A PEMF study is a device study with a long list of small events: sessions at home, settings, sham or active allocation, symptom diaries and adverse events. Capture holds all of it in one audited system, with ePRO, eConsent and randomisation, and you build it free in the sandbox.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Session
Session date
SESDATMinutes completed
SESDURPrescribed: 30 min
Programme used
PROGIDSession stopped early?
STOPFLSafety
Any new symptom since last session?
AEYNWhat a PEMF trial asks of its data system
The study type
Pulsed electromagnetic field (PEMF) and other magnetic field therapies are usually delivered by a device that a participant uses in a clinic, or increasingly at home, on a schedule set by the protocol. Studies take several forms: randomised sham-controlled trials, single-arm feasibility studies, post-market follow-up of a marketed device and wellness studies where the outcome is how people feel and sleep rather than a laboratory value. The therapy itself is not the topic of this page. The question here is practical: what does the data system need to hold so the results can be analysed and defended?
Three things make this study type different from a drug trial. First, the intervention is a repeated behaviour, so exposure is a stream of sessions rather than a dispensed pack. Second, the intervention is hard to blind, because a device can make a sound, a light or a sensation, so how the sham is built and how staff are kept unaware of allocation are part of the data design. Third, most outcomes are reported by the participant, which makes timing and completion rates the main data quality risks. The sections below take each in turn, with the Capture features that fit. For the wider device view, see EDC for medical device clinical trials and the post-market studies page.
Day 28 is an Arm B only visit, so Arm A participants never see it.
Data model
A starting list for protocol design. Your protocol decides what is required; these are the groups of data these studies usually need.
| Data group | What is recorded | How it fits in Capture |
|---|---|---|
| Device identification | Device serial or unit number issued to each participant, issue and return dates | Text and date fields on an eCRF form, with an audit trail on every change |
| Session log | Date, duration, programme or intensity setting, interruptions, who supervised | A repeating table on an eCRF form, or a short participant diary for home sessions |
| Prescribed versus actual dose | Planned minutes against completed minutes, per week | Calculated fields and edit checks that raise an auto-query when a value is out of range |
| Allocation | Active or sham, stratification factors, randomisation date | Randomisation with stratification; blinded roles never see arm values |
| Outcomes | Pain, sleep, fatigue, mood or function scores at set visits | ePRO questionnaires on the participant's phone, plus site-completed forms |
| Concomitant therapy | Medicines and other treatments during the study | Concomitant medications template as a repeating table |
| Safety | Adverse events, device-related events, discontinuations | Adverse event form, AESI form and SAE workflow |
| Wearable data | Sleep, heart-rate or activity data from the participant's own device | Native OAuth connection per study; the participant connects their own account |
Blinding
A sham device that looks and behaves like the active one is only useful if nobody who touches the data can tell which is which. In Capture, the randomisation is held in the system, and blinded roles are served masked database views, so treatment-arm values are not sent to them at all. That is stronger than hiding a column in a report. The sponsor decides who is unblinded, typically a statistician or a device manager, and everyone else works from coded participant IDs. See clinical trial randomization software for how allocation, stratification and kits are handled.
Think about the practical unblinding risks too. If the device unit number reveals the arm, the unit number field must be restricted or recoded. If a coordinator sets the programme on the device, the session form should record the programme code and not the arm. If an adverse event is device-related, the form should allow the safety reviewer to classify it without revealing allocation to the site. These are protocol decisions, and the sandbox is the place to test them before a single participant is enrolled.
Build the study, enrol five fake participants into each arm and ask a colleague with a blinded role to look for any screen, export or filter that leaks allocation. Fix it while the study is still in draft; approved forms are locked for live use, and changes after that follow your amendment process.
Free sandbox with every feature: forms, randomisation, ePRO, eConsent and the audit trail. No credit card, and you pay only when you go live.
Setup
A sequence that follows how these protocols are usually written.
Upload the protocol as PDF or DOCX and let the AI study builder draft the visits and forms. Nothing is saved until you review it. Or build the schedule of assessments by hand.
Decide whether sessions are recorded by site staff, by the participant or by both. Clinic sessions fit a repeating table on an eCRF form; home sessions fit a short ePRO entry the participant completes on a phone browser.
Start from the sleep diary, fatigue log or pain templates and adapt them. Set completion windows so a late entry is handled by a rule you chose.
Set arms, stratification factors and who is blinded. Test each role in the sandbox.
Use eConsent with email OTP and countersignature. If the protocol uses wearables, add the wearable data sharing consent template and enable the device for the study.
Enter sample data, trigger every edit check, then approve the forms. Approved forms are locked for live use.
Adherence
In a home-use device study, adherence is often the main secondary endpoint and the main source of missing data. A participant who stops using the device in week three looks like a dropout in a sham-controlled comparison, and the reason matters. Capture lets you record planned and completed sessions, derive weekly totals with calculated fields, and set edit checks that flag a participant who falls below a threshold your protocol defines. A flag can raise a query, so the coordinator contacts the participant while it still matters.
Reminders do part of the work. Participants get email or SMS reminders with a secure link that opens in the phone browser, with no app and no password. A reminder ladder sends up to three reminders per task and respects quiet hours in the participant's time zone. The ePRO compliance reminders page describes the controls, and ePRO software for clinical trials covers the wider feature set.
Record why a session was missed with a short coded list, not free text only: device fault, illness, travel, forgot, adverse event. A coded reason lets you separate technical failure from non-adherence in the analysis, which a free-text note rarely does.
Wearables
Many magnetic field studies track sleep, heart-rate variability or activity alongside questionnaires. Capture connects wearables natively: the study team enables a device for the study and can mark it required, and the participant connects their own account through the vendor's sign-in from a link in the study. Data then syncs automatically, with a last-sync time, a "Sync now" option and a disconnect control. Oura is the device named on the site today; confirm any other device with us before you plan around it.
Wearable data are context, not a replacement for the protocol's primary measure. Decide up front which wearable variable, if any, is an endpoint and which is exploratory, and pre-specify the handling of days with no data. The EDC for wearable device studies page covers this in more depth, and the HRV companion template and recovery tracker give participant-reported counterparts to the sensor data.
Before you start
Prescribed minutes, frequency and programme settings written into the protocol and mirrored in the session form.
Who is blinded, how the device unit number is handled and how unblinding is requested.
Primary and secondary instruments, their timing and completion windows.
Any questionnaire with licensing terms: record licence details on the participant questionnaire and keep site-form licences with your study documents.
Adverse event and SAE forms tried end to end in the sandbox.
Including consent for wearable data if used.
CSV or Excel with the data dictionary opens cleanly in your statistics package.
Regulatory and claims
Whether a given magnetic field device is regulated as a medical device, a wellness product or something else depends on its intended use and on the jurisdiction, and the study design follows from that. This page is not legal advice and makes no claim about the regulatory status or efficacy of any therapy. Ask your regulatory adviser which rules apply, and write the study to that standard. Capture provides 21 CFR Part 11-aligned controls, including a field-level audit trail, role-based access and electronic signatures, and the sponsor remains responsible for validated use. The audit trail software page describes what is logged.
A small feasibility study is a reasonable first step for a new protocol. How to run a pilot study and EDC for pilot clinical trials walk through scoping one so that the larger trial inherits a tested data set-up.
Yes. Sessions can be recorded by site staff as repeating rows on an eCRF form, or by the participant as a short phone entry. Each row carries its own audit trail.
Randomisation assigns active or sham, and blinded roles never receive treatment-arm values because masked database views enforce it. You decide who is unblinded and test each role in the sandbox first.
Yes. Questionnaires open in the phone browser through a secure link or a QR code, with no app to install and no password.
Yes, through a native connection the study team enables per study. The participant connects their own account. Oura is the device named today; confirm others with us.
No. Capture is data capture software. It makes no efficacy or regulatory-approval claim about any therapy, and this page is not legal advice.
With adverse event and AESI forms, an SAE workflow and a field-level audit trail. It is not a pharmacovigilance safety database, so use one for case processing if you need it.
Building and testing is free in the sandbox, with every feature, no credit card and no time limit. You pay only when the study goes live with real participants.
Keep exploring
EDC for medical device trials
The wider device-study view.
EDC for wearable device studies
Sensor data in the study record.
Clinical trial randomization software
Sham and active allocation.
ePRO software for clinical trials
Diaries and questionnaires on a phone.
Adverse event reporting software
AE and SAE capture.
EDC for longevity clinical trials
Healthy-ageing study designs.
Free sandbox with every feature. No credit card, and you pay only when you go live.