therefore, were not included in the data.
A more practical understanding of a near miss is apparent in the following illustration: “If pharmacy came to me and said, ‘Oh, Dr. Chaudhry, one of the physicians on your team wrote for drug X but the patient is also on drug Y and that’s a known contraindication so we don’t think this should be administered, do you agree?’ I’d say, ‘Yes, I agree.’ So the drug was never given, but it was a near miss for the way it was written if it hadn’t been picked up on.”
—Saul Weingart, MD, PhD
In fact there are two kinds of near misses. “One is where the error is intercepted by [a] pharmacy or somebody else before harm is done,” says Dr. Weingart. “The other kind is where the error reaches the patient—for instance, where he or she receives the antibiotic to which they are allergic—but no reaction takes place.”
An important element in the study was that both types of errors were identified during the course of routine, clinical care by Dr. Chaudhry and her colleague, Kolawole Olofinboba, MD. Errors first detected by other healthcare workers were also recorded. Of the 528 patients included in the study, 10.4% experienced at least one error: 6.2% a near miss and 4.2% an adverse event. Interestingly, most of the errors first detected by house staff, nurses, and laboratory technicians were adverse events, whereas most of the errors first detected by the attending hospitalists, pharmacists, and consultants were near misses. Drug errors were the most common type of error overall.
Although differences did not achieve statistical significance, Dr. Chaudhry says the data were illuminating and the hospital administration was anxious to clarify the processes of care that could be modified to make the workflow more efficient and the different pieces of each patient’s critical data more integrated.
“For instance,” she says, “at our hospital there was a lot of separation of data. … The vitals were right outside the patient’s room. The list of medications was kept in a central pharmacy location, and the medical chart with everyone’s notes was in a third location.”
After the study’s completion, the integration of data was facilitated by the adoption of an electronic medical record (EMR) and there is ongoing research at that institution concerning whether that has affected error rates.
—Sarwat Chaudhry, MD
“This was one of the first and, still I think, only studies of attending physicians reporting their own errors,” says Dr. Chaudhry, “as well as reporting the scope, type, and frequency of the errors.”
There were several lessons learned. The first is that near misses are very common—at least as common as adverse events and perhaps more frequent. The second is that the nature of near misses is similar enough to that of adverse events that they can still be informative in preventing harm to patients. The third lesson arises from the investigators’ review of the kinds of errors that were reported by the different providers. “Different types of providers are going to pick up on different kinds of errors,” says Dr. Chaudhry. “Pharmacy, of course, is going to be detecting drug errors. But nurses are going to be detecting a different kind of error than residents, and residents are detecting a different kind of error than attending physicians.”
Dr. Weingart, who was trained as an internist, says his sense is that “we often pick up errors that occur just upstream in the process of care. For instance, pharmacy finds doctor prescribing errors and nursing finds pharmacy dispensing errors.”
Drs. Chaudhry and Olofinboba discovered 17/47 (36%) of the errors on their own, and pharmacists, resident physicians, consulting physicians, nurses, and ancillary staff discovered the remaining 30/47 (64%). The hospitalists were more likely to detect therapeutic errors; house staff physicians were more likely to detect drug errors. House staff also detected procedural errors at a higher rate than did the hospitalists. Consequently, Dr. Chaudhry emphasizes, another highlight of these data shows that “engaging as many different kinds of healthcare providers in error reporting is very important to understand what’s going on.”
The two hospitalists in Dr. Chaudhry’s study underwent no formal training in error detection, and all errors identified were documented. This in itself is a remarkable aspect of this study for hospitalists to note.
“In most hospitals, the risk management department feels like breaking out a bottle of champagne to celebrate every time a physician submits an incident report,” says Dr. Weingart, “because in most hospitals physicians don’t view incident reporting as part of their responsibility. On the other hand, physicians constantly talk about the errors, near misses, and adverse events they encounter day-to-day. Unfortunately, formal incident reporting isn’t an activity that most physicians see as a part of their duty to deliver excellent care at the bedside.” By virtue of this, he says, “critical information is often lost to the system. Error experience is transmitted person-to-person but isn’t captured and analyzed.”
The good news here, however, is that by simply launching a study of this nature and capturing the data it is reasonable to expect that any hospitalist group could take it on and any hospital’s quality of care would be improved.
“There is a spectrum of practice review that is conducted by a variety of people,” says Dr. Chaudhry. “On one end of the spectrum, you have the uninvolved reviewer coming in and doing formal chart reviews … and our process was at other end of spectrum, where we were going about our daily business, the routine clinical care of patients.”
By maintaining the potential for errors at the forefront of their thinking, she says, it became the background against which they performed the constant daily review of patient data. “When I would go to the bedside in the morning,” she says, “to see the patient, talk to the patient, review the chart, look at notes, look at vitals, look at meds—that was my process for picking up on errors.”
—Saul Weingart, MD, PhD
Barriers and Opportunities
There are two major barriers to performing this kind of research in other institutions where hospitalists practice, but they are easily resolved. The first is the importance of standardizing definitions.
“We came to consensus before the study began of what things should count and what things shouldn’t count,” says Dr. Chaudhry. “Because there is a degree of subjectivity and especially when a patient experiences harm, it’s a bit easier to attribute that to an error—though not always. For instance, with a patient with asthma who ends up with fluid overload and gets intubated. Well, was that an error? … Or was that just the patient’s own asthma getting worse? How much did the fluid contribute to that?”
There is still a degree of judgment, even there, she says, “but as much as possible we wanted to come to consensus at least about our definitions, our terminology, and our categorization.”
The second barrier to performing this research is the real or perceived risk of litigation. There was a lot of concern at the beginning of her study, says Dr. Chaudhry (and again at the time of publication), regarding the aftermath of disclosing errors.
“As much as the Institute of Medicine and other accreditation and safety organizations talk about how error reporting is so important and that physicians have to be willing to come forward with their errors, there are really no formal, legal, protective mechanisms in place,” she says, adding that this was true three years ago, and she does not believe much has changed.
Consequently, when constructing the study, she reached out to the authors of other error studies and asked them how they had handled the risk of potential litigation. Most of them told her they did not think there was much risk of being sued, but disclosing errors in the literature is certainly not something an institution or individuals would be clamoring to do. If a multi-center study of this nature might be undertaken, she says, these barriers could be traversed and it would be worth it. “Because if we let that opportunity slip through our fingers, it will be taken on by nonphysicians,” she cautions, “by the external reviewers, by the accreditation organizations, by professional quality control people.”
Dr. Chaudhry’s overall recommendation to hospitalists involves the position they hold at this time. “Hospitalists are ideally poised to take a lead in promoting patient safety by becoming more involved and developing the systems for error reporting,” she says.
Dr. Weingart tells hospitalist audiences that the most important areas for detecting errors in their practice include safe prescription writing and error-proofing high-risk activities (such as hand-offs and sign-outs, and follow-up of test results). He also recommends that hospitalists play the role of responsible citizens in their institutions and be open to ways your patients can contribute by helping you detect errors in their care.4 These issues will be addressed in more detail in future articles in The Hospitalist.
Although Dr. Chaudhry is no longer practicing as a hospitalist, she believes that hospital teams working closely together in patient care have a strong advantage to identify any errors.
“As physicians, with hospital medicine expanding,” she says, “this is the opportune time to take on a leadership role, and this needs to be part of the hospitalist mission. Not everyone has to come out and report their adverse events, especially since the legal risks are still unclear. But a safe way and equally informative way to work on preventing harm is to focus on those near misses.”
Hospitalists who search for near misses, volunteer this information, and enroll their colleagues in doing the same contribute to reforming systems and processes that will ultimately reduce the potential for all errors. Those who pay close attention to standardized rules for prescription-writing, error-proofing high-risk activities, and their roles for institutional citizenship and encouraging patient contribution in regard to possible errors, are doing the best possible job to prevent harm to patients and the institution as a whole. TH
Andrea Sattinger will write about apology in medicine in the June issue.
- Kohn LT, Corrigan JM, Donaldson MS, Institute of Medicine Committee on Quality of Health Care in America. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 1999.
- Bleich S. Medical Errors: Five Years After the IOM Report. New York: Commonwealth Fund; July 2005: Available at: www.mihealthandsafety.org/pdfs/pub-4-commonwealth.pdf. Last accessed March 22, 2006.
- Chaudhry SI, Olofinboba KA, Krumholz HM. Detection of errors by attending physicians on a general medicine service. J Gen Intern Med. 2003 Aug;18(8):595-600.
- Weingart SN, Pagovich O, Sands DZ, et al. What can hospitalized patients tell us about adverse events? Learning from patient-reported incidents. J Gen Intern Med. 2005 Sep;20(9):830-836.