IIPatient safety

Unplanned re-attendance within 72 hours

The share of discharged patients who come back to the ED, unplanned, within 72 hours. A return — especially one that ends in admission — is a flag that the first visit may have missed something or discharged too soon.

Capture & escalation pipeline

How a discharge is linked to any 72-hour return, and how return-then-admitted cases trigger review.

How it’s measured

Numerator: Patients who re-attend the ED unplanned within 72 hours of discharge (with a sub-measure for those admitted on return).
Denominator: All ED patients discharged in the period.
Formula: (Numerator ÷ Denominator) × 100; report the 72-hour return rate and the return-with-admission rate.
Unit: % of discharges

Target

India

India: NABH tracks unplanned ED re-attendance; a commonly used operational range is ~2–3% within 72 hours, with the return-with-admission subset watched most closely.

Global

International: RCEM and ACEP use 72-hour (and 7-day) re-attendance as a standard safety/effectiveness metric.

Not every return is an error — but the return-then-admitted subset is the highest-yield group to review.

Who does what

The clinical chain of responsibility at the bedside.

Treating doctor (index visit)	Makes a safe disposition with clear discharge advice and safety-netting; documents it.
Registration	Captures the patient identifier reliably so a return can be linked to the index visit.
Quality cell / clinical lead	Identifies returns, reviews the index records, and classifies avoidable vs unavoidable.
Treating doctor (return visit)	Documents the reason for return so the pattern can be understood.

What to capture & how it’s automated

Who captures it

Medical records / HIM, by matching identifiers across visits.

What is captured

Index discharge date/time, return date/time, reason for return, and whether admitted on return.

Manual reality

Returns are spotted only if staff remember the patient or do a manual register search — so most are missed without a unique patient ID.

Automated in real life

The HMIS links visits by a unique patient/UHID, auto-detects any attendance within 72 hours of a discharge, and flags it for review — turning a hidden signal into a daily worklist.

Who in the hospital is involved

Beyond the bedside — the functions that make capture and improvement happen.

Medical records / HIM	Runs the visit-linkage report that detects 72-hour returns.
Quality cell / QI nurse	Reviews flagged returns, classifies avoidability, and feeds back themes.
ED clinical lead	Leads case review of return-with-admission cases as potential missed diagnoses.
IT	Enables UHID linkage and the automated 72-hour detection report.
ED manager	Acts on themes — crowding-driven early discharge, weak safety-netting, follow-up gaps.

Why it affects performance

Re-attendance is a NABH/board-tracked balancing measure: it stops the ED from gaming length-of-stay and LWBS by discharging too fast. The return-with-admission rate is a sensitive signal of diagnostic safety.

Why it affects patient care

Each avoidable return is a patient who was sent home with an unresolved or missed problem. Reviewing them catches the missed MI, the worsening sepsis, or the inadequate discharge advice before it becomes harm.

Capture pitfalls & gaming to watch for

No unique patient identifier, so most returns are never detected.
Counting planned reviews/dressing changes as unplanned returns.
Reviewing the rate but not the individual records, so no learning happens.
Treating all returns as errors and demoralising staff, instead of focusing on the avoidable subset.

What actually moves the number

A reliable unique patient ID (UHID/Ayushman/ABHA) to link visits.
Structured, written safety-netting advice at discharge.
Mandatory review of every return-then-admitted case.
Senior review of borderline discharges during crowding.
Clear follow-up and referral pathways so patients don't bounce back.

References

Freely citable

Compiled from contemporary emergency-medicine quality practice and freely citable accreditation and guideline standards. Educational use only.