Clinical Ethics in Anesthesiology. A Case-Based Textbook
1.Consent and refusal
14. Informed consent for preoperative testing: pregnancy testing and other tests involving sensitive patient issues
Gail A. Van Norman
A healthy 15-year-old girl presents for elective diagnostic ankle arthroscopy for ankle pain and swelling. She is accompanied by her mother. During the preoperative interview, she appears acutely uncomfortable with questions about whether she is sexually active (she denies it) and the timing of her last menstrual period. The anesthesiologist informs her that she will need to get a urine sample for a pregnancy test. The test is required by the anesthesia group’s policy of pregnancy testing female patients, and members of the group will not perform elective anesthesia on pregnant patients. The patient’s mother questions the necessity of the test, stating with confidence that “my daughter has never had sex.” The urine pregnancy test, however, is positive. State law prevents the anesthesiologist from informing anyone but the patient of her positive test, although the mother will surely guess the test results if the case is cancelled. Furthermore, the patient is below her state’s age of consent for sexual intercourse, and her pregnancy is therefore by legal definition the result of statutory rape according to state law – which also requires any doctor who suspects child abuse to notify state authorities.
While physicians often consider ethical issues concerning medical therapies, it is easy to overlook ethical issues regarding something as routine as a preoperative laboratory testing. Yet principles of beneficence (doing good) and nonmaleficence (avoiding harm) suggest that anytime we prescribe a medical test, ethical considerations may be relevant, since we doing such testing precisely because we hope to benefit patients and/or avoiding harm. Preoperative testing presumably benefits patients by identifying unrecognized or disguised conditions that might adversely affect anesthetic risk. But harms can also result from preoperative testing. Some harms include the risk of a false-positive test erroneously labeling a patient as having a condition they do not have; the risk of a false-negative test falsely reassuring a patient that they do not have a condition which they in fact do; the risk that erroneous results might lead to inappropriate therapy with its attendant complications; the risk that erroneous test results might deprive a patient of important therapy they would otherwise get; and the complications of performing the test itself, and monetary cost, to name a few.
Are all preoperative tests ethically equivalent? Some of the ethical problems that face the anesthesiologist in the case introducing this chapter may be obvious, and some may not. But is there really an ethical problem with obtaining an ECG, for example? This discussion will focus on issues related to common, routine preoperative tests, and also examine two preoperative tests with special social implications: HIV and pregnancy testing.
General ethical principles regarding medical testing
Physicians have ethical obligations based in principles of beneficence and nonmaleficence to make responsible and knowledgeable decisions about whether a preoperative test is even warranted. Principles of good medical practice require that physicians balance the cost of testing against the likelihood that testing will produce more benefits than harms. Physicians are also bound by an ethical principle of fidelity to their patients. Fidelity is the concept that physicians should be faithful and committed in providing good medical care, and not compromise that care in the interests of anyone else, not even for physicians’ personal interests.. This principle respects the vulnerability of patients in the doctor–patient relationship. Not only does the doctor have special medical knowledge and skills in which the patient must place their trust, but also the physician determines to a great degree how expensive medical testing and therapy will be.
The principle of nonmaleficence requires physicians to consider, in addition to the monetary costs of a test, both the medical and social harms that may result from unnecessary or poorly conceived testing. Medical harms include the discomfort and inconvenience of the test and the potential for false-positive or false-negative results that misdirect medical therapy in ways that create greater harms than benefits. Such misdirection can occur even with a simple ECG. Take, for example, a 40-year-old healthy man with no medical complaints who presents for knee arthromenisectomy. His surgeon orders a routine preoperative ECG as he has for the last few decades on all of his patients scheduled for surgery. The ECG demonstrates concerning but nonspecific ST segment changes, so the surgeon consults a cardiologist who orders stress cardiac imaging for further clarification. Imaging reveals a significant area of decreased apical uptake compatible with myocardial ischemia or possible attenuation artifact, so a cardiac catheterization is undertaken – which reveals normal coronary arteries. Ultimately, the patient suffers a femoral artery tear during catheterization and has to undergo emergency vascular surgery. The physical and financial cost to the patient is very high, although no medical decisions concerning the original surgery were ultimately altered and no surgical risks reduced as a result. In fact, this healthy patient’s risk of a major adverse event increased with each test his doctor ordered. The most recent guidelines for perioperative cardiac workup now indicate questionable utility of a preoperative ECG in this case. The subsequent stress test was also not indicated because it was unlikely to reveal anything that would favorably alter outcomes for a low risk surgery.
Good medical practice, both from ethical and medical standpoints, includes applying evidence-based guidelines in determining if a test should be done, rather than on individual experience and beliefs. Individual experiences suffer from bias, unique confounding factors, and situational conflicts of interest. Anecdotes may be useful when no systematic investigation has been undertaken that can advise physicians about the course of action most likely to lead to the best overall outcomes. But anecdotal experience, albeit a strong tradition in medical education, serves us best when it spurs systematic investigation that results in sound, evidence-based decision support for physicians., Once evidence-based algorithms are available, they should guide most decisions and replace “routine” or traditional patterns of ordering tests.
Preoperative HIV and pregnancy screening
Social risks associated with preoperative testing may not be as obvious as medical risks, but can be the source significant harm. Two examples of tests that can produce social harm but are of limited preoperative utility are HIV and pregnancy testing.
Adverse social consequences known to be associated with HIV seropositivity include employment discrimination, loss of insurance, and social isolation. Studies demonstrate that seropositive women experience high rates of marital break-up, abandonment, and verbal and physical violence when their HIV status is disclosed.1 Compulsory preoperative HIV testing is known to prevent some patients from seeking medical care. Recognition of these harms has led in the US to the inclusion of AIDS patients in the protections afforded under the Americans with Disabilities Act, and has resulted in legislation specifically protecting the privacy of a patient’s HIV status.
Revealing a positive pregnancy test may likewise have negative, even life-threatening consequences for vulnerable patients in social environments where their pregnancy is not accepted. Studies show that female patients and their fetuses are in some situations at risk of physical violence. Further, adolescent pregnancies are sometimes the result of child abuse, incest, and rape. Communication of a positive pregnancy test result to the parents of a pregnant minor can place the child in jeopardy of further physical harm, since it may be evidence of criminal behavior on the part of a family member, or family friend or acquaintance. Many states have statutory requirements for physicians to report evidence of child abuse, and some authorities recommend reporting pregnant minors to Child Protective Services for investigation of possible abuse.
In much of the US, a female patient of any age has the legal right to absolute privacy regarding reproductive matters. To reveal or even imply the results of a pregnancy test to a third party, even a parent or spouse, without the woman’s consent, would represent an overt violation of law. The anesthesiologist who discovers a pregnancy is therefore left with few comfortable legal options if they have not first obtained the patient’s voluntary informed consent for pregnancy testing and discussed both how the test results will be used and to whom they can be revealed.
Given that there are risks of both social and medical harms associated with HIV or pregnancy testing, is there evidence that routine preoperative testing for HIV or pregnancy alters outcomes in a sufficiently favorable way to justify risking such harms?
HIV testing is usually ordered by the surgeon or anesthesiologist to determine which patients may pose a risk to members of the operating room team, and therefore with which patients they should be particularly careful to avoid possible exposure. Studies show that most surgeons and anesthesiologists erroneously believe that: (1) compulsory routine HIV screening will reduce their personal risk of exposure; (2) ordering such tests is the prerogative of the physician; and (3) that such tests can be done without the patient’s consent.
The effort to protect fellow members of the operating room team may be laudable. The doctrine of self-interest, however, is not an ethical principle, and carries little or no weight ethically when balanced against issues that affect the patient and his or her rights. This is especially true when the risk to the physician is low, and the primary method of reducing risk is applied to all patients anyway (e.g., universal precautions).
HIV testing has not been demonstrated to improve operating room safety. When applied to a low-prevalence population, positive screening tests are more likely to represent false-positive results than if testing is selectively applied to a high-risk population. False-positive results are harmful to the patient, who is unnecessarily labeled with a serious illness and may undergo further testing or treatment related to the false result. Physicians, can be harmed by false-negative test results if they are falsely reassured by the result. Furthermore, true negative tests can occur early in the course of HIV infection when viral titers (and infectious risks) are actually at their highest. Even a true negative test may therefore falsely reassure the operating room team that a patient represents lower risk when the opposite is actually true. If the outcome is relaxed vigilance regarding universal precautions, such tests may paradoxically increase actual risk or exposure. In any case, preoperative HIV testing does not benefit thepatient, although they will bear the monetary and social costs of the test. It is therefore difficult to justify routine unconsented preoperative HIV screening on either ethical or medical grounds.
Routine Preoperative pregnancy screening
Anesthesiologists often cite three reasons for screening pregnancy tests: (1) a desire to avoid unnecessary anesthesia exposure that may affect fetal development and/or increase the risk of spontaneous miscarriage; (2) a desire to avoid litigation for fetal anomalies or miscarriages that might occur following elective anesthesia exposure, and (3) a belief that female patients may lie to them about their pregnancy status. Are any of these fears well founded?
Despite widespread belief to the contrary, large population-based studies have failed to show definitive differences in the rates of fetal anomalies or spontaneous miscarriage following anesthetic exposure during the first trimester of pregnancy. Exposure of the fetus to anesthetic agents in early pregnancy does appear to be associated with lower birth weight.2,3 More recently, concerns have been raised about the effects in primates of later fetal exposure to anesthetic agents and its possible effects on subsequent neurocognitive development. However, no studies have been done to test whether neurocognitive development in humans is adversely affected after fetal exposure to anesthetic agents.4
Preoperative pregnancy testing also does not constitute a “standard of care.” In one study, only about one-third of anesthesia practices required preoperative pregnancy testing.5 In 2003 the American Society of Anesthesiologists Task Force on Perioperative Testing and the American Society of Anesthesiologist Committee on Ethics issued a joint statement that anesthesiologists should offer preoperative pregnancy testing to any female patient who might desire one, but that medical evidence for requiring pregnancy screening prior to anesthesia and surgery was lacking.6
Is preoperative pregnancy testing even necessary? No study has directly tested how accurate female patients are in reporting their pregnancy status in elective conditions when patient privacy is strictly protected. A study of preoperative pregnancy testing in adolescent female patients revealed the patients were accurate in reporting if they could possibly be pregnant, and no unexpected pregnancies were detected in over 500 instances.7 In one recent study of female patients presenting for elective surgery at a facility that performed routine pregnancy testing, the rate of “unexpected” positive pregnancy test results was < 0.2%, although one of those positive tests was actually a false positive.8 Other studies have put the rate of positive preoperative pregnancy tests at between 0.9%.9 Studies that report the results of preoperative pregnancy testing fail to distinguish between positive test results that occurred in women who claimed they were not pregnant versus those who simply didn’t know and requested testing. The accuracy of urine pregnancy testing is estimated at 97%–99%. To put it another way, conducting a urine pregnancy test will result in a false negative in up to 1%–3% of cases. Asking a female patient presenting for elective surgery if she is pregnant results in a false-negative response < 1% of the time. It seems somewhat questionable whether a urine pregnancy test is actually superior to simply asking the patient whether she is pregnant. There is no evidence whatsoever that exploring a female patient’s sexual and menstrual history has any influence on pregnancy test outcomes, much less anesthesia and surgical outcomes, despite the bizarre invasion of privacy these questions represent.
Is undetected pregnancy a major litigation issue in the practice of anesthesiology? As of 2003, two cases had been described in the ASA Closed Claims database related to spontaneous miscarriage of a previously undetected pregnancy following elective anesthesia.10 In one case, the patient prevailed against her surgeon when it was determined that the patient should have had a pregnancy test prior to deciding to proceed to surgery, since a positive test would have explained her medical condition and eliminated the need to operate. In the second, the anesthesiologist prevailed against the patient, because the patient failed to demonstrate that anesthesia was the cause of a later miscarriage, or even that preoperative pregnancy testing by anesthesiologists was a standard of care.
The ethical principle of respect for patient autonomy requires physicians to respect the decisions of competent patients once they have been properly informed of the risks. In most studies of preoperative pregnancy testing, the finding of a positive test was highly correlated with a woman’s decision to postpone surgery, indicating that the test results represented important information most of the women considered relevant in deciding whether to have surgery. Anesthesiologists therefore have an ethical duty to explain what is known and what is not known about the risks of anesthesia and surgery in early pregnancy, and to offer pregnancy testing to women, in case their decision to have surgery would be affected by the information. An autonomous woman who is properly informed of these largely theoretical risks has the right to refuse testing, both by law and ethical principles. Coercing such a patient into having a test against her wishes violates patient autonomy, and in general, refusal to have a test should not result in a cancellation of anesthesia or surgical care.
The anesthesiologist in our case presentation faces several serious ethical and legal problems: (1) The laws of her state afford an absolute right to privacy for females of any age with regard to reproductive status. To reveal, or even imply to anyone other than the patient that she is pregnant is an explicit violation of state law and the patient’s rights. (2) It is impossible for the anesthesiologist to know if revealing the information will expose the patient or her fetus to risk of serious physical harm from others. (3) The anesthesiologist has no practical way to discuss the results without the mother’s knowledge, and simply canceling the surgery will certainly clue the patient’s mother in to the test results. (4) To make matters worse, state law also requires the anesthesiologist to report her findings to Child Protective Services as evidence of sex with a minor, which is illegal under child abuse statutes. (5) Failure to obtain informed consent for the test is also a separately actionable legal claim in that state, even if no medical harm occurs. Instead of reducing the possibility of litigation, the anesthesiologist’s management of this case has actually opened an entirely separate avenue of litigation against her.
As is often the case when a situation gets off to such a bad start, there is no simple way to easily extricate our anesthesiologist from the ethical and legal mess in which she finds herself. But this situation might easily have been avoided if ethical principles had been followed from the outset. In the first place, a policy of requiring routine preoperative pregnancy testing is questionable at best and based on little or no supportive medical evidence. It creates ethical and legal dilemmas and potential social harm, while failing to provide proven improved outcomes to patients. Because not everything is known, or likely will ever be known, about the effects of anesthesia on a developing fetus, patients should be privately informed of the theoretical concern that pregnancy outcomes could be affected by surgical stresses and anesthetic exposure. They should also be informed that true risks are unknown, and they should be offered pregnancy testing if they desire it. The practitioner should also determine how the patient wishes test results to be handled, whether positive or negative. The informed consent process requires respect for informed refusal, and with rare exceptions patients should not be coerced into undergoing screening pregnancy testing by threatening to cancel the case if they refuse.
Whether or not routine pregnancy testing is mandated by group policy, preoperative pregnancy testing requires informed consent. The anesthesia practice should have policies and procedures that at minimum assure the following; (1) reproductive information will be elicited in private; (2) if the patient refuses testing, the anesthesiologist will have a response to third parties (such as parents) that respects the patient’s right to privacy; and (3) the consent process includes a private discussion with the patient about to whom she wants test results to be revealed. Anesthesiologists need to be aware of the social and legal consequences of discovering a pregnancy in a minor female patient, and group practices should have a process in place to refer appropriate patients for counseling and prenatal care. In some cases, if the patient is a minor, Child Protective Services may need to be involved if a test returns positive. Finally, if a group practice does decide despite these difficulties to mandate pregnancy testing, it would be wise to provide that information prior to the day of surgery to all female patients contemplating anesthesia with that practice, so that patients can incorporate this information in their decision of whether to have the surgery done at their institution, or go elsewhere.
• Good medical practice, both from ethical and medical standpoints, includes applying evidence-based guidelines in determining if a preoperative test should be done.
• Social risks associated with preoperative testing may not be as obvious as medical risks, but can be the source significant harm – and may well outweigh any potential medical benefit.
• Pregnancy testing and HIV testing are examples of two tests with significant social implications, but little proven medical benefit as screening tests. Policies requiring such tests should be reconsidered in light of the ethical principles respecting patient autonomy and striving for beneficence and nonmaleficence.
• Patients should be informed of the risks and benefits of preoperative testing; informed refusals should in general be honored.
• If socially sensitive preoperative tests are mandated by policy, patients should be informed at the time of scheduling of surgery and anesthesia that these tests are required, so that they can make a determination whether, in the interest of their own privacy, they wish to have care elsewhere.
1* Lester, P., Partridge, J.C., Cheesny, M.A., and Cooke, M. (1995). The consequences of a positive prenatal HIV antibody test for women. J Acquir Immune Defic Syndr Hum Retrovirol, 10, 341–9.
2* Mazze, R.I. and Kallen, B. (1989). Reproductive outcomes after anesthesia and operation during pregnancy: A registry study of 5405 cases. Am J Obstet Gynecol, 161, 1178–85.
3 Reedy, M.B., Kallen, B. and Kuehl, T.J. (1997) Laparoscopy during pregnancy: A study of five fetal outcome parameters with use of the Swedish Health Registry. Am J Obstet Gynecol, 177, 673–9.
4* Loepke, A.W. and Soriano, S.G. (2008). An assessment of the effects of general anesthetics on developing brain structure and neurocognitive function. Anesth Analg, 106(6), 1681–707.
5* Kempen, P.M. (1997). Preoperative pregnancy testing: a survey of current practice. J Clin Anesth, 9, 546–5.
6* Practice Advisory for PreAnesthesia Evaluation. (2003) ASA Task Force on Preanesthesia Evaluation, Revised. American Society of Anesthesiologists, Park Ridge, IL.
7* Malviya, S., D’Errico, C., Renolds, P., et al. (1996). Should pregnancy testing be routine in adolescent patients? Anesth Analg, 83(4), 854–8.
8 Kahn, R.L., Stanton, M.A., Tong-Ngork, S., et al. (2008). One-year experience with day-of-surgery pregnancy testing before elective orthopedic procedures. Anesth Analg, 106(4), 1127–31.
9 Hennrikus, W.L., Shaw, B.A., and Gerardi, J.A. (2001). Prevalence of positive pregnancy testing in teenagers for orthopedic surgery. J Pediatr Orthop, 21(5), 677–9.
10 Personal communication to the author from ASA Closed Claims Database analyst.
Lawerence, V.A., Gafni,, A., and Kroenke, K. (1992). Preoperative HIV-testing: is it less expensive than universal precautions? J Clin Epidemiol, 46, 1219–27.
Nyrhinen, T. and Leino-Kilpi, H. (2000). Ethics in the laboratory examination of patients. J Med Ethics, 26(1), 54–60.