Design: National prospective observational cohort
Population: 7,839 adults having day-case surgery in the UK
Primary outcome: Pain severity and pain-related functional impairment on days 1, 3 and 7
Effect: Poor pain outcome 33.5% in week 1; functional impairment 26.3% at day 7

POPPY was a national prospective observational study of 7,839 adults having day-case surgery in the UK, following pain, function and quality of recovery on days 1, 3 and 7; 7,331 responded on day 7.

Pain scores fell over the week, but 33.5% had a poor pain outcome (score above 4/10) in the first 7 days, and 26.3% still reported pain limiting function on day 7. Poor pain outcomes were associated with treatment for anxiety or depression, chronic pain, frailty, larger operations and head and neck surgery. Poor functional outcomes were associated with orthopaedic surgery, larger operations, chronic pain and frailty. Older age and being opioid-naive were associated with better pain outcomes.

Day-case pathways usually end at discharge, with success judged on same-day departure. These data show recovery at home is incomplete for many, and that the patients most likely to struggle can be identified at pre-assessment. With day surgery growing in India, where post-discharge contact is often absent, a targeted follow-up call costs little.

Identify at pre-assessment patients with chronic pain, regular opioid use, frailty, anxiety or depression
Give these patients a written analgesia plan and a longer analgesic supply
Arrange a telephone follow-up on day 1 and day 3 for high-risk patients
Pay particular attention to orthopaedic and head and neck day cases
Audit pain and function at day 7, not only same-day discharge rates

Why it matters

Same-day discharge is counted as success, while a third of patients go on to have poorly controlled pain at home.

The statistics, in plain English

These are proportions from a large observational cohort, so the associated factors identify risk rather than causes. Denominators differed by outcome because not all patients answered every question, which could bias the estimates if those doing worse responded less.