Doccla virtual care platform reports 61% NHS bed-day reduction

TL;DR:

• European virtual care provider Doccla, which runs remote patient monitoring and virtual wards across NHS trusts, says its AI-enabled platform is driving a 61% reduction in bed days, an 89% drop in GP appointments and a 39% fall in non-elective admissions.
• The company cites £450 a day in savings versus the cost of an NHS hospital bed, and an estimated £3 of NHS savings per £1 spent on the technology compared with non-tech models.
• Resultsense view: the Doccla story is the second NHS AI deployment dataset published today, alongside Heidi’s documentation report. They tackle different parts of the same NHS pressure problem — Heidi reduces clinician paperwork, Doccla shifts care out of hospitals — and together they make the case that AI is now operating against multiple NHS bottlenecks at once, rather than as a single-issue productivity bet.

NHS England’s 7.25 million treatment waiting list, the move to community-led care under the “Fit for the Future” 10-year plan, looming doctor strikes and deepening staff shortages provide the backdrop against which both deployments are operating.

What Doccla does

Doccla provides remote patient monitoring and virtual wards designed to support earlier hospital discharge and prevent avoidable admissions, particularly for patients with long-term conditions. Machine learning models combine NHS and proprietary datasets to identify deteriorating patients, while continuous data from clinical-grade wearables — oxygen saturation, blood pressure, ECG — feeds early-warning detection. Doccla’s deputy CEO Michael Macdonnell, who previously worked inside the NHS, said the platform “lets clinical teams intervene sooner and safely manage far larger patient groups than would otherwise be possible”.

The deployment data

The headline 61% bed-day reduction and 89% GP-appointment drop come from Doccla’s deployments inside the NHS rather than independent evaluation. The £450-a-day saving and £3-per-£1 spent return are likewise vendor figures — useful as a deployment signal, but they will need independent evaluation alongside the broader NHS AI scaling programme.

LLMs sit alongside the predictive monitoring layer in Doccla’s stack: the company uses them to streamline clinical notes and present complex information to patients in more accessible language. Macdonnell argued this addresses two bottlenecks at once — administrative load on clinicians and information accessibility for patients — without removing clinical judgement from the loop.

The trust gap

The piece is candid about the unresolved question: clinical trust in AI remains low and will only grow through transparency and further evidence. Predictive models must demonstrate accurate and fair outcomes across diverse patient groups before scaling across real-world clinical settings. That is the work that NHS England’s national AI evaluation programme exists to do, and the gap between vendor-published deployment numbers and independent multi-site evaluations is where Royal Colleges and clinician unions will press the strongest.

UK context

Doccla’s claims sit alongside Heidi’s NHS documentation report and the Department of Health and Social Care’s wider push under the “Fit for the Future: 10 Year Health Plan for England” to move care out of hospitals. The deployment pattern is consistent across vendors: ambient AI inside the consultation, predictive AI between consultations, and virtual-ward infrastructure at the discharge boundary.

Looking forward

The credibility test for both Doccla and Heidi is independent multi-site evaluation that disaggregates time-saved and bed-day reductions from broader operational changes. UK integrated care boards now have sufficient deployment data to weigh procurement decisions but should expect NHS England guidance to evolve rapidly through 2026 as evaluation evidence matures.