The elephant in the room: What medicine misses when care becomes fragmented

After more than 12,000 endoscopies, I have learned that some of the most consequential decisions we make as gastroenterologists occur before we ever pick up an endoscope.

Before gastroenterology fellowship and before we learn to perform a colonoscopy, we build a foundation in internal medicine. Over three years, we develop the scientific knowledge, technical skills and clinical judgment to determine not only how to perform a procedure, but whether, when, and for whom it is appropriate.

Through practice, we learn that informed consent requires more than a signature, that doing good and avoiding harm are distinct responsibilities, and that meaningful discussions of procedural risk depend on understanding the evidence behind those estimates.

We learn to ask who was studied, which outcomes were measured, under what conditions, and whether the findings apply to the patient before us. Yet as colorectal cancer screening expanded into high-volume public-health pathways, the systems designed to deliver it did not always preserve the individualized clinical judgment needed to apply that evidence appropriately.

As prevention expanded, did we preserve individualized clinical judgment or simply become better at moving patients through an ever-evolving system?

When the moving parts changed

Internal medicine teaches us to identify disease, understand a patient’s physiologic limitations, optimize what we can and communicate risk. Building on that, gastroenterology similarly favors thoughtful pre-procedure assessment based on the patient’s history, examination, medications and procedural risk rather than routine testing.

These principles are not in conflict. The difficulty is that the systems surrounding the patient did not always evolve together. As colorectal cancer screening expanded, referrals, bowel preparation, authorizations, procedures and checklists became increasingly standardized. Each step serves a purpose, but completing every step does not necessarily mean the underlying clinical question has been answered.

The pathway can be complete. The clinical assessment may not be.

And the patient cannot be standardized.

When risk travels

“Cleared” once reflected the conclusion of an individualized assessment; yet too often, it is used as a substitute for the assessment itself.

I think of this as the Risk-Transfer Paradox: when fewer clinical checkpoints are equated with better access, the clinician who identifies uncertainty can seem to slow the system down, while the clinician who accepts that uncertainty allows it to keep moving.

The risk, however, has not disappeared. It has simply moved downstream.

Eventually, it reaches the gastroenterologist, anesthesiologist, or endoscopy unit—or the patient who arrives after bowel preparation and fasting with an unresolved concern that may change whether, when or where the procedure should occur. Sometimes postponing the procedure is the right decision. But by then, the patient may already have completed dietary restrictions, bowel preparation, fasting, missed work and travel arrangements. A justified cancellation may protect the patient, but it does not complete the care they came to receive.

Nor should the answer be to restore indiscriminate testing or require every patient to obtain another signature. The goal is to identify meaningful uncertainty earlier, address it and preserve a timely path to the procedure.

Each observation is real, but none captures the whole

In times of uncertainty, I am reminded of the parable The Blind Men and the Elephant. Several men encounter an elephant, each touching a different part. One feels the trunk and describes a snake. Another feels a leg and describes a tree. A third touches its side and describes a wall.

Each observation is real, but none captures the whole.

In medicine, the referring physician, gastroenterologist, anesthesiologist, procedural team and patient may each see a different part of the same clinical picture. No single perspective is sufficient on its own.

The blindfold is uncertainty. Recognizing that we are wearing it requires humility.

Clinical judgment, experience, guidelines, data, patient testimony and artificial intelligence can all inform our understanding, but none eliminates uncertainty. The work is to bring these perspectives together, test them against the evidence and remain willing to revise our conclusions.

Consider a familiar exchange: “Do you have any medical problems?” “No.” Does that mean none are present, or simply that none have yet been identified?

We should ask what an answer truly establishes before using it as the basis for a clinical decision.

Different perspectives will not always carry equal weight, nor should they. The goal is to ensure that relevant information enters the analysis before we draw a conclusion. A different perspective may change our conclusion substantially, barely move it or fail to survive scrutiny.

Responsible judgment requires that our model remain capable of reassessment.

When someone identifies a missing piece, does the system allow that information to change what happens next?

Beyond representation

This National Latino Physician Day (Oct. 1; #NLPD2026) and National Hispanic Heritage Month, our call to action should extend beyond representation. We should ask not only whose perspectives are missing, but whether those perspectives are allowed to shape what medicine sees, teaches, values and does.

Presence must have agency and consequence.

Medical education should teach us not only how to defend what we know, but also how to recognize when another perspective requires us to reconsider what we thought we knew. It was never only “include us,” but “let what we reveal change you.”

As professional and ethical physicians, our goal should be to bring scientific evidence, professional obligations, clinical judgment, patient experience and uncertainty into conversation—not to make every perspective equivalent, but to determine what each element contributes to the whole.

But bringing different perspectives into the room is only the beginning. For another perspective to change what we see, we must first accept that our own may be incomplete.

Humility allows me to admit: I may not have the whole picture. But trust allows me to consider: There may be something in your observation that belongs in mine. I cannot see what you see. But I trust enough to ask whether what you see should change what I think I know.

And in that relationship, a more complete picture becomes visible.

Related reading:

Reflections on the impact of the Latinx physician shortage in the US

Improving health care for Hispanic patients