Prospective Study vs RCT: Choosing the Right Clinical Study Design for Your Device
Should your device be studied in a randomised controlled trial or a prospective single-arm study? The answer affects cost, timelines, regulatory acceptance and market value. Here is how to decide, and how REGHELPS designs and runs both.
Why study design matters under EU MDR and FDA
EU MDR and the FDA both expect clinical evidence that is sufficient for your claims and proportionate to the device risk. The study design you choose decides whether that evidence is convincing, how many patients you need and how long it takes.
Choose a design that is too weak, and your Notified Body or FDA reviewer may reject the data. Choose one that is heavier than needed, and you spend years and budget you did not have to. The right design sits between the two.
Prospective study vs RCT: side by side
How the two designs compare on the points that matter most for device manufacturers.
| Aspect | Prospective study (single-arm or observational) | Randomised controlled trial (RCT) |
|---|---|---|
| Question answered | Does the device meet safety and performance goals in real patients? | Is the device better than, or as good as, an alternative? |
| Control group | None, or historical or literature-based comparator | Concurrent control, randomly assigned |
| Bias control | Moderate; depends on endpoints and comparator | High, through randomisation and blinding where possible |
| Sample size | Usually smaller | Usually larger |
| Cost and duration | Lower and faster | Higher and longer |
| Typical use | Performance goals, PMCF, lower-risk or well-understood devices | Comparative claims, novel high-risk devices, payer evidence |
| Regulatory acceptance | Widely accepted when justified | Strongest evidence for comparative claims |
Which study design does your device need?
The questions we work through with clients. The final design is always confirmed in the clinical investigation plan.
Where each design sits in the hierarchy of clinical evidence
Higher levels control bias better. Every level has a role in a complete clinical evaluation.
Study designs between a simple prospective study and a full RCT
Why RCTs are harder for devices than for drugs
These challenges explain why many device programmes use well-designed prospective studies.
| Challenge | What it means | How we handle it |
|---|---|---|
| Blinding | Patients and surgeons often know which device was used | Blinded outcome assessors and objective endpoints |
| Learning curve | Results improve as operators gain experience | Training phase and roll-in patients excluded from analysis |
| Operator skill | Outcomes depend on who uses the device | Site and investigator selection, stratification |
| Fast design changes | Device may change during a long trial | Shorter, adaptive or staged designs |
| Small populations | Rare conditions limit enrolment | Bayesian designs, registries and multi-country sites |
| Sham controls | Placebo procedures raise ethical concerns | Active comparators or standard of care controls |
What EU MDR and the FDA expect from study design
| Topic | EU MDR | US FDA |
|---|---|---|
| Basis | Article 61, Annex XIV and XV; MDCG guidance | 21 CFR 860.7 valid scientific evidence; 21 CFR 812 |
| Is an RCT required? | No; evidence must be sufficient for claims and risk | No; well-controlled investigations preferred for PMA effectiveness |
| Accepted alternatives | Single-arm studies, equivalence, PMCF, registries when justified | Partially controlled studies, objective trials, Bayesian and adaptive designs |
| Standard | ISO 14155:2020 | 21 CFR 812, ISO 14155 recognised |
How REGHELPS designs and runs your study
From choosing the design to delivering the final report.
- 1
Evidence and claims review
StrategyWe map your claims, risks and existing data to define the evidence gap.
- 2
Design options
Prospective or RCTWe compare designs on evidence strength, cost, time and regulatory acceptance.
- 3
Statistics and sample size
BiostatisticsWe set endpoints, comparators and sample size, including Bayesian or adaptive options.
- 4
Protocol and approvals
ISO 14155We write the CIP and manage ethics and regulatory submissions.
- 5
Study conduct
CROWe run sites, monitoring, data management and safety reporting.
- 6
Report and evidence use
CER · HTAWe report results and package them for your CER, FDA submission and market access.
Prospective study vs RCT FAQs
Is an RCT required for CE marking under EU MDR?
No. MDR does not require a randomised controlled trial for every device. It requires sufficient clinical evidence for your claims, proportionate to device risk. An RCT is expected mainly when you claim superiority or equivalence to another treatment, or when outcomes are subjective and bias must be controlled.
Is a prospective study the same as an observational study?
Not always. Prospective means subjects are followed forward in time from enrolment. A prospective study can be observational, where the device is used as in routine care, or interventional, where the protocol dictates treatment. An RCT is a type of prospective interventional study.
When is a single-arm prospective study enough?
When there are accepted objective performance criteria or performance goals for the outcome, when the device is similar to well-understood devices, or when randomisation would be unethical or impractical. The study design and comparator must be justified in the clinical investigation plan.
Why are RCTs harder for medical devices than for drugs?
Devices are often hard to blind, results depend on operator skill and learning curves, designs change quickly, patient groups can be small, and sham procedures raise ethical issues. These factors push many device programmes towards well-designed prospective studies or adaptive designs.
Does the FDA accept non-randomised studies?
Yes. Under 21 CFR 860.7, valid scientific evidence can include well-controlled investigations, partially controlled studies and objective trials without matched controls, depending on the device. The FDA also accepts Bayesian and adaptive designs when they are planned and justified.
Which design gives the strongest evidence for payers and HTA bodies?
Randomised controlled trials usually carry the most weight with payers and health technology assessment bodies because they control bias. Prospective studies and registries add valuable real-world evidence, and many strategies combine both.
Not sure which study design your device needs?
Share your device, claims and target markets using the form below and we'll recommend a design and budget.
