Aletho News

What It Actually Is

The phrase “defensive medicine” is not a critic’s label. It is the term physicians use themselves, in their own surveys, to describe diagnostic and treatment decisions made primarily to avoid legal liability rather than to benefit patients. The practice has been studied for more than four decades, appearing in peer-reviewed journals, in confidential surveys conducted by professional societies, and in geographic analyses showing that physicians in high-liability states order more diagnostic tests than physicians in states with malpractice tort reform.³

A 2005 study published in JAMA surveyed physicians in six high-risk specialties in Pennsylvania. Ninety-three percent reported practicing defensive medicine, and 43 percent reported using imaging technology in clinically unnecessary circumstances. The authors noted that defensive practice was reported across all specialties studied and was not confined to physicians with prior malpractice claims.⁴ The behavior described was not the conduct of a deviant minority. It was the operating culture of the specialties surveyed.

Researchers distinguish two registers. The first, positive defensive medicine, involves ordering tests, procedures, or consultations that the physician believes are unlikely to benefit the patient but that protect against accusations of inadequate diagnostic effort. The second, negative defensive medicine, involves declining to treat patients perceived as carrying higher litigation risk regardless of their clinical need. Both registers redirect clinical decisions away from the patient’s medical interest and toward the physician’s legal exposure.

The point is not the existence of defensive medicine but its normalization. Concealment from patients is not required because the practice is invisible by default. The patient cannot tell, sitting in the consulting room, whether a test was ordered to investigate their symptoms or to insulate the doctor against a hypothetical plaintiff’s attorney. The recommendation looks identical from the patient’s chair.

From the patient’s vantage point, what is visible is a competent professional, with medical training the patient does not have, who has determined that this particular test is required. What is invisible is whether that determination was made for reasons entirely indifferent to the patient’s medical situation.

The One-Way Ratchet

The legal system creates the conditions that produce defensive medicine, and it does so through a single asymmetry. Physicians face liability for under-diagnosis and almost none for over-diagnosis.

A man whose physician declines to order a PSA test, and who is later found to have advanced prostate cancer, has a malpractice claim. The claim may or may not succeed, but it is the kind of case malpractice attorneys take and insurance carriers settle. A man whose physician ordered the PSA test, who returned an elevated reading, who underwent a transrectal biopsy, who developed post-procedure sepsis from bowel bacteria driven into the bloodstream by the needle, who survived the sepsis but was left with persistent erectile dysfunction and urinary incontinence following a radical prostatectomy for a low-grade cancer that autopsy data suggests would have killed him only if he lived to ninety, that man has no claim. In most cases, he is grateful. His cancer was “caught early.” He was never shown the autopsy data on indolent prostate cancers, never told that 16 percent of American men receive this diagnosis but only 3 percent die from it, and is unaware that his cascade of harm originated in a screening decision medical evidence does not actually justify.⁵

The doctor who ordered the unnecessary test produced durable, measurable harm. The legal system records no exposure for that doctor. The doctor who did not order the test, and whose patient happened to be the rare one whose cancer was missed, faces a lawsuit regardless of whether ordering the test would have changed the outcome. From inside the legal architecture, the safest course is always to test more, because the asymmetry between under-diagnosis liability and over-diagnosis liability mandates it.

Studies of physician behavior confirm the prediction empirically. Doctors in high-liability states order significantly more diagnostic tests than doctors in states with malpractice reform. The variation tracks the legal climate, not the underlying disease burden.³ When Texas enacted comprehensive tort reform in 2003, defensive testing patterns in the state shifted measurably, although the magnitude and durability of the shift remain contested.³ The literature is consistent on the underlying point. Physician behavior is responsive to liability exposure, and the response is toward more testing rather than less.

This is the one-way ratchet, moving in a single direction over time. Settled malpractice cases raise the practical threshold of caution. New appellate decisions enlarge the set of investigations a prudent physician must order to satisfy the standard of care. The countervailing pressure that should exist from the harms produced by overtesting does not exist, because those harms generate no legal exposure for anyone.

From the patient’s vantage point, none of this is legible. The ratchet appears as escalating recommendations for tests, scans, and follow-up procedures. The patient assumes the escalation reflects an improving understanding of their personal medical risk. It does not; it reflects an evolving legal climate that has nothing to do with the patient’s biology.

The Standard of Care as Liability Shield

The medical profession refers to “the standard of care” as if it were a quality benchmark, a threshold below which clinical performance becomes inadequate. The standard of care in legal practice is something different. It is a defense position. Meeting the standard is what a malpractice defense attorney needs to establish to defend a doctor against a claim. The phrase describes legal adequacy, not clinical excellence.

The structural problem follows from how the standard is defined. The American Society of Anesthesiologists Closed Claims Project, the specialty’s primary instrument of self-examination for half a century, contains data that makes this explicit. When anesthesia care was judged by peer reviewers as appropriate, meeting the standard of care, payment was still made in more than 40 percent of malpractice claims. When care was judged substandard, payment was more likely but not certain.⁶

The standard of care, in other words, does not predict patient outcomes. It does not even reliably predict legal outcomes. What it predicts is the defensibility of a physician’s practice in a malpractice proceeding.

The circular logic closes off accountability at the definitional level. If the standard of care is what the majority of practitioners do, and the majority of practitioners do not obtain pre-operative cognitive baselines before general anesthesia in elderly patients, do not disclose post-operative cognitive dysfunction risk before consent, and do not select anesthetic agents based on the mechanistic evidence suggesting that some agents are worse than others for cognitive outcomes, then none of these omissions constitute substandard care. The standard is self-referentially defined by the mean of existing practice. The mean of existing practice cannot, by definition, be substandard.⁷

Harm that occurs in conformity with standard practice is not negligence. Negligence requires deviation from standard practice. Standard practice is what most practitioners do. What most practitioners do is whatever protects them most reliably from claims, which is whatever the standard of care currently is. The loop closes on itself.

The result is a body of harm that the accountability system cannot see by its own architecture. The Closed Claims Project, for all its methodological rigor, measures only what generates a claim. Acute, visible, legally contested events register: death, brain damage, esophageal intubation. The reduction in respiratory-related anesthetic deaths between 1985 and 2000 followed exactly this mechanism. The harms were acute and visible, pulse oximetry made them detectable in real time, malpractice exposure existed, the insurance system promoted the monitoring standard, and uptake was rapid.⁶ Chronic, diffuse, legally uncontested events do not register in any meaningful volume. Post-operative cognitive dysfunction, the slow cascade of overdiagnosis triggered by routine screening, the accumulation of pharmaceutical harm over decades: these generate no claims. They produce no signal in the accountability system the specialty uses to identify and correct its harms.

The Closed Claims architecture, the standard of care doctrine, and the malpractice insurance system together form a coherent legal-administrative apparatus whose function is to defend physician behavior against external challenge. Protecting patients from the iatrogenic consequences of defensive practice was never within the apparatus’s design brief, and it does not perform that protective function in practice.

The Cascade

What patients experience inside the defensive system is rarely a single unnecessary test. It is a sequence of tests, each triggered by the previous, each individually defensible, none of them collectively defensible.

Consider a fifty-year-old executive who opts in to a whole-body CT scan offered by a private screening clinic. The scan reveals a small lung nodule. The nodule is asymptomatic and almost certainly benign, since incidental findings of this kind appear on roughly 25 percent of all CT scans.⁸ The clinic recommends a follow-up CT in three months to assess for growth. The follow-up shows the nodule unchanged but identifies a small adrenal mass in the same imaging field. A PET scan is recommended to characterize both findings. The PET scan is negative for malignancy but introduces ambiguity about a mediastinal lymph node. A bronchoscopy is performed to sample tissue. The pathology returns benign scar tissue. The bronchoscopy is complicated by a pneumothorax requiring chest tube placement. The chest tube placement is uneventful. The executive returns to work three weeks later, having paid approximately $14,000 out of pocket, with no disease detected, no condition improved, and a measurable cumulative radiation exposure that did not exist before the cascade began.

The pattern is not exceptional. Incidentalomas trigger an average of $1,100 to $3,500 per patient in additional testing, with no associated improvement in mortality.⁸ Lung cancer screening trials found that one-third of patients had false-positive findings after two scans, and that one in fifteen underwent unnecessary invasive procedures as a result.⁸

The PSA cascade follows the same architecture in a different organ. An elevated PSA reading, which can result from prostatitis, benign prostatic hyperplasia, recent ejaculation, a urinary tract infection, or simply a larger-than-average prostate, triggers a transrectal biopsy. The biopsy punctures the bowel wall and introduces fecal bacteria into otherwise sterile tissue. Post-procedure sepsis is a documented complication, with antibiotic-resistant E. coli identified as the most commonly cultured organism, and hospitalization rates have risen alongside global antibiotic resistance. The 2022 GRAM Report published in The Lancet attributed approximately 1.3 million annual deaths worldwide to what it terms antimicrobial resistance, identifying E. coli as the most significant contributing organism.⁹ Tuncel and colleagues found that 41 percent of men reported erectile dysfunction at one month post-biopsy, with 15 percent still affected at six months.¹⁰ Cardiovascular events increase by a relative risk of 1.3 in the first year after diagnosis. Suicide risk rises by a relative risk of 2.6, with the highest risk in the first week.¹¹ These complications do not appear on the consent form.

The thyroid cascade follows the same architecture in a different specialty. A postpartum woman is screened for thyroid dysfunction. Her labs are abnormal during the transient hypothyroid phase of postpartum thyroiditis, a condition that typically resolves within twelve to eighteen months without intervention. She is started on levothyroxine and told she will need it indefinitely. The ultrasound performed to evaluate her thyroid identifies nodules, which are common and usually benign. The nodules require monitoring. Monitoring reveals minor changes. The changes prompt biopsy. The biopsy is indeterminate. A partial thyroidectomy follows. The woman who had a self-resolving postpartum condition is now missing part of her thyroid, on lifelong hormone replacement, and entered as a chronic patient in the disease registry.¹²

No single doctor along the cascade feels responsible for the cumulative outcome. Each one acted on the previous finding, documented their reasoning, and met the standard of care for their step in the sequence. Stopping the cascade at any point would have required a particular doctor to accept the legal risk of letting an uncertain finding go uninvestigated, and that doctor would be the patient’s doctor, not the patient. The asymmetry of where risk lands is the engine of the cascade. Legal risk is borne by the physician’s career, clinical risk by the patient’s body.

The cascade is the natural product of an accountability system that punishes physicians for missed diagnoses but not for downstream harms. It is what the system is structurally calibrated to produce. From the patient’s chair the cascade looks like a sequence of reasonable medical decisions; from the physician’s, the same sequence looks defensible at every step. The two perspectives never meet, and the gap between them is where the iatrogenic harm accumulates.

What the Doctor Will Not Say Out Loud

The defensive arrangement depends on a structural information asymmetry. A patient cannot make an informed decision about a recommended test if the test’s purposes, risks, and statistical performance are not disclosed. The published data are consistent across multiple settings and countries: this information is not disclosed.

Among American academic medical centers, only 15 percent routinely inform patients about radiation risks before CT scanning. By contrast, 84 percent of the same institutions discuss the risk of contrast reactions, which is a less significant long-term risk for most patients.¹³ Only 9 percent of surveyed sites inform patients about non-radiation alternatives to CT that might avoid radiation exposure entirely.¹³

The communication gap has been measured directly. In one emergency department study, 78 percent of physicians reported that they routinely discussed CT radiation risks with patients. Twenty percent of patients in the same encounters recalled being told.¹⁴ Both groups were describing the same conversations.

In 1999, Federman and colleagues published a study in Effective Clinical Practice examining whether informed consent actually occurred before PSA testing. Of the men who had received PSA tests, 31 percent were unaware their physicians had ordered the test. Of those who were aware, only 47 percent recalled any discussion of risks and benefits.¹⁵ Some men were tested without their knowledge. The study is more than twenty-five years old. Subsequent investigations show that the pattern has not changed. A qualitative analysis of men’s lived experience after PSA-driven diagnosis found that patients felt inadequately prepared and unsupported from the point of screening through biopsy and treatment decisions, and that the harms of overdiagnosis and overtreatment were rarely explained.¹⁶

The Australian and New Zealand College of Anaesthetists publishes a position statement on informed consent (document PS26(A)) that sets out what the specialty formally requires of its practitioners. The standard is specific: patients must be provided with information that a reasonable person in their position might wish to know and to which they might attach significance. The document includes the threshold that risks should be disclosed when “an adverse outcome is rare but the detriment severe, and an adverse outcome common but the detriment slight,” states explicitly that a surgical consent form does not constitute informed consent for anesthesia, and notes that time pressure does not eliminate the requirement for genuine consent.¹⁷

Applied to a documented condition such as post-operative cognitive dysfunction in elderly surgical patients, measurable in approximately one in four at one week after general anesthesia and one in ten still impaired at three months, with no known treatment, the ANZCA standard produces an unambiguous conclusion. The risk satisfies both disclosure thresholds, the detriment is significant, and the standard requires that the disclosure be made. Disclosure does not occur in clinical practice in the overwhelming majority of cases.¹⁸

The disclosure standard exists in writing. The disclosure itself does not occur in the consulting room.

The reason is not that physicians lack knowledge of these risks. The relevant studies appear in journals to which the practicing physicians subscribe and which they cite in their own work. The information has accumulated for more than two decades in some cases, more than seventy years in others. The reason disclosure does not occur is simpler than ignorance. Disclosing risks creates two adverse consequences for the physician. First, some patients decline the recommended intervention, which reduces the procedure rate and exposes the physician to liability if the rare missed diagnosis occurs. Second, patients who proceed and experience the disclosed complication have documented grounds for claiming they were misinformed about the relevant probabilities if the disclosure was incomplete. The path of least legal resistance is a generic, time-pressured consent process that satisfies the formal documentation requirement without producing a genuinely informed decision.

The consent form records that the patient was informed; the evidence shows they were not. Both statements are simultaneously true within the legal and clinical architecture as currently structured.

The Trapped Clinician

A complete account of defensive medicine has to address what is happening to the doctor inside the structure, because reducing the practice to physician greed or laziness misses what the evidence shows. Most doctors who order defensive tests are not behaving unethically by the standards of the system they trained in. They are responding rationally to the incentive structure they are embedded in.

That structure has several layers. The first is education. American medical training increasingly emphasizes adherence to clinical practice guidelines, many of which are written by panels with substantial financial ties to the manufacturers of recommended products. A widely cited analysis published in the BMJ found that the majority of panelists on the National Cholesterol Education Program guidelines had current or recent financial ties to pharmaceutical companies whose products the guidelines would mandate.¹⁹ Roughly two-thirds of American medical school department chairs hold financial relationships with industry of some kind.²⁰ A physician trained to follow guidelines that emerged from such panels is trained to order more tests, prescribe earlier, and intervene more aggressively, and to consider this evidence-based medicine.

The second layer is employment. Physician practice in the United States has shifted dramatically from independent practice to employment by health systems. The employed physician is subject to productivity metrics that reward volume: number of patient encounters per day, number of procedures per quarter, billing per visit. A physician who orders fewer tests, sees patients longer, or recommends conservative observation is not rewarded, and in many systems is penalized. Performance reviews and bonuses turn on whether the physician’s practice patterns align with institutional throughput targets, which are set to amortize the cost of expensive imaging equipment and staffed procedural suites.²¹

The third layer is time. A primary care visit in the American system is typically scheduled for fifteen to twenty minutes. That visit is meant to address whatever the patient came in for, plus all the screening reminders generated by the electronic medical record: colonoscopy overdue, bone density scan overdue, mammogram overdue, annual physical overdue. The path through this consultation that minimizes liability and maximizes billing is to order the indicated screenings, prescribe what the algorithm suggests, and move to the next patient. The path that protects the patient requires more. It requires taking time to ask whether each screening is appropriate for the particular person sitting in the chair, discussing harms as well as benefits, and documenting an informed refusal where indicated. That work adds twenty to thirty minutes of unbillable time to a visit that the system already says is too long.

The fourth layer is the broader commercial logic. A 2018 Goldman Sachs analyst report titled The Genome Revolution, written by Salveen Richter, included a section addressing the financial implications of curative therapies. The question, posed in writing within an investment research document published by one of the largest investment banks in the world, was direct: “Is curing patients a sustainable business model?” The report’s analysis indicated that one-time cures produce short revenue arcs, while chronic disease management produces long ones.²² The economic logic surrounding the physician, including the institutions that employ them, the products they prescribe, and the diagnostic technologies they order, is calibrated to manage disease rather than to cure it. The defensive medicine arrangement is consistent with this calibration. A patient subjected to a cascade of testing and treatment generates recurring revenue. A patient told that observation is reasonable does not.

None of this absolves the doctor. The structural pressures are real and explain why the system produces the outcomes it does, but the patient sitting across the desk is still being subjected to risks they have not been told about. A doctor who orders a defensively motivated CT scan, who fails to disclose its radiation risks, who refers an incidentaloma for follow-up imaging that initiates a cascade, is participating in iatrogenic harm whether they intend to or not. The structural account explains how the participation occurs without absolving the physician of having participated.

The honest version of the consultation would be something close to the following. I am going to order this test primarily because I will face legal exposure if I do not, and missing the rare cancer it might detect would be a career-ending event for me. I have no good way to predict whether the test will produce useful information in your case. If it produces a false positive, I will refer you for further investigation, and the cascade that follows may harm you. You can decline the test, and I will document your decision.

That consultation does not take place. Conducting it would expose the arrangement as it actually operates, and the arrangement depends on the patient’s continued belief that the recommendation reflects clinical necessity rather than legal self-preservation.

There is a question every doctor reading this should be willing to ask themselves before their next clinic. Of the tests, procedures, prescriptions, and referrals I will recommend today, how many would I recommend if there were no liability exposure attached to declining to recommend them? The honest answer, for most physicians in most American practices, would substantially shrink the volume of recommended interventions. That difference is the iatrogenic harm. The volume of intervention that exceeds what would be recommended on purely medical grounds is the volume the patient is being asked to absorb on the doctor’s behalf.

What to Do in the Consulting Room

The structural problem is not solved at the individual level. A single patient cannot reform American medical training, the employment structure of physicians, or the asymmetry of the legal system. What a single patient can do is recognize the arrangement, refuse to participate in the parts of it that produce harm, and document the refusal.

Three questions interrupt almost any cascade at the point of entry.

The first is what happens if we wait? Most diagnostic decisions are not emergent. The doctor will typically describe a worst case, because the legal exposure forces them to describe a worst case, but the relevant question is what happens in the average case if observation is chosen over intervention. For most screening tests, the answer is that nothing happens. The natural history of the condition, in the absence of detection, would not have caused symptoms or harm within the patient’s lifetime. The cascade was triggered by the test, not by the underlying disease.

The second is what are the alternatives? Most diagnostic algorithms admit multiple paths. Ultrasound and MRI can substitute for CT in a wide range of clinical situations; active surveillance is a documented option in many cancers where biopsy is reflexively recommended; lifestyle modification has demonstrated effects in conditions for which pharmaceutical intervention is the default. The default path is rarely the only path. It is the path of lowest legal exposure for the physician, which is not the same thing as the path of lowest harm for the patient.

The third is is this required, or recommended? “Required” is almost never the correct answer. The patient who receives a flat recommendation often assumes the test is mandatory. It is not. It is recommended. The distinction matters because a recommendation can be declined.

Declining is the operative skill. Pressure converges on the patient from multiple sources to maximize the rate of test acceptance: electronic medical record reminders for “overdue” screenings, clinic staff who raise the issue at every visit, automated reminder letters, partner and family pressure, the doctor’s recommendation itself, and the patient’s own fear of missing something serious. None of these constitutes medical necessity. They constitute the operating environment of a system that profits from compliance.

Specific phrases work. I have reviewed the evidence on this test and decided the harms outweigh the benefits for someone in my situation. Please note my informed refusal in my record. I understand you recommend this. Can you tell me the number needed to screen and the overdiagnosis rate? I would like to make an informed decision. *I am not declining medical care. I am declining this specific test based on the published evidence. I would like to discuss what, if anything, I should monitor instead.*²³

The documented refusal protects the patient and, by an arrangement the system rarely advertises, also protects the doctor. A patient who has formally declined a recommended test, with the refusal documented in the medical record alongside the doctor’s recommendation and the discussion of evidence, has produced exactly the consent process that the system claims to require. The doctor’s liability exposure for the refused test is substantially reduced because the patient now owns the decision. The patient’s exposure to the iatrogenic harms of the unnecessary test is eliminated because the test does not occur.

This outcome is available to any patient who insists on it. The system does not offer it spontaneously because it is the system’s least profitable configuration.

Closing

The trauma center study identified 416 patients out of 1,097 who received CT scans ordered for reasons that had little to do with their medical condition. Each of those patients sat through the procedure believing the scan had been ordered because the doctor’s clinical judgment required it. Each absorbed, on average, 8.8 millisieverts of ionizing radiation. A small percentage of them will develop cancers from that exposure decades from now, and by that point the link between cause and effect will be unrecoverable. None of the doctors who ordered the scans will be sued, none of the institutions that performed them will be sued, and the eventual cancer patients will not know which encounter caused the disease. If they did know, the standard-of-care defense would close the case before it began.

What is happening in the American consulting room, and increasingly in clinical encounters across every country that has adopted the American model, is a form of risk transfer. The legal risk that the physician would otherwise carry is transferred to the patient in the form of unnecessary radiation, biopsies, cascades of investigation, lifelong prescriptions, and the diminished autonomy of being managed rather than cared for. The transfer is invisible from the patient’s chair because it is denominated in a currency the patient cannot read.

The standard of care functions as a liability defense rather than a clinical benchmark, and the consent form that accompanies it operates as an administrative shield rather than a record of genuine disclosure. A defensive test recommended on these foundations is calibrated to the doctor’s legal exposure first and to the patient’s clinical situation second. The cascade that follows a positive result is not the pursuit of the patient’s recovery; it is the completion of a documentation chain demonstrating that the system did everything it was supposed to do.

You can decline. The right to decline is the last patch of ground in the consulting room that the system has not yet enclosed. The document the doctor wants you to sign is consent to a process calibrated for everyone except you. The document you can produce, a written and dated refusal of a specific recommended intervention with your reasoning attached, is the only piece of paper in the room that says what it claims to say.

How I’d Explain This to a Six-Year-Old

Pretend your school has a rule. If you fall down at recess, even just a little bit, the teacher has to send you to the nurse. Not because the teacher thinks you are hurt. The teacher knows you are fine. But if the teacher doesn’t send you and your knee turns out to be a bit bruised, your mom might call the principal, and the principal might get cross with the teacher.

So you go to the nurse. The nurse looks at your knee. To be safe, the nurse takes a picture of inside your leg with a special machine. The machine uses a tiny bit of bad stuff that goes into your body. Just a little. Not enough to hurt today. But every time the machine takes a picture of you, a little more bad stuff goes in, and after lots of pictures the little bits add up.

The picture shows your knee is fine. The nurse already knew that. But the picture also shows a tiny spot somewhere else that is probably nothing. To be safe, the nurse sends you for more pictures. The more pictures show another small thing. Now you are scared. Your mom is scared. Nobody can stop, because if they stop and the small thing turns out to be something, the nurse and the doctor and the school could get into trouble.

You came in with a scraped knee. You leave with three pictures of your insides, a worried mom, and a bit of the bad stuff inside you. Nothing was wrong. Nothing got better. All the grown-ups feel safer, because they did everything they were supposed to do.

That is what happens to grown-ups at the doctor every day. There is a name for it. The name is defensive medicine. It means the people taking care of you are mostly trying not to get into trouble themselves, and those two things, looking after you and not getting into trouble, almost never point the same way.

Here is the thing nobody tells you. You can say no. When the grown-up says “I think we should do this test,” you are allowed to ask, “what happens if we wait?” You are allowed to say, “no thank you, not today.” You are allowed to leave with a scraped knee and nothing else.

References

1. The prospective study at a Level I trauma center identifying 38 percent (416/1,097) of CT scans as defensively motivated, with associated radiation exposure of 8.8 mSv per patient and a 2.2 percent management-change rate, is documented in Unbekoming, The Screening Trap, v1.0, 2026, chapter on CT scan overuse. The $501 million annual estimate for defensive CT imaging in the United States is drawn from the same source, which synthesizes figures from Mello MM, Chandra A, Gawande AA, Studdert DM. “National Costs of the Medical Liability System.” Health Affairs 2010;29(9):1569–1577.
2. Smith-Bindman R, Miglioretti DL, Lacson R, et al. “Projected Lifetime Cancer Risks From Current Computed Tomography Imaging.” JAMA Internal Medicine 2025. The study projects approximately 103,000 future cancers from the 93 million CT scans performed in the United States in 2023.
3. On geographic variation in physician test-ordering tracking liability climate: Kessler DP, McClellan M. “Do Doctors Practice Defensive Medicine?” Quarterly Journal of Economics 1996;111(2):353–390; Kessler DP, McClellan M. “How Liability Law Affects Medical Productivity.” Journal of Health Economics 2002;21(6):931–955. On the specific effect of Texas tort reform (enacted 2003) and subsequent damage cap analysis: Paik M, Black B, Hyman DA. “The Receding Tide of Medical Malpractice Litigation: Part 2, Effect of Damage Caps.” Journal of Empirical Legal Studies 2013;10(4):639–669.
4. Studdert DM, Mello MM, Sage WM, et al. “Defensive Medicine Among High-Risk Specialist Physicians in a Volatile Malpractice Environment.” JAMA 2005;293(21):2609–2617. The study surveyed physicians in six high-risk specialties in Pennsylvania; 93 percent reported defensive practice, and 43 percent reported imaging used in clinically unnecessary circumstances.
5. National Cancer Institute, “Prostate-Specific Antigen (PSA) Test,” cancer.gov. The 16 percent lifetime probability of diagnosis and 3 percent probability of mortality figures are the standard reported values. Autopsy data on indolent prostate cancers: Bell KJL et al. “Prevalence of incidental prostate cancer: A systematic review of autopsy studies.” International Journal of Cancer 2015;137(7):1749–1757.
6. American Society of Anesthesiologists Closed Claims Project data summarized in Unbekoming, Before You Go Under, v1.0, 2026, chapter on the consent gap and the standard of care. Original source: Domino KB, Posner KL, Caplan RA, Cheney FW. “Awareness during anesthesia: a closed claims analysis.” Anesthesiology 1999;90(4):1053–1061. Additional Closed Claims data on standard-of-care judgments and payment patterns from the ASA Closed Claims Project as documented through the early 2000s.
7. The circular logic of standard-of-care doctrine and the self-referential definition of negligence-by-deviation is analyzed in detail in Unbekoming, Before You Go Under, v1.0, 2026, chapter on the closed claims ceiling. The argument applies generally across surgical specialties.
8. Incidentaloma rates and associated downstream costs documented in Unbekoming, The Screening Trap, v1.0, 2026, drawing on Berland LL, Silverman SG, Gore RM, et al. “Managing incidental findings on abdominal CT: white paper of the ACR incidental findings committee.” Journal of the American College of Radiology 2010;7(10):754–773. Lung cancer screening false-positive rates: Croswell JM et al. “Cumulative incidence of false-positive results in repeated, multimodal cancer screening.” Annals of Family Medicine 2009;7(3):212–222.
9. GBD 2019 Antimicrobial Resistance Collaborators. “Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis.” The Lancet 2022;399(10325):629–655. The report attributes approximately 1.3 million deaths directly and 5 million deaths in association with antibiotic-resistant bacterial conditions.
10. Tuncel A, Toklu H, Belenli O, et al. “The impact of transrectal prostate needle biopsy on sexuality in men and their female partners.” Urology 2008;71(6):1128–1131.
11. Fang F, Keating NL, Mucci LA, et al. “Immediate risk of suicide and cardiovascular death after a prostate cancer diagnosis.” Journal of the National Cancer Institute 2010;102(5):307–314. Discussed in Unbekoming, The Screening Trap, v1.0, 2026, and in Unbekoming, The PSA Trap, 2026.
12. Postpartum thyroid cascade documented in Unbekoming, Medicalized Motherhood, edition 1.0, 2026, chapter on postpartum screening. See also Stagnaro-Green A et al. “Guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and postpartum.” Thyroid 2011;21(10):1081–1125, for the conventional clinical pathway.
13. Lee CI et al. “Diagnostic CT scans: assessment of patient, physician, and radiologist awareness of radiation dose and possible risks.” Radiology 2004;231(2):393–398. The 15 percent / 84 percent and 9 percent figures are widely cited and appear in subsequent surveys of academic medical centers.
14. Caoili EM et al. “Patients’ knowledge and perception of CT scan radiation risks.” American Journal of Roentgenology 2014;202(3):W232–W237. The 78 percent / 20 percent physician-patient recall gap is documented in emergency department settings and is broadly consistent across studies of CT consent.
15. Federman DG, Goyal S, Kamina A, et al. “Informed consent for PSA screening: does it happen?” Effective Clinical Practice 1999;2(4):152–157.
16. Hersch JK, Nickel BL, Ghane A, et al. “Resisting recommended treatment for prostate cancer: a qualitative analysis of the lived experience of possible overdiagnosis.” BMJ Open 2019;9(5):e026960.
17. Australian and New Zealand College of Anaesthetists. PS26(A) Position Statement on Informed Consent for Anaesthesia or Sedation. Melbourne: ANZCA; 2021. The thresholds and standards quoted are from the publicly available document.
18. Post-operative cognitive dysfunction (POCD) prevalence figures and the disclosure gap are documented in Unbekoming, Before You Go Under, v1.0, 2026, drawing on Li LY, Staffaroni AM, Whitlock EL. “Subjective Cognitive Complaints and Anecdotal Descriptions of Postoperative Cognitive Decline: Missing Pieces of the Postoperative Neurocognitive Disorder Puzzle.” Advances in Anesthesia 2024;42(1):27–40; and Moller JT et al. “Long-term postoperative cognitive dysfunction in the elderly: ISPOCD1 study.” The Lancet 1998;351(9106):857–861.
19. Lenzer J. “Majority of panelists on cholesterol guidelines have current or recent ties to drug industry.” BMJ 2004;328(7452):8. See also Abramson J, Wright JM. “Are lipid-lowering guidelines evidence-based?” The Lancet 2007;369(9557):168–169.
20. Campbell EG, Weissman JS, Ehringhaus S, et al. “Institutional academic-industry relationships.” JAMA 2007;298(15):1779–1786. The two-thirds figure for medical school department chairs holding industry relationships is widely cited and is documented in Unbekoming, The Architecture of Deception, v1.0, 2026.
21. Casalino LP et al. “External incentives, information technology, and organized processes to improve health care quality for patients with chronic diseases.” JAMA 2003;289(4):434–441. The broader literature on the shift toward physician employment and productivity-based compensation in American health systems is extensive.
22. Goldman Sachs Equity Research, Salveen Richter. The Genome Revolution. Industry note, April 2018. The “sustainable business model” question and the report’s analysis of revenue arcs for curative therapy are widely cited; the document was subsequently the subject of multiple public commentaries. Discussed in detail in Unbekoming, Chronic Conditions, v1.0, 2026.
23. Specific patient phrases for documented refusal adapted from the patient toolkit in Unbekoming, The Screening Trap, v1.0, 2026.