Proof of concept stage · Inviting NHS co-design partners

Know when every discharge step has to start.

Patients who are well enough to leave hospital often wait because the steps around their discharge start too late. Okas reads what the patient record already holds, gives each step a done-by time, and shows every team its own part.

  • Designed to be read-only
  • No extra data entry
  • No clinical decisions
Bed 12 · made-up patientLive example
Now Tue 08:00 Going home Thu 14:00
    Illustrative example. Time runs fast; the patient and times are made up.

    One discharge, many teams. Okas gives each one its own view.

    Teams: nurse in charge, ward nurses and doctors, physiotherapy, occupational therapy, pharmacy, discharge coordinators, social care, district nursing, patient transport and care transfer hubs.

    The case for change

    Well enough to go home, and still in a hospital bed.

    Delayed discharge is a national problem in England. These figures come from the National Audit Office. They describe the NHS as a whole and say nothing about Okas.

    408,350

    bed days used in April 2026 by patients who were fit to be discharged, 14.7% of all bed days that month.

    £229m

    NHS England’s estimate of what those delays cost in April 2026 alone.

    14.2%

    of hospital stays in 2025–26 involved a delayed discharge.

    Source: National Audit Office, Managing the flow of patients through hospital from A&E (HC 588, 9 September 2026). The NAO notes limits in national discharge data, including that reasons for delay are recorded only for stays of seven days or more.

    What this looks like on a ward

    Steps happen one after another

    Therapy, equipment, medicines, transport and community referrals each depend on the one before. When one starts late, the rest slip.

    Nobody can see how late is too late

    Records show whether a step has been done. They do not show when it needed to start for the patient to leave on the planned day.

    Staff spend time chasing

    Nurses ring round teams to find out where things are, often when the planned date has already gone.

    The product

    One patient record. A clear view for each team.

    Okas is software for hospital operations. It works out when each discharge step needs to be done, then shows each team only what it needs to act on. The views below are from the proof of concept, using made-up patients.

    The whole ward at a glance

    For the nurse running the ward and the board round

    • Every patient with a planned discharge, and which of their steps need attention today.
    • What is running late, while there is still time to act.
    • Where the record has a gap, such as no planned discharge date.
    Ward boardIllustrative
    • Bed 4home Wed 15:00 · 1 step lateAct now
    • Bed 9home Thu 11:00 · 2 steps dueDue today
    • Bed 12home Thu 14:00 · all in handOn track
    • Bed 15no discharge date recordedCheck record

    What the ward needs to send, and by when

    For nurses and doctors looking after the patient

    • Referrals and bookings the ward still needs to send, in order of urgency.
    • Take-home prescriptions and the time pharmacy needs them.
    • What has been sent and is waiting on another team.
    To sendIllustrative
    • Bed 4 · therapy referralsend by Mon 15:00Running late
    • Bed 9 · transport bookingbook by Wed 14:00Due today
    • Bed 9 · take-home prescriptionwrite by Thu 09:00On track

    Assessments in the order they matter

    For physiotherapy and occupational therapy

    • Referrals with the time each assessment needs to be finished.
    • What is on its way but has not been sent yet.
    • Which assessments other steps are waiting on.
    With your teamIllustrative
    • Bed 4 · assessmentcomplete by Tue 12:00Due today
    • Bed 12 · assessmentcomplete by Tue 16:30Done
    • Bed 20 · assessmentnot sent yet · due Wed 10:00Coming

    Take-home medicines, spread through the day

    For hospital pharmacy

    • Which patients are going home, and when their medicines need to be ready.
    • Prescriptions not written yet that will be needed soon.
    • Timing only. Okas does not read what the medicines are.
    Medicines to have readyIllustrative
    • Bed 9ready by Thu 10:00On track
    • Bed 12ready by Thu 13:00On track
    • Bed 4ready by Wed 14:00 · not prescribedRunning late

    The handover to community care

    For discharge coordinators and care transfer hubs

    • Referrals raised by wards that need forwarding to community teams.
    • How much time is left for each one to arrive in time.
    • Referrals by discharge pathway.
    To forwardIllustrative
    • Bed 12 · district nursingforward by Wed 12:00Due today
    • Bed 9 · care packagecomplete by Thu 11:00On track

    Notice of who is coming home

    For district nursing and patient transport

    • Status only: the patient, the planned time, and whether notice has reached the team.
    • Nothing from the clinical record.
    • Time to plan a first visit or a journey.
    Coming homeIllustrative
    • Patient Ahome Thu 14:00Notice received
    • Patient Bhome Fri 11:00Notice expected

    Where it sits in the discharge pathway

    AdmissionDischarge planningReferrals and bookingsGoing homeCommunity handover

    Okas is designed to sit alongside the electronic patient record. It does not replace it or any existing patient flow system.

    What Okas is designed never to do

    • Write to the patient record
    • Ask staff to enter anything extra
    • Question a clinical decision or decide who is fit to go home
    • Name or score individual staff
    • Read clinical notes, diagnoses or what a medicine is
    • Guess a missing date. It reports the gap

    How it works

    Three steps, using information the hospital already records.

    1. Reads

      Okas receives a copy of the timing information a hospital already records about a discharge: when the patient is planned to go home, and when each referral was sent and completed.

    2. Works out

      For each patient, it works out the time by which each step needs to be done for them to leave as planned. Where the record has a gap, it says so.

    3. Shows

      Each team sees its own part: what is with them, what is coming, and what is running late while there is still time to act.

    Want to see it working? We will walk you through Okas with made-up patients. Book a demonstration.

    Who is building Okas

    Khushi Rai, founder of Okas

    A nurse saw the delays on the ward. So she built Okas.

    Khushi Rai

    Founder, Okas Health · Registered nurse (NMC)

    Khushi trained as an adult nurse on NHS wards in London, from medical and surgical wards to intensive care and day surgery. Before nursing she worked in business development, events and operations. She designed, built and tested the Okas proof of concept herself.

    • BSc (Hons) Adult NursingKingston University
    • Clinical placementsAcute medicine and surgery, intensive care, coronary care, theatres and day surgery
    • Commercial backgroundBusiness development, events and operations

    Proof of concept

    What the proof of concept set out to test.

    The proof of concept tests one thing: whether Okas gets the timing right. Benefit to patients, staff and services is measured later, in a pilot and an independent evaluation, against measures agreed with our NHS partner.

    1. Proof of conceptWhere Okas is now
    2. Co-designNext
    3. PilotPlanned
    4. Independent evaluationPlanned
    5. DeploymentPlanned
    6. Routine NHS adoptionPlanned
    How each result was checkedReplay
    1. Physiotherapy assessmentTue 16:30
    2. Care referral sentTue 15:00
    3. Transport bookedWed 14:00
    4. Take-home medicines readyThu 13:00
    5. No planned discharge dateGap expected
    Illustrative replay with a made-up patient. It shows the method: every Okas result was compared with an answer key worked out by hand before the build.
    What was tested
    Whether Okas produces the right done-by time and status for each discharge step, including for records that are incomplete or inconsistent.
    Data format
    Built to read data in FHIR R4, the international standard NHS systems use to share records.
    Where and when
    On a test system outside any NHS network, in November 2026. All patients were made up. No real patient data was used.
    Who took part
    Built and tested by the founder, a registered nurse, using made-up patients only. NHS staff are involved from the co-design stage onwards.
    What we measured
    Each time and status Okas produced was compared with an answer key worked out by hand before the build.
    What we found
    Every timed step matched the hand-worked answer key. Where a record lacked the information needed, Okas reported the gap instead of producing a time.
    What the next stages measure
    How Okas works with a hospital’s live systems, how staff find it in daily use, its safety in practice, and its effect on delays and cost. These are assessed in a pilot and an independent evaluation, not claimed in advance.

    Evidence and context

    The national picture Okas responds to.

    Independent sources on the scale of the problem and on how digital health products should be assessed. They inform our approach; they are not about Okas specifically.

    National audit reportNational Audit Office9 September 2026

    Managing the flow of patients through hospital from A&E

    Reports the scale and estimated cost of delayed discharge in England. Recommends that NHS trusts “more consistently use end-to-end hospital flow management tools and approaches, building on the rollout of electronic patient records”.

    Relevance: sets out the problem Okas is aimed at. The report does not assess any specific product.

    National planNHS England2025/26

    Urgent and emergency care plan 2025/26

    Asks that “all settings should eliminate any internal delays to discharge of more than 48 hours” and that trusts set local discharge targets by pathway.

    Relevance: internal delays are the kind Okas is designed to make visible. Its effect on them is what a pilot would measure.

    National planDepartment of Health and Social Care3 July 2025

    Fit for the future: 10 Year Health Plan for England

    Sets three shifts for the NHS: from hospital to community, from analogue to digital, and from sickness to prevention.

    Relevance: general policy direction only. The executive summary does not discuss discharge timing.

    Evidence frameworkNICEPublished 10 December 2018, updated 9 August 2022

    Evidence standards framework for digital health technologies

    Sets out the evidence digital health technologies should show, including user acceptability, the current and proposed pathway, real-world evidence that claimed benefits can be realised in practice, and budget impact.

    Relevance: we are planning our evidence against this framework from the outset.

    Guidance for adoptersAI and Digital Regulations ServiceLast reviewed 11 October 2024

    Piloting digital technologies in a health or care service

    Says a pilot needs a defined scope, a data-collection plan, clear criteria for success or failure and a defined time period, and should “generate results that help you decide whether to continue using the technology”.

    Relevance: this is the structure we are using to plan our first pilot.

    Where we go from here. The next step is co-design with one NHS ward and its digital team, followed by a pilot with success measures agreed in advance and an independent evaluation.

    What comes next

    From test data to a live ward, one agreed step at a time.

    Each stage starts only when the one before has met a test agreed with our NHS partner.

    1. Proof of concept In progress

      Logic built and tested on made-up patients. Options for connecting to hospital systems are being worked through.

    2. Co-design with one ward Next

      Ward, therapy, pharmacy, community and social care staff shape each view. The trust’s digital, clinical safety and information governance leads agree how data would be shared.

    3. Learn from historical data Planned

      Okas’s times are checked against what actually happened in past discharges, under an agreed data protection arrangement.

    4. Run quietly Planned

      Okas runs alongside normal practice without staff acting on it, so its accuracy can be checked before anyone relies on it.

    5. First ward pilot Planned

      A time-limited pilot on one ward, with scope, measures and success criteria agreed in advance.

    6. Independent evaluation Planned

      An academic or health innovation partner assesses effect on discharge timing, staff workload, usability and cost. Decisions about wider use follow the evidence.

    What NHS deployment involves

    Before any NHS deployment, the relevant clinical, data protection, security, regulatory, usability, integration and evaluation requirements must be assessed with the participating organisation. The list below reflects the areas still to be worked through; it is not a claim of approval or readiness.

    • Connection to a trust’s patient record, agreed with its digital team
    • Clinical safety case under the NHS standards DCB0129 and DCB0160
    • Digital Technology Assessment Criteria (DTAC)
    • Data Security and Protection Toolkit
    • Data protection impact assessment with the trust
    • Confirmation of regulatory status
    • Usability testing with the staff who would use it
    • Independent evaluation of effect and cost

    For partners and funders

    Three ways to work with us now.

    We are looking for a small number of partners for the next stage. Tell us which fits and we will reply with a short briefing.

    NHS trusts and systems

    Help shape Okas on one ward, then host a first pilot.

    • One ward with a discharge improvement goal
    • A named operational sponsor
    • Time with your digital, clinical safety and information governance leads
    Discuss a pilot

    Evaluation partners

    Universities and health innovation networks who can design and lead an independent evaluation.

    • Agree measures before the pilot starts
    • Independent analysis and reporting
    • Joint funding applications
    Explore an evaluation

    Funders and investors

    Support the build, assurance and evaluation work listed above.

    • Development for a first NHS pilot
    • Clinical safety, security and data protection assurance
    • Independent evaluation
    Request a briefing

    Trust and governance

    Governance and assurance.

    Where Okas stands today on the questions NHS organisations ask first. Every statement here can be backed up.

    Company
    Okas Health Ltd, registered in England and Wales, company number 17580519. Registered office: 195–197 Wood Street, London E17 3NU.
    Founder
    Khushi Rai, registered nurse (Nursing and Midwifery Council).
    NHS relationship
    An independent company building with NHS staff in mind. Okas is not an NHS service.
    Regulatory status
    Okas is designed for operational scheduling and makes no clinical decisions. Its formal classification is confirmed before any live use.
    Clinical safety
    Hazard log in place from the proof of concept stage. The full clinical safety case is built with the partner trust ahead of live use.
    Assurance
    Completed with the participating organisation before live use, including DTAC, the Data Security and Protection Toolkit and clinical safety standards. See what NHS deployment involves.

    Privacy and security

    These principles are built into Okas from the start. Each is tested and independently assured with the participating organisation before NHS use.

    Patient data to date
    None. The proof of concept used made-up patients only.
    Read-only by design
    Okas is designed to receive a one-way copy of information, with access that the hospital sets and can withdraw.
    Minimum information
    Okas is designed to use timing information only. It is not designed to read clinical notes, diagnoses or the content of prescriptions.
    This website
    This website does not set cookies or use analytics. The contact form does not store or send anything by itself: it prepares a message for you to send from your own email.

    Questions

    Questions we are asked most.

    Okas has been built and tested as a proof of concept, using made-up patients. The next step is co-design with one NHS ward, followed by a pilot and an independent evaluation. It is not yet in use in any NHS organisation.

    No. Okas is designed to sit alongside the electronic patient record and read from it. Existing systems show what has been done. Okas adds when each step needs to be done.

    It is designed not to. Okas is designed to be read-only, with no function that writes back. How that would be enforced in practice is agreed with each trust’s digital team and tested before any live use.

    Timing information about a discharge: admission, ward, the planned discharge date and time, the discharge pathway, when referrals were sent and completed, and when take-home medicines were requested and ready. It is not designed to use clinical notes or diagnoses.

    Okas does not guess one. It reports that the date is missing, so the gap is visible to the ward.

    Okas is designed for operational scheduling and makes no clinical decisions. Its formal regulatory classification is confirmed before any live use.

    One ward, a named sponsor, and time from your digital, clinical safety and information governance leads to agree data sharing and approvals. The scope, measures and success criteria are agreed before it starts.

    Pricing depends on scope and is agreed with each partner. Pilot terms are discussed directly.

    Yes. We can walk you through Okas working with made-up patients, online or in person. Book a demonstration.

    Get in touch

    Talk to us about a first ward pilot.

    Tell us who you are and what you would like to explore. We will reply with a short briefing and suggest a call.

    Or write to us directly:

    This form does not store or send anything. It prepares a message for you to send from your own email. Please do not include any patient information.