Free USMLE Step 2 CK Ethics and Patient Safety practice questions

Availability bias is judging the likelihood of a diagnosis by how easily examples come to mind. A recent, vivid, emotionally charged case, especially one the clinician personally missed, makes that disease feel far more common than it is. Aortic dissection is rare among emergency patients with chest pain, but after her painful experience this physician now overestimates its probability in every patient, leading to overtesting, with its costs, contrast and radiation exposure, incidental findings, and delays for other patients. The opposite face of the same bias is forgetting rare diseases that one has not seen recently. Debiasing strategies include using explicit pretest probability and validated tools (for example, the aortic dissection detection risk score with D-dimer in selected low-risk patients), reviewing base rates, and peer feedback; supportive debriefing after a bad outcome also matters, because the "second victim" experience often drives this pattern.

A. Anchoring is failing to adjust a first impression in a single patient when new data arrive; here the issue is a probability estimate shifted across many patients by a memorable case. B. Confirmation bias is selectively seeking or overweighting data that support a favored diagnosis; she is testing to exclude dissection, not filtering evidence to confirm it. C. Premature closure is accepting a diagnosis before it is verified and stopping the search, which would produce undertesting, not overtesting. D. Sunk cost bias is reluctance to abandon a plan after investing in it, which does not describe a pattern triggered by a prior case.

Medical malpractice is a civil tort that requires the plaintiff to prove all four Ds by a preponderance of the evidence: duty (a physician-patient relationship), dereliction (care below the standard of a reasonably prudent physician), direct causation (the breach more likely than not caused the injury), and damages (actual physical, emotional, or financial harm). If any element is missing, the claim fails.

Here the pediatrician clearly had a duty to the child, and prescribing a drug to which the child had a documented allergy, despite an alert, falls below the standard of care. However, the child suffered no injury: no reaction occurred, and the infection was treated successfully. Without damages there is nothing for the breach to have caused and nothing to compensate, so a malpractice claim would fail. An error that causes no harm is still an error, however. It must be disclosed honestly, as the pediatrician did, and it should be reported through the practice's safety reporting system so that the cause (an overridden allergy alert) can be analyzed and prevented.

A is wrong because the pediatrician examined and treated the child, which establishes a physician-patient relationship. B is wrong because ignoring a documented allergy is a departure from accepted practice. C is wrong as the best answer because causation is evaluated only in relation to an injury; the fundamental gap is that no injury occurred. D is wrong because the standard of care is the yardstick for judging dereliction rather than a separate element, and the care here did fall below it.

Acetaminophen is the most common co-ingestant in intentional overdoses, it is hidden in many combination products (cold remedies, "PM" sleep aids, opioid combinations), and early poisoning causes few or no symptoms, so the history and examination cannot exclude it. Patients also give unreliable histories after a suicide attempt. For these reasons a serum acetaminophen concentration is recommended for every patient with an intentional ingestion, regardless of the reported agent, and a salicylate level is commonly added. A missed level delays N-acetylcysteine, which is nearly completely protective when started within 8 hours.

Here the error was a system failure: the right test depended on each physician remembering to order it. The most effective fix is a standardized overdose order set (or default panel) that automatically includes an acetaminophen level for all intentional ingestions. Standardization and forcing functions are stronger than education or vigilance in the hierarchy of safety interventions because they make the correct action the default rather than relying on memory under time pressure.

A. Retraining one physician is a weak, person-focused intervention and does not change how the department orders tests. B. Urine drug immunoassays and most comprehensive urine screens do not provide the quantitative serum level needed to apply the nomogram; acetaminophen is not detected on standard immunoassay panels. C. Mandatory toxicology consultation for every overdose is impractical, delays care and still does not guarantee that the level is drawn. D. Ordering the level only when the history suggests acetaminophen is exactly the practice that failed, because histories are unreliable and the drug is often unrecognized in combination products.

Valid informed consent requires that the patient understand her diagnosis, the proposed treatment, its risks and benefits, and the reasonable alternatives, including their risks and benefits. For a small, unifocal, clinically node-negative breast cancer, breast-conserving therapy (lumpectomy plus whole-breast radiation) and mastectomy give equivalent overall survival. Many patients request mastectomy because they believe it is safer, as this patient's fear after her aunt's death suggests. A choice based on a misunderstanding of the alternatives is not an informed one.

The surgeon should acknowledge her fear, then briefly and clearly explain that both operations offer the same chance of cure, and outline what each involves: lumpectomy requires radiation and has a slightly higher local recurrence rate, whereas mastectomy is a larger operation, may involve reconstruction, and has its own complications. Patients may decline detailed information, but the physician should first make sure the patient knows that an equally effective, less extensive option exists. If she still prefers mastectomy after understanding the options, that is a legitimate choice that should be respected, because patient preference is an accepted indication for mastectomy. This is shared decision-making.

A. Proceeding without mentioning breast conservation means her consent would lack a key alternative, even though mastectomy itself is a reasonable choice. C. Refusing her preferred operation disregards her autonomy; mastectomy is an accepted option for any patient who prefers it after counseling. D. Bilateral mastectomy offers no survival benefit for an average-risk woman with negative genetic testing, and recommending it would add morbidity based on fear rather than evidence. E. Fear after a relative's death is an understandable reaction, not evidence of a psychiatric disorder or of lack of capacity, and referral would delay care.

The Children's Health Insurance Program (CHIP) covers uninsured children in families whose income is too high to qualify for Medicaid but too low to afford private coverage. Like Medicaid, it is jointly funded by the federal and state governments and run by each state within federal rules, so income limits and program names vary from state to state. CHIP covers well-child care, office visits, hospital care and prescription drugs, which would give this girl reliable access to her asthma controller medication and follow-up, the gaps that led to her emergency visit. Gaps in coverage and unaffordable cost-sharing are well-recognized drivers of missed controller therapy and avoidable emergency department use in children with asthma.

A. Medicare is a federal program based on age 65 or older, end-stage renal disease, ALS, or 24 months of Social Security disability; a child with asthma does not qualify. B. Medicaid is means-tested, and this family's income is above the state limit for children. D. An individual marketplace plan without subsidy would carry premiums the family cannot afford, and a child who is eligible for CHIP is generally not eligible for marketplace premium tax credits. E. Hospital charity care can reduce individual bills but does not provide ongoing insurance coverage for primary care, controller medications and prevention.

The positive likelihood ratio (LR+) is sensitivity divided by (1 minus specificity): how much more likely a positive test is in a patient with the disease than in one without it.

From the table, 60 children had catheter infection (51 + 9) and 140 did not (14 + 126). Sensitivity = 51/60 = 0.85. Specificity = 126/140 = 0.90, so the false-positive rate is 14/140 = 0.10. LR+ = 0.85/0.10 = 8.5. A likelihood ratio between 5 and 10 produces a moderate to large increase from pretest to posttest probability, so a DTP of at least 2 hours meaningfully supports the catheter as the source. Clinically, a positive DTP in a child with a high-risk organism such as S. aureus, Pseudomonas or Candida, or with sepsis or a tunnel infection, would prompt catheter removal, whereas coagulase-negative staphylococcal infection of a needed long-term catheter may be treated with systemic antibiotics plus an antibiotic lock.

A. 0.17 is the negative likelihood ratio: (1 minus sensitivity)/specificity = 0.15/0.90. It indicates how much a negative DTP lowers the probability of catheter infection. B. 0.78 is the positive predictive value, 51/65; unlike likelihood ratios, predictive values change with the prevalence of disease in the study population. C. 0.85 is the sensitivity, the proportion of children with catheter infection who had a positive DTP. D. 3.6 is the raw ratio of true positives to false positives (51/14); it ignores that the two groups differ in size (60 versus 140), which is why the likelihood ratio uses proportions (the true-positive rate and the false-positive rate) rather than counts.

The positive predictive value (PPV) is the probability that a person with a positive test has the outcome: true positives divided by all positives. Here, 8 of the 80 carriers developed SJS/TEN, so PPV = 8/80 = 10%. Put another way, 9 of every 10 carriers would have tolerated carbamazepine, yet testing is still recommended because the outcome is catastrophic (TEN mortality is high) and safe alternatives exist. The test is valuable mainly for its high sensitivity and very high negative predictive value: 8 of the 10 cases (80%) occurred in carriers, and only 2 of 920 noncarriers (0.2%) developed SJS/TEN. FDA labeling and pharmacogenomic guidelines (CPIC) recommend HLA-B*15:02 testing before carbamazepine in patients of Asian ancestry and avoiding carbamazepine, and preferably oxcarbazepine and phenytoin, in carriers. A negative result lowers but does not eliminate risk, so patients are still counseled to stop the drug and seek care at the first sign of rash, especially in the first few months.

A. 1% is the cumulative incidence of SJS/TEN in the whole cohort (10/1000), which is closer to the pretest probability than to the PPV. C. 80% is the sensitivity (8/10): the proportion of SJS/TEN cases who tested positive. D. 92.7% is the specificity (918/990): the proportion of people without SJS/TEN who tested negative. E. 99.8% is the negative predictive value (918/920): the probability that a noncarrier will not develop SJS/TEN.

A DNR order limits only cardiopulmonary resuscitation (and a DNI order only intubation); it is not 'do not treat.' Patients with DNR orders remain eligible for any indicated treatment, including palliative surgery such as hemiarthroplasty for pain control and mobility. The operating room, however, raises specific questions: anesthesia routinely involves airway management, vasopressors and treatment of arrhythmias that overlap with resuscitation. The American College of Surgeons and the American Society of Anesthesiologists therefore reject automatic suspension and call for required reconsideration: before surgery, the surgeon and anesthesiologist meet with the patient to decide whether the order will be fully suspended, kept in full, or modified (for example, intubation for anesthesia allowed but no chest compressions, or a goal-directed approach in which reversible anesthetic events are treated). The agreed plan, and when the original order resumes, is documented and communicated to the whole team.

A. Automatic suspension overrides the patient's autonomy without her input; it is the outdated practice that guidelines specifically reject. B. A DNR order does not make indicated, beneficial surgery inappropriate, and she has consented to the operation. D. A capacitated patient makes her own decisions; family are involved only if she wishes. E. Unilaterally keeping the order and dictating the anesthetic technique also skips the conversation; she may well accept intubation for the procedure while still declining CPR, and the anesthetic plan should follow her informed choice.