Medication Reconciliation at Admission and Discharge

A best possible medication history is a systematic history taken from more than one source, not a copy of whatever the last admission recorded. The standard approach uses a structured patient or carer interview supported by at least one corroborating source: the medicines the patient brought in, previous discharge summaries, the referring clinician's records, community pharmacy dispensing history, or the patient's own repeat prescription list. One source is a claim; two sources is a history.

The interview technique determines the yield. Open questions about what the patient takes, when, and whether they have missed any recently uncover far more than reading a list and asking for confirmation, because patients confirm lists. Specific prompts are needed for the categories patients routinely omit: inhalers, eye drops, topical preparations, injections including insulin and depot medications, weekly or monthly doses, contraceptives, over-the-counter analgesics, and ayurvedic, homeopathic, or other traditional medicines taken alongside allopathic treatment.

That last category deserves emphasis in the Indian context. Concurrent use of traditional preparations is common and frequently unreported unless asked about directly and without judgement, and some carry genuine interaction potential. A history that does not ask has not established what the patient is taking.

At admission the history is compared against the admission orders, and every difference is classified. Some differences are intentional and clinically correct — the anticoagulant held before surgery, the antihypertensive stopped because the patient is hypotensive — and these need to be documented as intentional so that the next clinician does not helpfully restart them. Undocumented intentional changes are indistinguishable from errors two days later, and they generate exactly the same rework.

The remaining differences are unintentional discrepancies: omissions, dose or frequency changes, duplications, and formulation substitutions made without a decision. These go to the prescriber for resolution. Prioritise by risk rather than by list order — an omitted antiepileptic, antiplatelet, immunosuppressant, or insulin needs attention within hours, while an omitted vitamin supplement does not.

Timing is a design decision. Completing reconciliation within a defined window of admission — commonly twenty-four hours, sooner for high-risk patients — is the standard most accreditation frameworks and NABH-aligned medication management policies expect, and it is achievable if the responsibility is assigned rather than assumed. When reconciliation is 'everyone's job', it is done last.

Internal transfers get far less attention than admission and discharge, and they produce a characteristic error. A patient moves from intensive care to the ward and the infusion-based regimen is converted, or moves from the ward to theatre and back with the pre-operative holds never reversed. Nobody omitted anything deliberately; the list simply changed hands while the reasons stayed behind.

The fix is to treat transfer as a reconciliation point with the same structure as admission: compare the pre-transfer list against the post-transfer orders, classify every difference, resolve the unintentional ones. The comparison is quicker because both lists are already in the system, but it is not automatic.

Two specific checks catch most transfer errors. First, every medicine held for a procedure needs an explicit restart decision recorded when the patient returns, not an assumption. Second, every intensive-care-specific route or formulation needs a deliberate conversion decision, with the ward-appropriate equivalent prescribed rather than inherited.

Discrepancy resolution is a conversation, and the way it is structured determines whether it happens. A pharmacist who presents an unsorted list of eleven differences during a ward round gets a distracted acknowledgement; one who presents the two that matter clinically, with the specific question and a recommendation, gets decisions. Prioritisation is not a courtesy to the prescriber — it is what makes the high-risk items get resolved at all.

Every resolution needs to land in the record with a reason. The minimum useful documentation is the medicine, what the pre-admission regimen was, what was decided, why, and by whom. That reason field is what prevents the same discrepancy being raised again at the next transition, and it is what makes the discharge summary intelligible to the family physician.

Escalation matters when a high-risk discrepancy cannot be resolved because the prescriber is unreachable. Define a timeframe and a route — the covering consultant, the duty medical officer — rather than leaving the pharmacist to keep trying. Unresolved omission of a critical chronic medicine is a patient safety incident, and it should be reportable as one.

Discharge reconciliation compares three things: the pre-admission medicines, the medicines the patient was on during the stay, and the intended discharge regimen. Each pre-admission medicine must be explicitly continued, stopped, or changed, with a reason; each new medicine must have an indication and a duration; and any medicine started for the acute episode with no ongoing indication must be actively stopped rather than carried forward by default.

That last point causes a disproportionate share of downstream harm. Acid suppressants started for stress ulcer prophylaxis in intensive care, sedatives started for a disturbed night, and analgesics started post-operatively all have a tendency to persist for years because no discharge decision was recorded to stop them. The discharge summary should name each such medicine, its indication, and its stop date.

The discharge medication list must reach the people who will use it: the patient, the family physician, and any community pharmacy involved. A structured, coded list shared through the patient's ABHA-linked record makes that transfer far more reliable than a printed summary that reaches the next clinician only if the patient remembers to bring it. HealUDoc can carry reconciliation status and the reason for each change through into the discharge summary, so the receiving clinician sees not only what the patient is on but why it changed.

A reconciled list is only useful if the patient can act on it. Discharge counselling should cover what each medicine is for, how and when to take it, what has changed and why — with particular emphasis on anything stopped, since patients typically have the old supply at home — what to watch for, and what to do about a missed dose. Written material in the patient's own language reinforces the conversation; it does not replace it.

Teach-back is the technique that reveals whether counselling worked. Asking the patient to explain their regimen back in their own words, rather than asking whether they have understood, surfaces the misunderstandings that a yes conceals. It takes an additional minute or two and it is the only part of counselling that produces evidence of comprehension.

Attend to the practical barriers, because they determine adherence more than knowledge does. Cost of the new regimen, whether the medicines are available locally, dexterity for inhalers or injections, literacy, and who at home will actually manage a complex schedule are all more predictive of what happens next week than how well the pharmacology was explained. Where a barrier exists, it needs a plan before the patient leaves, not a note in the file.

Pharmacist-led reconciliation produces the most accurate histories, but few hospitals can staff a pharmacist at every admission. The realistic model is risk-stratified: pharmacist-led histories for the patients where the payoff is greatest — the elderly, those on many medicines, those with high-risk agents such as anticoagulants, insulin, or antiepileptics, and readmissions — with trained nursing or physician-taken histories for the rest, subject to pharmacist review of the high-risk subset.

Measure the process, not the paperwork. Useful measures include the proportion of admissions with a documented history from two or more sources, median time from admission to completed reconciliation, unintentional discrepancies identified per patient, the proportion resolved within the target window, and the proportion of discharge summaries stating a reason for every change. HealUDoc can report reconciliation status and elapsed time by ward and admission type, which makes the overdue cases visible while they are still actionable. Completion-rate metrics alone are the ones most easily gamed by signing a form.

Feed the findings back to the teams generating them. A monthly summary showing which discrepancy types recur, and where, turns reconciliation from an audit into an improvement loop — and it is usually how a hospital discovers that its real problem is not the reconciliation process at all, but the admission clerking template that never asked about inhalers.