A condition known as “Mysterious Anesthesia Reaction” or MAR is associated with a significant risk of complications upon exposure to an anesthesiologist. The condition affects 0.5% of the population overall. The test has a false positive rate of 5%, and a false negative rate of zero. Your preoperative clinic director decides that all patients must be tested preoperatively for the condition. The test costs $20. You test your first 5000 patients:
(a) Testing in a low-prevalence population
Number of people tested
Number of people with MAR
Number of people without MAR
Total
Positive test result
(True positive)
25
(False positive)
250
275
Negative test result
(False negative)
0
(True negative)
4725
4725
Total
25
4975
5000
Total cost of testing: $100,000.00
Cost to detect a true positive (true cost of avoiding one complication): $4000.00
Number of people labeled as positive: 275, or 15 × the actual number of people who are positive. 250 of 5000 people will be falsely labeled as having the condition
(b) Testing in a high-prevalence population
It turns out that MAR only occurs in natural redheads, who represent 2% of your clinic population. MAR occurs in 10% of redheads. So you decide to only test redheads. Of the next 5000 preoperative evaluations, the numbers look like this:
Number of people tested
Number of people with MAR
Number of people without MAR
Total
Positive test result
(True positive)
10
(False positive)
5
15
Negative test result
(False negative)
0
(True negative)
85
85
Total
10
90
100 (2% of 5000)
Total cost of testing: $2000.00
Cost to detect a true positive: $200
Number of people labeled as positive: 15, or 1.5 × of the number of people who are actually positive. 5 of 5000 people will be falsely labeled as having the condition
Total savings in testing strategy B: $88,000.00
Total absolute reduction in false positive tests: 345
“Targeted” tests, in contrast, evaluate a specific problem in a particular patient at risk. Myocardial imaging studies may be performed prior to surgery in a patient with a history of chest pain suggestive of angina, for example. Targeted tests have a much higher probability of having true positives and true negative results, since they are directed at patients who have a higher prevalence of disease (Table 2.1b).
All medical tests, whether screening tests or targeted tests, carry significant risks that must be weighed against the probability of the test providing real, relevant, and important information. Furthermore, not all medical tests are ethically equivalent. Some, such as HIV and pregnancy testing, can introduce serious social risks and create specific and avoidable harms that other tests do not.
The Case Presentation, Continued
The anesthesiologist in the perioperative clinic is reviewing charts late one night and discovers that our patient’s activated partial thromboplastin time (aPTT) is 56. All other coagulation studies are within normal limits. She finds no evidence of factors that would elevate the patient’s risk of bleeding and calls the surgeon to alert her of the results. The surgeon responds that she wants another aPTT drawn in the preoperative holding area in the morning. The next day, the phlebotomist makes 5 attempts before obtaining blood for a repeat aPTT. The blood clots on the way to the lab, and the phlebotomist returns to the patient to make 2 more attempts. The repeat aPTT is 56. The operating room is delayed for 2 h, and the surgeon decides to proceed with surgery despite the second abnormal lab result. With no peripheral veins for access, the anesthesiologist elects to place a central line, causing a large pneumothorax that requires treatment of hypotension and the placement of a chest tube.
False positive and false negative results can label a patient as having a condition they do not have, or falsely reassure that the patient does not have a condition that they in fact do. In the former case, further testing and subsequent interventions may increase expenses and subject the patient to potential complications of the testing, or to unnecessary treatments with more attendant risks. True positive test results may be clinically insignificant but can also lead to unnecessary further testing with its accompanying risks. Further testing always increases cost–through the cost of the test itself and the cost of managing any complications or treatments that may result from it. In our clinical case, further testing had the unintended consequence of exhausting peripheral intravenous access, necessitating a more invasive procedure that was then attended by a potential life-threatening complication.
Systematic over-testing increases the cost of health care to all patients in an already over-burdened health care system. This type of expenditure also diverts funds that might be spent on productive aspects of medical care to enterprises that have little or no hope of being beneficial to this patient or to patients in general.
The Role of Evidence-Based Testing in Ethical Care of Patients: Addressing the Balance of Risks and Benefits
The principles of medical ethics demand that risks of preoperative testing (potential maleficence) be weighed against the importance of the information to be gained (potential beneficence). The modern practice of medicine is firmly rooted in belief in scientific evidence: using scientific data to inform patient care is what distinguished the ancient practice of medicine from that of sorcery. Currently, we speak of practicing “evidence-based medicine” (EBM). In evidence-based practice, clinical experience is inextricably wedded to the conscientious and explicit use of the best available medical evidence when making clinical decisions. Nonsystematic (anecdotal) clinical experience and untested hypotheses are insufficient grounds for determining a physician’s actions in all but the rarest of circumstances–and then only when no credible evidence exists. EBM emphasizes respect for patient autonomy, by requiring that the patient be informed about the risks and benefits of proposed tests and therapies. Informed consent and recognition of the patient’s values and goals are also emphasized in the practice of EBM [5–7].
Evidence shows that many commonplace medical practices are not only unhelpful, but actually cause serious harm. Examples include administration of albumin for treatment of shock, which has now been shown to increase mortality [8]. Screening mammography prolongs the life of 1 in 2000 women over a 10-year period, but leads to a false diagnosis of cancer and institution of cancer therapy in 10 women during the same period, leading many to question its benefit-risk ratio [9].
Despite strong medical evidence regarding harms, many practitioners are unwilling to relinquish long-standing practices [10]. Patients’ expectations of the benefits and harms of screening tests and of treatments are often unrealistic, and may further encourage physicians to be injudicious in ordering them [11]. Reluctance to give up traditional but potentially harmful practices may have deep psychological roots, leading some to propose legislative action to compel doctors to use the evidence presented in such studies [12]. Movements such as the “Choosing Wisely” campaign in the United States (US) are gaining ground, and, “Choosing Wisely” has already identified unnecessary preoperative testing as a target area for practice improvement in anesthesiology and surgery [13].
Once we understand that ethical preoperative testing involves (1) identifying the true benefits to patients and (2) balancing them against evidence of harms, (3) determining which tests actually provide not only accurate, but clinically relevant information (i.e., information that will lead to effective preoperative optimization and perioperative risk-reduction strategies), (4) providing this information to patients in the informed consent process and (5) pursuing in partnership with the patient a clinical strategy designed to maximize potential benefits and minimize potential harms in accordance with their values, we can draw clearer and more cost-effective clinical plans for perioperative care.
What Is the Evidence for Performing Preoperative Screening Tests?
In order to appropriately inform our patients about preoperative tests and rationally determine whether tests should be ordered, all physicians involved in perioperative care are ethically obliged to have at least passing knowledge of the evidence regarding any tests they order, and to use that knowledge in a rational way. Fortunately for anesthesiologists and surgeons, the literature on routine preoperative tests is both abundant, and quite consistent. Sadly, studies show that anesthesiologists and surgeons are commonly ignorant of the evidence regarding the tests usefulness, and/or willfully ignore the evidence at hand. Some studies show that in a significant number of cases, the physician never even examines the results of the tests they order [14].
Electrocardiogram
In 1993, Atkins and Roizen, in an argument prescient to EBM, provided a powerful case for giving up one of our most beloved preoperative screening exams: the ECG [15]. The ECG is an insensitive and nonspecific examination that has little role in preoperative management, and yet is one of the most difficult to convince anesthesiologists and surgeons to relinquish.
Patients with abnormal preoperative ECGs have a greater risk of cardiovascular death than those with normal ECGs (1.8% vs. 0.3%, odds ratio 4.5). Although that may appear to be a meaningful difference, for low-to-intermediate risk surgery, the absolute difference in cardiovascular death is clinically insignificant, and the value of this test as a predictor of negative outcomes is virtually nil [16]. Furthermore, the test does not lead to changes in management that might account for any risk mitigation. In an unselected preoperative population, almost half of all ECGs are abnormal. Coronary revascularization prior to noncardiac surgery is now based on clinical findings and the patient’s current medical therapy and not the ECG, and is rarely indicated in the truly asymptomatic patient [17]. A comprehensive literature review found that the predictive power of preoperative ECGs is weak at best and that there is no evidence to support the value of “baseline” ECGs preoperatively [18]. Age over 65 independently predicts the presence of an abnormal ECG, but does not predict a modifiable risk factor for surgical patients. In one study, of 1149 ECGs reviewed, only 0.44% of patients without clinical risk factors had abnormalities on their ECG, irrespective of age [19]. In patients over age 70, up to 75% have abnormalities on a preoperative ECG, but these abnormalities do not independently predict postoperative cardiac complications when other clinical risk stratification is done such as the American Society of Anesthesiologists (ASA) physical status and the American Heart Association (AHA) cardiac risk index [20]. The predictive power of an abnormal ECG for perioperative cardiovascular events is slightly greater for patients with cardiovascular risk factors than those without, but the difference is not statistically significant. In fact, in one study a normal ECG had the same predictive value for such events [21]. According to AHA guidelines, preoperative ECGs are not indicated in asymptomatic persons of any age undergoing low-risk procedures [22]. The ASA Task Force on Preanesthesia Evaluation found no compelling evidence for routine ECG testing for age indications alone, but recommended possible testing for those at risk due to underlying disease, or due to clinical findings at a preoperative visit [23].
The total burden of harm of screening ECGs is not known, but in many patients with either falsely abnormal or clinically insignificant abnormal findings, additional testing is often undertaken with its attendant economic cost and medical risks. How can we argue ethically or medically that ECGs should be routinely ordered in low-to-intermediate risk patients when (1) there is no evidence that the vast majority of patients benefit, (2) there is no evidence that care is changed, and (3) there is evidence at least from an economic standpoint that both individual patients and patients collectively suffer harm?
Electrolytes
Recent literature on routine preoperative screening of serum electrolytes is scant, but results parallel that of ECG testing.
In one study, not only was the incidence of abnormal findings on chemistry tests low, but also in no patient was the anesthesia management modified as a result of the test [24]. In another systematic review, routine preoperative biochemistries discovered unexpected abnormal sodium or potassium levels 1.4% of the time and abnormal creatinine levels in up to 2.5% of patients, but clinical management was rarely modified [18]. A survey of studies from the National Health Service in Britain failed to identify any evidence about the effectiveness of such tests [25]. Furthermore, several studies demonstrated that chronic hypokalemia is not associated with increase risk of perioperative arrhythmias in the noncardiac surgical patient [26, 27]. Fritsch et al. found that abnormal electrolytes occurred in about 1.6% of patients preoperatively, but were not predictors of perioperative complications [28]. Johnson and Mortimer reported that preoperative management was altered in response to only 0.2% of results in 100 patients. No complications arose that were attributable to the test results [29].
Complete Blood Count
In one review of preoperative complete blood count (CBC) testing in otherwise asymptomatic patients, abnormal levels of hematocrit (HCT) or hemoglobin (Hgb) were found in about 5%, but Hgb was rarely below 9 g/dl, well above currently recommended thresholds for transfusion. CBC findings lead to a change of management in 0.1–2% of patients. Unexpected low platelet counts were rare (<1%) and rarely if ever resulted in management changes [18]. Another study of routine preoperative tests found that 60% had no indications and only 0.22% found abnormalities that might have changed perioperative management. However, those abnormalities were never acted on, and no complications resulted [30].
Coagulation Screening
Routine preoperative testing of coagulation parameters in asymptomatic patients in an effort to predict intra- and post-operative bleeding is well studied. For patients undergoing noncardiac surgery who do not have a history or physical findings suggestive of an increased risk of bleeding (Table 2.2), preoperative testing does not identify those at increased risk, is expensive, and exposes patients to the risks of further unnecessary testing.
Table 2.2
Factors that identify risk for increased perioperative bleeding
Patients generally at risk for increased bleeding can generally be ruled out by the following five questions; in the absence of a positive response to at least one of the questions, the patient should be considered low risk for bleeding |
1. Do you personally have a diagnosed bleeding disorder? |
2. Does anyone in your family have a diagnosed bleeding disorder? |
3. Do you bleed or bruise easily (e.g., bruising with minimal or no trauma, bleeding from the gums for a full 5 min after stopping brushing teeth, history of transfusion for minor trauma or surgery, very heavy menses requiring intervention, heavy bleeding after prior surgery or vaginal delivery) |
4. Do you take a blood thinning medication such as coumadin, lovenox, heparin, Plavix or Pradaxa? |
5. Do you have a medical condition that predisposes to bleeding such as advanced liver or kidney disease? |
Multiple studies involving adults and children undergoing low risk ambulatory surgery, major general surgery, major orthopedic surgery, major neurosurgery, spine surgery, endoscopic procedures, and ear, nose, throat (ENT) procedures demonstrate that coagulation screening does not predict bleeding risk and that patient history is at least as effective in predicting perioperative bleeding [31–37]. The ASA Task Force on Preanesthesia Evaluation does not recommend routine coagulation screening [23].
Related posts:
Pediatric Patients: Do Not Resuscitate Decisions
Perioperative Considerations of Do Not Resuscitate and Do Not Intubate Orders in Adult Patients
Informed Consent and the Disclosure of Surgeon Experience
Professionalism in the Operating Room
Futility and the Care of the Perioperative Patient
Honesty in the Perioperative Setting: Error and Communication
Full access? Get Clinical Tree
