Background

There is increasing attention to safety culture, teamwork and communication practices in medicine [1, 2]. This focus has lead to large initiatives focused on effective communication skills [3]. As a result of such initiatives, the contribution of dysfunctional behavior in the perioperative setting is increasingly being scrutinized as a potential source of adverse outcomes and healthcare worker job dissatisfaction [4, 5]. Among these dysfunctional behaviors may be intentional physician-to-physician deception.

The phenomenon of intentional misrepresentation of the truth among physicians is little studied. Physicians occasionally misrepresent the truth to insurance companies to gain coverage for their patients [6–8], and there is a modest literature on the evolution of physician attitudes towards misrepresentation to patients with poor prognoses [9]. But almost nothing is known about whether physicians ever intentionally deceive each other in the course of patient care.

Anecdotally, anesthesiologists and surgeons may misrepresent the truth to each other in the course of perioperative care, such as by misreporting blood losses, the urgency of a case, or the volume of fluids given. Research has shown that surgeons are often inaccurate in their prediction of the duration of cases [10–12], but it is unknown whether this inaccuracy is intentional or not. Similarly, documentation in the anesthetic record can vary considerably when record keeping is automated as compared to manual documentation by an anesthesia provider [13, 14], but the reasons for these discrepancies and whether they are intentional or not are also unknown.

Methods

To test the hypothesis that some degree of intentional misrepresentation might occur among physicians in the perioperative setting we conducted a pilot survey of anesthesiologists and surgeons nationally.

After obtaining approval from the Partners Human Research CommitteeFootnote 1 we mailed a self-administered survey to 1130 anesthesiologists and 1130 surgeons drawn from the American Medical Association (AMA) Masterfile, which includes all physicians practicing in the United States. After removing ineligible recipients, the sample of anesthesiologists was 1084 and for surgeons it was 1090. As an incentive, respondents were entered into a raffle to win an iPod. Six months after the initial mailing a follow-up post-card was mailed requesting that the original survey be returned or directing respondents to a secure website to enter responses.

To assess the frequency of misreporting information among anesthesiologists and surgeons, each survey included a set of potential facts (such as the amount of blood loss, the urgency of the case, or minor complications not apparent to the other physician) that could be misreported.  Each list was tailored to anesthesia or surgical practice (complete lists are available from the authors).  Survey recipients were asked to report how often, if ever, they had “knowingly mis-reported (or chosen not to report) the following intra-operative information to any member of the surgical/anesthesia team?” Responses on a 6-point likert scale ranged from 1) daily, 2) once or twice a week, 3) once or twice a month, 4) once or twice a year, 5) less than once a year, 6) and never.

Next, a list of potential justifications for misreporting was presented and respondents were asked “How important are the following possible justifications for deciding to mis-report (or not report) information to the surgical/anesthesia team.” Responses on a 5-point likert scale ranged from 1) very unimportant, 2) somewhat unimportant, 3) somewhat important, 4) very important, and 5) not applicable or I never mis-report. Respondents were also asked “in the last year, how often have you been concerned that your surgical/anesthesia colleagues have misrepresented the truth to you?” with 6 response options ranging from”daily” to “never”.

Respondents were informed in the cover letter that their responses would remain strictly confidential and that the Partners Human Research Committee approved the study protocol. Because of the sensitive nature of the survey questions, each subject was randomly assigned a unique identifier. Mailings were prepared in Chicago but return-mailed to the Boston-based investigators. The key linking unique identifiers to subjects was kept in Chicago, thereby preventing the Boston based team from linking responses to subjects.

Survey results were tabulated and demographics of responders and non-responders were compared by cross-indexing the unique identifier to demographic data in the AMA Physician Masterfile.  Descriptive statistics were presented for demographics and for survey results. Continuous variables were summarized in means and 95% confidence intervals while categorical variables were reported as frequencies and proportions. Demographics and survey responses of responders vs. non-responders, and anesthesiologists vs. surgeons were compared using two-sample Student’s t-test or chi-square test. All tests were two sided with a significance level of 0.05.

Results and discussion

Given the highly sensitive nature of the questions we asked, it is not surprising that this pilot survey generated a relatively low response rate (11 %, 252/2260); 128/1084 by anesthesiologists and 124/1090 by surgeons. By the same token, we suspect our exploratory findings on the frequency of anesthesiologist-surgeon deception may be conservative, given the socially undesirable nature of admitting to intentional deception of colleagues. More importantly, despite being far from definitive, our results raise provocative questions about truth-telling practices among physicians in the perioperative setting.

There were no significant differences between respondents and non-respondents for age or years since graduation, although more male surgeons did not respond to the survey (P = 0.009) (Table 1). Amongst respondents, there were no significant differences between anesthesiologists and surgeons for age, sex, years in practice, hospital size, English as a first language, or self reported religious status (Table 2).

Table 1 Respondents vs non-respondents demographics from AMA Masterfile
Full size table
Table 2 Respondent demographics–self reported
Full size table

Among the 252 respondents to our survey, up to 27 % of anesthesiologists and 7 % of surgeons admitted to misrepresenting the truth to each other at least once a month regarding a number of factors that can influence perioperative care (Tables 3 and 4). The justifications reported typically involved concerns around the adverse consequences of being truthful; in particular, paternalistic concerns about disclosing truthful information to a colleague who might not understand or who would then demand unreasonable steps (Table 5). Such concerns were the drivers of intentional misrepresentation much more often than worries about personal blame, chastisement or legal consequences.

Table 3 Most commonly misreported events by Anesthesiologists (a)
Full size table
Table 4 Most commonly misreported events by Surgeons (a)
Full size table
Table 5 Justifications considered “Very Important” when misrepresentation was acknowledged (*)
Full size table

When clinical decision making is based on erroneous information, a cascading series of events can occur, leading to harm to patients and adverse consequences for physicians. This is indicated by the fact that 7 % of responding anesthesiologists and 2 % of surgeons said that a patient had been harmed by their misrepresentation of the truth to a colleague (data not shown in tables). Alternatively adverse consequences may occur to the physician who is misrepresenting or receives misrepresented information. In this survey 1 % of anesthesiologists and 4 % of surgeons reported suffering adverse consequences themselves due to an episode of misrepresentation (data not shown in tables). The nature of harms to patients and consequences to physicians cannot be identified from the present survey and requires further investigation.

The greatest impact of dishonest communication between colleagues in the perioperative setting may be that it damages teamwork among professionals engaged in high-stakes activities.  Candid and open communication and shared decision-making are foundational to optimizing teamwork. Studies have shown correlations between such team skills and actual or potential improved outcomes for patients [15, 16].

What can be learned from these pilot data? First, the motivations for intentional misrepresentation of the truth to colleagues likely represents a spectrum, ranging from gaming of the perioperative care system to optimize the ability to accomplish what physicians feel to be in the best interests of their patient to willful misrepresentation and deception (or lying) with a view to obfuscating facts to obtain personal benefits. Our results suggest the former motivation predominates, though future research should aim to better address the issue of motivation.

Second, the apparent asymmetry of our survey results requires further investigation. Anesthesiologists were both more likely to report misrepresenting facts to surgeons and more likely to be concerned about surgeons misrepresenting facts to them (Table 6). Perhaps of greater concern, more anesthesiologists (36 %) than surgeons (8 %) reported that they had seen their teachers engage in misrepresentation of the truth during their training. It is possible that anesthesiologists are more willing to be honest about misrepresenting issues to surgeons on a survey, or it could be that intentional misrepresentation really is more common within the anesthesia community. This question deserves focused attention, given its potential import for the profession, especially if in some cases deliberate misrepresentation is being taught, explicitly or implicitly, to trainees, perhaps as an anticipatory strategy for controlling or gaming the perioperative environment.

Table 6 How often have you been concerned your counterpart is misrepresenting information (in the past year)?
Full size table

Third, we hope these preliminary results will spark more formal study of these issues. This poll was intended to generate conversation and hypotheses. After all, if there is a possibility that misrepresentation is becoming normative, on the assumption that one’s anesthesiology or surgeon counterpart is also engaging in a similar practice, this deserves direct and focused attention. In this regard, our results are especially troubling if our estimates for prevalence are low. In addition, while some might argue that minor misrepresentations comprise a normal social practice, our results also suggest that a potential for harm to patients and physicians exists and that further discussion, exploration and investigation is warranted.

Fourth, with regard to future research, this pilot work suggests that any future survey research on this topic will need to include robust efforts to obtain high response rates; and it is possible that the very sensitive nature of this issue will always preclude obtaining high levels of response. Other approaches to this area of enquiry might yield valuable insights, such as using qualitative data or the examination of automated data collected during surgery.

Finally, while our study focused on misrepresentations between surgeons and anesthesiologists, it is reasonable to be concerned that misrepresentation of medical practice might occur between other health professionals. For example, there are case reports of nursing staff intentionally hastening a patient’s death without explicit authority to do so and in the absence of formal documentation [17]. Residents have also reportedly engaged in sham resuscitations when they believed orders for resuscitation were not warranted [18]. These examples suggest that misrepresentation is probably more widespread in medicine than just between physician colleagues in the perioperative setting.