Three hundred and sixteen articles, representing 254 unique first authors, were included (109 reviews, 74 editorials/commentaries, 69 empirical studies, 42 letters, 18 personal narratives, 4 interviews). Content analysis of the articles identified 91 factors (53 impeding and 38 facilitating), as shown in Appendix B
. Reliability testing showed that the majority of the most commonly cited factors were reproducibly identifiable within and between investigators (κ statistics for the 10 most commonly cited facilitating and impeding factors ranged between 0.47 and 0.96).
The literature review identified 91 factors (53 impeding and 38 facilitating), as shown in Appendix B
. The 10 most frequently cited facilitating factors were as follows: accountability, honesty, restitution, trust, reduce malpractice risk, consolation, fiduciary relationship, truth-telling, avoid “cover up,” and informed consent. The 10 most frequently cited impeding factors were as follows: professional repercussions, legal liability, blame, lack of confidentiality, negative patient/family reaction, humiliation, perfectionism, guilt, lack of anonymity, and absence of a supportive forum for disclosure. The 3 most common contexts for error disclosure were as follows: (1) reporting errors to institutions to improve patient safety; (2) discussion of errors among physicians to enhance learning; and (3) informing patients about errors as part of patient care. Statistical analyses (32) showed that the most commonly cited facilitating factors, except for accountability, were more frequently mentioned in articles focusing on disclosing errors to patients, as opposed to institutions or colleagues (P
< .001). By contrast, impeding factors showed no consistent differences in the frequency of citation based on the context of disclosure.
Final Taxonomy, with Selected Annotations from Exploratory Focus Groups
The 27 factors identified from the focus groups were combined with the 91 factors from the literature review for a total of 118 facilitating and impeding factors. All the factors were reviewed by 1 of the investigators (L.C.K.) and those factors that appeared directly related were aggregated into mini-groups, as shown in Appendix B
, reducing the number of factors and factor mini-groups to 61. In the pile-sorting task, participants placed the 61 factors and factor mini-groups into an average of 7 piles. In the hierarchical cluster analysis, plots of cluster statistics did not reveal a definitive jump in the values that would suggest an obvious cluster solution. Using Ward's method, we reviewed printouts of 5 different clustering solutions (factors clustered into 5, 6, 7, 8, and 9 clusters) and determined that the 8-cluster solution (R2
=.586) was the most satisfactory in terms of clinical interpretation; this solution established the number of domains for the taxonomy as well as the initial contents of the 8 domains. As a result of the confirmatory focus groups, 4 items were relocated, 1 item was deemed redundant and dismissed, and 1 item (“restitution”) could not be categorized due to competing interpretations that resisted consensus. Two expert ethicists found the taxonomy to be comprehensive and recommended changes to the first 4 domain names to increase descriptive clarity and movement of 2 items from 1 domain to another. The final taxonomy comprised of 4 domains of facilitating factors () and 4 domains of impeding factors ().
Factors that facilitate physician disclosure of medical errors
Factors that impede physician disclosure of medical errors
Responsibility to patient focuses on the physician's fundamental respect for the patient as a person, through open communication and ongoing care. The core of this domain was suggested by a student who said “It boils down to just how you view other people. Do you view them as worthy of knowing?” An intern spoke similarly “That's really the focus of what we're doing here: patient care … it comes down to what's happened with this particular patient.”
Responsibility to self focuses on personal and professional values that derive from the physician's character, commitments, and desire for integrity. An intern stressed the importance of “being accountable for our errors and not being a weasel or arrogant or denying that we ever make errors.” A faculty physician acknowledged the need for courage “If you don't have the guts to say, “I screwed up” to a patient, you're in the wrong business.” A resident commented that “in order to receive forgiveness you have to admit to your wrong,” and another expressed the need to “make amends” with a harmed patient in order to move forward. Participants saw the need to accept fallibility and to be willing to be vulnerable. Such willingness, a student observed, will drive “a lot of your desire to report to anybody because you are going to be vulnerable when you say “I made a mistake.” Some participants articulated spiritual or religious motivations, such as the student who said: “I should be motivated by love and also I'm ultimately responsible to God for my actions … whether I'm deceptive with patients or whether I tell them the truth about what's going on.”
Responsibility to profession focuses on the physician's desire to improve the medical profession through sharing lessons learned, modeling disclosure skills, fostering a culture of disclosure, and providing support to colleagues who are involved with errors. A faculty physician described the need for role modeling in discussing and disclosing errors “If I as a faculty member can't express my own fallibility … how can the learner learn?” A resident said “When people have come out and told the patient, have taken responsibility—it's usually based on just a need to do the right thing and the need to be a good role model for those who are training under you.” A student spoke of the need for support “There's a catharsis in being able to say to your colleagues, ‘This is what happened’ and then to be able to hear, ‘I made that same mistake, I've been there, I know how you feel, this is what I did to correct it.’”
Responsibility to community focuses on the physician's desire to improve the quality of care for all patients, to enhance society's trust in physicians and the medical profession, and to educate the community about medicine's complexities and imperfections. One student's remark about reporting errors to improve systems of care was representative “Your one case may not seem to make a difference, but if there are trends at a certain hospital or in a certain area of the country, this is how we get demographic information, this is how we improve our care … and there are public health implications of reporting, if you feel there's a duty to improve for the greater good.”
Attitudinal barriers focus on a range of attitudes that may hamper disclosure. Perfectionism was a persistent theme in the focus groups. An intern explained “Even though I know it's not logical for me to think that doctors aren't going to make errors, I hold doctors to that standard, that we're going to be perfect and we're not going to make errors.” Participants also drew a connection between silence about errors and the competitive nature of medical training.
A student observed:
[As a student] you're competing within your class, competing with yourself, and trying to reach the academic goals that you want. As a resident you're competing to attain that certain fellowship position. You don't get points for making mistakes; in fact, you get points taken away. It's like the SATs. So admitting to mistakes doesn't exactly help your career…. It's the inherent competition within our career that kind of fuels a lot of people who want to put their mistakes under the carpet and just show off their achievements and try to put themselves in the best light possible.
Uncertainties focus on doubts about how to disclose errors, which ones to disclose, what constitutes an error, and disagreements between clinicians about whether an error occurred. A student expressed the struggle to discern the difference between a complication and an error:
There's a risk that you're going to cause a pneumothorax when you do a thoracentesis…. But if I am the one that causes that pneumothorax, is it because I was an idiot? Do I say, “You know, I collapsed your lung, I'm really sorry, I made a mistake” or do I just present it as, “It's one of the risks, you signed informed consent.” I really struggle with how you even define some of the errors.
A faculty physician described the difficulty of determining whether an error is significant enough to disclose “At what level does the error become big enough that now something needs to be done about it?” A resident spoke of conflicting views between supervisors and trainees “Medicine is vague enough sometimes that even though I feel [an error occurred], there's no higher power for me to appeal to if the higher power within my group feels that the right thing was done.”
Helplessness focuses on dissatisfactions with the process, context, follow-up, and outcome of error disclosure, as well as not having the power necessary to improve the system of care. A student complained of not knowing what will “happen with what you [report], the path [the information] is going to take, and who's going to be reading it,” suggesting that reported errors may be “going down a black hole” and may result in “retributions that come back to you.” An intern described discussion forums at a prior institution that were demoralizing:
Morbidity & mortality conferences were just brutal. We wouldn't go, we wanted nothing to do with them. The students would actually sometimes go to see the residents they didn't like just get toasted.
Participants were disappointed by lack of feedback after reporting errors. A resident complained “So far as I know, [the report] goes to some dead space out there and it's vaporized….”
Fears and anxieties focus on a range of potential negative consequences of error disclosure. Participants spoke about profound personal struggles related to their identities as healers. An intern said “[Patients are] coming in here, they're sort of putting their life in my hands, and they're trusting me and I've violated this trust.” Another intern remarked “Disclosing to the patient makes you admit to yourself that—what's that first tenet of our oath, “First do no harm?”—well, we did harm.” A faculty physician opined “I'm delivering bad news to the patient about something, but I'm also delivering bad news about myself because I have been the cause of that bad news.” A resident articulated the difficulty of apologizing for negligence:
Saying “I'm sorry” has got to be some of the toughest words in any language and we, as physicians, take a lot of pride in the fact that we're pretty smart and capable people…. To make a mistake that acknowledges my own [fallibility] is in a way saying that I'm not as good as I could be.… If it is something like you forgot to deflate the catheter that ends in the patient dying, that's a pretty, pretty serious outcome. Like, if you're flying a jet and you drop the bomb on the wrong person. Those things live in your memory forever….
A resident feared the loss of reputation: “There's the fear of other people saying, ‘Boy, he dropped the ball, he screwed up, he's not really a good doctor, he really doesn't know what he's doing.’ You don't want people pointing fingers at you. It's enough to be pointing fingers at yourself, but you don't want other people to say ‘he doesn't belong among us.’”