Drug Allergy Records: Separate Unknown from None Known
Record uncertain allergy histories with source-preserving examples, dated assessments and checks across prescribing, referral and export.
New patient-submitted information should remain identifiable as a report until the responsible clinician reviews it. A newly reported allergy or medicine change needs visible attention before the relevant prescription is issued; opening the intake form is not proof of review. Preserve the source, review decision and unresolved details, and define a clinician-led exception path so a software gate does not obstruct urgent care.
Indian doctors and clinic teams using online intake forms, reception-assisted histories and pre-consultation questionnaires.
An intake form establishes what someone reported, when it was submitted and, where known, who supplied it. It does not independently establish a diagnosis, an allergy classification or the patient’s current medicine use. A family member filling the form from a shared phone may be describing an earlier event or another person’s experience.
Keep the original report separate from the clinically accepted record. The clinician may confirm it, clarify it, identify a duplicate or leave part unresolved. This separation allows staff to collect useful information without turning every typed phrase into a definitive structured fact.
NICE’s drug-allergy guidance recommends checking and confirming allergy status before prescribing and keeping the record updated. This is a UK clinical reference, not an Indian software-blocking law. The review states and timing examples below are original proposals for a clinic intake workflow.
Source: NICE CG183: Drug allergy, documenting and sharing information
“Received”, “seen” and “reviewed” are often used interchangeably, but they describe different events. A receptionist can confirm receipt and identify the patient. A doctor can open a form without resolving its contents. Clinical acceptance requires a decision about the reported information and its relevance to current care.
Use a small set of states with a named next action. An unresolved report should not disappear because somebody has viewed it. Equally, a duplicate submission should not require pretending that a second clinical event occurred just to clear the queue.
On a narrow screen, scroll the table horizontally to read all columns.
| State | What is known | Next responsibility |
|---|---|---|
| Submitted | A patient or helper supplied information | Match the patient and route the report |
| Identity/source clarified | The intended patient and reporter are understood | Clinician reviews the relevant content |
| Clinically accepted or corrected | A clinician has recorded the interpretation and decision | Use the accepted information in the current plan |
| Unresolved | A relevant detail remains uncertain | Name the clinician and clarification or care action |
| Duplicate or wrong submission | The submission is linked to its actual context | Preserve the explanation without creating a false new clinical fact |
Suppose a fictional intake response says, “Reaction to the tablet given last winter; name not known.” The useful first record is that statement with its source and uncertainty. Converting it to “no known allergies” because the medicine is unidentified discards the concern. Selecting a likely medicine from the catalogue invents certainty in another direction.
The clinician asks the questions needed for the present decision and records what can be established. A previous prescription, pack or clinical note may help, but it should not be treated as confirmation unless it actually resolves the reported event. Uncertain details remain visible with a plan for clarification.
Do not make the patient navigate a technical catalogue to submit a concern. A free-text route can preserve their language, while the review process supplies the clinical interpretation. The form should explain that sending information does not mean the doctor has already assessed it or that it is an emergency contact channel.
In a fictional timing exercise, a doctor reviews intake version 1 at 08:42 and starts a prescription. A family member submits a medicine update at 08:45. At 08:47 the doctor is ready to issue the draft. The earlier review should not silently stand for information that arrived three minutes later.
The system needs a way to distinguish the version reviewed from the latest relevant submission. Show what changed and route it for review before the affected output is issued. This is a workflow requirement to evaluate, not a claim that an existing product detects every race or conflict.
The same problem arises with two clinicians or two browser tabs. A review badge in one view may be stale in another. In a rehearsal, submit a change after the draft is opened and verify the final issuing path, including a saved draft reopened later. Merely testing the intake screen misses the boundary that matters.
Extend the earlier timing exercise as fictional case V-17. All times are on the same day in IST and are used only to reconstruct the sequence. At 08:42 the clinician reviews submission R1 and begins a draft. At 08:44 the draft is saved as D1, still based on that information. At 08:45 a helper sends R2 with a new report about a medicine. At 08:46 reception corrects the contact address in R3. When issuing is attempted at 08:47, R3 is the latest submission, but the clinically relevant medicine change in R2 has not been reviewed.
A single badge saying “latest form reviewed” would conceal two different questions: which submission is newest, and which clinical information the doctor assessed. The address correction does not prove that the medicine report was considered, nor should its administrative nature be used to dismiss R2. Show the actual changed fields, their sources and the earlier review they post-date. This is a requirement to evaluate in a workflow, not a claim about a particular software implementation.
At 08:49 the clinician reviews the new account and records both what was clarified and what remains uncertain. That decision can support a revised draft D2 under the clinician’s plan without pretending that every unknown was resolved. The previous D1 must not remain available as an apparently current, separately printable prescription. Record whether any output had already been issued; in this exercise none had been. If it had, checking only the saved draft would miss the need to assess the issued version and its recipients.
There are three intake submissions, two draft versions and two clinical review events in this example. They are not three consultations or five prescriptions. The five-minute interval from the initial review to the attempted issue explains why the old review was stale; it is not a safe window during which changes can be ignored. A later receipt, even seconds before an action, needs the relevance and ownership decision required by the actual clinical context.
On a narrow screen, scroll the table horizontally to read all columns.
| Time, IST | Event | Version evidence | Implication |
|---|---|---|---|
| 08:42 | Clinician reviews and begins drafting | Review of R1 | Original information considered |
| 08:44 | Draft saved | D1 based on R1 | A saved draft is not an issued prescription |
| 08:45 | Helper reports a medicine change | New submission R2 | Relevant change awaits clinical review |
| 08:46 | Reception corrects contact address | New submission R3 | Administrative edit does not resolve R2 |
| 08:47 | Issue attempted | D1 still reflects R1 | Show the unresolved change before the affected action |
| 08:49 | Clinician reviews R2 and records decision | Second review; revised draft D2 | Keep remaining uncertainty and current output status explicit |
An intake update is often a mixture of facts, recollections and administrative corrections. A patient’s statement that a medicine was stopped, a photograph of a packet and a receptionist’s address correction should not acquire the same authority simply because all three appear in one form. Keep the original wording and the relevant date: a photograph taken today may show a pack from an earlier course, and the upload time is not the time the medicine was last used.
For a medicine change, the clinician needs to decide whether the submitted account describes current use, a previous prescription or a proposed change that has not happened. For an unidentified reaction, preserve uncertainty rather than assigning the nearest searchable medicine. When one detail is clarified, limit the correction to that detail; identifying a product name does not automatically establish the reaction mechanism or the accuracy of a reported dose.
A helper using a shared family phone may discover that an answer belongs to another person. Verify the intended patient and affected records through the clinic’s process. Do not copy the clinical report into every linked family chart, or simply delete the mistaken answer if it already influenced an output. The responsible clinician needs to know whether the error remained in an unreviewed submission, entered a structured record or appeared on an issued prescription.
Where staff translate an account from Hindi or another language, retain the patient’s meaning and mark the assistance. “I felt unwell” is not automatically equivalent to a confirmed allergy. If the meaning cannot be established, record the unresolved question for the doctor. This gives reception a useful task—accurate capture and routing—without asking staff to make a clinical classification.
On a narrow screen, scroll the table horizontally to read all columns.
| Reported change | Evidence to clarify | Decision boundary |
|---|---|---|
| Medicine reportedly stopped | Who reported it, actual use and timing | Clinician distinguishes current use from intended change |
| Unidentified tablet reaction | Original account, event history and reliable source material | No invented medicine name or automatic “none known” |
| Contact address corrected | Patient identity and verified contact detail | Administrative correction does not acknowledge clinical content |
| Answer was for another family member | Intended patient and where the mistaken answer travelled | Correct affected records with history; do not propagate by phone match |
A review-complete checkbox can be misleading when the clinician has assessed the report but an external detail remains unknown. Keep “review performed” separate from “all facts confirmed”. The record should say what the clinician did with the uncertainty and who will obtain the missing information. Do not leave the next doctor to infer the decision from an empty field or an unlabelled attachment.
In V-17, a useful handover names R2’s medicine report, the decision recorded at 08:49, the replaced draft D1 and the current D2 status. It also states that no output had been issued at that checkpoint. If the patient later sends a clearer pack image, it is a new source event: relate it to the outstanding question and route it for assessment rather than claiming the earlier review anticipated the new evidence.
During downtime, record the source and review sequence on the approved fallback record and reconcile it when the system returns. Preserve the actual consultation and submission times alongside the later digital entry. Staff should not manufacture a neat digital sequence in which the report appears to arrive before it did. Urgent care remains a clinician-led process while the information record is repaired.
Identify which unresolved information should require attention before a particular action. A new clinically relevant report can justify a review gate; an unrelated administrative preference should not automatically stop all clinical work. The clinical owner should determine the scope rather than letting a generic “form incomplete” rule make that decision.
If a clinician needs to proceed in an urgent situation, the workflow must support appropriate care, a documented reason and continued ownership of unresolved information. Reception staff should not use an administrative override to make a clinical warning disappear. A recorded exception is not proof that the underlying uncertainty has been resolved.
Keep the patient’s care pathway available during network or software failure. Use the clinic’s established direct assessment and documentation process, then reconcile the record with original and later entry times. A screen that cannot be cleared is not a substitute for clinical judgment.
If a receptionist enters answers spoken by a patient, record that assistance instead of making the form appear self-completed. Preserve whose account it is and the language used where it matters. Staff can capture the report, but should not translate a vague symptom into a diagnostic label or decide that a medicine reaction was insignificant.
When a family member is helping, confirm which patient the submission concerns and whether the patient’s own account differs. The same mobile number may identify several family records. A form link sent for one person should not automatically become authority to edit everyone associated with that phone.
Corrections need an explanation. If the helper reports that an earlier answer was for the wrong family member, retain enough history to reconstruct the mistake and assess whether it affected any issued output. Quietly overwriting the form can conceal the need for clinical follow-through.
For a fictional morning with 12 submissions, suppose nine have documented clinical decisions, two remain awaiting review and one is a confirmed duplicate linked to an already reviewed submission. That accounts for all 12 submissions, but it does not establish 12 distinct patients or nine completed consultations.
Inspect the two awaiting review by relevance and current clinical activity. One may belong to a future appointment, while the other may concern a prescription being issued now. A single count cannot describe their urgency. The responsible clinician determines the response to the content and context.
Review a sample of accepted entries against their source. A low pending count is not useful if staff achieve it by marking every submission reviewed. Look for clear decisions, retained uncertainty and evidence that later changes remain visible at the point of use.
In the fictional 08:42–08:47 exercise, a new medicine report arrives after review and before prescription issuing. The clinic checks whether the changed content is visible at the final action, which version the doctor acknowledges and who owns any unresolved clarification. The earlier review timestamp is retained rather than relabelled as a review of information not yet received.
Use these prompts in your clinic’s own notes. Record the result, responsible person and next action for each item.
No. The newest submission and the latest clinical review are different records. An administrative edit should not hide an unreviewed medicine or allergy report. Show the changed clinical information and the review decision relevant to the current action.
Yes, if the record accurately describes what the clinician assessed, what remains unknown and the plan for that uncertainty. Review does not mean every reported fact is confirmed. Preserve the clarification owner and the effect on drafts or issued information.
No. Preserve it as a report with its source. Clinical acceptance, correction or unresolved status requires an attributable review decision.
No. Opening proves only that the view was accessed. The record should distinguish review of the relevant content and the version considered.
No. Keep the report and uncertainty visible for the responsible clinician. A missing medicine name does not establish absence of a concern.
Make the changed information and review version visible before the affected output is issued. Do not silently apply an old acknowledgement to a new submission.
A software gate must not replace appropriate clinical care. Define a clinician-led exception and documentation route that preserves unresolved information and assigns follow-through.
Updated on . This article combines the cited references with Bractis’s original workflow analysis and illustrative examples. Product capabilities depend on the configured scope. No independent clinical or legal review is claimed.
UK guideline published September 2014. Official indexed recommendation 1.2.6 was checked on 28 October 2026; direct source access was unavailable during this check. Used narrowly for allergy-status confirmation, not Indian law or a software gate. The version examples are original.
Record uncertain allergy histories with source-preserving examples, dated assessments and checks across prescribing, referral and export.
Compare medicine strips, discharge papers and reported use with a three-source case, Hindi intake question and discrepancy handover.
Record patient-specific alert decisions, distinguish repeat displays from new episodes, and audit corrections, deferred reviews and decisions to proceed.
Handle shared family phones with separate patient records, worked matching cases, purpose-specific contact changes and queued-message checks.
Use a fictional intake update arriving between draft creation and issuing to inspect review versions, visible uncertainty and clinician-led exception handling.
Request a clinic demo