One in five older patients needing a discharge action met an error in general practice

Design

Retrospective record review in 7 general practices

Population

263 patients aged ≥65 discharged from hospital, 2022–2023

Primary outcome

Errors and harms in actioning discharge summaries; 90-day readmission

Effect

Error in 20.6% and harm in 4.4% of patients needing an action; readmission 17.5% (13.7–23.0)

This BJGP study, published on 1 October, reviewed the GP records of 263 patients aged 65 or over discharged from hospital to seven West Midlands practices over one year. It looked at what discharge summaries asked the practice to do, and what happened.

The 160 summaries that requested action contained 551 actions between them — medication changes, blood tests, referrals, reviews. Among patients needing an action, 20.6% experienced an error, meaning a requested action was not completed, and 4.4% came to harm. Most harms led to readmission, and three of eight were judged preventable in primary care. Related readmission within 90 days occurred in 17.5%. Errors and harms were concentrated in patients with dementia or a recorded carer.

This is a small, purposive sample from one English region, but the problem is general. In Indian practice, where discharge summaries often go with the patient rather than to a named doctor, the gap is likely wider.

• Have a named person in the practice read every discharge summary and list the actions it requests.
• Track each requested blood test, medication change and referral to completion, not just to filing.
• Prioritise patients with dementia, or who rely on a carer, for an early post-discharge contact.
• Reconcile medicines at the first post-discharge visit; started, stopped and changed doses are where errors cluster.
• When a summary is unclear, contact the hospital team rather than guessing.

Why it matters

The handover from hospital to family doctor is where actions quietly fail, and the harms land as readmissions.

Don't overread it

Seven purposively chosen English practices cannot give a national error rate.

The statistics, in plain English

The 95% confidence interval for 90-day readmission (13.7% to 23.0%) is wide because the sample is small; the true rate in similar practices could lie anywhere in that range.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.