47–50%
of before and after measurements end up linked, even in England's national programme where patients score their own outcomes (PROMs), for hip and knee.
NHS PROMs, 2018–2019For specialists who charge fee supplements
Your patient records store consultations. Kliva turns them into evidence.
Kliva is an outcomes registry: it measures how your patients do after each treatment, so you can show in numbers what your care delivers.
- Questionnaires go out on their own, at the right moment.
- Results per treatment, with n and method.
- A dataset that an ethics committee and a journal will accept.
Your consultations stay exactly as they are. Kliva works afterwards, with the software you already use.
Tinnitus, tVNS
Symptom score, lower is better
0%
complete measurement pairs
Built around the software you already use
- Mediris
- Synops
- Vesalius.ai
- Jotform
If you charge more than the national tariff, you should be able to show what it delivers.
In the Belgian debate on fee supplements, nobody argued with outcome data. By the end of July 2027 the sector has to propose a ceiling itself. Practices that can then show results with n and method will be in a stronger position. Today almost no one can, and that is not the doctors' fault.
8.5–18.5 h
of work on the ethics submission for a single retrospective study, spread across eight forms.
Linderkamp et al., 202632×
the same data is typed in again in that submission. In a practice, matching questionnaires to patients is still done by hand, in Excel.
Linderkamp et al., 2026These figures come from published sources outside Belgium and describe the context. They are not Kliva results.
A questionnaire sent at the wrong moment is a measurement you have lost.
Kliva reads your calendar. It knows when a treatment starts and ends, and sends the right questionnaire at intake, at the start, at the end and at follow-up. Nobody has to remember.
Links sent by hand, by email. Answers matched in Excel by name and date of birth.
From consultation to publication, in four steps.
Your daily work stays the same. Kliva works after the consultation.
-
1
Connect
Kliva connects once to your calendar, patient records, questionnaire tool and report software. Patients are matched to their answers automatically. Doubtful matches go to quarantine, so a person checks them.
- A separate, isolated instance per practice, in Belgium
- Encryption, audit log and retention periods
-
2
Measure
The right questionnaire goes out at the right moment, based on your calendar. Free text such as “tVNS session 3” is recognised as a treatment. Patients get a personal progress report as a PDF. There is no app.
- Reminders for missing follow-ups
- Live view of complete pairs per treatment
-
3
Prove
Results per treatment, with n, change and response rate. One click gives a pseudonymised dataset with data dictionary and methods text, plus an ethics pack for a retrospective study.
- Export to SPSS, R or Castor
- Opt-out enforced in every export
-
4
Show
A public results page for your practice, such as “results of our tinnitus treatments, n=312”. Built from the same data as the paper, at aggregate level, and online only after you approve it.
- The same n and method as in your publication
- Groups too small to show stay hidden
One dataset. Four things you do with it.
The dashboard, the paper and the public page come from the same source. So the number on your website has the same n and method as the number in the journal.
| Treatment | n | Complete pairs | Mean change in symptom score |
|---|---|---|---|
| tVNS | 118 | 91% | −14.2 |
| tDCS | 74 | 86% | −9.8 |
| Botox | 41 | 78% | −6.1 |
| IMES | 23 | 61% | too little data |
See what each treatment does
For each treatment, the mean change, the number of patients and how many before and after measurements are complete. You see right away when a treatment has enough data for a study.
See the full demo dashboardFictitious example data.
| Variable | Type | Description |
|---|---|---|
| pid | text | Pseudonym, generated anew for each export |
| age_cat | category | Age in 10-year groups |
| treatment | category | tVNS, tDCS, botox, IMES |
| score_t0 | number | Symptom score at intake, 0 to 100 |
| score_t2 | number | Symptom score after 12 weeks, 0 to 100 |
| opt_out | excluded | Patients who opted out are not in this export |
SPSSRCastor
A dataset that is ready for research
Pseudonymised before reuse, as the Belgian framework law of 30 July 2018 requires. With data dictionary and methods text. The ethics pack contains a protocol template, patient information, an opt-out text and the entry for your record of processing activities.
Draft, to be edited by the author
Transcutaneous vagus nerve stimulation for chronic tinnitus in a private practice: a retrospective cohort study
Methods
We analysed routinely collected data from all patients who received tVNS between January and December. A validated tinnitus questionnaire was administered at intake and twelve weeks after the start of treatment. Patients who objected to the reuse of their data were excluded
- Study design
- Setting
- Participants
- Variables
- Bias
- Study size
A first draft of methods and results
Written from aggregate tables, with a STROBE checklist alongside. The interpretation and the paper stay yours. You edit, you sign.
Results of our tinnitus treatments
n = 312
patients with before and after measurement
12 weeks
time to follow-up
Symptom score
validated questionnaire
Method, response rate and excluded patients are listed at the bottom of this page.
A result a competitor cannot copy
Anyone can write a slogan like “91.5% feel better” without n or source. Nobody can fake a page with n, questionnaire, follow-up time and response rate. You approve every publication before it goes online.
Fictitious example data.
AI that works with your numbers, not decisions about your patients.
Kliva uses a European language model hosted in the EU, or an open model inside your own instance. The model only sees aggregate or pseudonymised data, does not train on your data and keeps nothing.
Question
−13.6 points on average, n = 64, 89% complete pairs. Two out-of-range scores are flagged for review.
Fictitious example. Answers come from aggregate tables.
Kliva does
- Answer questions about your results at aggregate level
- Suggest free text from your calendar as a treatment, for a person to confirm
- Flag missing follow-ups, impossible scores and duplicate patients
- Write a first draft of methods, results and the ethics file
Kliva is not
- A patient records system, it works alongside yours
- A patient app or messaging service
- Advice or alerts about individual patients
- A replacement for the ethics committee
Privacy is built into the product, not added in an appendix.
Reusing health data for research comes with strict rules. Kliva follows them in every installation and every export, so you do not have to work them out yourself.
See what is built and what comes next- Your own instance in Belgium
- Each practice gets an isolated environment. Your data does not sit in a shared database.
- Pseudonymise first
- Anonymous where possible, otherwise pseudonymised before reuse, in the order set by the framework law of 30 July 2018.
- An opt-out that works
- A patient who objects disappears from every export and every chart. Not only from a list.
- Small groups stay hidden
- A result is only shown when the group is large enough that no one can be recognised.
- You stay the owner
- Kliva processes data on your behalf, under a data processing agreement. You can export everything and leave.
One paper a year included. And the registry keeps running afterwards.
A retrospective study done by hand easily costs a practice €6,300 to €28,400 in time and fees. Kliva costs €6,648 in the first year, setup and one publication package included.
Founding practice
€429 per month
Core price (or Plus at €649), €1,500 setup free. For two practices.
- Everything in Core
- Direct contact with the team, and a say in what gets built
- In return for a written reference and a case study
2 places
Core
€429 per month
Up to 3 doctors. One-time setup €1,500.
- Automatic questionnaires and reminders
- Results per treatment
- Research dataset and ethics pack
- Public results page
- One publication package per contract year
Plus
€649 per month
4 to 8 doctors. €60 per extra doctor.
- Everything in Core
- Up to 3 integrations
- Several indications and questionnaire sets
Publication package: one per contract year included, €1,200 per extra study. Frozen dataset, completed ethics pack, first draft of methods and results, one round of revisions. Statistician and publication fees not included.
Prices exclude VAT. Annual contract, billed monthly or yearly, indexed yearly. Extra support beyond normal use at €95 per hour. SMS invitations at cost plus 25%.
What doctors ask us first
Do I have to change my records software or calendar?
No. Kliva connects to what you already use and works after the consultation. Your front desk and your doctors change nothing about their day.
Do I still need an ethics committee?
Yes. Kliva prepares the submission: protocol, patient information, opt-out text and record entry. The review itself, and any committee fee, remain. Practices without their own committee can use an external one, such as the committee of ZNA in Antwerp.
Who owns the data?
You do. Kliva is the processor and works on your behalf, under a data processing agreement. Your data sits in its own instance in Belgium and you can export all of it at any time.
Is the AI a medical device?
Kliva gives no advice about individual patients and writes no summaries for care. The AI works on aggregate numbers for quality and research. That keeps Kliva outside the rules for medical devices.
What if patients do not answer?
Kliva fixes questionnaires that were never sent or arrived too late, and it sends reminders. Patients who choose not to answer will still exist. You see per treatment how many pairs are complete, so you know how strong a result is.
How long does setup take?
We install your instance, connect one calendar or patient records system, prepare the questionnaires for one indication and import your historical data. Expect meaningful results per treatment after six to twelve months of measurements.
Become one of the two founding practices.
A half-hour conversation. We look at how you measure today, which treatment you want to follow first and whether Kliva fits your practice.