ClinicOS / Physiotherapy & rehab
No reassessment, no outcome. No approval, no session.
From the fourth quarter of 2026, every outpatient rehab provider licensed in Abu Dhabi reports its outcomes to the Department of Health each quarter, patient by patient, from a documented baseline and a day-42 reassessment. Across the UAE, insurer-funded physiotherapy already runs on prior approvals and session caps. ClinicOS puts both on the physiotherapist’s board, writes the note from the readings they take, and drafts the insurer’s request from the record. It sits beside your EMR, not instead of it.
ClinicOS for rehab in 90 seconds. The people are characters; the screens are the product, running on demo data.
Four people work on one recovery. Each sees a different part of it.
The treatment is rarely what goes wrong. What goes wrong is everything around it: a reassessment nobody booked, an approval that ran out, a home programme dropped in the second week.
Does the exercises when he remembers.
Most of rehab happens at home, on a sheet of paper. When it stops working, the first sign is usually a missed visit, and then no visit at all.
Twelve patients a day, then twelve notes at night.
The readings are taken in the session. The note, the progress report and the insurer’s request are written afterwards, from memory, in time that should be rest.
Replaces the knee, then hears nothing for weeks.
What happens after the operation decides whether it worked. The outcome usually comes back at the follow-up visit, if the patient comes.
Loses sessions to approvals that quietly ran out.
An approval that ends before the plan does is a session not paid for, or not given. The extension request is a report someone has to write, usually at night.
Measure. Type. Sign.
2 min 4 s. Arjun Menon, a physiotherapist, goes from writing notes at 8:55 pm to leaving at five; Sarah Whitfield, his senior, sees the whole team’s caseload.
The day starts with what needs attention, and the paperwork is done before the patient leaves.
- The board sorts the day. Approvals running low or used up, reassessments due or overdue, home practice slipping, a pain score that rose at home. Each patient carries the reason, in words, not just a colour.
- Sessions at risk, in dirhams. Visits booked in the next 14 days that no approval covers yet, priced from the payer’s list, and the extensions signed this week.
- Three readings in, the SOAP note out. Knee bend, pain on the stairs, Timed Up and Go. The note is written from those readings; every number in it is one the physiotherapist measured, and it says which changes are clinically meaningful.
- The extension request, drafted from the record. Sessions approved, used and left, each measure against baseline and goal, attendance and home practice. The physiotherapist adds a sentence in their own words and signs.
My own stairs
1 min 57 s. Saeed Al Nuaimi, six weeks after a knee replacement, and his daughter Mariam, who follows his programme from work.
The programme moves from a sheet of paper to his phone, and what happens at home reaches the clinic.
- His home programme, in Arabic or English. Each exercise, how many and how often, in the words his physiotherapist wrote. A tick when it is done.
- A pain score that travels. After the exercises he gives his pain a number and, if he wants, a line about it. A rise above his usual raises a flag on his physiotherapist’s board the next morning.
- “My family”. A family member linked to his record can see his programme and his next visit, without having to ask him.
- Progress against his own goal. Eighty degrees at the start, one hundred and four now, one hundred and fifteen to go.
What the evidence supports
No single feature moves adherence much on its own; the effect comes from goals, daily logging and a therapist who checks in, together. That is the combination ClinicOS puts on the patient’s phone and the physiotherapist’s board.
See the recovery too
1 min 38 s. Dr Khalid Al Mazrouei, an orthopaedic surgeon, follows the patient he operated on through rehab.
The surgeon owns the operation. Now they can see the recovery, and the group can see every site.
- The episode on one page. Knee flexion, pain on the stairs, Timed Up and Go: each reading against the baseline and against the goal.
- Clinically meaningful, by a threshold. A change counts as meaningful when it passes the minimal clinically important difference for that measure: published values where they exist, set by your clinic where they do not.
- Why the sessions are needed. The signed extension request shows the referrer exactly what was asked of the insurer, and on what numbers.
- The Jawda view, site by site. Baselines documented, reassessments done by day 42, 60 or 84, pain and strength improved, each against the target, with the patients pulling the figure down listed first.
What Abu Dhabi counts, from Q4 2026
| Indicator | Who is counted | Counted as improved | Reassessed at | Target |
|---|---|---|---|---|
| REH001 Pain | Musculoskeletal and orthopaedic, aged 12+, physiotherapy or OT | Pain down 2 points or more (NRS, or VAS equivalent) | Day 42 or discharge | ≥ 70% |
| REH002 Strength | Physiotherapy or OT, all ages | Muscle strength up more than one grade | Day 42 (MSK) · day 84 (neuro) | ≥ 70% |
| REH003 Balance | Adult neurological physiotherapy | Functional Reach up 3.7 cm or more | Day 42 | ≥ 70% |
| REH004 Mobility | Ambulatory adult neurological | Timed Up and Go faster by more than 2.9 s | Day 60 | — |
Four of the indicators in the set; the guidance also covers paediatric goal attainment, speech and swallowing, vestibular and cardiorespiratory rehab. Each needs a documented baseline before therapy; a patient with no reassessment in the quarter stays in the denominator and is not counted as improved. ClinicOS’s Jawda view is modelled on this guidance; definitions are confirmed against the published specification before any submission.
Source: Department of Health – Abu Dhabi, “Outpatient Rehabilitation Service JAWDA Performance Indicators”, Version 1, issued May 2026, effective Q4 2026, doh.gov.ae.
For the quality lead: run your own numbers
A missed reassessment does not drop out of the figure. It stays in the denominator and counts against you. Move the sliders to your own quarter and see what the pain indicator would report.
Illustrative arithmetic on REH001’s counting rule, not a measured result: reported = reassessed on time × improved among those reassessed. At the defaults, a team that helps 85% of the patients it reassesses still reports 68%.
What changes, person by person
| Who | Before | With ClinicOS |
|---|---|---|
| Patient | A printed sheet of exercises and a pain they mention at the next visit. | The programme on their phone in Arabic or English, a tick a day, and a pain score that reaches the physiotherapist. |
| Physiotherapist | Notes and extension reports written at night, from memory. | Readings typed in the session, a SOAP note written from them, an extension request drafted from the record; they read and sign. |
| Senior physiotherapist | Finds out an approval ended when the patient is at the desk. | The team’s caseload by what needs attention: approvals, reassessments, home practice. |
| Referring surgeon | Hears about the outcome at follow-up, if the patient comes. | Each measure against baseline and goal, the extension request and the home practice, from the record. |
| Finance | Sessions given beyond an approval, found when the claim is rejected. | Sessions at risk in dirhams before they are given, and extensions signed this week. |
| Quality lead | The Jawda figure assembled at quarter end from whatever was documented. | Baselines and reassessments by site as the quarter runs, with the patients pulling the figure down listed first. |
What the AI does not do
A group should know where the line is before it lets software write to an insurer or into a patient’s record. In ClinicOS it is drawn like this.
It never supplies a number.
Every reading is typed by the physiotherapist and checked against the measure’s range. The model is given words, not values; a draft whose figures do not match the record is rejected and written in the standard layout instead.
It never signs.
Notes and extension requests are drafts until a clinician reads and signs them, and a signed request says so: “AI-assisted draft — reviewed by” the physiotherapist who signed it.
It never decides.
Whether a change is clinically meaningful is a threshold set per measure, not a model’s opinion. The plan, the request and the discharge stay with the clinician.
Signed documents are kept as PDFs, Arabic included.
Your EMR stays the record. ClinicOS runs rehab.
A hospital EMR handles rehab as forms and charges. What a rehab service needs on top is the episode: the approval and its cap, the measures against baseline, the reassessment clock, the home programme. That is the part ClinicOS runs.
Your EMR
- The master patient record
- Orders, results and inpatient care
- The group’s revenue cycle
ClinicOS Rehab
- Episodes, approvals and session caps
- Outcome measures and the Jawda view
- Home programmes the patient keeps
- AI drafts the physiotherapist signs
Where the EMR is live at a site, ClinicOS sends each signed document back to it, over FHIR where the EMR allows. Where it is not live yet, ClinicOS runs the rehab service now and the data moves with it when the EMR arrives. Malaffi, NABIDH and Riayati document feeds go through your existing integration route.
Many sites, three regulators
Abu Dhabi, Al Ain, Dubai and the Northern Emirates in one view, each site against its own emirate’s payers and caps.
Clinic, home and long-term care
One plan of care follows the patient from the hospital gym to the home visit.
A new site is configuration
A new site or brand is set up in the same instance, not bought as a new system.
Measured against your own numbers
From an extract of your rehab data, at no cost: reassessments done on time, extension requests and how they ended, sessions lost to expired approvals.
One or two sites you choose, connected to your EMR route. Your physiotherapists, your patients.
One Jawda reporting quarter, compared with the baseline on the three numbers you name.
Then buy it the way that suits your group. Take a licence for your own instance, single-tenant, with the source code as an option. Or take it under Outcome-as-a-Service: no licence fee up front, a fee per billed physiotherapy session, the unit your rehab service is paid on, and payment from go-live. A quiet month is a small invoice. Your data leaves in standard formats (FHIR) whenever you ask, and scope changes in the pilot are not charged.
Built on the Xamun platform
More than 50 systems built on it in its first two years; delivery by BlastAsia, est. 2000.
No Questions, for three years
If an agreed outcome number slips, we fix it without asking why.
Deliberately small in the UAE
Delivery is founder-led from DIFC, so we take on a limited number of new engagements each quarter.
Questions rehab teams ask
Who is this for?
ClinicOS for physiotherapy is for outpatient rehab: hospital physiotherapy departments, rehab clinic groups across several sites, and home physiotherapy services. It was designed around the UAE: Abu Dhabi’s outpatient rehab Jawda indicators, insurer prior approvals and session caps, and patients who read in Arabic or English.
Does it replace our EMR?
No. ClinicOS sits beside it. The clinic chooses whether the EMR or ClinicOS holds the clinical record; where the EMR holds it, each signed note and report is queued for it as a document, and sent over FHIR where the EMR allows. Where the EMR is not live yet at a site, ClinicOS runs the rehab service now and the data moves with it when the EMR arrives.
Does the AI write the notes on its own?
No. The physiotherapist types the readings they measured; ClinicOS writes the SOAP note from them and states which changes are clinically meaningful. The model is given words, not numbers, and a draft whose figures do not match the record is rejected. The physiotherapist reads, edits and signs.
How does it help with insurer approvals?
Each programme carries its approval: sessions approved, used and left, and the expiry date. The board flags approvals running low and caps reached, and shows sessions booked beyond an approval in dirhams. When an extension is needed, the request is drafted from the record, with the measures against baseline and goal, attendance and home practice, and the physiotherapist signs it.
What does the Jawda view show?
For each site and quarter: baselines documented, reassessments done by day 42, 60 or 84, and patients with pain and strength improved, each against the target, with the patients pulling the figure down listed first. It is modelled on the Department of Health’s Outpatient Rehabilitation Jawda guidance V1; definitions are confirmed against the published specification before any submission.
What does the patient see?
Their own home programme on their phone, in Arabic or English: each exercise, how many and how often, a tick when done, and a pain score with an optional note that reaches their physiotherapist. A family member linked to their record can see the programme and the next visit.
How do we buy it?
Most groups start with a pilot: a baseline from an extract of your rehab data at no cost, then one or two sites live, then one Jawda reporting quarter compared with the baseline on three numbers you name. After that, a licence for your own single-tenant instance, or Outcome-as-a-Service: no licence fee up front and a fee per billed physiotherapy session, from go-live.
Can we try it?
Yes. The demo at physio.clinicos.xamun.ai runs on a fictional three-site rehab group, with sign-ins for the group admin, the senior physiotherapist, the physiotherapist, the referring surgeon, finance and the patient. Or book a walkthrough and bring one site’s numbers.
All sources cited on this page:
1. Department of Health – Abu Dhabi, “Outpatient Rehabilitation Service JAWDA Performance Indicators”, Version 1 (issued May 2026, effective Q4 2026), doh.gov.ae.
2. Insurance System for Advancing Healthcare in Dubai (ISAHD), General Circular 03/2024, Essential Benefits Plan, isahd.ae.
3. Balinski et al., “Factors Associated With Physical Therapy Attendance Rate and Self-Discharge in People With Chronic Pain”, JOSPT Open 2(4):332–339 (2024), doi:10.2519/josptopen.2024.1136.
4. Ley & Putz, “Efficacy of interventions and techniques on adherence to physiotherapy in adults: an overview of systematic reviews and panoramic meta-analysis”, Systematic Reviews 13:137 (2024), doi:10.1186/s13643-024-02538-9.
5. Taylor, Bottrell, Lawler & Benjamin, “Mobile telephone short message service reminders can reduce nonattendance in physical therapy outpatient clinics: a randomized controlled trial”, Archives of Physical Medicine and Rehabilitation 93:21–26 (2012).
Name three numbers. We bring the baseline.
Tell us the three numbers that would make rehab a success for your group, in your own words. We build the baseline from an extract of your data and show your sites running in ClinicOS, from the physiotherapist’s board to the Jawda view.