Diagnostic Red Flags: Epistemic Risk Factors
These are structural features of a diagnostic landscape, never accusations against individual clinicians or researchers. They represent situations where the evidence itself, or the conditions around it, warrant extra scrutiny, more questions, and sometimes a second perspective.
Surface 1-3 of the most relevant flags per user situation. Each flag should be accompanied by a concrete suggested action.
The Flags
1. Understudied condition
What it means: Limited published research. Few or no RCTs. Clinical guidelines based on small studies or expert consensus rather than robust evidence. Why it matters: When evidence is sparse, clinical practice varies more. What one doctor recommends may differ substantially from another, and neither may have strong evidence backing their approach. Suggested action: "Ask your clinician: How strong is the evidence behind the recommended treatment? Is there consensus in the clinical perspective on this, or do different doctors approach this differently?"
2. Common symptom, many possible causes
What it means: The primary symptom(s) could be explained by many different conditions across different body systems (e.g., fatigue, chronic pain, dizziness, GI distress). Why it matters: When a symptom has dozens of possible explanations, early diagnostic decisions can narrow the search prematurely. The first specialist you see may look within their domain and miss causes outside it. Suggested action: "Ask your clinician: What else could explain these symptoms that we haven't tested for yet? Should I see a specialist in a different area?"
3. Anchoring risk
What it means: An initial diagnosis or hypothesis may be influencing all subsequent clinical thinking, even when new evidence doesn't fully support it. Why it matters: Anchoring is one of the most common cognitive biases in medicine. Once a diagnosis is in your chart, subsequent clinicians often start from that assumption rather than re-evaluating from scratch. This is a feature of human cognition, not a failure of any individual clinician. Suggested action: "Consider asking a new clinician to evaluate your symptoms without seeing your previous diagnosis first. You can also ask your current team: If we started from scratch today, would we reach the same conclusion?"
4. Known demographic disparities in diagnosis
What it means: Research shows that diagnosis rates, time-to-diagnosis, or treatment quality differ by demographic group (sex, race, age, weight, etc.) for this condition. Why it matters: These disparities are well-documented in the medical literature. They don't mean any individual clinician is biased, but they mean the system as a whole has patterns that can affect your care. Being aware of them helps you advocate more effectively. Suggested action: "If relevant, mention awareness of these disparities to your clinician. Many clinicians actively work to counteract these patterns when they're made visible."
5. Exclusion diagnosis
What it means: This condition is diagnosed by ruling everything else out rather than by a definitive test. There's no single biomarker or scan that confirms it. Why it matters: Exclusion diagnoses are inherently uncertain. The diagnosis is only as good as the thoroughness of the workup that preceded it. If the differential was too narrow, the exclusion may be premature. Suggested action: "Ask your clinician: What conditions did we rule out to reach this diagnosis? Are there any we haven't tested for yet? What would change your confidence in this diagnosis?"
6. Symptom overlap with a more common condition
What it means: Your symptoms closely match a condition that is much more common than the one being considered (or vice versa — you may have been diagnosed with the common condition when a rarer one fits better). Why it matters: Clinicians are trained to consider common conditions first ("when you hear hoofbeats, think horses, not zebras"). This is usually good medicine, but it means rarer conditions with overlapping symptoms can be missed or delayed. Suggested action: "Ask: What's the less common explanation for these symptoms? Has it been considered and ruled out, or just not considered?"
7. Long average time-to-diagnosis
What it means: For this condition, the published literature or patient advocacy data shows that the average time from symptom onset to diagnosis is years, not months. Why it matters: A long time-to-diagnosis usually signals that the condition is hard to recognize, lacks a definitive test, or overlaps with other conditions. It means your experience of "not getting answers" may be typical of the condition itself, not a reflection of your credibility or the quality of your care. Suggested action: "Mention this pattern to your clinician. Ask whether they've seen cases of this condition and what the diagnostic pathway typically looks like."
8. Self-report as primary evidence
What it means: For this condition, the patient's description of their symptoms is the primary or sole diagnostic evidence. There's no definitive lab test, imaging finding, or biomarker. Why it matters: Self-report-based diagnosis depends on the clinician believing and carefully listening to the patient, and on the patient being able to clearly describe their experience. Communication quality matters enormously. Having a structured symptom inventory can make a significant difference. Suggested action: "Bring a written symptom log with frequency, severity, timing, and functional impact. This gives the clinician structured data rather than relying entirely on what you remember to say in the appointment."
9. Treatment is highly personalized
What it means: Individual treatment and responses to treatment are highly idiosyncratic. Treating this condition often requires working closely with a provider to determine whether a treatment plan is working. Different individuals may respond better or worse to the same treatment. There are known issues in treatment effect heterogeneity (the fact that the same treatment can work very differently in different people) depending on individual characteristics. Why it matters: Medical research is heavily weighted toward population averages and generalizing from samples to average effects. Personalized medicine is still a new and emerging frontier, and does not always get translated into individual healthcare. For many conditions, clinicians need to develop a personalized plan with the patient. This may mean that even though the research is valid, your individual response does not match the average result found in studies. Suggested action: "Ask your clinician: How will we know if this treatment is working for me specifically? What's the timeline for seeing a response, and what's the plan if I don't respond as expected? Are there factors about my situation that might predict how I'll respond?"
10. Metascience concerns
What it means: There are controversies in the research literature itself — conflicting meta-analyses, clinical guidelines that haven't been updated to reflect recent evidence, known issues with replication, or documented gaps between evidence and clinical practice. Why it matters: Sometimes the problem isn't that evidence is missing, but rather that the evidence base itself is contested, or that clinical practice hasn't caught up with the best available evidence. This may mean you may encounter conflicting recommendations, and it may be particularly important for patients who are understudied or in a context with less access to cutting-edge clinical care to be aware of these issues. Suggested action: "Ask your clinician: Are there recent studies or updated guidelines that might change the approach? If you get conflicting advice from different specialists, ask each one what evidence they're basing their recommendation on."
How to use these flags
These flags are tools for thinking, not diagnoses of your medical care. Their purpose is to help you:
- Ask sharper questions — each flag comes with specific questions you can bring to your clinician
- Calibrate your confidence — more flags active means more reason to seek additional perspectives
- Understand your situation — sometimes knowing "this condition is hard to diagnose" is itself valuable information that helps you be patient with the process while still advocating for yourself
Include the 1-3 most relevant flags in the structured artifact, with the associated suggested actions framed as discussion points for the medical team.