Medication reconciliation at discharge
Discharge medication reconciliation is the comparison of what a patient was taking before admission, what they were given during it, and what they should take afterwards — resolving the differences deliberately rather than by default. It is the step that determines a patient’s medication for months, and it is performed in the most compressed part of their stay.
The discharge prescription is treated by everyone downstream as the current truth, which makes an error in it unusually durable.
Why does it fail so often?
Because it requires three sources of information at a moment when the pressure is to complete the discharge. The pre-admission list is frequently incomplete or inaccurate, the inpatient record contains changes made for inpatient reasons, and the person doing the reconciliation is often the most junior clinician available and the one least likely to know why a medication was stopped three days ago.
The structural problem is that inpatient decisions and discharge decisions have different logic. A medication held during an admission — because the patient was nil by mouth, because a level was awaited, because it interacted with something given acutely — was held for a reason that expired. Whether it should restart is a clinical decision that requires knowing why it stopped, and the record frequently does not say.
So the default behaviour takes over: generate the discharge prescription from the active inpatient orders. That is fast, defensible, and produces two of the four error types below every time it is used without a check.
What are the four error types?
Each has a different cause and a different consequence.
Unintended omission
A long-term medication held on admission and never restarted. The patient goes home without a drug they had been on for years, and nobody notices until something happens. The most common and the most consequential.
Unintended continuation
A medication started for an inpatient reason — a proton pump inhibitor for stress ulcer prophylaxis, a sedative for one bad night — that continues indefinitely because it appeared on the discharge prescription. The patient may take it for years.
Duplication
The inpatient formulary substitute and the patient’s own brand both end up on the list, or a drug appears under two names. The patient takes both, in good faith.
Dose discrepancy
A dose adjusted for an inpatient reason — renal function during an acute illness, an interaction with something now stopped — carried home unchanged. Correct at the time, wrong on the day the patient leaves.
What would actually fix it?
An accurate pre-admission list, which is upstream of everything and is the real constraint. Reconciliation cannot be better than the list it reconciles against, and in most systems that list is assembled from the patient’s memory at the worst moment of their year.
Recording the reason a medication was stopped, at the moment it was stopped. This is a small documentation change with a disproportionate effect: a discharge reconciliation that can see "held — nil by mouth" versus "stopped — caused bradycardia" is a completely different task from one looking at an absence.
A pharmacist involved in the reconciliation, where that is possible. This is well established as effective and it is a staffing question, which is why it is not universal.
And someone confirming what the patient is actually taking after they get home. This is the part almost nobody does, and it is where the reconciliation is finally tested — the patient who has both the old bottle and the new prescription and is taking both, or has neither because the pharmacy did not have one of them, is invisible to any process that ends at the door.
Why is reconciliation incomplete without post-discharge contact?
Because the reconciliation produces an intended list, and the patient produces an actual one. The gap between them is created by things no reconciliation process can anticipate: a medication the pharmacy could not supply, a cost the patient could not meet, an old supply at home that seemed wasteful to throw away, and instructions delivered to someone who had just been discharged from hospital.
That last point is worth taking seriously. Discharge instructions are given at a moment of relief and exhaustion, frequently with a family member half-present, and retention is poor for entirely ordinary reasons. A list confirmed as understood at the bedside is not a list confirmed as understood a week later.
A structured contact in the first days after discharge — asking what the patient actually has, what they are actually taking, and whether anything is confusing — is what converts an intended list into a verified one. It is also the point at which the two halves of the medication loop join up: the hospital knows precisely what was given, and the contact establishes what continued.
Frequently asked
- What is medication reconciliation at discharge?
- The comparison of what a patient was taking before admission, what they were given during it, and what they should take afterwards — resolving the differences deliberately rather than by default. It determines a patient’s medication for months and is performed in the most compressed part of their stay.
- What errors does discharge reconciliation produce?
- Four types: unintended omission, where a long-term medication held on admission never restarts; unintended continuation, where a drug started for an inpatient reason continues indefinitely; duplication, where a formulary substitute and the patient’s own brand both appear; and dose discrepancy, where a dose adjusted for an inpatient reason is carried home unchanged.
- Why does discharge medication reconciliation fail?
- Because it needs three information sources at the moment of maximum pressure to complete the discharge, the pre-admission list is often inaccurate, and the person doing it is frequently the most junior clinician available and least likely to know why a drug was stopped three days ago. The default — generating the discharge prescription from the active inpatient orders — is fast and produces two of the four error types every time.
- Is medication reconciliation complete at discharge?
- No. Reconciliation produces an intended list; the patient produces an actual one. The gap comes from medications the pharmacy could not supply, costs the patient could not meet, old supplies at home that seemed wasteful to discard, and instructions given at a moment of relief and exhaustion. A structured contact in the first days after discharge is what converts an intended list into a verified one.