Everyone has a list of decentralized trial tools. Few say how to decide which layers a given protocol needs. This is a layer-by-layer view for small and mid-sized sponsors, with an honest note on what is and is not known about which components lasted.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Layers needed
4 of 7
Systems
1 platform
Audit trails
1
The short version
Framing
Plenty of 2026 articles list DCT tools. This one does not, because tool lists age fast and say nothing about fit. It follows from our state of decentralized clinical trials in 2026 review, which covers the regulatory timeline and evidence, and it does not repeat that material. For operating guidance, see the decentralized clinical trial guide.
The question to answer is this: for my protocol, which layers must exist, which can wait, and how many systems will I end up validating? We use seven layers, drawn from how 2026 guides describe the stack: data capture, electronic consent, patient-reported outcomes, randomization and supply, connected devices, telehealth and home services. One 2026 guide notes that most studies decentralize two or three of these, not all.
What we know
The title of this post is a question, and a fair answer starts with a limit. We searched for studies or sponsor post-mortems that measure which DCT components hybrid trials retained, and found none. What exists is descriptive: vendor and consultancy guides listing layers, press releases about suites, and a few outcome analyses of specific site models.
What can be said with some confidence comes from three observable facts. First, the regulators did not ask for remote conduct; they set out conditions for it. FDA's September 2024 final guidance speaks of decentralized elements, and ICH E6(R3) Annex 2 (Step 4, 3 June 2026) addresses decentralised elements without endorsing an operational approach. Second, a pure-play DCT provider, Science 37, was acquired by eMed in a deal announced in January 2024 at about USD 38 million, after revenue growth stalled. Third, in 2026 vendor guides present eConsent, eCOA and connected devices as parts of a unified platform rather than as separate products. Together they suggest that the elements persisted and the stand-alone full-service model did not. That is an inference, not a measurement.
The layers
| Layer | Our default | Why | Watch out for |
|---|---|---|---|
| Data capture (EDC, eCRF) | Keep | Everything else feeds it. It is the system of record for the study. | Audit trail depth, query handling, exports in analysis-ready form |
| Electronic consent | Keep for remote enrolment | Remote consent is among the most routinely cited decentralized elements, and FDA summaries stress that the investigator remains responsible for obtaining it. | Identity checks, countersignature, record of the exact document signed |
| Patient-reported outcomes (ePRO) | Keep if endpoints are patient-reported | Diaries completed at home are the most common remote data source. | Question types, reminders, licensing of validated instruments |
| Randomization and supply | Keep if randomised | Allocation must stay concealed and traceable, with or without remote elements. | Stratification, blinding, drug supply for home use |
| Connected devices and wearables | Add only for a named endpoint | Useful when passive measures replace a visit-based one. Otherwise they add data without a use. | Participant account linking, sync gaps, device validation, FDA DHT guidance |
| Telehealth visits | Add if remote assessment is justified | Fits assessments that do not need physical examination. | Local licensure, source documentation, ethics committee view |
| Home nursing and direct-to-patient shipment | Handle with care | High operational load and country-specific rules. The EU paper has a national-requirements appendix for exactly this reason. | Product suitability, chain of custody, safety escalation |
Defaults are our editorial view for small and mid-sized sponsors, not a finding from a source.
Capture covers data capture, eConsent, ePRO and randomization under one audit trail. Try all of it in the free sandbox: every feature, no credit card, pay only when you go live with real participants.
Consolidation
Every added system brings a new place where a record can be wrong, a new inspection surface and a new validation task. The regulatory texts do not forbid multiple systems, but they put weight on traceable data and on computerised systems being fit for purpose. Our validation checklist shows how much work that is for even one system.
Practical consequences of a multi-vendor stack include reconciling consent status with enrolment, matching participant identifiers across systems, and answering an inspector who asks for the history of one value that moved through three tools. A single platform with one audit trail does not remove those tasks, but it makes them smaller.
Large, multi-country studies with specialised needs, such as imaging, long-standing site infrastructure or a sponsor that has already validated a specific tool, may reasonably keep separate systems. The point is not that one platform is always right. It is that the burden of integration should be counted when you compare options.
Method
A visit-based exam, a diary, a lab, a sensor. Remote elements should replace a measurement or a burden, not decorate the design.
Ask whether it needs equipment or judgement that cannot be done at home, and whether the ethics committee and regulator will accept the remote version.
Use the table above. If a measurement needs a layer you do not have, decide whether to add it or change the measurement.
Fewer is easier to validate and inspect. Check whether one platform can cover the core layers before adding more.
Consent, first questionnaire, device linking and a reminder, on a phone, as a participant. If it is painful in a sandbox, it will be worse at scale.
Devices
Wearables sit in most DCT diagrams, but 2026 guides treat passive sensor data as an add-on to the core stack rather than a requirement. That seems right. A device is justified when an endpoint depends on continuous or frequent measurement, and it is costly when it only adds a stream nobody analyses.
If you do add one, plan the participant side first. In Capture, the study team enables a device per study and can mark it as required; participants connect their own account through the vendor's sign-in and data syncs into the study automatically, with the last sync time visible. Oura is the example device we show. Read best EDC for wearable data for evaluation criteria, and FDA's digital health technologies guidance for the regulatory view.
Where we fit
Capture is a self-serve, all-in-one platform for Phase 1 through Phase 3: EDC, eCRF builder, ePRO, eConsent, randomization, an AI study builder, safety reporting and monitoring workflows, all under one audit trail. Sites can enrol participants by QR code with no app install, and exports come as CSV or Excel with a data dictionary, or as SDTM datasets (SAS XPT with Define-XML). Data is hosted in the EU (Frankfurt) or the USA (N. Virginia).
Home nursing, courier logistics and telehealth video visits are outside what we describe as the platform, so if your protocol needs them, plan for another provider and plan how that provider's records reach the study database. See decentralized clinical trial software for what to look for, and EDC for decentralized trials for the data side.
Typically data capture, electronic consent, patient-reported outcomes, randomization and supply, connected devices, telehealth and home services such as nursing and drug shipment. 2026 guides note that most studies use only two or three of these.
No source we found measures this. What is observable: regulators framed decentralization as elements, vendors now sell suites, and a pure-play DCT provider was acquired in 2024. Our inference is that the elements persisted inside mainstream operations, but it is an inference.
Not always. If your remote elements are consent, diaries and data capture, an integrated platform can cover them. Telehealth, home nursing and shipment usually need additional providers.
No. 2026 guides treat them as an add-on. Add a device only when an endpoint depends on continuous or frequent measurement.
As few as the protocol allows. Each extra system adds an integration, an audit trail and a validation task. Start by checking whether one platform covers data capture, consent, ePRO and randomization.
Those are outside what we describe as the platform. Capture covers EDC, eCRF, ePRO, eConsent, randomization, safety reporting and monitoring workflows, plus wearable sync through participant sign-in. Telehealth, home nursing and courier services come from other providers.
FDA finalised its decentralized-elements guidance in September 2024 and ICH adopted E6(R3) Annex 2 on 3 June 2026. Both set conditions for remote conduct rather than requiring it. Check the current text for your region.
Keep exploring
State of decentralized trials 2026
Regulation, adoption and evidence.
Decentralized clinical trial guide
How to run a decentralized study.
Decentralized clinical trial software
Capabilities to look for.
eConsent for decentralized trials
Remote consent in practice.
Best EDC for wearable data
Evaluating wearable data capture.
ICH E6(R3) Annex 2 guide
The decentralised annex explained.
Free sandbox with every feature. No credit card, and you pay only when you go live with real participants.