|Home | About | Journals | Submit | Contact Us | Français|
In the United States, medical errors cause approximately 44,000 to 98,000 unnecessary deaths each year and as many as 1,000,000 excess injuries.1 Disclosure of these errors is a highly charged ethical and legal issue. While physicians correctly perceive an ethical duty to disclose an error to a patient,2,3 deterrent factors like lawsuits and other punitive actions cause a “disclosure gap.” Although most patients want their physicians to disclose harmful medical errors, this happens less than half of the time.4
The bounds of appropriate behavior are even less clear for physicians who discover medical errors made by their colleagues.5 The following vignette and discussion will address the ethical and legal issues faced by physicians who identify a medical error made by a professional colleague, and discuss the factors they must consider in determining when to disclose the error and to whom.
A 36-year-old patient with cancer moves to a new city and is referred to an oncologist at a local hospital.
Eight months previously the patient felt a lump in her left armpit. Her primary care physician referred her to a surgeon. The 2 × 2 cm left axillary node was excised, and identified by a pathologist as a highly undifferentiated malignant neoplasm, most probably of breast origin. No cyto- or immunocytochemistry studies were done. Bilateral mammography and physical examination of the breasts were normal. A chest x-ray and a computed tomography of the chest revealed a left hilar lymph node measuring 3 × 2 cm, and two left lower lobe round parenchymal infiltrates 1.5 cm in diameter each. The remainder of the physical examination, radiologic, and laboratory findings were normal.
The patient was diagnosed with “metastatic breast cancer” and referred to a local oncologist. The oncologist initiated therapy with paclitaxel every 3 weeks. The patient did not tolerate the drug well. She developed severe paresthesias and weakness in the legs, and severe inflammation of her nailbeds. After the fourth course of paclitaxel, the patient had a slight decrease in size of the pulmonary parenchymal infiltrates, persistence of the hilar lymph nodes, and the appearance of a new lymph node under the scar of the prior lymph node resection. She was then switched to intravenous doxorubicin (60 mg/m2 every 3 weeks), which she tolerated quite well. After the fourth course, the hilar node decreased to 1.5 × 1 cm, and the axillary node and parenchymal infiltrates disappeared.
At the time of her first visit to the new oncologist, the patient shuffled her feet and ambulated very slowly. She complained of fatigue. On physical examination, she had no palpable lymphadenopathy. She had skin hyperpigmentation, especially on the back of her arms and in the creases of her palms. There was marked decrease in tactile sensation of both upper and lower extremities, with weakness of both upper and lower extremities. All of the patient's reflexes were markedly decreased. The remainder of the examination was normal.
The patient was mildly anemic, and had a mild elevation in her lactate dehydrogenase.
The chest x-ray revealed a left hilar lymph node measuring 2 cm in its largest diameter. The computed tomography confirmed the hilar adenopathy, normal-sized liver and spleen, a round defect measuring 2 cm in diameter near the splenic hilum, and no other lymphadenopathy. Bilateral mammograms were normal. A multiple-gated acquisition (MUGA) scan revealed a left ventricular ejection fraction (LVEF) of 45%.
The blocks of the original axillary lymph node were received from the hospital where the initial lymphadenectomy had been done. The H&E stained slides revealed an undifferentiated malignant neoplasm. Immunocytochemistry revealed the tumor to be negative for cytokeratin and positive for CD45 and CD20. The diagnosis was changed to B-cell lymphoma, diffuse, large cell type, stage IV.
Due to the patient's reduced LVEF, rituximab plus cyclophosphamide, mitoxantrone, vincristine, and prednisone (R-CNOP) chemotherapy was administered. After three courses, the hilar lymph node and the splenic defect disappeared. However, the LVEF had decreased to 40%. Because of this, the mitoxantrone was discontinued, and the patient received three additional courses of retuximab-cyclophosphamide, vincristine, and prednisolone chemotherapy.
One and one-half years after completion of chemotherapy, the patient remains free of symptoms ascribable to her lymphoma. Her latest MUGA scan revealed a LVEF of 50%. However, she still has significant problems ambulating and performing fine manual labor. She asks her new oncologist about the treatment she received from her referring physicians.
The first errors in this case were made by the pathologist. The pathologist failed to obtain immunohistochemistry, and did not classify the patient's lesion even though it was “highly undifferentiated.” Furthermore, the status of estrogen receptor/progesterone receptor/HER2-neu was not evaluated, as it should be in any malignancy suspected of breast origin. The patient is young, with high likelihood of estrogen receptor/progesterone receptor–negative or HER2-neu over-expressive breast cancer. Treatment with trastuzumab hormonal agents might have been indicated, if the appropriate studies had been done.
Second, the patient was given the wrong drug, paclitaxel, because she was misdiagnosed with breast cancer. Regardless, the paclitaxel should have been discontinued when the patient demonstrated progressive neuropathy. This patient was already thought to have metastatic disease with lung nodules. Continuing aggressive treatment in the face of progressive toxicity, without confirming benefit, is questionable. Furthermore, this toxicity left the patient with an enduring, life-limiting disability.
Third, when switching the patient to doxorubicin, the patient's prior oncologist did not check her baseline cardiac function or monitor her for additional toxicities. Because doxorubicin was likely given with palliative intent, it was particularly important to spare the patient unnecessary toxicity.
The medical errors in this vignette are readily apparent. However, in real life cases the facts are often more ambiguous. Physicians may question whether a problematic course of treatment is an actual error or a legitimate difference in medical opinion. Physicians know that medical opinions are based on many factors, including community-based standards, personal knowledge and expertise, and previous experience. Reasonable physicians could treat a patient in very different ways without committing an error. This point is demonstrated by a recent study on second opinions in breast cancer, in which changes to patient's initial treatment plans were recommended more than 50% of the time.6 This is not to say that more than half of the patients involved experienced a medical error. Rather, this study shows that diagnosis and treatment plans recommended by physicians can vary and still be within normal bounds.
When distinguishing between legitimate treatment variations and actual errors, it may be helpful to consult the definition of error provided by the Joint Commission on the Accreditation of Healthcare Organizations: a medical error is an “unintended act, either of omission or commission, or an act that does not achieve its intended outcome.”7 It may also be helpful to ask the patient's prior physician about the rationale for his or her treatment decisions before deciding on a course of action that presumes an error occurred.
Medical practitioners are accountable to patients and, in some cases, institutions, to reveal observations of error. Disclosure of a harmful medical error reflects respect for the patient's autonomy and upholds high standards for health care quality and patient safety.
Talking to a colleague who may have made a medical error can present significant challenges for the identifying physician.
First, discussing a prior physician's treatment may seem counterintuitive for physicians who focus instinctively on his/her patient's present and future care. However, if an error was made, educating the prior physician presents an opportunity to improve the quality of care received by other patients that should not be overlooked.
Second, telling another physician he or she made a harmful error can be uncomfortable, especially when the physician who made the error is a colleague. The identifying physician will want to choose his or her words carefully to keep the tone of the conversation collegial. The identifying physician could say “I believe I have discovered an error in your diagnosis of a patient that I would want to know about if it were my error.” The physicians may also review the patient's medical history together.
In this case, the identifying physician should disclose, in a thoughtful discussion with the patient, both the errors made by her previous physician and the injuries she has suffered as a result. This includes misdiagnosis based on the pathologist's incomplete evaluation and the neurological and cardiac damage caused by the previous physician's chemotherapy choices.
Disclosure will enable the patient to make informed decisions about changes to her treatment plan, and gives her the opportunity to follow-up with her previous physician if she thinks it would be beneficial.
Whether a physician must report a harmful error made by a colleague can vary from state to state. Some states require licensed health care professionals to report errors that constitute actual misconduct, including gross negligence, incompetence, and moral unfitness.8 Treating patients while impaired by alcohol or drugs and abandoning or neglecting a patient in need of immediate care are examples of misconduct that must be reported. Misunderstandings about diagnoses and treatments are not.9 On a finding of misconduct, state licensing boards can restrict or revoke a physician's license,10 and may have to report to the National Practitioner Data Bank.11 State laws may offer protection from civil damages for those who report errors in good faith and without malice.12
For institutions, Joint Commission on the Accreditation of Healthcare Organizations patient safety guidelines require policies that foster disclosure of unanticipated outcomes to patients.13 However, the implementation of these provisions can vary greatly. For a physician considering reporting a medical error made by a colleague at a different institution, it is unlikely that the physician's own institution's policy will apply. If the identifying physician chooses to report a medical error, the institution where the error took place is responsible for investigating that error, reporting it to the appropriate authorities, and taking a peer review action against the physician if necessary.
Physicians who report medical errors should be mindful of applicable state and federal privacy laws, including the Health Insurance Portability and Accountability Act (HIPAA). Under the HIPAA regulations, a patient's protected health information (PHI) cannot be disclosed without permission unless disclosure is in furtherance of treatment, payment or health care operations.14 While disclosing PHI for error or adverse event reporting is likely permissible for purposes of quality improvement,15 physicians who are considering discussing a medical error should either disclose without identifying the patient, or ask for the patient's authorization to speak with his or her previous physician and hospital officials. Physicians should also minimize the amount of PHI actually disclosed. It is important to note that state privacy laws may vary, and can be more stringent than HIPAA.
In addition, physicians who report medical errors made by their colleagues may be asked to play a role in various institutional and legal proceedings. It is thus important that physicians document their identification of a medical error and any follow-up actions.
Tremendous variability exists regarding physician approaches to disclosure.16 Physicians may be reluctant to disclose a colleague's medical error to patients, institutions, and state licensing boards because they fear severe repercussions for the physician who made the error and for themselves. However, the overriding consideration should be the opportunity for enhancing patient safety, quality of care, and professional and educational responsibilities. Institutions and practices can help facilitate this process by developing innovative error reporting systems, educating physicians about the relationship between disclosure and quality improvement, and encouraging supportive exchanges with physicians who have experienced making a medical error.17,18
The authors gratefully acknowledge the contribution of the ASCO Ethics Committee, of which they are members.