When the product is a drug and a device together, such as an autoinjector, prefilled pen or inhaler, a single injection can generate a pharmacology result, a use experience and possibly a malfunction. This page explains how to capture all three without mixing them up.
Free sandbox · No credit card · 21 CFR Part 11 aligned
| Device problem? | Participant harmed? | Record as | |
|---|---|---|---|
| Cap would not release | Device deficiency | ||
| Needle shield failed, finger prick | Deficiency and AE | ||
| Dose delivered, injection site rash | Adverse event | ||
| Held too briefly, partial dose noted | Use difficulty, deficiency review | ||
| Instructions misread, wrong day | Use error, protocol deviation review |
The short answer
Why this is different
A drug trial asks whether the drug works and whether it is safe. A device study asks whether the device performs and whether it can be used safely. A combination product study often asks both, with a third question on top: can the intended users, at home and without a nurse, operate the device correctly to receive the full dose? A perfect drug in a device that people cannot use will fail, which is why use-related data count.
Regulators treat combination products as a category with its own rules. In the US, FDA assigns the product to a lead centre based on its primary mode of action, and regulations in 21 CFR Parts 3 and 4 cover assignment, current good manufacturing practice and postmarketing safety reporting for combination products. Clinical studies may run under an IND, an IDE or both, depending on the product and the regulatory path, and guidance on human factors and usability applies to the device part. In Europe and elsewhere the rules differ. Do not take this page as advice on classification; take it as a data design checklist, and let your regulatory advisor decide the path.
On the data side, the same study often generates the clinical dataset (visits, labs, adverse events), a use dataset (observations and diaries about handling) and a device dataset (identifiers, deficiencies, returns). Hold them in one system with one participant ID so they can be linked. For general device study design, see how to run a medical device clinical study, EDC for medical device clinical trials and early feasibility and pivotal device studies.
Data map
| Data | Example fields | Who records it | When |
|---|---|---|---|
| Device identification | Device type and model, lot or batch reference, expiry, number of the unit used | Site staff, for supervised doses; participant, for home doses | Every dose |
| Training | Date trained, trainer, method, participant demonstrated correct use (yes or no), repeat training needed | Site staff | Baseline and when retrained |
| Supervised dose observation | Steps completed in order, difficulty rating, observer comments, time to complete | Trained observer | First doses, per protocol |
| Home dose diary | Date and time, injection site, heard click or saw indicator, complete dose, any problem, free-text note | Participant (ePRO) | Each home dose |
| Use difficulties | Step affected, type (unable, delayed, wrong), consequence, whether help was needed | Participant or site staff | Whenever it happens |
| Device deficiency | Description, date, device reference, consequence, device returned (yes or no), return date | Site staff | On report |
| Adverse events | Term, onset, severity, seriousness, relatedness to drug and to device, action, outcome | Site staff | Every visit and on report |
| Device return and analysis | Return date, result of the manufacturer's analysis, date received | Study coordinator | After return |
Field names and timing are illustrative. Use your protocol and device documents to settle them.
Home use
Most combination product studies include home doses, so the diary carries the use-related data. With Capture ePRO the participant opens a link on a phone or computer, with no app to install, and answers a few questions after each dose. The question types cover what you need: single choice for the injection site and whether the click was heard, a time field for the dose, a dropdown for problem type and a long text for anything that went wrong.
Autoinjector home dose diary
Participant 002-0019 · Dose 7 of 12
Time of injection
INJTMInjection site
INJSITEDid you hear a click and see the indicator change?
CLICKDid you have any problem with the device?
PROBTYPEStudy team notified
Describe what happened
PROBDESCKeep the device if you can
Create both in the free sandbox, run a test participant through twelve doses and see what staff see. Every feature, no credit card, no time limit.
The hard part
In device studies, a device deficiency is generally understood as an inadequacy of the device in identity, quality, durability, reliability, usability, safety or performance, which includes malfunction and use error, and the standard ISO 14155 on clinical investigation of medical devices uses this idea. An adverse event is any untoward medical occurrence in a participant. The two overlap: a deficiency may or may not cause an event, and an event may or may not involve a deficiency. Your protocol should say how each is reported, to whom and by when, and it should define the terms in the same way for sites, monitors and the safety team.
Practical design follows. First, separate forms: the adverse event form (reusing the standard template) and a device deficiency form. Second, cross-reference fields: the AE form asks "was a device deficiency involved?" and, if yes, which report number; the deficiency form asks "did this lead to an adverse event?" and which one. A pair of edit checks keeps them consistent, so a "yes" on one side without a link on the other raises an auto-query. Third, relatedness to each part: record relatedness to the drug and to the device separately, because the answer drives different follow-up. The AE causality assessment software page covers assessment logic, and adverse event reporting software covers the basics.
Use errors need the same discipline. If the participant gave the wrong dose because the instruction was ambiguous, that is a use-related observation that may feed a human factors analysis and also, under your protocol, a protocol deviation. The protocol deviation tracking software page covers the second half. Be consistent: classify with the same decision tree every time and document how it changed if the protocol is amended.
Report received
Diary entry or site call: needle shield issue
Device deficiency form opened
Description, date, device and lot reference
Participant harmed?
Yes: minor finger prick
Adverse event form opened and linked
Severity mild, related to device
Device return and analysis
Returned unit logged; manufacturer analysis awaited
Reporting decision
Per protocol and regulatory advice
Two examples
The device differs; the structure of the data capture does not.
Capture supervised doses on a staff observation form at the first visits, then home doses by diary. Key use data: cap removal, placement, hold time, click and indicator, completeness. Key device data: lot, malfunction, leak, needle shield. Key drug data: pharmacokinetic sampling visits and injection site reactions. See EDC for medical device clinical trials.
Capture technique checks at visits (shake, exhale, inhale, hold) on a staff checklist, plus a diary with symptom scores and puffs used. Key device data: counter reading, blocked mouthpiece, dose counter failure. Key drug data: lung function and exacerbations. The [asthma peak flow diary template](/templates/asthma-peak-flow-diary-epro) is a starting point for the participant diary.
Read moreIf the comparator uses a different device, build forms that work for both. Do not rely on the device type to classify a problem; collect the same use-related questions in each arm so the comparison is fair.
Setting up
List the drug fields, use fields and device fields. For each, decide who completes it and when. Prefer controlled choices to free text.
Start from the adverse event, concomitant medication, vital signs and demographics templates and add the device forms. Use the AI study builder to draft the visit schedule from your protocol, then review it.
Add the linking fields between AE and deficiency forms and set the edit checks that raise a query when links are missing.
Decide who can see treatment assignment, who completes deficiency forms and who can approve closure. Blinded roles never receive treatment-arm values.
Decide how used devices are returned and who confirms receipt. Capture records the return as data fields; physical tracking is handled in your own process.
Walk a mock participant through training, supervised dose, home doses and a deliberately reported problem. Check that the deficiency and AE forms link and the export shows both.
Analysis
Use-related summaries are often tabulated by step: for each step of the instructions, how many doses were performed correctly, with difficulty, or not at all. Device summaries tabulate deficiencies by type, severity of consequence and whether a returned unit was analysed. Safety summaries show adverse events by relatedness to drug and to device. To produce these, the export needs the device reference, dose number, step results and links between forms. Capture exports CSV and Excel in wide format with a data dictionary and can filter by site or date; SDTM datasets can be produced as SAS XPT with Define-XML for the standard clinical domains. See export EDC data to R, SAS and SPSS.
Capture does not supply a human factors analysis engine, an automatic device failure classification or a device inventory; if you need those, plan for them outside the platform. Keep the blank eCRF PDF for submission packages, and record device documents in your trial master file.
Home doses expected
720
Diaries received
701
Use difficulties
14
Device deficiencies
3
finger prick
It is a product made of two or more regulated components, such as a drug and a device, combined or packaged together. Autoinjectors, prefilled syringes, inhalers and drug-eluting devices are common examples.
No. Capture has no device or companion-app integration for these. Use observations and device identifiers are recorded on forms by staff or participants.
Use a device deficiency form with no linked adverse event. Record the description, device and lot reference, whether the unit was returned and the date.
Not by default. Use errors are use-related observations; they become adverse events only if a participant suffers an untoward medical occurrence. Your protocol decides how each is classified and reported.
In the US, 21 CFR Part 3 covers assignment to a lead centre, and Part 4 covers cGMP and postmarketing safety reporting for combination products. The clinical study may run under an IND, an IDE or both. Ask your regulatory team which applies; this page is not advice.
Yes. ePRO runs in the browser from a link, with no app to install, and reminders can be set for each scheduled dose.
Yes. Capture is suitable for Phase 1 to 3, and device-related data design does not depend on phase. The sample size for use data is set in the protocol.
Use one participant ID, one dose number and a device reference on each form, then export a wide file with a data dictionary. Cross-link fields between AE and deficiency forms make the connection explicit.
Keep exploring
EDC for medical device clinical trials
Device study data capture.
How to run a medical device clinical study
Design and operations basics.
Early feasibility and pivotal device study EDC
Device study phases.
AE causality assessment software
Relatedness to drug and device.
Protocol deviation tracking software
Where use errors meet deviations.
ePRO in Capture
Browser-based diaries with reminders.
Open the free sandbox, create a home dose diary and a linked device deficiency and adverse event pair, and walk a mock participant through them. No credit card, and you pay only once you go live.