Skip to main content

Drug Allergy Records: Separate Unknown from None Known

An allergy record needs both a clinical status and an information history. “No known drug allergy,” “unable to establish the history,” and “not yet reviewed” are different states. Record the reported medicine and reaction, identify the source, and preserve the clinician’s assessment separately. This prevents an empty field or an incomplete import from being mistaken for a completed allergy review.

By Bractis · Updated · 14 min read

For Indian OPD teams designing allergy intake, reviewing imported charts or evaluating how an EMR presents prescribing information.

The problem is often missing context, not a missing red banner

A patient registers using a relative’s phone number. Reception opens a new chart, the allergy field is blank, and a prescription template prints “No known allergies.” Nothing in that sequence establishes that anyone asked the patient about a reaction. The software has converted an absence of information into a clinical statement.

A warning banner cannot repair this if the underlying status is wrong. Start by deciding what the record is allowed to say, who supplied it and when it was last assessed. Then decide how to display it during prescribing and handover. This order matters when importing paper histories, merging duplicate charts or using an older prescription as the starting point for a new visit.

The distinction is also useful at a busy desk. Staff can report that they have collected an uncertain history without being expected to resolve it. The doctor can see what remains uncertain instead of receiving a polished but misleading “complete” intake form.

Separate clinical status from review status

NICE CG183 describes drug-allergy status using recorded allergy, none known and unable to ascertain, and distinguishes allergy information from other adverse reactions. It recommends documenting the medicine, reaction and timing and checking the status before prescribing. This is UK clinical guidance; here it informs record design rather than asserting an Indian legal requirement.

Add an operational review state alongside that clinical information. “Collected by reception” can be true while “reviewed by the doctor for this encounter” is false. Combining these into a single green tick hides the handover. The following model is an implementation example, not a diagnostic classification.

On a narrow screen, scroll the table horizontally to read all columns.

Two dimensions of an allergy record
Information in the chartWhat it tells the next userWhat it does not establish
Patient reports a reactionThere is a history for the clinician to assess, with a source and description.It does not by itself prove a mechanism, severity classification or safe alternative medicine.
Clinician-recorded none knownA clinician has recorded this status in a stated context and at a stated time.It is not a lifetime guarantee or permission to skip a later review.
History uncertain / unable to obtainThe information gap is explicit.It should not be silently displayed as none known.
Not yet reviewed this encounterAn operational review remains outstanding.It does not automatically erase or contradict previously documented information.

Source: NICE CG183: Drug allergy, documentation and sharing information

What to collect when a patient reports a reaction

Ask for the name as the patient knows it and retain their description of what happened. A strip, old prescription or hospital document can help identify the source, but an uncertain name should remain uncertain. Do not turn a partially remembered brand into a precise coded medicine because the dropdown requires a selection.

Separate the event date from the date the history was entered. “Reported today; reaction happened several years ago” is a different statement from “reaction today.” If only an approximate period is known, preserve that approximation. Record who supplied the information when it comes from a caregiver rather than the patient.

Useful context includes which supporting documents are available, what details could not be established and which clinician will review the entry. The purpose of intake is to carry the history faithfully to that reviewer. It is not to decide whether the patient should receive a related medicine or to perform an allergy assessment at reception.

Worked intake example: “that tablet caused a problem”

In this fictional OPD conversation, a patient says, “A tablet I took for an infection caused a rash several years ago.” The name is unknown, the old prescription is unavailable and the relative at the desk remembers a different year. This is useful information, but it does not establish which medicine caused the event or its clinical classification.

A faithful intake note could say: “Patient reports a rash after an unidentified tablet taken for an infection; event several years ago, exact date unknown. Relative recalls a different year. No supporting prescription available today. Clinical assessment pending.” The note preserves the account without turning the word infection into a confirmed antibiotic identity or the word rash into a confirmed allergy diagnosis.

Compare that with “antibiotic allergy, severe, confirmed today.” Every added assertion changes the meaning: the medicine class was inferred, severity was supplied without evidence, the report became a diagnosis and today became the apparent event date. A required dropdown should not force any of those substitutions. Provide a way to record the unresolved history and send it to the clinician.

If the patient describes a current problem instead of a remote history, route that through the clinic’s clinical assessment process. Do not use this historical documentation exercise to delay assessment, classify symptoms or decide urgency at reception.

Keep the report, assessment and later correction distinguishable

A patient’s account and the clinician’s conclusion can legitimately differ. The record should support both without making staff choose between deleting the history and retaining an unsupported label forever. Preserve the original report, then document the assessment and any change with its author and reason.

For example, a record imported from another clinic may contain a broad medicine-class label but no reaction details. The receiving doctor may need further history before deciding how to interpret it. A migration tool should not silently make the label more specific, while a receptionist should not remove it merely because the patient cannot remember the event.

A later correction is a new piece of record history. Make it visible to the people relying on the chart, including the next clinician and anyone preparing a referral. This is especially important where a structured allergy field and an old free-text note disagree: resolve the discrepancy through review rather than allowing different screens to present different apparent truths.

An old “none known” entry and a new report need a chronology

The newest timestamp is not always the strongest clinical evidence. A file imported today may contain a much older assessment, while a new patient account may still require review. Record when the event reportedly occurred, when someone documented it, and when the clinic received or assessed the information.

The following fictional history A-16 shows why a single last-updated field is inadequate. An earlier none-known entry can remain a truthful record of what was documented then without being used to suppress a later report. Equally, the later report does not authorize reception to supply a diagnosis.

On a narrow screen, scroll the table horizontally to read all columns.

Fictional allergy history A-16: event time, record time and current review
EvidenceWhat the timestamp meansHow it should inform the current encounter
12 March 2024: clinician recorded none knownDate of that earlier assessment; not a lifetime guarantee.Retain it as historical context, with its author and source.
28 October 2026: patient reports an older reactionDate the account was collected; the event itself remains only approximately dated.Make the new report and uncertainty visible for assessment.
Current encounter: clinical review still pendingNo new clinical conclusion has yet been recorded.Do not print the old none-known phrase as if today’s question were resolved.
After clinical reviewA separately attributed assessment, whenever it actually occurs.Show the clinician’s recorded disposition and any unresolved follow-through without erasing the earlier sources.

What can go wrong when allergy information is migrated?

Migration often exposes a mismatch between data models. One system may store a single free-text field; another expects a coded medicine, reaction and status. A blank value can mean “not collected,” “not exported” or “intentionally empty.” Those possibilities cannot be resolved by treating every blank as none known.

Before accepting an import, build a small set of deliberately different fictional records. Include a detailed reaction, an uncertain name, an explicit none-known entry and an unreviewed chart. Compare the source and destination, then inspect the prescribing and export views. Count how many records need manual reconciliation rather than reporting only how many rows imported.

On a narrow screen, scroll the table horizontally to read all columns.

Allergy migration checks that expose lost meaning
Source situationFailure to look forAcceptance evidence
Blank source fieldDestination prints a reassuring default.Unresolved information remains visibly unresolved.
Uncertain medicine nameImport chooses a precise product without evidence.Original wording and uncertainty can be retrieved.
Old reaction plus later assessmentOnly the latest string survives, with no context.Reviewer can distinguish report, assessment and change history.
Conflicting charts during mergeOne entry silently overwrites the other.Conflict is visible and routed for authorized review.

Evaluate visibility at the point where it matters

Ask the doctor to open the prescribing view from several paths: a new consultation, a repeat visit and an imported chart. The relevant status and uncertainty should be understandable without opening a hidden administration panel. An alert that appears only after the document is issued has a different value from information available during review.

Also inspect the referral letter and record export. A local banner may disappear when the next clinician receives a PDF. Conversely, a printed phrase may be more definite than the chart it came from. Compare the actual wording across outputs; do not assume that shared colours or icons mean shared semantics.

Software alerts depend on the product’s knowledge base, configuration and available data. Ask what they cover and what they do not. An absent alert is not clinical clearance, and this guide does not provide rules for cross-reactivity or alternative prescribing. The useful software question is whether the doctor can find, assess and communicate the information accurately.

Make “reviewed” refer to a clinical action and a particular record

Opening a chart, importing a file or checking that mandatory fields are filled should not silently record that a doctor assessed the allergy history. Those are different events. In A-16, an assistant may complete collection while the clinical review remains outstanding; the information can be organized without claiming the decision has been made.

A review entry should identify the evidence considered and the clinician’s disposition. If the assessment leaves the medicine identity uncertain, that uncertainty can remain explicit. The software should not require a precise medicine code or a reassuring status merely to complete the encounter’s administrative fields.

A later document may change the evidence available. Link it to the unresolved question and identify who will assess it, including cover where appropriate. Do not retrospectively attach a new source to an old review in a way that suggests the earlier doctor had already considered it. The review time and the document-receipt time need to remain distinguishable.

Follow one history through the chart, prescription and referral

Use the unresolved A-16 history to inspect the actual documents that leave the consultation. The aim is consistent meaning, not identical formatting. A short patient-facing statement and a detailed clinical record can differ in length while preserving the clinician-approved status and relevant uncertainty.

Test an output created before the review and another generated after it. A saved old PDF may remain historical, but its wording must not be mistaken for the current document. A freshly generated referral should not continue using a cached none-known phrase after the doctor has recorded a different disposition.

On a narrow screen, scroll the table horizontally to read all columns.

Allergy information across three views: a focused acceptance check
ViewEvidence to inspectReason to return the case for correction
Current prescribing screenThe report, source and review state are available to the doctor.A blank code or recent import hides the unresolved account.
Newly generated patient documentThe displayed wording matches the clinician-approved current record.The template inserts none known even though that is not the current assessed status.
Referral or record exportThe recipient can distinguish the relevant assessment, source history and outstanding question.Only a reassuring default or an unsupported confirmed label survives the export.

A workable division of responsibility in a small clinic

Reception or an assigned assistant can collect a report and identify missing documents within the clinic’s permitted role. The responsible clinician assesses its meaning and decides how it affects care. The clinic administrator can check that access, imports and output templates preserve the information. These are different responsibilities even when one person handles more than one administrative task.

Make the unresolved handover explicit: who needs to review this entry, at which point in the encounter, and what happens if the person is unavailable? Do not measure success solely by the percentage of fields populated. A confidently filled but unsupported field is worse information than an honest uncertainty that reaches the doctor.

Worked example: an old reaction, a shared phone and an incomplete import

In this fictional scenario, two family members share a contact number. One reports a reaction to an unidentified tablet years earlier; the other has a clinician-recorded none-known status. The clinic first confirms which person is attending. It retains the first patient’s account as uncertain and sends it for review. During an import rehearsal, the team notices both records have become “none known” because the export lacked a structured allergy code. They reject that mapping, restore distinct states and retest the prescribing and referral outputs. The defect was not a missing alert: it was loss of patient identity and meaning during data conversion.

Allergy Record Review Checklist

Use these prompts in your clinic’s own notes. Record the result, responsible person and next action for each item.

Meaning
Can staff distinguish reported information, uncertainty and a clinician’s reviewed status without relying on colour alone?
Provenance
Can the doctor see who supplied the history, when the event reportedly occurred and when it was entered or assessed?
Change handling
Test imports, chart merges and later amendments; retain conflicts and reasons instead of silently overwriting.
Clinical handover
Inspect prescribing, referral and export wording and identify the owner of every unresolved review.

Questions doctors ask

What if an old chart says no known allergy but the patient now reports a reaction?

Retain the earlier assessment with its date and source, record the new account without inventing missing details, and make the current review need visible to the clinician. Neither automatic deletion nor simply choosing the latest timestamp resolves the clinical question.

What should staff enter when the patient cannot identify the medicine?

Preserve the name or description as reported and explicitly record that identity is uncertain. Keep supporting documents and unanswered questions linked for clinician review; do not guess a product or class to satisfy a required dropdown.

Should “no known drug allergy” be a default on a new patient record?

A new chart does not establish that history was obtained. Keep the collection or review state explicit, and let the responsible clinician record the assessed status through the clinic’s process.

Can reception enter an allergy reported by a patient?

Assigned staff can faithfully record the report, its source and uncertainty within their role. That entry should remain distinguishable from a clinician’s assessment; it should not automatically select treatment or a substitute medicine.

Should every reported side effect be placed in the allergy field?

Preserve the reported experience and route classification to the clinician. The data model should allow the account, other adverse reactions and the assessed allergy information to be represented without forcing staff to make a clinical conclusion.

What does an absent EMR allergy alert mean?

Only that the system did not show that alert in that configuration and context. It does not establish safe prescribing. Review the history and the product’s supported alert scope with the responsible doctor.

What is the most important allergy check during an EMR migration?

Verify meaning, not just row counts. Test that uncertain histories, none-known entries, reports and later assessments remain distinguishable and appear correctly in prescribing and exported documents.

Sources and scope

Updated on . This article concerns faithful documentation and information handover. It does not diagnose drug allergy, classify an individual reaction or recommend alternative treatment. No independent clinical or legal review is claimed.

  • NICE CG183: Drug allergy, documentation and sharing information

    Section 1.2 was checked through indexed text from NICE on 28 October 2026; direct retrieval returned an access error. UK documentation guidance is not presented as Indian law. The A-16 history, review record and output checks are original fictional workflow examples, not a diagnostic protocol.

Try the workflow
with Bractis.

Use four fictional histories—reported reaction, uncertainty, reviewed none known and an empty import—and compare how they appear during prescribing, referral and export.

Request a clinic demo