Your product is software, but the evidence still has to look like a clinical trial. Capture gives a lean DTx team an audited EDC, ePRO in the phone browser and eConsent in one system, so the pivotal study does not need a five-vendor stack or an enterprise sales cycle.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Randomized
84 / 120
Week-8 PRO complete
91%
Open queries
6
Reminders sent to 4
What a DTx team needs from a data system
The problem
Digital therapeutics teams usually come from product engineering. They already have data: sessions, module completions, in-app scores. What they lack is the clinical-grade record an FDA reviewer, a notified body or a payer expects to see alongside it: who consented and when, who was randomized to what, which outcome was collected under which conditions, and who changed which value afterwards. Prescription digital therapeutics are regulated as software as a medical device, and those that have reached the US market have generally done so through the De Novo or 510(k) routes, supported by controlled clinical data. Your regulatory strategy decides the exact evidence bar; your data system should not be the thing that limits it.
The usual options are awkward for a team of this shape. A spreadsheet stack has no audit trail. An enterprise EDC is built for sponsors with data-management departments and arrives with an implementation project. Stitching together a survey tool, an e-signature tool and a randomization script produces four logins, four audit logs and an integration problem you now own. Capture puts the clinical study in one place: EDC, ePRO and eConsent share one participant record and one field-level audit trail.
Patient-reported outcomes
Most digital therapeutic trials measure how people feel or function: sleep, mood, pain, craving, symptom burden. Participants receive a link by email or SMS, complete the questionnaire in their phone browser with no second app to install, and each response is timestamped against the visit window the protocol defines. Reminders chase late responses, and the dashboard shows who is overdue.
Evening diary
Question 3 of 8
How would you rate your fatigue today?
0 = no fatigue, 10 = worst imaginable
Randomization and blinding
A sham-app or waitlist control only protects the result if the people assessing outcomes cannot see allocation. Capture randomizes participants inside the study and gives blinded roles masked views, so treatment-arm values are never sent to a blinded user. The restriction is enforced in the database layer, not only hidden in the interface.
Study coordinator
Subject 01-004
Randomized · R-0012
KitK-10418
Arm columnNot sent
Unblinded administrator
Subject 01-004
Randomized · R-0012
KitK-10418
Arm columnA
Same subject, same screen. The arm never reaches blinded roles.
Create the schedule, the primary PRO and the randomization, then complete a visit on your own phone. Every feature, no credit card, no time limit.
Architecture
Your app generates usage data all day: logins, lesson completions, exercise counts. That data matters for dose-response and engagement analysis, but it belongs to your product stack, and it is not the same thing as the case report form. A clean design keeps them apart and joins them on the coded participant ID at analysis time.
Capture does not claim a built-in connection to your app's telemetry, so plan that join as an export-and-merge step: your engineers export engagement summaries per participant, the data manager exports the clinical dataset from Capture, and the statistician merges on the participant number. Capture exports CSV and Excel with an automatic data dictionary, and SDTM datasets as SAS XPT with Define-XML for a regulatory submission; see exporting EDC data to R, SAS and SPSS and CDISC SDTM export.
Consent and eligibility, randomization, clinic-measured outcomes, participant-reported outcomes, concomitant treatment, adverse events and device deficiencies or product issues that your protocol asks you to log. Safety events recorded in the eCRF carry the same audit trail as every other field.
High-volume interaction logs, A/B test infrastructure and anything that is not a protocol-defined data point. Decide before the first participant which engagement summaries the statistical analysis plan will use, and export only those.
How a study runs
Upload the protocol and let the AI study builder draft the visit schedule and forms, or build by hand. Nothing is saved without human review.
Enter sample participants, trigger edit checks, sign forms, and read the audit trail. Your clinical lead and your regulatory consultant can review the same sandbox.
Approved forms are locked for live use. Invite sites or enrol participants by link and QR code. You pay only once real participants are in.
Consent and screening, randomization, visit-linked PROs with reminders, queries on out-of-range values, and a dashboard for completion and open queries.
Resolve queries, export the analysis dataset with its data dictionary (and SDTM with Define-XML where required) for your statistician.
Is it a fit
| DTx need | In Capture |
|---|---|
| Remote consent for app-based enrolment | eConsent with on-screen signature, email OTP confirmation and countersignature; signed PDF export |
| Primary PRO on a schedule | ePRO in the phone browser, visit windows, reminders, completion dashboard |
| Sham or control app, blinded assessor | Randomization in the study; blinded roles receive masked views |
| Clinic or remote assessor visits | Site forms with edit checks, auto-queries and calculated fields |
| Audit-ready record | Field-level audit trail: user, time, old value, new value, reason for change |
| Statistical analysis | CSV/Excel with data dictionary; SDTM XPT with Define-XML |
| Data residency | Hosted in the EU (Frankfurt) or the USA (N. Virginia) |
Capture is the data-capture system for the study. Your quality system, risk management file and regulatory submissions remain your own responsibility.
Capture suits Phase 1 to 3 style studies, including a controlled pivotal trial. It provides consent, randomization, PRO and site data capture and a field-level audit trail. What evidence your regulator expects is a matter for your regulatory strategy.
Not as a built-in integration we claim. Plan a join on the coded participant ID at analysis time: export engagement summaries from your product and the clinical dataset from Capture, then merge.
Give assessors a blinded role. Blinded roles never receive treatment-arm values, enforced by masked database views, and unblinding events appear in the audit trail.
No second app is needed for questionnaires. Participants open a link on their phone and complete the form in the browser.
Yes, if you hold the right licence. You enter the licence details on the patient questionnaire; Capture does not supply the instrument or verify the licence with the rights holder.
No. The sandbox includes every feature with no credit card and no time limit. You pay only once you go live with real participants.
In the EU (Frankfurt) or the USA (N. Virginia), depending on your study.
Keep exploring
Build the schedule, the primary PRO and the randomization in the sandbox. You pay only once you go live.