Diagnosis Crisis: How Diagnostic Errors Shape Patient Outcomes and What Clinicians Do Differently

Episode snapshot

In this episode of Short Wave, NPR probes the reality of diagnostic errors in modern medicine. Through patient experiences and conversations with experts, the show examines how misdiagnoses happen, the systemic pressures that hinder accurate diagnosis, and the human strategies clinicians use to improve diagnostic accuracy. The dialogue highlights humility as a key trait for clinicians who aspire to better diagnose complex cases and to learn from feedback.

• Key insight: diagnostic errors are common and multifactorial, influenced by time constraints, fragmented records, and communication gaps.
• Key insight: master diagnosticians practice humility and continuous learning, seeking feedback and reflecting on practice.
• Key insight: humans remain central to diagnosis; AI can support but not replace the human elements of listening, empathy, and clinical judgment.
• Key insight: real improvements require better systems for feedback and accountability within healthcare.

Introduction and scope

The podcast centers on the Diagnosis Crisis, a discussion about diagnostic errors and their impact on patient outcomes. It weaves together a patient story, Diana Sejas, a pediatric neurologist, and Alexandra Sifferlin, an editor at the New York Times who wrote a book on the topic. The conversation situates diagnostic error as a pervasive problem in healthcare that stems from a combination of educational gaps, time scarcity, and systemic design flaws. By interviewing clinicians who specialize in diagnostic processes, the podcast illuminates how misdiagnosis can alter the trajectory of a patient’s life and health, sometimes with tragic consequences.

Understanding diagnostic error

The National Academies of Sciences, Engineering, and Medicine have highlighted that nearly every person will experience at least one diagnostic error in their lifetime. The episode discusses how an incorrect, delayed, or missed diagnosis can arise, emphasizing that diagnoses are not the result of a single moment of insight but the culmination of information gathering, analysis, and judgment under real-world constraints. The discussion also probes whether the problem is rooted in medical education, clinical workflow, or broader health-system structures that hamper thorough evaluation.

Different viewpoints on the diagnostic process

Gurpreet Dhaliwal, a professor at UC San Francisco, is described as a master diagnostician who rejects the title while embracing a philosophy of ongoing self-improvement. His ER work reveals a key truth: feedback loops are often incomplete, and clinicians may not routinely learn from misdiagnoses. Dhaliwal’s approach—reviewing recent cases, identifying where the reasoning went wrong, and using that knowledge to improve future practice—illustrates a practical form of accountability that many clinicians lack. The episode underscores that such feedback is essential to becoming better at diagnosis and that humility is not a weakness but a core component of good diagnostic practice.

Patient perspective and trauma-informed care

Diana Sejas’s account provides a visceral example of misdiagnosis. She describes how a neck mass was dismissed for years and only after extensive testing was it identified as a tumor. The resulting medical complications—stroke, paralysis, and long-term deficits—underscore the stakes of diagnostic delays and errors. The host and guests discuss how empathy, trauma-informed care, and inclusive communication can help restore trust and improve the diagnostic encounter for patients and families facing uncertainty.

AI, humanity, and the future of diagnostics

The conversation turns to artificial intelligence as a tool with potential benefits and clear limitations. While AI may surface rare diagnoses by sifting vast datasets, it cannot replace the human elements of history taking, physical examination, and the nuanced understanding of patient context. The guests agree that AI should augment rather than supplant clinician judgment, acknowledging the risk of algorithmic bias and the need for human oversight. Alexandra Sifferlin notes that AI may be especially helpful for rare diseases thanks to large data volumes, but the human-centered relationship remains essential.

Implications for practice and policy

The discussion closes with reflections on accountability and feedback mechanisms in healthcare. The participants emphasize the need for better systems that encourage clinicians to learn from diagnostic errors, create opportunities for collegial input, and support clinicians in delivering trauma-informed, patient-centered care. The take-home message is that improving diagnostic accuracy requires both personal humility and structural reform that fosters continuous learning within clinical teams.

Takeaways

• Diagnostic error is widespread and multifactorial, necessitating systemic improvements and enhanced feedback loops.
• Humility and continuous learning are hallmarks of effective diagnostic practice.
• Human clinicians remain central to diagnosis, with AI serving as a supportive tool rather than a replacement.
• Trauma-informed, patient-centered communication can improve diagnostic encounters and patient trust.