Key Takeaways
Demand evidence, because a doctor's charm and reputation can hide poor care.
Appearances mislead patients. Makary introduces two surgeons whose reputations point in opposite directions from their clinical strengths. Hodad is warm and adored by patients, yet colleagues witness poor judgment and avoidable harm. The Raptor is abrasive and feared, yet surgeons and nurses trust his technical skill. Patients cannot easily see the difference: they may mistake confidence, prestigious credentials, or gratitude for proof of good care.
Ask people who know. Hospital staff understand what patients cannot observe during an operation. When Makary asked employees where they would seek comfort care, many chose Hodad. For surgery, they chose the Raptor. The distinction is the book's warning: satisfaction matters, but it cannot reliably measure technical quality.
What's striking is that the story separates two dimensions of care that are often collapsed into a single rating: technical competence and the experience of being treated well. Consumer reviews can capture communication and courtesy, but they are weak instruments for judging an unseen procedure. Staff knowledge offers a valuable counterweight because workers observe repeated performance, though their views can also reflect workplace politics or personal loyalties. The practical lesson is not to replace one authority with another, but to triangulate: ask about experience, outcomes, teamwork, and how a clinician handles limits.
Publish adjusted outcomes, and hospitals start competing on survival rather than scenery.
Public numbers changed behavior. In 1989, New York required hospitals to report death rates for coronary bypass surgery, adjusted for how sick their patients were. The first results ranged from 1 percent to 18 percent. Hospitals with poor results faced lost business and scrutiny, then examined their programs, replaced unsafe practices, and sometimes stopped poorly performing surgeons from operating.
The results improved. One hospital reduced mortality from 9 percent to 2 percent after building a dedicated heart team. The state's heart-surgery death rate fell 41 percent during the program's first four years. Makary argues that public reporting redirected competition away from advertising and amenities toward results. The public data also gave hospitals a reason to listen to staff and fix local problems.
The New York example makes transparency more than a consumer convenience. It changes what organizations consider worth investing in. In principal-agent terms, patients cannot reward quality if quality is invisible; public measures make neglected performance consequential to hospital leaders. The evidence is especially persuasive because hospitals chose many of their own fixes rather than receiving a detailed government prescription. Still, a public score is only as fair as its definitions and risk adjustment. Poorly designed comparisons can encourage hospitals to avoid complex patients, so transparency requires careful measurement, not merely publication.
Choose the specialist who repeats your exact procedure, not the famous one.
Fame is not relevant experience. The Shah of Iran was treated by Michael DeBakey, a celebrated heart surgeon, for a spleen problem. DeBakey had little experience with spleen surgery and did not use a standard drain to guard against a pancreatic leak. The resulting infection was diagnosed late, and the Shah died. Makary's point is not that DeBakey lacked brilliance, but that expertise in one operation does not automatically transfer to another.
Ask for specific volume. Research cited in the book found that low-volume heart surgeons had mortality rates four times the state average. Ask how many times a doctor has performed the exact procedure, and seek another opinion when the answer is unclear. Volume is not a complete quality measure, but it is more relevant than titles or general prestige.
The case illustrates a basic principle of expertise: skill is often task-specific. A surgeon may have exceptional judgment and dexterity in one domain while being less practiced in another. The same logic applies to selecting a lawyer for a specialized case or a mechanic for a particular vehicle system. Volume alone, however, can mislead. A high count does not prove good outcomes, and the strongest choice combines relevant experience with team quality, complication data, and a clinician willing to refer when the case exceeds their expertise.
When payment rewards more treatment, restraint needs structural protection.
More care is not always better. Makary describes patients offered major operations when a less invasive option, or no intervention, might have served them better. He links such variation to incentives: doctors and hospitals often earn more from procedures, tests, and complications than from careful restraint. In one Texas comparison, similar cities had nearly double the health spending in one, with more medical activity rather than worse population health explaining much of the gap.
The incentives are concrete. Back surgery expanded sharply, while device makers paid some surgeons millions in royalties and consulting fees. Cancer drug markups and payments tied to specific products can also influence borderline recommendations. Makary contrasts these pressures with salaried models, which reduce the direct reward for doing more, while recognizing that doctors still need fair compensation.
This is a classic principal-agent problem: the professional advising the patient may also benefit financially from the recommended service. Disclosure helps, but it cannot fully neutralize a payment structure that rewards volume. Salary models can reduce that conflict, although they can create different pressures, such as doing less or seeing patients too quickly. The strongest safeguard is therefore layered: disclose financial ties, compare alternatives and outcomes, and make team review and referral routine. Incentives shape judgment without proving that every clinician who benefits is acting in bad faith.
Make it safe for nurses to interrupt the person holding the scalpel.
Teamwork is a safety measure. A confidential survey of staff at 60 hospitals found wide differences in teamwork and safety culture. In some hospitals, fewer than one in five employees reported good teamwork; elsewhere, nearly everyone did. The scores tracked with patient outcomes. One revealing finding: doctors rated teamwork with nurses at 88 out of 100, while nurses rated teamwork with doctors at 48.
Speaking up prevents harm. Makary describes a near miss in which a nurse questioned which side of a patient's chest was to be operated on just before surgery began. Her intervention stopped a wrong-site operation. Briefings that introduced the operating-room team and reviewed the procedure helped staff feel safer speaking up and improved awareness of the patient and plan.
The key resource here is not simply a checklist. It is permission to interrupt authority when something looks wrong. Organizational research describes this as psychological safety: the shared belief that raising a concern will not bring humiliation or retaliation. In medicine, that belief can determine whether a small discrepancy becomes a catastrophe. Brief introductions matter because they reduce anonymity and give colleagues a human connection. Yet surveys capture perceptions, not every hazard. Their value is greatest when leaders treat poor scores as prompts for investigation and repair, rather than as branding problems.
Prevent harm by repairing unsafe systems, while still confronting dangerous clinicians.
Errors have conditions. A study cited by Makary found that as many as 25 percent of hospitalized patients were harmed by medical mistakes. He recounts an error he made while exhausted and covering an intensive-care unit alone. A nurse corrected the ventilator settings before the patient suffered an immediate crisis. The episode left him with a question larger than personal guilt: why were so few clinicians responsible for so many critically ill patients?
Accountability must reach individuals too. Doctors at a conference unanimously knew a physician who should not be practicing, yet fear of retaliation often kept colleagues silent. The book describes weak oversight of impaired doctors and poor communication between hospitals and state boards. System failures need correction, but impairment and repeated unsafe practice also demand a credible reporting and review process.
A useful distinction is between blame and accountability. Blame asks who can be punished; accountability asks what failed, who had responsibility, and what prevents recurrence. Safety engineering uses layered defenses because people inevitably make mistakes, but those defenses do not excuse reckless or impaired practice. A mature system needs both learning and boundaries. The tension is real: punitive cultures suppress reporting, while consequence-free cultures leave patients exposed. Independent review, protected reporting, and fair investigation can help avoid both extremes.
Judge medical innovation by outcomes, and measure enough to prevent gaming.
New technology needs proof. Makary examines the surgical robot, a device costing more than $2 million. At the time he reviewed the evidence, studies had not convincingly shown better patient outcomes than standard minimally invasive surgery for most uses. Yet hospitals promoted robotic surgery as superior, and patients sometimes requested it because the marketing made advanced equipment sound synonymous with better care.
One metric can distort care. Hospitals rewarded for faster treatment of heart attacks sometimes performed angioplasties without enough attention to whether each patient needed one. Makary's proposed public dashboard therefore combines readmissions, complications, avoidable events, staff safety-culture scores, hospital volumes, and access to records. Independent collection and risk adjustment help make comparisons credible.
This section connects medical measurement to Goodhart's law: when a measure becomes a target, it can stop representing the thing it was meant to measure. Speed, infection rates, or procedure counts can each reward a narrow behavior while obscuring appropriateness and long-term results. A balanced dashboard reduces that vulnerability, but more indicators can also overwhelm consumers and invite statistical noise. The goal should be a small set of understandable measures backed by transparent methods, independent audits, and context about patient risk. Sophisticated equipment is not a substitute for evidence of benefit.
Turn consent into a real choice by asking about alternatives and recovery.
A signature is not understanding. Makary recalls patients being urged toward procedures without a clear account of risks, alternatives, or what recovery would feel like. One patient asking about life after heart surgery exposed how little a trainee knew beyond the expected hospital stay. In another case, an elderly woman who did not want cancer treatment was pushed toward tests and a biopsy, then suffered a serious complication.
Make the decision shared. Ask what happens without treatment, what less invasive options exist, how often the clinician performs each approach, and what recovery at home involves. The book also describes open notes, patient corrections to medical records, family participation in hospital care, and connecting prospective patients with people who have undergone the same procedure. These practices help patients compare medical advice with their own priorities.
Shared decision making is a response to unequal access to knowledge, not a demand that patients become their own doctors. A person can understand a risk in statistical terms yet still need to know what it means for cooking, work, pain, or independence. Peer conversations and accessible records add that lived context. The approach has limits: patients may be overwhelmed, and choices can feel like responsibility shifted onto them. Good consent therefore combines plain-language information, a clear professional recommendation, time for questions, and respect for a patient's informed refusal.
Analysis
Unaccountable is both a medical memoir and an argument about institutional design. Its narrative begins with a moral shock: the author sees a patient undergo an unwanted procedure and die, then enters a profession where colleagues privately recognize unsafe care but rarely challenge it. The stories of Hodad, the Raptor, the Shah, and patients harmed by unnecessary or poorly coordinated treatment make the system legible before the policy proposal arrives. The recurring contrast is between what patients can observe, such as bedside manner and reputation, and what they cannot, such as technical skill, complication patterns, or team communication.
The central claim is that transparency can change incentives where professional norms and internal oversight fail. New York's cardiac-surgery reporting supplies the book's strongest demonstration: public comparisons prompted hospitals to investigate and improve, while adjusted outcomes offered a practical alternative to prestige-based choice. The book also broadens accountability beyond outcome rankings. Staff surveys, independent data collection, procedure volume, open records, video, and checklists each reveal a different failure mode. Together, they suggest that quality is not a single score but a system of observable practices and results.
The argument is strongest when it treats transparency as infrastructure for learning and competition, not as public shaming. Its most important qualification is also implicit in its examples: raw metrics can mislead, and a poorly chosen target can encourage gaming or avoidance of difficult patients. Fair comparisons require risk adjustment, independent collection, and multiple measures. Moreover, patient choice cannot correct every failure when care is urgent, insurance limits options, or local access is scarce. Transparency is therefore necessary but insufficient. It must be paired with protected reporting, competent regulators, sensible payment, and leadership that responds to frontline concerns. The book's lasting contribution is to recast patient safety from a private professional matter into a public accountability question: if hospitals possess usable evidence about quality, patients should not have to choose in the dark.
Review Summary
Unaccountable receives high praise for exposing flaws in the healthcare system and advocating for transparency. Readers appreciate Makary's insider perspective, highlighting issues like medical errors, overtreatment, and profit-driven decisions. The book offers practical advice for patients, including questions to ask before procedures. While some criticize certain sections, most find it eye-opening and essential reading. Many express hope that increased transparency will lead to improved patient safety and reduced healthcare costs. The book's relevance persists, though some wish for an updated edition given its 2012 publication.
People Also Read
Glossary
Danger zones
Unsafe hospital unitsMakary's term for hospital departments or services where care is especially risky because of poor teamwork, weak staffing, outdated practices, or known patterns of bad outcomes. The danger may exist within an otherwise respected hospital, which is why hospital-wide reputation alone can obscure the quality of a particular unit.
Teamwork Culture Score
Staff-rated teamwork measureA score derived from confidential staff surveys that assess communication, coordination, willingness to raise safety concerns, and whether employees would accept care in their own unit. Makary describes scoring hospital-wide and department-level responses. The measure is intended to expose local workplace conditions that may be difficult for patients and administrators to see.
National Surgical Quality Improvement Program
Risk-adjusted surgical outcomes registryA national approach to measuring surgical results using standardized definitions of complications, patient risk factors, and independent data collection. Nurses review records and may contact patients to identify outcomes. Risk adjustment helps distinguish performance from differences in case complexity, making comparisons more credible to clinicians and hospitals.
Never events
Preventable serious errorsMedical events that should not occur when basic safety practices are followed, such as operating on the wrong patient or body part, or leaving an instrument inside a patient. Unlike many complications, these events are considered avoidable. Makary argues that hospitals should track and publicly report them to strengthen prevention.
Eat what you kill
Pay tied to individual volumeA compensation model in which a doctor's earnings rise with the amount of care or number of procedures personally delivered. The model can reward productivity, but it may also create financial pressure to recommend more treatment or retain patients who would benefit from referral. Makary contrasts it with fixed-salary arrangements.
Fred Flintstone care
Outdated medical practiceMakary's informal label for care that relies on an older method when safer or less invasive approaches are available. A doctor may continue using the familiar technique because newer methods were not learned or adopted. Patients can ask about alternatives and seek another opinion when a recommended approach seems unusually invasive.
Shared decision making
Patient and clinician decide togetherA model of care in which clinicians explain reasonable treatment options, benefits, risks, and uncertainties, then help patients choose in light of their goals and circumstances. The clinician still offers expertise and a recommendation. The patient is not simply asked to sign a form or comply with a directive without understanding the tradeoffs.
Comprehensive Unit-Based Safety Program (CUSP)
Local team safety-improvement programA unit-level approach that brings clinicians, nurses, and hospital leaders together to identify safety problems and work on practical solutions. Makary describes these groups as a way to make frontline knowledge actionable, improve communication, and address hazards within a specific specialty or hospital unit rather than relying only on broad, top-down rules.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.