When Clinical Language Causes Harm: A Case Study in Endometriosis Risk

Clinical language is often treated as neutral. In practice, it shapes diagnostic pathways, escalation decisions, and how seriously patient-reported symptoms are taken. When language is inaccurate or biased, it can unintentionally embed risk into care delivery — particularly for conditions that disproportionately affect women.

During the COVID-19 pandemic, I identified several examples of clinical language related to endometriosis that misrepresented risk, normalized severe symptoms, or framed the condition in ways that delayed appropriate escalation. What followed offers a clear case study in how language itself can function as a patient safety issue.

The Problem: When Language Becomes a System Failure

Endometriosis affects millions of women, yet diagnosis is often delayed for years. One contributing factor is how symptoms are framed in clinical guidance and public-facing hospital materials.

In several instances, symptom descriptions:

• minimized the severity of pain
• normalized prolonged suffering
• embedded demographic assumptions that were clinically irrelevant
• subtly discouraged escalation or further evaluation

While these framings were not malicious, they reflected deeper system blind spots — particularly how women’s pain is interpreted and prioritized.

The risk was not theoretical. Language influences:

• how patients describe symptoms
• how clinicians triage and document concerns
• how quickly conditions are taken seriously

Left unexamined, these patterns compound over time.

The Intervention: Public Identification of Risk

In response, I publicly documented specific examples of problematic symptom framing and explained why the language created clinical and reputational exposure for hospitals.

The analysis focused on:

• how language shaped patient expectations
• how it signaled acceptable thresholds for pain
• how it could delay diagnosis and treatment

The intent was not critique for its own sake, but risk identification — highlighting how small wording choices could produce outsized downstream harm.

Institutional Response: Rapid Correction

Following this public analysis, multiple hospitals revised their online clinical language related to endometriosis. Changes were made quickly once leadership recognized that the framing posed risk — both to patient safety and institutional credibility.

Notably, Johns Hopkins Hospital updated language shortly after the issues were raised.

The speed of response demonstrated an important truth:
When institutions clearly see how language embeds risk, they act.

Key Insight for Healthcare Leaders

This case underscores a broader systems lesson:

Clinical language is not just educational.
It is operational.

Language shapes:

• escalation thresholds
• clinician bias
• patient self-advocacy
• institutional liability

When language fails to reflect risk accurately, systems fail patients — often quietly, and often repeatedly.

Why This Matters for Maternal Health

The same patterns that delay diagnosis in endometriosis — normalization of pain, dismissal of symptoms, and biased escalation — also appear in maternal care, particularly for Black women.

In maternal health, the stakes are even higher.

This is where my current advisory work is focused: helping hospitals identify where language, protocols, and escalation systems unintentionally reproduce risk — and redesigning those systems before preventable harm occurs.

Moving From Intent to Safety

Hospitals increasingly express commitment to equity and improved maternal outcomes. The challenge is translating that intent into systems that hold under pressure.

That work begins not with training alone, but with examining the foundations:

• language
• protocols
• accountability
• escalation pathways

When these are aligned, outcomes follow.

How to Move Forward

Preventable harm rarely begins with a single failure. It begins with small, accepted norms that go unchallenged.

Clinical language is one of those norms — and one of the fastest places institutions can intervene when safety and trust are at stake.