Which medical error to disclose to patients and by whom?

Background

Disclosure of near miss medical error (ME) and who should disclose ME to patients continue to be controversial. Further, available recommendations on disclosure of ME have emerged largely in Western culture; their suitability to Islamic/Arabic culture is not known.

Methods

We surveyed 902 individuals attending the outpatient's clinics of a tertiary care hospital in Saudi Arabia. Personal preference and perceptions of norm and current practice regarding which ME to be disclosed (5 options: don't disclose; disclose if associated with major, moderate, or minor harm; disclose near miss) and by whom (6 options: any employee, any physician, at-fault-physician, manager of at-fault-physician, medical director, or chief executive director) were explored.

Results

Mean (SD) age of respondents was 33.9 (10) year, 47% were males, 90% Saudis, 37% patients, 49% employed, and 61% with college or higher education. The percentage (95% confidence interval) of respondents who preferred to be informed of harmful ME, of near miss ME, or by at-fault physician were 60.0% (56.8 to 63.2), 35.5% (32.4 to 38.6), and 59.7% (56.5 to 63.0), respectively. Respectively, 68.2% (65.2 to 71.2) and 17.3% (14.7 to 19.8) believed that as currently practiced, harmful ME and near miss ME are disclosed, and 34.0% (30.7 to 37.4) that ME are disclosed by at-fault-physician. Distributions of perception of norm and preference were similar but significantly different from the distribution of perception of current practice (P < 0.001). In a forward stepwise regression analysis, older age, female gender, and being healthy predicted preference of disclosure of near miss ME, while younger age and male gender predicted preference of no-disclosure of ME. Female gender also predicted preferring disclosure by the at-fault-physician.

Conclusions

We conclude that: 1) there is a considerable diversity in preferences and perceptions of norm and current practice among respondents regarding which ME to be disclosed and by whom, 2) Distributions of preference and perception of norm were similar but significantly different from the distribution of perception of current practice, 3) most respondents preferred to be informed of ME and by at-fault physician, and 4) one third of respondents preferred to be informed of near-miss ME, with a higher percentage among females, older, and healthy individuals.

Background

In healthcare, it is not uncommon that patients are exposed to risks of harm. Some risks are predictable, at least at statistical level, and an informed consent is obtained. Other risks, such as those occurring because of medical errors (ME) are in a sense unpredictable and an informed consent can not be obtained. An ME is defined as an act or omission that would have been judged wrong by knowledgeable peers at the time it occurred [1]. Some ME may not materialize into harm; a near miss is an event that under slightly different circumstances could have been an accident, either because the error was detected and corrected in time or because the patient was just lucky [2]. When an ME occurs, two actions should be considered: reporting it to the healthcare system (and through it to potential future patients) and disclosing it to the patient involved. Reporting ME is paramount for quality and safety improvement and should include near miss ME [1,3]; compared to reporting harmful ME, reporting near miss ME offers greater frequency and fewer barriers to data collection [3].
Full disclosure of ME includes an explicit statement that an error (rather than just a "complication") occurred, basic description of the error, who committed the error, why it did happen, how recurrences will be prevented, and an apology [4,5]. Non disclosure of harmful ME is considered a violation of ethical principles from both deontological and consequentialist perspectives [6]. A policy of open disclosure standard that demands disclosure of critical events by the provider or the institution [7] was promulgated in 2001 by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and is now reflected in similar initiatives in the UK, Canada, and Australia [8-12]. These regulations and theoretical ethical considerations are consistent with the results of empirical ethics studies in Western cultures showing that patients overwhelmingly desire full disclosure of harmful ME [5,13-17] and that full disclosure is likely to have a positive or no effect on how patients respond to ME [18].
Disclosure of near miss ME to patients is a matter of controversy [19,20] and an issue on which current guidelines are silent. Disclosure is recommended by some [21-24] but not all authors in the field [2,25]. The American Society for Reproductive Medicine states that if there is clearly no adverse effect of a ME, disclosure may not be obligatory if it may unnecessarily increase patient's stress [26]. In the few empirical studies that have specifically addressed near miss ME, 88% to 92% of patients desired disclosure [15,27]. In contrast, most physicians opposed near miss ME disclosure [5]. Determining who should disclose ME is another matter of controversy. According to Liang model [21] and the policy described by Kraman and colleagues [28], risk management committee should be responsible for disclosure. Others believe that the responsibility for disclosure belongs to the physician [10,11,22,29]. JCAHO standard requires that a responsible licensed independent practitioner or his/her designee explains the outcome [7].
Current disclosure literature contains important but unanswered questions such as how patient' preferences for disclosure vary along cultural and other dimensions [10,11]. An individual's ethical decision-making is based on his/her values and beliefs. Although major ethical values are rather universal, ethical values are subject to individual interpretation and people naturally differ in their values' hierarchy and in their beliefs. Autonomy is placed at the top of the "moral mountain" and is given a "place of honor" in Western but not other ethics [30]. Further, there are several meanings of autonomy along a spectrum from a negative or anti paternalistic model to a positive mandating model [31]. Furthermore, it has been argued that respect should be for the person rather than purely for autonomy [31,32]; trust in providers, treatment with respect, and dignity were more closely associated with patients overall evaluation of their hospitals than adequate involvement in decision-making [33]. To our knowledge, there is no study on patients' views on disclosure of ME that has been conducted in Arabic/Islamic countries or that compared preference (a statement about the person who has the preference) and perception of norm (a statement about the thing which is being judged).
The aim of this study was to obtain empirical evidence on public views on disclosure of ME in the outpatient's setting at a tertiary care hospital in Saudi Arabia. We examined preference, perception of norm, and perception of current practice on two topics, which ME to be disclosed and who to disclose ME.

Methods

This cross sectional survey was conducted in accordance with the ethical principles contained in the Declaration of Helsinki and after approval of the Research Ethics Committee of the King Faisal Specialist Hospital and Research Center (KFSH&RC) in the period from November 2007 to March 2009. All respondents gave verbal consent.
Two sets of three questionnaires addressing personal preference, perception of norm (what is appropriate in general/should be done), and perception of current practice at KFSH&RC regarding which medical error is disclosed to patients (set one) and by whom (set two) were developed by the authors in Arabic language based on literature review. After initial development, the questionnaires were presented for comments to 6 physicians and revised accordingly (minor changes in language usage to have consistency throughout the questionnaires). Face validity was assessed by interviewing 10 respondents after completing the questionnaires. The final version was pilot tested on 10 other individuals for clarity and stability (2-3 days) and found suitable. An English translation (accuracy confirmed by back translation) of the two questionnaires on personal preference is shown in Table 1. Similar statements with appropriate modifications were used for the questionnaires on perceptions of norm and current practice. For example, we used the phrase "I prefer" combined with "to be" and "my/me" to indicate personal preference and "I think" combined with "should be" and "patient/his" to indicate perception of norm. For perception of current practice questionnaires, "I prefer" was omitted and "is" was combined with "patient/his". The statements in each questionnaire were arranged from least to most demanding. Before completing the questionnaires, participants were given the following introductory information on medical errors: "Clinical practice, just like any other beneficial practice, could hardly be completely free from harm. Such harm can be divided into two types: 1) harm that can be predicted and thus can be avoided, e.g. anaphylactic shock caused by penicillin administration to a person known to have penicillin allergy, and 2) harm that can't be predicted/avoided, e.g. inflammation of the bowel after some antibiotics treatment. The first type is called medical error. A medical error is defined as failure to complete a planned medical action as intended, or the use of a wrong plan to achieve an aim. Medical errors may or may not cause harm, for example, penicillin could be wrongly prescribed by a physician but not given to the patient because the error is discovered and corrected in time by a pharmacist or nurse. Physicians may not disclose medical errors to patients for a variety of reasons that are related to patient's interests or physician's interests or because they may think it is useless to do so. Similarly, some patients wish to be informed about medical errors and some do not. Disclosing medical errors to patients is an issue separate from disclosing them to hospital administration. In this study we are interested in disclosing medical errors to patients. We would like to know your views on: 1) which medical error to be disclosed to patients, and 2) who to disclose medical errors to patients. There are three groups of statements for each of these two questions. The first is on what you personally prefer, the second is on what you think is best in general, and the last is on what you think reflect the current practice at KFSH&RC". For each questionnaire, participants were asked to choose the most representative statement. The six questionnaires are available in Additional file 1.