Overall, 102 (87 per cent) of 117 orthopaedic surgeons (forty-two attending surgeons and sixty residents) who were given the survey completed it. All twenty-seven residents and ten of the thirteen attending surgeons at site 1 completed the survey, and thirty-three (92 per cent) of the thirty-six residents and thirty-two (78 per cent) of the forty-one attending surgeons at site 2 completed it. There were ninety-two men (90 per cent) in the sample. The residents were a mean of twenty-nine years old (range, twenty-six to thirty-five years old), and the attending surgeons were a mean of forty-seven years old (range, thirty-three to eighty-six years old). Forty-nine (82 per cent) of the residents had received formal ethics education, whereas only fourteen (33 per cent) of the attending surgeons had (Table I). Of the residents who had had ethics education, thirty-nine (80 per cent) had received it in medical school and thirteen (27 per cent), during residency. Thirteen of the attending surgeons who had received ethics education had done so in medical school. With the numbers available, no significant differences (p > 0.3) could be found between the residents at the two sites or between the attending surgeons at the two sites.
The respondents answered a mean of nineteen (73 per cent) of the twenty-six questions correctly. Overall, 84 per cent of the responses or more were correct in the subindices of confidentiality, economic aspects of care, truth-telling, and the approach to an incompetent colleague. Respondents performed less well with regard to the sections on end-of-life decision-making (a mean of 78 per cent of the responses were correct), physician-patient relationship (a mean of 72 per cent of the responses were correct), and informed consent (a mean of 58 per cent of the responses were correct). There were no significant differences between the two study sites with regard to these seven subscale scores. However, the attending surgeons scored significantly higher than the residents on the subscales of economic aspects of care (95 compared with 87 per cent, p = 0.01) and the approach to an incompetent colleague (88 compared with 83 per cent, p = 0.03). The residents scored significantly higher than the attending surgeons on the subscale of confidentiality (86 compared with 80 per cent, p = 0.006) (Table II). The overall mean composite ethics score, with all seven subindices equally weighted, was 79 per cent.
There were no differences in the composite ethics score according to site (79 per cent for site 1 compared with 80 per cent for site 2, p = 0.7) or when attending was compared with resident status (80 compared with 79 per cent, p = 0.7). In a multivariable regression model of the composite ethics score, attending compared with resident status, age and gender of the physician, and ethics training were not found to be significantly related to the ethics score, with the numbers available.
Confidentiality
Eighty-four per cent of the respondents answered the confidentiality questions correctly. They recognized that information may be released to the parents of a minor and to the designated decision-maker for an incompetent patient but not to a patient's supervisor at work without the permission of the patient. However, only fifty-two respondents (51 per cent) correctly responded that information about a patient who has a new diagnosis of gonococcal arthritis could be released to the health department without a patient's consent. Almost all respondents recognized common clinical situations that might breach patient confidentiality, including discussing a patient's case in a crowded elevator (100 per cent) and releasing information to an unauthorized individual (93 per cent). The percentage of correct responses to the confidentiality questions was higher for the residents (86 per cent) than for the attending physicians (80 per cent) (p = 0.006).
Informed Consent
The respondents were less knowledgeable with regard to issues of informed consent. Sixty-eight respondents (67 per cent) correctly recognized that, when a patient cannot make decisions for himself or herself, "what the patient would want" should guide care; however, fifteen respondents answered that what the family thought was best was most important and seven respondents answered that what the physician thought was best was most important. The remaining twelve respondents answered that the patient's diagnosis or prognosis was most important. When defining the capacity to consent, most respondents recognized that a patient must be able to understand the important risks and benefits as well as the indications for and alternatives to a procedure. However, nineteen (19 per cent) of the 102 respondents thought that a patient must be able to read the consent form, forty (39 per cent) thought that the patient must understand all of the risks of the procedure, twelve (12 per cent) did not think that the patient needed to be able to communicate in some fashion, and thirty (29 per cent) indicated that the patient must not be taking any medication with psychoactive effects.
Misunderstanding regarding informed consent and analgesia was also demonstrated by the responses to the following question.
A rugby player is brought alone to the emergency room with a severely comminuted grade-IIIa femoral fracture with an overlying penetrating wound. He agrees to emergent open reduction with internal fixation after you explain the indications, risks, and benefits of the procedure. He pleads for pain relief. The narcotic analgesic is ready for administration, but the nurse has not yet located the surgical consent form. What is the appropriate action?
A: Administer Toradol (ketorolac tromethamine) intramuscularly but no narcotic analgesic until the consent form is located and is signed by the patient.
B: Administer the narcotic analgesic and document that the patient understood the indications, risks, and benefits of the procedure. When it is available, have the patient sign the consent form if he is capable.
C: Administer the narcotic analgesic. When the consent form is located, have two physicians sign the form because the procedure is emergent and the patient cannot provide consent because he has received narcotics.
D: Administer no analgesic until the consent form is signed.
Twenty-seven respondents (26 per cent) chose A, forty-four (43 per cent) chose B, nine (9 per cent) chose C, and twenty-two (22 per cent) chose D. The appropriate response is B. In this case, the patient fulfills the elements of informed consent and agrees to the procedure. The analgesic for this individual, who is suffering, should not be withheld until the consent form is located.
Other cases demonstrated that the use of implied consent and therapeutic privilege were understood by eighty-one and eighty-three (79 and 81 per cent) of the respondents, respectively. With the numbers available for study, we could detect no significant difference in the proportion of correct responses between the residents and the attending surgeons (p = 0.11).
Truth-Telling
Ninety-six respondents (94 per cent) indicated that they would tell the truth concerning a therapeutic mishap. Eighty-six (84 per cent) indicated that they would not write an untruthful medical excuse for a patient who was seeking reimbursement for an airline ticket, two indicated that they would write the note, and fourteen indicated that they would subvert the truth by writing that "the patient reports recurrent back pain preventing airline flight." Attending surgeons and residents answered the truth-telling questions equally well (87 compared with 91 per cent, respectively, of the questions were answered correctly; p = 0.42).
Physician-Patient Relationship
The respondents demonstrated an excellent understanding of the obligations to a patient who is seen on a continuing basis and to a patient who is infected with the human immunodeficiency virus. Ninety-nine respondents (97 per cent) indicated that if a patient who was an intravenous drug user and had been seen on a continuing basis was found to have an infection they would provide emergency care despite the fact that the patient had violated an agreement to stop using drugs.
The respondents were less accurate regarding the economic aspects of the physician-patient relationship, as presented in the following case.
You are planning to perform a hip arthroplasty on a forty-two-year-old athlete who has debilitating pain due to osteoarthrosis. From the many prostheses available, you choose one that you designed. Since you receive a royalty on each prosthesis used, what should you do?
A: Not use this prosthesis.
B: Reveal that you receive a royalty on this prosthesis.
C: Not discuss with the patient that you receive a royalty on the prosthesis but include it in the text of the informed-consent form.
D: Use the prosthesis only if you credit the patient the amount of your royalty.
E: Use the prosthesis and never raise the issue of royalties.
One respondent (1 per cent) chose A, seventy-three (72 per cent) chose B, three (3 per cent) chose C, four (4 per cent) chose D, and twenty (20 per cent) chose E. One respondent did not answer the question. The appropriate response is B. The physician must reveal that he or she receives a royalty on the prosthesis because it is not a fact that would be readily known by the patient and the revelation is necessary to avoid both real and apparent conflict of interest.
With regard to fee-splitting, seven of the 102 respondents were willing to return a portion of a surgical fee as a gift and one was willing to pay for some of the expenses of the referring physician's practice.
When presented with a different type of patient-physician conflict of interest, many physicians did not act according to a patient's preferences, as shown by their answers to the following question.
One day, while covering for the Orthopaedic Service at University Hospital, you speak with a family that is distraught concerning the condition of their seventy-nine-year-old mother who has been in the intensive-care unit for four weeks. An infection developed after an operation for a hip fracture sustained during a fall. The patient became septic, went into atrial fibrillation, and suffered a massive stroke. She is now unresponsive; is being ventilated; and is receiving pressors, dialysis, and amphotericin B for fungal sepsis. The family has explained that they are sad regarding the outcome but their mother had not been well before the fall. They are most upset that she is being kept alive because she had made them promise that she would never be kept alive on machines if she could not return to a functional state: a conclusion that the neurologists came to with certainty two weeks ago. What should be done?
A: A psychologist or social worker should be called to help the family to deal with their grief.
B: The hospital risk manager should be called.
C: The patient should be made comfortable and be allowed to die.
D: The patient should be transferred to a long-term-care facility.
Twenty-two respondents (22 per cent) chose A, twenty-six (25 per cent) chose B, fifty-three (52 per cent) chose C, and none chose D. One respondent did not answer the question. The appropriate response is C. This patient's wishes should be respected, despite the fact that such action will result in a perioperative mortality.
With the numbers available, there was no significant difference between the proportion of correct answers given by the attending surgeons (74 per cent) and that given by the residents (70 per cent) to the questions concerning the physician-patient relationship (p = 0.27).
Economic Aspects of Care
Ninety-one per cent of the responses regarding ethical conflicts posed by managed care were correct. For example, the respondents were asked to consider the following case.
You see managed-care patients with the understanding that all procedures must be preapproved. Late one afternoon, you evaluate a managed-care patient with chronic back pain who had a sudden onset of dysfunction of the bowel and bladder. You call to get approval for magnetic resonance imaging, but the primary-care physician is unavailable and the approval office of the health maintenance organization denies the request, stating that this is not on their list of conditions requiring magnetic resonance imaging. They suggest that you send the patient to the primary-care physician in the morning. What do you do?
A: Arrange for magnetic resonance imaging first thing the next morning and send a fax to the primary-care physician's office.
B: Send the patient for the magnetic resonance imaging, but tell her that she may need to pay for it if it is not approved retrospectively.
C: Document the attempts to get care for the patient and arrange for her to be seen the next day.
D: Write in your note that there are no neurological symptoms or signs so the magnetic resonance imaging is not apparently indicated but arrange for an early follow-up examination.
One respondent (1 per cent) chose A, ninety-nine (97 per cent) chose B, two (2 per cent) chose C, and none chose D. The appropriate response is B. Ninety-nine of the respondents recognized their responsibility to the patient and elected to arrange emergent magnetic resonance imaging despite the response from the health maintenance organization.
The duty to advocate for the patient's best interest is fundamental to the physician's role, and the American Medical Association Council on Ethical and Judicial Affairs notes that "when managed care plans place restrictions on the care that physicians in the plan may provide to their patients … regardless of any allocation guidelines or gatekeeper directives, physicians must advocate for any care they believe will materially benefit their patients."7 The physician should inform the patient of his or her potential financial responsibility for the procedure; however, if the insurer refuses to pay for the test because of an inappropriate guideline, "the physician's duty as patient advocate requires not only a challenge to any denials of treatment … but also advocacy at the health plan's policymaking level to seek an elimination or modification of the guideline."7
In another scenario, all 102 respondents chose to take steps to change or appeal a primary-care physician's objection to an expensive new, but proved, treatment modality, and none were willing to tell the patient that nothing could be done, as the primary-care physician preferred.
However, concerning ownership of an imaging unit, a substantial number (twenty-six respondents [25 per cent]) were willing to conceal the financial conflict of interest, as demonstrated by the following case.
You and two lucky colleagues have a practice that is 100 per cent fee-for-service. Together you own a magnetic resonance imaging machine in a freestanding diagnostic center. The three of you purchased it in order to provide better service to your patients than that offered by the medical imaging monopoly in town. What must you say to a patient when ordering a magnetic resonance image?
A: "This magnetic resonance imaging is more convenient than that in other places."
B: Nothing.
C: "This magnetic resonance imaging is more convenient, and I am a part-owner of the scanner."
D: You need to reveal that you are part-owner of the scanner only if the magnetic resonance imaging is not completely covered by insurance.
One respondent (1 per cent) chose A, twenty-one (21 per cent) chose B, seventy-six (75 per cent) chose C, and four (4 per cent) chose D. The appropriate response is C.
Overall, the attending surgeons answered 95 per cent of the questions regarding economic aspects of care correctly compared with 87 per cent of the residents. The difference was significant (p = 0.01).
End-of-Life Decision-Making
The level of knowledge regarding terminal-care issues was lower than that in most other areas. Fifty-three respondents (52 per cent) were aware that decision-making based on a preference for care written in a living will was better than reliance on a family member's interpretation of a patient's wishes (a substituted judgment) and that this, in turn, was better than a best-interest decision made by the family and the physician. However, thirty-nine respondents (38 per cent) rated a best-interest decision as being better than one based on the patient's written preference or on a substituted judgment. Eighty respondents (78 per cent) answered that withholding and withdrawing a ventilator are not ethically equivalent acts, despite the well accepted opinion that they are1,9. The lack of understanding about how to temper the many aspects of life-prolonging care at the end of life was exemplified in the responses to the following question, for which respondents were asked to choose all of the appropriate answers.
A thirty-four-year-old man was severely injured in a boating accident. He was submerged for more than five minutes before rescuers arrived. After successful resuscitation and stabilization, he was taken to the operating room for treatment of a bilateral femoral fracture. Postoperatively, he was managed for anoxic brain damage by the neurology department, but the family turns to you for decisions. After a week of observation, the patient remains in a deep coma and it is clear that he will never recover mental function, although his lungs are improving to the point where he might be weaned from the ventilator and all other organ function is good. All of the members of the man's family agree that he would never want to live in this state and that he should be allowed to die. What maneuvers are acceptable?
A: Write a do-not-resuscitate order.
B: Remove the patient from the ventilator.
C: Start a morphine drip for the purpose of suppressing respiration.
D: Stop total parenteral nutrition.
E: Move the patient out of the intensive-care unit.
Ninety-seven respondents (95 per cent) chose A, thirty-nine (38 per cent) chose B, two (2 per cent) chose C, forty-one (40 per cent) chose D, and forty-seven (46 per cent) chose E. Answers A, B, D, and E are all appropriate responses. It is ethically acceptable to write a do-not-resuscitate order; to remove all therapies, including the ventilator and total parenteral nutrition; and to move the patient out of the intensive-care unit. The use of narcotics with the intent of suppressing respiration is not acceptable.
There also was confusion regarding the meaning of a do-not-resuscitate order, as may be noted in the responses to the following question.
What does a do-not-resuscitate order mean? (Choose all that apply.)
A: Do not intubate.
B: Do not move into an intensive-care unit.
C: Comfort measures only.
D: Do not perform cardiopulmonary resuscitation.
E: Do not perform an operative procedure.
Sixty-six respondents (65 per cent) chose A, twenty-three (23 per cent) chose B, twenty-one (21 per cent) chose C, ninety-nine (97 per cent) chose D, and fifteen (15 per cent) chose E. The appropriate response is D, and in most institutions A would also be appropriate. Do-not-resuscitate orders have no effect on other treatment modalities.
Similarly, nearly all respondents understood the usefulness of advance directives that designate an individual to make decisions for the patient should the patient be unable to do so, and all respondents knew that the document may be retracted. However, many held misconceptions regarding advance directives, as noted by the responses to the following question.
Which of the following are true of an advance directive? (Choose all that apply.)
A: May designate an individual to make decisions for the patient should the patient be unable to do so.
B: Requires the assistance of an attorney in order to be valid.
C: May encourage a discussion about future treatment decisions among the patient, the surrogate decision-maker, and the physician.
D: Delineates who should make financial decisions for the patient.
E: May be retracted, but only in writing.
F: Must be followed, even if the patient changes his or her mind.
Ninety-seven respondents (95 per cent) chose A, twenty (20 per cent) chose B, eighty-three (81 per cent) chose C, twenty-seven (26 per cent) chose D, fifty-three (52 per cent) chose E, and none chose F. The appropriate responses are A and C.
Overall, there was no difference between the percentage of correct answers given by the attending surgeons (77 per cent) and that given by the residents (80 per cent) regarding end-of-life decision-making (p = 0.19).
Incompetent Colleague
Ninety-nine respondents (97 per cent) stated that they would intervene in the practice of the orthopaedic surgeon in the following case.
A colleague of yours has begun to drink heavily. Although you are unaware of any clinical problems that have arisen, you noted that he was inebriated during surgery today and that the senior resident was forced to do the whole procedure. Must you intervene?
Yes
No
Ninety-nine respondents (97 per cent) chose yes, and three (3 per cent) chose no.
You discuss with this colleague his use of alcohol. He denies having a problem. Which of the following is a reasonable next step? (Choose all that apply.)
A: Advise the hospital credentials committee.
B: Do nothing.
C: Warn his patients by telephone.
D: Advise the departmental quality-assurance committee.
E: Tell his friends.
Thirty-four (33 per cent) chose A, two (2 per cent) chose B, none chose C, ninety-three (91 per cent) chose D, and eleven (11 per cent) chose E. Answers A and D are appropriate responses.
Overall, the percentage of correct responses to the questions regarding how to deal effectively with an incompetent colleague was 88 per cent for the attending surgeons and 83 per cent for the residents. This difference was significant (p = 0.03).
This survey demonstrated that orthopaedic surgeons and residents are quite knowledgeable in many critical areas of clinical ethics. Most of the respondents indicated that they would appropriately handle cases involving difficult ethical questions concerning economic issues in clinical orthopaedics, truth-telling, confidentiality, and the approach to an incompetent colleague, and they correctly answered questions testing knowledge. This is particularly important with regard to economic issues, given the level of concern that has been voiced regarding the pressures on orthopaedic surgeons in the current health-care economic environment8,10,17,19. In the present study, both the attending orthopaedic surgeons and the residents responded that they would protect the interests of their patients when presented with managed-care dilemmas that placed a patient at risk of harm or inadequate care. Similarly, most respondents answered that they would preserve the truth in dealing with a medical mistake and in writing a letter for a patient. These physicians also recognized their role with regard to approaching and, if necessary, reporting an incompetent colleague.
It should be noted that appropriate responses to questions and scenarios do not necessarily translate into appropriate behavior. However, a lack of knowledge concerning issues of medical ethics makes appropriate behavior unlikely. Thus, there should be concern about the small number of respondents who did mishandle these hypothetical cases. One-quarter (twenty-six) of the respondents were willing to conceal ownership of a diagnostic imaging unit from their patients. Sixteen (16 per cent) were willing to write a note for a patient that either was an outright lie or violated the spirit of the physician's authority to evaluate the presence of illness. Although it is reassuring that most respondents knew how to approach confidentiality, truth-telling, and cases of economic conflict of interest, a small number of orthopaedic surgeons need additional education or intervention.
The findings of this survey show that there was greater misunderstanding of proper ethical conduct in the areas of the physician-patient relationship, informed consent, and end-of-life decision-making. It was encouraging that respondents answered that they would not abandon a difficult patient in need of emergent care. As shown in a previous survey based on self-reporting by physicians2, these orthopaedic surgeons stated that a patient's infection with the human immunodeficiency virus would not stop them from providing optimum quality care. However, analogous to the willingness not to reveal a potential conflict of interest concerning an imaging unit, one-fifth (twenty) of the respondents in the present study said that they would not tell patients about royalties to be obtained from a total joint prosthesis. If a surgeon will gain royalties from the use of a prosthesis, it is incumbent on him or her to reveal this potential conflict of interest to the patient14,23. In the case regarding fee-splitting, eight respondents (8 per cent) were willing to pay a kickback to a referring physician. A third form of conflict of interest arose in a case in which a patient was kept alive against her wishes rather than being allowed to die and being counted as a perioperative mortality. Forty-eight (47 per cent) of the respondents were unwilling to let the patient die. These responses probably demonstrate the discomfort of orthopaedic surgeons in dealing with death more than a difficulty in handling this type of conflict of interest. However, taken together, the responses in these cases show that the orthopaedic surgeons in the sample surveyed are in need of additional education about conflict of interest in the physician-patient relationship.
Given that obtaining informed consent is integral to the daily practice of most orthopaedic surgeons, the deficits in knowledge revealed by this survey were unexpected. The elements of informed consent and the capacity to consent were not well understood by many respondents. In addition, how to orchestrate decision-making for a patient who does not have the capacity to consent was misunderstood by a substantial number of physicians. Perhaps of greatest concern was the fact that twenty-two respondents (22 per cent) said that they would withhold all analgesics from a patient in severe pain who has consented to the operative procedure but who has not signed the informed-consent form because it has not yet been located. An additional twenty-seven respondents (26 per cent) stated that they would provide only a non-narcotic analgesic for such a patient. In this case, informed consent had been obtained by the physician. The informed consent process should be described in the medical record, along with the patient's decision-making capacity and the unusual circumstance of the consent form being unavailable. Optimally, this note in the chart should be countersigned by the patient and should be adequate to document the patient's consent. However, in many hospitals, a signed consent form is needed before the patient can be taken to the operating room and given anesthesia. Under such circumstances, a surrogate's signature or perhaps two physicians' signatures may be needed on the form to document an implied consent. From an ethical perspective, these latter two options should be avoided because each disregards the patient's autonomously provided consent and implies that another decision-maker is needed after a valid consent has been tendered. However, these options are preferable to leaving the patient in pain while waiting for the informed-consent form to be signed.
Orthopaedic surgeons had an inadequate fund of knowledge about end-of-life issues. The answers to questions regarding these scenarios revealed that they did not universally recognize that a patient's preferences concerning end-of-life care should be respected even when the care strategy would permit death. There was substantial misunderstanding concerning the meaning of a do-not-resuscitate order. Such an order means only that a patient should not receive cardiopulmonary resuscitation. The order can be given for a variety of reasons, including the patient's perception that resuscitation may not be beneficial or would result in an unacceptable quality of life, the patient's unwillingness to receive invasive medical care to maintain his or her current quality of life, or his or her preference to avoid certain procedures. An understanding that the do-not-resuscitate order does not affect any care other than resuscitation is necessary in order to properly manage a patient for whom such an order has been given. The misconception that do-not-resuscitate orders have a much broader meaning than intended leads physicians to underuse such orders.
Similarly, a lack of familiarity with advance directives makes these documents less likely to be used. While nearly all of the respondents knew that the documents could be used to designate a surrogate decision-maker, twenty (20 per cent) thought that an attorney's assistance was needed to complete the document and more than half thought that such documents could be retracted only in writing. These perceptions are incorrect and might decrease the likelihood that a physician would encourage a patient to use such a document to state preferences in advance. Given that orthopaedic surgeons are performing an increasing number of procedures on elderly and functionally impaired patients and given the increase in palliative orthopaedic procedures, a better understanding of the ethical issues concerning end-of-life decision-making is important for orthopaedic surgeons.
Attending surgeons demonstrated significantly better knowledge of ethical economic issues and the handling of an incompetent colleague, while residents scored significantly better in the area of confidentiality. This difference may represent the greater experience that attending orthopaedic surgeons have with contemplating and dealing with economic issues and incompetent colleagues. Since residents reported more formal training in ethics, perhaps confidentiality was covered in this education. However, overall there was no difference in the composite ethics score between the attending surgeons and the residents. This finding may reflect the strong influence of attending role models on residents' knowledge and perceptions of appropriate behavior. It is also of interest that no relationship was found between formal ethics education and the composite ethics score. Ethics education in medical schools has increased substantially15, and a survey of individuals who had had such training indicated that they considered the courses valuable16. Previous studies have shown that ethics education in medical school can improve a student's moral reasoning20,21. Furthermore, a study of orthopaedic surgeons suggested a relationship between moral reasoning and clinical behavior: surgeons who had higher scores for moral reasoning had had fewer malpractice claims during the eight years preceding the study3. It is possible that the areas of clinical ethics tested in the present survey are not the focus of the ethics courses that medical students and residents receive. It is also apparent from our findings that the attending surgeons will have difficulty teaching the material to their residents since they have similar deficits in knowledge.
Considering the ethical issues addressed by this survey raises the question of whether there are correct answers to ethical dilemmas. On the basis of the principles of autonomy, beneficence, justice, and truth-telling that are fundamental to medicine and to the physician's profession, guidelines have emerged to direct behavior. These guidelines, which are often embodied in consensus statements (several of which have been referenced in this paper), are for the most part in concert, forming a structure from which to solve most ethical issues. Physicians may wonder if laws would provide the same guidance. In general, laws provide only a basic level of direction while ethical reasoning may help to resolve specific complex situations involving specific patients14. Certainly, physicians must be aware of state and federal laws, and in rare instances there will be a conflict between the ethically appropriate act and the law. However, far more often ethical behavior is consistent with the law, as seen in the discussion of informed-consent issues in the present report. Misunderstanding of the law may lead to inappropriate ethical behavior, as demonstrated by several of the responses to the end-of-life decision-making questions.
This study has substantial limitations. We surveyed only academic orthopaedic surgeons and their residents. These results do not represent orthopaedic surgeons in private practice or those practicing in other venues. In addition, we surveyed only two residency programs. However, the fact that there were no significant differences between the scores at the two sites suggests that it may be possible to generalize these findings to other orthopaedic residency programs. It should be noted that, at the time of this survey, neither program offered courses in clinical ethics. Individuals from programs that provide such clinical ethics training may demonstrate better knowledge.
The greatest problem with this study is that we presented scenarios and questions regarding knowledge rather than studying behavior. A response to a scenario may not reflect actual behavior13. Methods relying on scenarios underestimate deficiencies because a response reflecting adequate knowledge is assumed to mean that the behavior associated with that knowledge would be carried out. According to the model developed by Rest, moral behavior requires several steps: moral sensitivity, judgment of whether actions are moral, motivation, and development of a moral character. These steps move beyond the level of questions in our survey. In focusing on knowledge of ethics and then asking about intended behavior, the present survey focused primarily on the respondent's moral sensitivity. Yet, it is important that we detected deficits in knowledge at this level, as moral sensitivity is the first in a hierarchical series of steps to moral behavior. This survey methodology demonstrates areas that are in need of improvement. Furthermore, the survey used in this study focused on common clinical ethics domains and may be improved by redirection and expansion to evaluate better the full spectrum of an orthopaedic surgeon's practice.
We identified, among both faculty and residents at two orthopaedic surgery training programs, a number of clinical ethical issues about which orthopaedic surgeons should be more knowledgeable. Whether educational programs will succeed in remedying these gaps in knowledge must be the subject of future studies.
NOTE: The authors thank Judith Wilson Ross, M.S., and David Blake, Ph.D., J.D., for advice regarding the survey cases, and Honghu Liu, Ph.D., for assistance with the statistical analysis.