Health Economics Metrics

Topic 30 of 76 · Health system operational economics

Readmission Rate

The 30-day readmission rate is the percentage of discharged patients who return as an emergency within 30 days. It is the health system's canonical quality-of-discharge metric — and it carries direct financial penalties.

Why it matters

A readmission means the first discharge didn't stick: premature discharge, failed medication handoff, no follow-up, or missing social support. Payers penalize it explicitly — the US Hospital Readmissions Reduction Program docks up to 3% of a hospital's Medicare payments; the NHS has historically not paid for avoidable 30-day emergency readmissions. So readmission avoidance is one of the few benefit categories that is directly cash-relevant to a provider, not just capacity.

The math

Readmission rate = emergency readmissions within 30 days / index discharges × 100

Risk-standardized comparisons adjust for case mix; penalty programs
compare observed vs expected for like hospitals.

Value of avoidance = avoided readmissions × (cost per readmission spell
                     + penalty exposure per readmission)

Worked example

A heart-failure discharge-support app (symptom tracking, weight alerts, medication reminders, nurse escalation): 2,000 discharges/year, baseline readmission rate 18%, trial shows 14% with the app.

Avoided readmissions = 2,000 × (0.18 − 0.14) = 80/year
Cost per readmission spell ≈ £3,500 → £280,000/year avoided treatment cost
Plus penalty/non-payment exposure on those spells.
App cost: 2,000 × £60 = £120,000/year

Net ≈ +£160,000/year, before any QALY claim for avoided deterioration.

The number to defend is the 4-percentage-point effect: it must come from a controlled comparison, because readmission rates swing with case mix and season.

Software engineering connection

Readmission is the health system's change failure rate (see DORA metrics): work that "shipped" and bounced back within 30 days. The analogies run deep — reopened tickets and regression incidents indicate poor "discharge quality" (weak verification, premature closure, missing handoff docs); penalty-style accounting (the fixing team pays, not the receiving team) changes behavior; and both fields learned the same lesson, that pushing raw throughput (faster discharge, faster shipping) without investing in the handoff simply converts visible queues into invisible rework. A "30-day reopen rate" belongs on any team dashboard that celebrates cycle time.

Pitfalls

  • Gaming by re-labeling: readmissions coded as observation stays or new conditions; audit the definition.
  • All-cause vs related-cause: 30-day all-cause includes genuinely unrelated events; penalties usually use all-cause precisely because "related" is gameable.
  • Case-mix blindness: a hospital serving sicker, poorer populations readmits more for reasons no app fixes — risk-adjust before comparing.

Sources