A medical record should remember what clinicians learned—and what they were still trying to find out.
Imagine opening your medical record and finding a condition nobody ever told you that you had.
You remember a doctor mentioning it as a possibility. There were tests, a follow-up appointment, and eventually a different explanation. You thought the first possibility had been set aside.
Yet there it is, listed under your medical history.
At the next appointment, someone asks how long you have had it. You explain that you aren't sure you ever did. Now the visit contains a disagreement between you and your chart.
The chart looks certain. You sound uncertain.
That is an uncomfortable reversal. The uncertainty may have belonged to the original diagnosis all along.
The patient who never had the disease
This kind of persistence is more than a hypothetical concern. In a 2012 case published by AHRQ's Patient Safety Network, an emergency physician found a history of pulmonary embolism in a patient's record. The patient denied ever having one.
Looking backward, the physician discovered that “PE” had originally meant physical examination. The abbreviation had apparently been misplaced and misinterpreted, then carried forward as a history of pulmonary embolism. The error helped prompt a CT order. The account reports no resulting harm, and the record was corrected.
One case cannot establish how often this happens. It does expose a disturbing possibility: later documentation can appear to corroborate an earlier statement while merely repeating it.
Five notes do not necessarily represent five independent judgments. They may represent one judgment copied four times.
Repetition can make a medical history look better established without adding any evidence.
How a possibility loses its question mark
An obvious transcription error is only one route. A more subtle problem arises when the condition is plausible, but its degree of certainty becomes unclear.
Consider this hypothetical sequence:
First assessment“Symptoms could be consistent with condition X. Further evaluation needed.”
Later summary“Being evaluated for condition X.”
Subsequent history“History of condition X.”
Each version is shorter. The last is also a different claim.
The first describes a possibility. The second describes an investigation. The third presents an established past diagnosis. Without the original context, a reader cannot tell whether the certainty increased because evidence arrived or because qualifications disappeared.
Clinical summaries are necessary. Nobody can reread an entire chart before every decision. The difficult design question is what a summary must preserve. A diagnosis's name may be less useful than it appears if the reader cannot see whether it was confirmed, suspected, excluded, or simply never revisited.
AHRQ's 2024 discussion of diagnostic documentation identifies uncertainty as a challenge in medical records and describes opportunities including diagnostic reassessment and patient review of recorded problems. The concern here follows from that challenge: certainty should not increase merely because information has been compressed.
A diagnosis changes what comes next
Diagnoses are indispensable. They organize knowledge, make treatment possible, and allow one clinician to learn from another's work. A well-supported diagnosis can save a patient from having to start over at every appointment.
But a diagnosis is also an instruction about where to look.
If a clinician believes a symptom already has an explanation, the next question may concern management rather than discovery. That can be appropriate. Reopening a securely established diagnosis at every encounter would waste time and could expose patients to unnecessary investigations.
The danger arises when an uncertain explanation receives the privileges of an established one. New information may be interpreted within an account that has not earned that confidence.
Imagine a patient whose earlier episodes of breathlessness were attributed to anxiety. That history is relevant. It should not, by itself, settle the explanation for a later episode with a different pattern. Equally, the possibility of a new explanation does not establish that the earlier assessment was wrong.
The clinical task is to decide which features belong to the familiar pattern and which require a fresh look. A label should help with that comparison. It should not make the comparison unnecessary.
The right kind of uncertainty
“We don't know yet” can sound like an absence of care. It may instead be an accurate description of the evidence.
The difference depends on what accompanies it.
Uncertainty without a plan leaves the patient waiting. Uncertainty with a plan identifies the leading explanation, what remains unresolved, and what should prompt reconsideration.
A working diagnosis can be useful before it is certain. Medicine often requires action while information remains incomplete. The obligation is to make clear what kind of claim is being made, rather than withholding every judgment until doubt disappears.
This matters to the patient as well as the next clinician. “This is our current explanation, and here is what would make us revisit it” gives someone a different understanding from “This is what you have.”
Changing that explanation later need not mean the earlier clinician was incompetent. An assessment can be reasonable on the information available and still require revision when the information changes.
A record that preserves that development shows clinical reasoning over time. A record that flattens every stage into certainty makes revision harder to understand.
Three questions worth bringing to a visit
Patients should not have to audit their own records to receive safe care. But if you encounter an unfamiliar or doubtful diagnosis in your chart, a focused conversation can help clarify what it means.
- “Is this confirmed, a working diagnosis, or something that has been ruled out?” Ask what status the entry is intended to convey.
- “What evidence supports it?” There may be a result or specialist assessment you have not seen—or a reason to review the original entry.
- “Does it still describe my situation, and can the record reflect that?” The appropriate change may be clarification, an updated status, or a correction rather than deletion.
A chart entry alone is not a reason to change treatment. The purpose of these questions is to establish a shared understanding with the care team.
The practical goal is a record that distinguishes the history of a disease from the history of considering one.
A record that can change its mind
Better records should preserve the source and status of important diagnostic claims. They should make corrections visible where clinicians are likely to encounter the old information. They should support reassessment when the course of illness no longer fits the explanation.
These are design proposals, not a claim that every diagnosis needs an expiration date. A remote diagnosis may remain essential for decades. Another may need review within weeks. The relevant question is what kind of evidence justifies its continued role in care.
There is a human reason to get this right. A patient should be able to say “That isn't what happened” without their uncertainty automatically counting against them. Their account may be incomplete, just as the record may be incomplete. The discrepancy is something to investigate.
The chart and the patient are sources to reconcile. Treating either as incapable of error prevents that work.
A medical record should help the next clinician understand the patient. It should leave room for the patient to change that understanding.
References
Hirschtick, Robert E. “Sloppy and Paste.” PSNet, Agency for Healthcare Research and Quality, 1 July 2012, https://psnet.ahrq.gov/web-mm/sloppy-and-paste.
Agency for Healthcare Research and Quality. “Challenges and Opportunities for Improvement in Diagnostic Documentation.” Documenting Diagnosis: Exploring the Impact of Electronic Health Records on Diagnostic Safety, July 2024, https://www.ahrq.gov/diagnostic-safety/resources/issue-briefs/dxsafety-ehr-impact4.html.
Comments
Post a Comment