The three labels describe where trial activities happen, not how good the science is. This page compares them activity by activity, shows what each one asks of your data system, and ends with a short method for choosing a model for your own protocol.
Free sandbox · No credit card · 21 CFR Part 11 aligned
| Traditional | Hybrid | Decentralized | |
|---|---|---|---|
| Consent | At the site | Remote or at site | Remote |
| Questionnaires and diaries | Often at visits | At home | At home |
| Physical exam and labs | Site | Site | Home, local clinic or telehealth |
| Investigational product | Site pharmacy | Site or shipped | Shipped or local |
| Data entry | Site staff | Site staff and participant | Staff, participant and devices |
The three models in one page
Definitions
People use these three words loosely, so it helps to pin them to something checkable: the location of each protocol activity. In a traditional trial, the participant comes to the site for screening, consent, procedures, questionnaires, dosing and follow-up. The site team collects data, usually into an electronic data capture system, and a monitor reviews it centrally or on site.
A decentralized trial moves some or many of those activities out of the clinic. The FDA guidance on conducting clinical trials with decentralized elements, finalised in September 2024, frames it this way: trial-related activities occur at locations other than traditional sites, for example through telehealth visits, in-home visits by trial personnel, or visits with local health care providers. Read the guidance itself for the precise wording, and for what it asks of sponsors on oversight and data.
A hybrid trial mixes the two. A study might consent people online, collect weekly symptom diaries on their own phones, and bring them to the site only for a physical exam and blood draw at baseline and week 12. Regulators have not agreed one formal definition of "hybrid" in the texts we checked, and the EMA and HMA recommendation paper on decentralised elements (December 2022, non-binding) talks about elements rather than models. So when you write a protocol, name each element and where it happens. That description is what ethics committees and inspectors will read.
A truly site-less study is realistic mainly where the endpoint can be measured at home, such as a symptom diary, a wearable signal or a simple self-collected sample. Many trials called decentralized still rely on a principal investigator, a medical monitor and a physical site as the legal and clinical anchor. Treat "decentralized" as a point on a scale, not a switch.
Side by side
The descriptions are general. A particular protocol, indication or country can change any row.
| Dimension | Traditional | Hybrid | Decentralized |
|---|---|---|---|
| Where activities happen | Almost all at the investigative site | Split: site for procedures, home or remote for the rest | Home, local providers, telehealth, with limited or no site visits |
| Who can take part | People who can reach a site regularly | Wider reach; some site travel remains | Widest geographic reach; needs a phone or device and reliable connectivity |
| Consent | On paper or on a site tablet, in person | Either, often remote with a site conversation | Remote with identity checks and a way to talk to staff |
| Patient-reported data | Collected at visits, sometimes on paper diaries | Collected at home between visits | Collected at home, on the participant's schedule |
| Clinical assessments | By site staff at scheduled visits | Site for those that need it; telehealth for others | By telehealth, home visit staff or local clinics |
| Monitoring | Site visits and central review | Central review plus targeted site visits | Mostly central; needs strong data checks and visit-level oversight |
| Main operational risk | Missed visits and slow recruitment | Handoffs between site and remote steps | Data from many sources, identity, local provider quality and technology failure |
| Data system demands | EDC with a good site workflow | EDC plus consent and participant-facing forms in one place | All of the above, plus devices, reminders and remote oversight |
Choose the model per endpoint and visit, not per study name. The label follows from the design, not the other way round.
Trade-offs
Traditional designs earn their place through control. Procedures are done by trained staff in a known environment, equipment is calibrated, and the monitoring approach is familiar to every ethics committee. The costs are travel burden, which affects who enrolls and who drops out, and the dependence on a small number of sites for recruitment.
Decentralized elements reduce travel and can widen who is able to join, but they move work rather than remove it. Someone still has to verify identity, handle a shipped product, check that a home measurement was taken correctly and deal with a participant who cannot get a signal. Data arrives from more places, so quality checks need to run earlier and more automatically. A sponsor has to be able to show oversight of activities that happen outside its direct view.
Hybrid models try to take the best of both, and they bring a specific risk: the seams. A participant who completes a diary at home, attends a visit, then receives a call from a local nurse produces three streams of data that must land against the same visit and the same participant ID. The more you can keep those streams in one system with one audit trail, the fewer reconciliation errors you carry into database lock. That is also why a platform that combines EDC, ePRO and eConsent matters more in hybrid designs than in purely site-based ones.
Diary completion, last 7 days
86%
Open queries
14
Consents awaiting countersignature
3
Visits out of window
1
What the platform covers
Capture is a self-serve clinical trial platform, so the same study can hold the site-entered CRFs and the participant-entered forms. That is the part of hybrid and decentralized designs a data system can help with. It does not supply home-visit nurses, couriers or telehealth video.
Enrolls from QR code
Phone, no app install
Signs consent
Countersigned by site staff
Baseline visit at site
Coordinator enters eCRFs
Weekly diary at home
ePRO reminder
Week 12 visit at site
Edit checks run on entry
Build the hybrid version of your schedule in the free sandbox: site forms, a participant diary and a consent flow. Walk through it as a coordinator and as a participant, and see which handoffs feel fragile.
Choosing
Work from the endpoints outward. The best model is usually the least complicated one that still gets reliable data from the people you need.
Consent, screening, each assessment, dosing, sample collection, questionnaires, safety follow-up. One line each, with who does it today.
Procedures that need equipment, trained staff, a controlled setting or direct observation stay at the site. Be honest about this; it is the anchor of the design.
For each remaining activity ask whether the data is as reliable at home, who checks it, and what you do when it fails. Diaries and consent usually pass; complex measurements often do not.
Who are they, how far do they live from a site, and can they use a phone? If the population has limited connectivity, a fully remote design may exclude the people you most need.
Decide what central checks, reminders and reviews will show you that remote activities are happening correctly. Write these into the monitoring plan.
Run a practice participant through each setting, with practice data, before approval. Fix the handoffs while it costs nothing.
Data system questions
Whatever you use, put these questions to the vendor and test them in a sandbox.
Site-entered and participant-entered data attach to the same participant and visit. See EDC for decentralized clinical trials.
Identity checks, a way to ask questions and a countersignature step. See remote patient consent.
Questionnaires reachable from a phone browser. See ePRO software without app development.
Every change, whoever made it, in the same history, exportable for inspection.
Coded data for researchers, identity only where needed, as the consent form promised.
Edit checks and auto-queries catch implausible values the moment they arrive, which matters when you cannot look over the shoulder of a participant.
Data and a data dictionary your statistician can use without reverse engineering the system.
Regulatory context
The headline point from FDA's final decentralized trial guidance is that the requirements are the same whether or not a trial includes decentralized elements. That means the sponsor stays responsible for participant protection, for the reliability of the data and for oversight of everyone who performs trial tasks, including local providers and technology vendors. The guidance was published in the Federal Register on 18 September 2024 and finalised a May 2023 draft. Read the current text for the details on investigator oversight, data from digital health technologies and drug shipment.
In the EU, the Clinical Trials Regulation applies to all trials, and the EMA and HMA recommendation paper on decentralised elements (December 2022) is non-binding. It points readers to national provisions, which differ between member states. If you recruit in more than one country, check each one before assuming a remote step is accepted. Our decentralized clinical trial guide goes deeper on running one, and the electronic informed consent regulations guide covers the consent side. Nothing on this page is legal advice.
Go deeper
What decentralized elements are and how to plan one.
Read more SoftwareWhat to look for in a platform for mixed designs.
Read more SoftwarePlatform needs when most activity is remote.
Read more How-toHow participants send data from home.
Read more GuideA closer look at the first activity most teams move out of the clinic.
Read moreA decentralized trial moves trial activities to places other than traditional sites, such as the participant's home, local providers or telehealth. A hybrid trial keeps some activities at the site and moves others out. There is no single regulatory definition of hybrid in the sources checked, so describe the split activity by activity in the protocol.
Neither is better in general. Traditional designs give more control over procedures and are familiar to reviewers. Decentralized elements can reduce travel and widen who can take part. The choice depends on the endpoints, the population and what can be measured reliably away from a site.
FDA states that its regulatory requirements are the same for trials with and without decentralized elements. Good clinical practice, informed consent, data integrity and safety reporting still apply. In the EU, national provisions can differ, so check each country where you recruit.
Yes, and many do in practice. Moving consent and questionnaires to the participant's phone while keeping physical assessments at the site is often the simplest hybrid design and needs no special logistics beyond a data system that handles both site and participant entry.
At minimum, an EDC for site data, participant-facing forms for diaries and questionnaires, and electronic consent, ideally in one system with one audit trail so the home and site data attach to the same participant and visit.
Capture covers the data side: site eCRFs, participant ePRO forms, eConsent with on-screen signature and countersignature, QR-code enrollment, role-based access and a field-level audit trail. It does not provide home-visit staff, couriers or telehealth video, which you arrange separately.
Not necessarily. It can lower participant travel and some site costs, but it adds work in remote oversight, logistics and technology support. Compare the full cost of each design for your protocol rather than assuming savings.
Keep exploring
Decentralized clinical trial guide
Plan and run decentralized elements.
EDC for hybrid clinical trials
Data capture for mixed designs.
EDC for decentralized clinical trials
Data capture when most activity is remote.
Hybrid clinical trial software
Platform checklist for hybrid designs.
eConsent for decentralized trials
Remote consent in practice.
What is eConsent?
How electronic consent works and what the rules say.
Remote patient monitoring
Data from home and devices.
Free sandbox with every feature. No credit card, and you pay only when you go live with real participants.