How do I predict bed blockers 48 hours in advance?

Patients who become bed blockers share identifiable characteristics at admission: age over 75 with 2+ comorbidities, social care needs flagged but not yet assessed, and no confirmed discharge destination by day 3. Identifying these patients at day 1 and triggering early discharge planning reduces average delayed transfers of care by 30–40%.

Bed blockers — patients medically fit for discharge but unable to leave due to unresolved social, care, or logistical arrangements — are predictable from admission data if the right fields are examined. The prediction model uses five admission-level variables: NHS number–linked social care history (prior community care package, district nurse dependency, care home placement history), previous DTOC episodes in the past 24 months, age above 75 with two or more comorbidities documented at clerking, absence of a confirmed discharge destination in the admission record, and flagged psychiatric needs requiring specialist assessment. Patients scoring on three or more of these criteria have a 73% probability of DTOC lasting more than 3 days.

The clinical reason for the delay is almost always system response time rather than clinical complexity. Social care assessment turnaround times vary from 24 hours to 14 days depending on Local Authority capacity and weekend staffing. Care home availability in the patient's local area is finite and unpredictable. District nurse package commissioning requires GP involvement in many areas. None of these downstream steps can be accelerated if they are not initiated until the patient is clinically ready for discharge — typically day 4 or 5 of an admission that could have been resolved in 2–3 days with earlier initiation.

The DTOC data published in NHS England SitRep returns is a lag indicator. It counts beds that are already blocked today — patients already past their discharge-ready date. This data is useful for reporting and funding negotiations with Local Authorities but useless for preventing the next cohort of bed blockers. By the time a patient appears in the DTOC count, the opportunity for early intervention has passed.

Predictive use requires combining admission data with social care assessment wait times, which typically live in a different system entirely — either the Local Authority's social care case management platform or a paper-based referral process. Most Trusts have no real-time visibility of how many social care assessments are outstanding for current inpatients, what stage each assessment is at, or what the current Local Authority turnaround time is this week. This information gap means discharge planning operates reactively rather than proactively, adding 1.5–2.5 unnecessary bed-days per DTOC patient on average.

Implement an at-admission DTOC risk screening tool embedded in the clerking workflow. When a patient is admitted, the five risk factors are checked automatically against the patient record and referral data. High-risk patients receive a day-1 social care referral — not a referral when the clinician decides discharge is approaching, but a referral within 24 hours of admission. This gives social care assessors maximum runway and signals to the ward team that discharge planning begins at admission, not at discharge readiness.

Track discharge destination from day 2 for all high-risk patients. If no discharge destination is confirmed by day 2, an automated alert goes to the discharge coordination team to actively source options — care home vacancies, rapid community packages, return-to-residence assessment. Pairing this with a discharge-to-assess pathway, endorsed by Integrated Care Board commissioning, allows patients to be discharged to an interim setting while longer-term arrangements are finalised. This approach, aligned with NHS England discharge-to-assess guidance and Local Authority partnership frameworks, reliably reduces average DTOC duration from 6+ days to 3–4 days.