What FDA's final guidance on DHTs for remote data acquisition asks of sponsors and investigators, section by section, and what it means for the data system that receives the readings. General information, not legal advice.
Free sandbox · No credit card · 21 CFR Part 11 aligned
DHT records the measurement
Sensor, wearable or app
Data and metadata transmitted
Secured in transit, alerts on failure
First durable electronic data repository
FDA treats this copy as source data
Investigator review per safety plan
As specified in the protocol
Retention and inspection
Human-readable, with metadata
Key points
The document
FDA finalized the guidance on 22 December 2023, after a 2021 draft. It responds in part to section 3607 of the Food and Drug Omnibus Reform Act of 2022, which asked FDA to issue guidance on DHTs in clinical investigations. It applies to investigations of drugs, biological products and medical devices, and it is addressed to sponsors, investigators and other stakeholders, including IRBs.
Three other FDA documents are usually read with it. The Framework for the Use of Digital Health Technologies in Drug and Biological Product Development (March 2023) sets out how CDER and CBER organize their DHT work; it is a program framework, not guidance. Conducting Clinical Trials With Decentralized Elements (final, September 2024) covers remote visits and local providers. And Electronic Systems, Electronic Records, and Electronic Signatures in Clinical Investigations: Questions and Answers (final, October 2024) gives FDA's current Part 11 recommendations, including access controls and audit trails for data captured remotely. Our 21 CFR Part 11 compliance checklist covers that side in detail.
Whether a DHT is itself a device matters for the paperwork, not for the core expectations. The guidance explains when an investigational device exemption may be needed, for example for a significant risk DHT used in a drug trial, and when a cleared device used within its indications needs no IDE. Verification and validation are expected either way.
Section by section
Section IV of the guidance holds the practical recommendations. This table maps each part to what a sponsor typically writes down.
| Guidance section | What FDA recommends | What you document |
|---|---|---|
| A. Selection and rationale | Match the DHT to the clinical event, the population and the trial design; set minimum technical and performance specifications | Selection rationale, model and version list, population considerations |
| B. Description in a submission | Explain why the DHT is fit-for-purpose and describe the data flow to the first durable repository | DHT description, data flow diagram, access controls, data management plan |
| C. Verification, validation, usability | Objective evidence that the DHT measures accurately and captures the event in your population; usability evaluations | V&V reports (yours or the manufacturer's), usability findings |
| D and E. Endpoints and statistics | Justify novel endpoints; plan the analysis and how missing data are minimized and handled | Endpoint justification, statistical analysis plan |
| F. Risks | Clinical, cybersecurity and privacy risks, including end-user licence terms that share data | Risk assessment, consent language |
| G. Record protection and retention | Transfer data and metadata securely to a durable repository; keep it human-readable | Retention plan, source data definition |
| H. Other considerations | Sponsor and investigator roles, training, updates, loss and error procedures, closeout | Training materials, risk management plan, safety monitoring plan, update log |
Summary only. Read the full guidance on fda.gov for the exact recommendations and footnoted regulations.
Selection and validation
The guidance separates two ideas. Verification confirms that the device measures the physical parameter, such as acceleration or temperature, accurately and precisely. Validation confirms that it measures the clinical event or characteristic, such as step count or heart rate, in your population. FDA gives the example of a step algorithm built on healthy volunteers that may not hold for people with a shuffling gait. Evidence can come from the manufacturer, from device labeling or from a right of reference to another submission, but the sponsor has to judge whether it fits the trial.
Usability gets its own attention. Sponsors should look for use errors before the trial starts, and feed what they learn into training. Selection should also consider who will use the device: age, language, education, health and technical skill all affect whether participants can use it as the protocol intends.
Sponsors may let participants use their own DHT or their own smartphone, which can reduce burden. The guidance asks sponsors to weigh this, notes it may not suit highly specialized devices, and says sponsor-provided devices and connectivity should be available so that people without their own are not excluded. The BYOD ePRO guide covers the questionnaire side of this choice.
What it means for your data system
Section G is the part that lands on the data system. Data captured by the DHT, with relevant metadata such as when each measurement was made, should be securely transferred to and kept in a durable electronic data repository as part of the trial record. For data collected directly from participants, FDA considers the data in the first durable repository they reach to be the source data, and that copy should be available for inspection. FDA generally does not intend to ask for raw machine data such as voltages that need processing to be understood, but the data must be retained and human-readable.
That has practical consequences. Your submission should name the repository and draw the flow from device to repository. You need to know whether the first durable copy sits with the device maker, a middleware vendor or your EDC, because that is where retention, access control and audit trail expectations apply. The investigator must also be able to review source data where the protocol says so, and FDA generally does not expect continuous monitoring; the safety monitoring plan should say how often continuous data are reviewed and how abnormal readings are handled.
Section H then lists what the sponsor should plan: training for staff and participants before use, technical support, a risk management plan covering misuse, loss, malfunction and updates, confirmation that data actually reached the repository, and closeout steps such as ending transmission and revoking access. For the wider integrity picture, the ALCOA+ data integrity checklist is a useful companion.
Devices in scope
1
Open items
2
Build the visits, ePRO and device settings in the free sandbox, test the audit trail and exports, and pay only when you go live.
Risks, consent and change
Taken from sections F and H of the guidance.
Some devices and phone platforms have licence terms that let the maker access data. Consider negotiating study terms, and tell participants who may access their data and for how long.
Explain what the device collects, how it is used and monitored, what to do about a concerning reading, privacy limits and any added costs such as data charges.
Keep a record of the timing and nature of updates per device. If an update could change measurements, check before and after data and plan sensitivity analyses before unblinding.
Have replacement procedures and, where possible, an alternative way to record data while a device is out of use.
Protect data at rest and in transit, and consider FDA's cybersecurity information where a compromised reading could harm a participant.
Define when transmission ends and when system access is revoked at the end of the study.
Where Capture fits
Most of the guidance is sponsor work: choosing and validating the device, justifying endpoints and writing the plans. A study platform helps with the record those plans depend on. In Capture, a study team enables a wearable per study and can mark it required; the participant connects their own account through the vendor's sign-in (for example Oura), and data then sync automatically, with a last-sync time, a Sync now option and a disconnect control. Device data sit in the same study as site eCRFs, ePRO and eConsent.
Every change to study data is written to a field-level audit trail with the user, timestamp, old value, new value and reason, and the trail is append-only and hash-chained. Access is role-based, with participant names visible to site staff and coded IDs for researchers. Data are hosted in the EU (Frankfurt) or the US (N. Virginia), encrypted with AES-256 at rest and TLS in transit, and export to CSV or Excel with a data dictionary, or to CDISC SDTM datasets with Define-XML. These are 21 CFR Part 11 aligned controls; your validation, procedures and DHT evidence complete the picture. Whether Capture is the first durable repository for your device depends on your data flow, so draw it and confirm it with your team. See remote patient monitoring and the decentralized clinical trial guide for adjacent topics.
Yes. Digital Health Technologies for Remote Data Acquisition in Clinical Investigations was issued as final guidance in December 2023, replacing the 2021 draft. Like all FDA guidance it is nonbinding, and alternative approaches can be used if they satisfy the regulations.
FDA defines a DHT as a system that uses computing platforms, connectivity, software and/or sensors for health care and related uses. Wearable sensors, connected devices and eCOA apps on phones or tablets are typical examples.
It is the system where DHT data and metadata are securely stored as part of the trial record. FDA considers the data in the first durable repository that receives them to be source data, and does not intend to inspect individual devices if data reach the repository as the sponsor planned.
It depends. A cleared or approved device used within its indications generally does not, and nonsignificant risk use can follow abbreviated requirements. A significant risk DHT may need an IDE. Discuss your case with the relevant FDA center.
The guidance allows it after weighing pros and cons, but sponsor-provided devices and connectivity should be offered so that people without their own are not excluded.
Sponsors should log updates for each DHT, check that the device remains fit-for-purpose, compare data before and after if measurements may have changed, and describe how differences will be handled in the analysis before unblinding.
No. Device verification and validation is the sponsor's responsibility, often using manufacturer evidence. Capture provides the study record: synced device data, audit trail, access controls and exports.
Keep exploring
Best EDC for wearable data
Choosing a platform for a wearable study.
EDC for wearable device studies
Setting up a study with device data.
21 CFR Part 11 compliance checklist
Electronic records and signatures.
Digital biomarker validation studies
Running a validation study.
Decentralized clinical trial guide
Remote elements around the DHT.
FDA AI guidance for trials (2026)
FDA's related AI guidance.
Build and test the whole study in the free sandbox with every feature. No credit card; pay only when you go live.