Pediatricians' Reports of Their Education in Ethics

Objective  To study pediatricians' assessments of the quality of their ethics education, the impact of various learning methods, and their confidence in confronting ethical dilemmas arising in pediatric practice.

Design  Cross-sectional survey.

Participants  Two hundred fifty physicians who completed pediatric or medicine/pediatric residency programs in 2004 were randomly selected from the American Medical Association Physician Masterfile. Evaluable responses were received from 150 of 215 eligible pediatricians (70%).

Results  Of 150 respondents, 44.7% rated their ethics education during residency as fair or poor. More than 80% reported that informal discussions with fellow residents and attending physicians had a moderate or major effect on their ethics education, whereas 53.3% reported that formal teaching conferences had a moderate or major impact. Most respondents (>60%) reported confidence in addressing 4 of 23 ethical challenges, a moderate proportion (40%-60%) reported confidence in addressing 8 of 23 ethical challenges, and fewer (<40%) reported confidence in addressing 11 of the ethical challenges. Areas associated with low confidence included ethics in end-of-life care and research ethics.

Conclusions  Efforts are needed to augment formal and informal ethics teaching during residency. Additional studies at both the individual physician and residency program levels are needed to improve the ethics education that pediatricians-in-training receive.

During residency, trainees in pediatrics regularly face an array of ethical challenges relevant to general and subspecialty pediatric practice. The ability to approach ethical dilemmas systematically is a skill that must be learned and cultivated over time.¹ Residents in pediatrics have 3 years in which to acquire this proficiency, although the learning process inevitably continues beyond residency. Indeed, residency training may provide an irreplaceable foundation for ethical care on which pediatricians can build throughout their careers. Without deliberate efforts to teach residents how to approach ethical questions, many young pediatricians may be unable independently to develop this skill.²

Recent guidelines released by the Accreditation Council for Graduate Medical Education (ACGME), which accredits residency training programs, support a rigorous examination of ethics education.³ One of the 6 ACGME core competencies, “professionalism,” relates to the way physicians demonstrate respect, compassion, integrity, and a commitment to ethical principles. The ACGME requires residency programs to “define the specific knowledge, skills, and attitudes required and provide educational experiences as needed in order for their residents to demonstrate [professionalism].”³ Despite this mandate, the ACGME offers little guidance as to how to achieve the goals of teaching professionalism and ethics. In addition, studies have not yet been conducted to explore whether ethics education during residency adequately meets pediatricians' needs. To address this deficit, we conducted a cross-sectional survey of pediatricians who had finished residency training within the preceding year. Our objective was to explore the confidence these newly trained pediatricians report as they confront an array of ethically challenging clinical situations intrinsic to pediatric practice. We also aimed to assess respondents' ratings of the ethics education that they received and to determine which learning methods most influenced their education in ethics.

We identified physicians listed in the American Medical Association Physician Masterfile (1) whose primary or secondary specialty was pediatrics, (2) who had completed a US-based pediatric or combined medicine/pediatric residency training program in 2004, and (3) who had a US mailing address. This list included 2742 physicians. After limiting the list to the 1876 physicians who graduated from medical school from 1999 through 2001, we randomly selected 250 pediatricians for inclusion in our survey.

The survey instrument consisted of 16 questions grouped into multiple domains. Questions used either multiple choice or ordinal scale formats. First, respondents were asked to rate the impact that various learning methods (eg, formal teaching conferences, discussions with other residents) had on their education in ethics. Response options included “no impact at all,” “minimal impact,” “moderate impact,” and “major impact.” Next, respondents were asked to rate the ethics education they received in residency, as well as the extent of support for education in ethics from both residency leadership and supervising attending physicians. Response options included “excellent,” “very good,” “good,” “fair,” and “poor.” Finally, respondents were asked, “In your practice as a pediatrician, how confident are you in your ability to confront the ethical challenges that may arise in the following situations?” This question was followed by descriptions of 23 ethically challenging clinical situations that might arise during pediatric clinical care. Responses were on a 5-point ordinal scale (“not at all confident,” “a little confident,” “moderately confident,” “confident,” and “extremely confident”). The ethical challenges included in the survey were selected based on discussion among the authors and input from the staff of the Office of Ethics at Children's Hospital Boston. The survey also collected demographic information about the respondent's primary professional role, current practice setting, and time spent in clinical duties.

We pilot tested the survey with 6 senior (postgraduate year 3) pediatric residents in April 2005. After completing the survey, pilot participants were interviewed by an investigator (J.C.K.) about the clarity of questions, the ease of responding, and the relevance of the 23 ethical dilemmas to their residency experience.

We conducted the first study mailing in May 2005. The study packet included a cover letter, survey instrument, self-addressed stamped return envelope, $5 Starbuck's gift card, and reply postcard. The survey contained no identifiers; the postcard contained a coded identifier. Respondents returned the postcard separately from the questionnaire, to notify us that they had completed the survey. Alternately, they could check an opt-out box on the postcard indicating that they did not wish to take part.

Three to 4 weeks after the initial mailing, we sent a second packet to nonrespondents. This packet was sent via Federal Express to permit delivery tracking and to facilitate identification of inaccurate addresses. Packets addressed to a post office box (ie, that were ineligible for Federal Express shipping) were sent instead with a “return receipt” via US mail. Of 250 pediatricians to whom questionnaires were mailed, addresses were incorrect for 31. An additional 4 reported not having participated in pediatric or medicine/pediatric training programs and were coded as ineligible. Completed questionnaires were received from 150 of the remaining 215 pediatricians (69.8%).

Data from the completed questionnaires were manually entered into a Microsoft Access database (Microsoft Corp, Redmond, Washington). Before analysis, data were verified by comparing all electronic records with the paper originals.

The study was approved by the Children's Hospital Boston Committee on Clinical Investigation, which considered return of the questionnaire as evidence of informed consent.

Analyses were primarily descriptive. The main study end point was physicians' confidence in confronting the ethical challenges that arise in their practices. For clarity of presentation, we collapsed the 5 response options into 3 categories (“not at all confident” or “a little confident”; “moderately confident”; and “confident” or “extremely confident”). We also combined responses to the confidence questions to obtain an average confidence score, after ensuring that the internal reliability of these questions was sufficiently high (Cronbach α = 0.89). This score could range from 1 (least confident) to 5 (most confident). We then conducted exploratory bivariate analyses to identify candidate predictor variables associated with greater average confidence in confronting the ethical challenges that respondents face.

Sample size calculations were based on the width of the confidence intervals around the estimates of proportions falling into each response category. With 100 responses, 95% confidence intervals around estimates of proportions would be no wider than ±10%. We estimated that we would need to mail questionnaires to 250 pediatricians to obtain 100 responses, based on an anticipated 50% response rate, a 10% ineligibility rate, and a 10% undeliverable rate.

Analyses were conducted using Stata 8 statistical software (Stata Corp, College Station, Texas).

Pediatricians' mean age was 31.8 years, and 63.1% were female (Table 1). Of 150 respondents, 90.0% had graduated from a US medical school, and 82.4% had completed pediatric residency training programs. The sample reported spending more than 75% of their time on patient care in 82.0% of cases. When asked about their current practice setting, 41.6% reported working in a private office, while 39.6% practiced in a teaching hospital. The pediatricians sampled reported primary care practice as their current professional role in 58.4% of cases.

Respondents were asked to rate the impact of various learning methods on their education in ethics during residency (Table 2). Discussions with fellow residents and with supervising attending physicians had the largest effect on respondents' ethics education; 120 respondents (80.5%) and 134 respondents (89.3%), respectively, reported a moderate or major impact. Formal teaching conferences, involvement in ethics consultations, and discussions with hospital ethicists were rated as having moderate or greater impact by 41.6% to 53.3% of respondents. Readings in medical journals, ethics texts, and the lay press were rated as having the least impact.

Respondents were asked to rate the quality of the ethics education they received during residency training, the degree of support for ethics education offered by the leadership of their residency program, and the level of attention paid by attending physicians to the ethical dimensions of patient care (Table 3). Sixty-seven respondents (44.7%) rated the quality of the ethics education they received during residency as “fair” or “poor,” and 72 (48.0%) rated the support that their residency leadership offered for ethics education as “fair” or “poor.” The level of attention paid by supervising physicians to the ethical dimensions of patient care was rated “fair” or “poor” by 34 respondents (22.7%).

We created a “quality of ethics education” composite score by averaging the responses to the 3 questions listed in Table 3 (Cronbach α = 0.87). In bivariate analyses, respondents who reported that formal teaching conferences had a large effect on their education in ethics rated the quality of their ethics education higher than did other respondents (Spearman rank correlation ρ = 0.45, P<.001). Similarly, respondents who reported that discussions with other residents (ρ = 0.24, P = .004), discussions with supervising attending physicians (ρ = 0.39, P<.001), involvement in ethics consultations (ρ = 0.22, P = .009), and discussions with hospital ethicists (ρ = 0.26, P = .002) had a large effect on their education in ethics rated the quality of their ethics education more favorably than did other respondents. The remaining learning methods were not significantly associated with residents' ratings of the quality of their ethics education.

Finally, respondents were asked about their confidence in their abilities to confront 23 ethical challenges arising in pediatric clinical practice (Table 4). More than 60% of respondents rated themselves as “confident” or “extremely confident” in their abilities to confront 4 ethical challenges. These included identifying the proper decision maker for a pediatric patient, delivering bad news, deciding whether to respect an adolescent patient's request to withhold information from his or her parents, and discussing newborn screening with parents.

A moderate proportion of respondents (40%-60%) rated themselves as “confident” or “extremely confident” in their abilities to confront 8 ethical challenges. These included using opioids near the end of life, discussing do-not-resuscitate orders with parents of a terminally ill child, obtaining informed consent from adolescent patients without parental involvement, deciding whether an adolescent qualifies as an emancipated minor, deciding about the appropriateness of genetic testing, weighing the cost of therapy in deciding treatment recommendations, and responding to offers of gifts or receiving medical information from pharmaceutical representatives.

Fewer respondents (<40%) rated themselves as “confident” or “extremely confident” in their abilities to confront 11 ethical challenges. These included deciding about withdrawing assisted ventilation, deciding about withdrawing artificial nutrition and hydration, requesting permission for organ donation, requesting permission for autopsy, deciding whether to respect an adolescent's refusal of recommended care, discussing whether to attempt resuscitation for a premature infant near the margin of viability, making decisions about life-sustaining therapies for infants with severe neurocognitive disabilities, obtaining parents' permission to enroll a child in a clinical trial, obtaining assent from an average 10-year-old to enroll in a clinical trial, performing a blood draw on a young child for research purposes, and weighing authors' financial relationships with study sponsors when reading reports of clinical trials.

The mean confidence score was 3.4 (SD, 0.8; range, 1.2-5.0). Among the demographic variables, sex was significantly associated with confidence; males had higher mean confidence scores than females (3.6 vs 3.3; P = .02). Also, pediatricians who reported working in academic settings had significantly higher mean confidence scores than other respondents (3.7 vs 3.3; P = .009). Higher confidence was associated with higher ratings of the overall quality of ethics education (ρ = 0.38, P<.001), with greater support for education in ethics from residency leadership (ρ = 0.27, P = .008) and greater attention paid to ethics by supervising attending physicians (ρ = 0.23, P = .005).

We surveyed recent pediatric and medicine/pediatric residency graduates to explore their perceptions of their ethics education and to evaluate their confidence in confronting ethically challenging situations in pediatric practice. Three major findings emerge. First, many respondents rated the quality of the ethics education they received in residency as either “fair” or “poor.” Second, although more than half of respondents reported that formal teaching conferences had a large effect on their education in ethics, informal discussions with fellow residents or supervising attending physicians were rated as having greater impact. Third, respondents reported limited confidence in confronting several ethically challenging situations, especially with respect to issues that arise in pediatric end-of-life care and research ethics.

The results of this survey also suggest several provocative associations that merit further investigation. First, respondents who said that formal teaching conferences in ethics had a moderate or major effect on their education also reported that the quality of the ethics education they received was high. Second, respondents who gave positive ratings to the quality of their ethics education, and to the support that residency leadership and supervising attending physicians provided for teaching in ethics, reported greater confidence in their abilities to address the ethical challenges arising in pediatric practice.

Previous work conducted in diverse specialties indicates that residents, although receptive to education in ethics, may be dissatisfied with the ethics training they receive.^4,5 In addition, on entering training, residents typically lack knowledge and confidence in ethics.^6,7 Our data suggest that deficits in confidence persist even on completion of residency, thus reinforcing the concern that education in ethics during residency does not meet learners' needs. In particular, despite the fact that end-of-life care has been identified by experts in multiple disciplines as a major learning objective for resident ethics education,⁸ the uncertainty respondents expressed about addressing ethical challenges in end-of-life care is consistent with the findings of others.⁶ For example, Solomon et al⁹ described limitations in knowledge regarding ethical guidelines for pediatric end-of-life decision making. Our data suggest the need to develop new strategies to achieve this objective.

Many authors, including Downing et al⁴ and Goold and Stern,⁸ have attempted to address this need and to guide curricula by elucidating the ethics content that would improve teaching in this domain. Also, several ethicists and educators have described efforts to implement novel curricula to address concerns about ethics education during residency.^10-12

Our study has several limitations. First, the survey relies on pediatricians' self-reports about their confidence and their educational experiences. It is therefore subject to recall, social desirability, and other biases associated with self-report. Second, the use of confidence, a subjective domain that is difficult to validate externally, as a primary outcome poses some challenges. Notwithstanding this concern, confidence remains a promising domain on which to focus because it has proven both predictive of behavior and responsive to educational interventions.^11,13 Third, there is room for debate about whether all pediatricians should be confident in approaching the entire array of ethical dilemmas inherent in pediatric practice. It is possible that the lack of confidence in areas such as research ethics and end-of-life care represents either a lack of interest on the part of residents who know their careers will not involve these topics, limited experience with the topic during residency, loss of confidence among residents who no longer require such skills in their work, or the intrinsic difficulty of these subjects. The study raises the question of whether every pediatrician needs to be confident in addressing issues related to areas they are unlikely to face in their work.

Fourth, our data on the impact of various learning methods do not clarify whether a particular learning method had a lesser impact because it was not available to the respondent, or because it was available but was ineffective in achieving learning goals. Finally, owing to the anonymity of our respondents, we cannot assess variation among the particular programs our participants attended.

Nevertheless, our research findings have several important implications for residency program directors and others who develop ethics education for pediatric residents. First, they highlight the difficulty in fully evaluating whether residency program graduates are adequately trained in ethics so long as the objectives and content for this teaching remain undefined. Thus, there is a need for educators to clarify the core content of an ethics curriculum. The domains identified in our survey may prove useful as a starting point for developing such curricular objectives.

Second, our study demonstrates a need to augment the formal teaching in ethics that residents receive. The positive association between the impact of formal teaching conferences and reported quality of residents' ethics education suggests the hypothesis that formal ethics teaching, including preplanned learning opportunities such as didactics, interactive case discussions, or small group learning, is one proper avenue to meeting pediatricians' needs. A constructive response to these findings would be the development of a core curriculum for ethics and professionalism that residency programs nationwide could adopt for pediatric residents. Leadership from the American Academy of Pediatrics Committee on Bioethics, the Association of Pediatric Program Directors, and the Resident Review Committee could convene to define such a standardized curriculum. Such a curriculum should focus, among other topics, on ethical dilemmas related to pediatric research and to end-of-life care.

At the same time, the substantial impact that respondents ascribe to informal learning methods, such as discussions with other residents and with supervising attending physicians, highlights the need to understand and enhance these less structured aspects of the curriculum. Since supervising attending physicians clearly play a powerful role in ethics education for residents, providing opportunities for faculty development and learning in ethics will be important as well.

Finally, our study highlights the need to determine aspects of individual residency training programs that influence the ethics education received by trainees. Enhanced understanding of programmatic variables associated with better or worse outcomes in the domain of ethics is an essential next step. Such an understanding will form the foundation for interventions, at the residency program level, to increase the effectiveness of ethics education provided to residents.

Correspondence: Steven Joffe, MD, MPH, Department of Pediatric Oncology, Dana-Farber Cancer Institute, 44 Binney St, Boston, MA 02115 (steven_joffe@dfci.harvard.edu).

Accepted for Publication: September 21, 2007.

Author Contributions: Drs Kesselheim and Joffe had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Kesselheim, Johnson, and Joffe. Acquisition of data: Kesselheim and Joffe. Analysis and interpretation of data: Kesselheim, Johnson, and Joffe. Drafting of the manuscript: Kesselheim and Joffe. Critical revision of the manuscript for important intellectual content: Kesselheim and Johnson. Statistical analysis: Joffe. Obtained funding: Kesselheim. Administrative, technical, and material support: Joffe. Study supervision: Joffe.

Financial Disclosure: None reported.

Funding Support: This study was supported by the Lovejoy Grant for Resident Education and Research (Dr Kesselheim) and the McGovern Scholarship (Dr Kesselheim) from Children's Hospital Boston. The study sponsor had no involvement in the study design; collection, analysis, or interpretation of data; writing of the report; or the decision to submit the paper for publication.

Additional Contributions: Tina Gelsomino, MSW, helped with data management. Jennifer Mack, MD, Robert Truog, MD, Mildred Z. Solomon, PhD, and the participants in the 2006-2007 Ethics Fellowship in the Division of Medical Ethics, Harvard Medical School, performed thoughtful reviews of the manuscript.