Phase III clinical trials: population size, diversity and risk

What is a Phase III clinical trial?

A Phase III trial is a large, late-stage study designed to confirm whether an investigational treatment works. It also compares the treatment against a standard treatment or placebo in the population it is meant to help^[1]. These trials are sometimes called pivotal studies. Their results are what a regulatory agency reviews when deciding whether a treatment can be approved^[2].

By the time a treatment reaches Phase III, it has already passed through smaller Phase I studies. These early studies focus on safety and dosing and usually involve 20 to 100 people. It has also passed through Phase II studies. These look for early effectiveness signals and typically involve up to a few hundred people^[3]. Phase III is the stage where enough people are treated, for long enough, to detect effects that smaller studies could not reliably show^[2]. This is also the stage at which the treatment is tested closest to the conditions of everyday clinical practice. It is compared directly against whatever is already used to treat the condition, rather than against no treatment at all^[1]. Regulatory decisions rest heavily on these comparisons. That is why the design of a Phase III trial is scrutinised so closely before it ever begins recruiting patients^[2].

How many people take part, and for how long?

Phase III trials typically enrol 300 to 3,000 volunteers who have the condition the treatment targets, and they run from 1 to 4 years^[2]. Some late-phase trials are considerably larger: the reSURFACE 1 trial for a plaque psoriasis treatment enrolled 1,772 patients, and reSURFACE 2 enrolled 1,090 patients^[4].

Trials are usually conducted at multiple centres at once, sometimes across several countries, so results reflect a broader range of clinical settings than a single-site study could^[3]. As a participant, this means your data will be pooled with data from people treated at other hospitals or clinics, which is part of what makes the eventual result more reliable. Cooperative research groups exist specifically to organise this kind of multi-centre recruitment across borders. Those coordinating international cancer trials, for example, pool expertise and patients that no single institution could gather alone^[5]. The larger the number of participating centres, the more likely it is that the trial's population will include the kind of variation in age, disease severity and prior treatment history that reflects real clinical practice. This is more representative than relying on a narrow slice of it^[3].

Phase	Main purpose	Typical number of participants	Typical duration
Phase I	Safety, dosing	20–100	Several months
Phase II	Effectiveness signals, side effects	Up to a few hundred	Several months to 2 years
Phase III	Confirming efficacy, monitoring adverse reactions	300–3,000	1–4 years

^[2]

Why does participant diversity matter for how results apply to you?

Clinical trial populations have not become substantially more diverse over the past 25 years, even though the number of countries participating in pivotal trials submitted for drug registration has nearly doubled. This matters because how a treatment behaves in your body can depend on both intrinsic factors and extrinsic factors. Intrinsic factors include genetics, age and organ function. Extrinsic factors include diet, other medicines you take, or where you live. These sources of variability are recognised in current regulatory guidance^[6].

Regulatory agencies, patient advocacy groups and clinical researchers have introduced recruitment targets and broader eligibility criteria specifically to make trial populations better reflect the people who will eventually use the treatment. Despite these efforts, certain groups remain underrepresented in development programmes. Representation can also vary from country to country within the same global trial. This uneven representation means that a treatment's average result, as reported for the whole trial population, may not describe what happens for a group that made up only a small fraction of participants. This can be true even when that group is well represented in the disease overall^[6].

Some trials address diversity of disease biology rather than only demographic diversity. Rare gynaecological cancers, for example, are heterogeneous both in how they present clinically and at the molecular level. This heterogeneity has driven the design of trials focused on specific molecular subgroups rather than mixed populations. The PAOLA-1 trial, a large Phase III study in high-grade ovarian cancer, showed that benefits from a maintenance treatment were particularly marked in patients with a specific tumour characteristic (HRD-positive disease). This illustrates why knowing your own disease subtype can affect what a trial's results mean for you. Building trials around such molecular subgroups requires sustained collaboration between research networks, patient groups and specialist centres. Rare cancers, by definition, produce fewer eligible patients at any single site^[7].

Why can only Phase III find some long-term or rare risks?

Because Phase III trials are larger and last longer than earlier phases, they are more likely to reveal long-term or rare side effects that smaller, shorter studies could miss. Phase III trials provide the bulk of the safety data used to describe a treatment's known risks once it reaches approval^[2].

In the reSURFACE 1 and reSURFACE 2 trials of tildrakizumab, the proportion of patients with at least one adverse event by Week 12 ranged from 44.3% to 55.1% across treatment and placebo groups. One death occurred during the trials, with the cause considered undetermined. This kind of detailed adverse-event tracking, across more than 1,000 patients per study, is what allows researchers to distinguish side effects genuinely linked to a treatment from events that would have happened anyway. Without a comparison group receiving placebo or an existing standard treatment, it would be far harder to know whether a given symptom reported during the trial was caused by the investigational drug. It could instead simply reflect the natural course of the underlying condition^[4].

Some organisations run long-term follow-up studies to track how a treatment affects patients years after the original trial ended. The EORTC 10801 trial in breast cancer, for instance, had a 20-year follow-up confirming that breast-conserving therapy could be offered as standard care alongside mastectomy, based on data collected over decades. If you take part in a trial, you may be asked to continue reporting your health status well beyond the main treatment period for exactly this reason. This kind of extended monitoring is particularly important for treatments intended to be used for years. Risks or benefits that only emerge after prolonged exposure would otherwise remain invisible to shorter studies^[5].

What do randomisation and blinding mean for what treatment you receive?

Most Phase III trials are randomised, meaning you are assigned by chance, not by choice, to receive either the investigational treatment, a standard treatment, or a placebo. Most are also double-blind, meaning neither you nor your treating clinician knows which treatment you are receiving. This prevents that knowledge from influencing how symptoms are reported or assessed^[3].

In a single-blind study, only you as the participant are unaware of which treatment you are receiving^[3]. In the reSURFACE 1 trial, patients initially assigned to placebo were switched to an active dose of tildrakizumab after 12 weeks. This approach, called a crossover design, allows placebo participants to eventually access the treatment being studied. Crossover designs of this kind are one way trial sponsors try to balance the scientific need for a placebo comparison against the ethical concern of leaving participants without active treatment for the whole length of the study^[4].

How does Phase III evidence compare with real-world evidence?

Real-world evidence, drawn from electronic health records and disease registries, can complement Phase III trial evidence but is not considered a substitute for it. Real-world data can capture information from patients who are typically older, in poorer health, and more diverse in ethnicity, gender or treatment adherence than those enrolled in randomised trials. This helps identify adverse effects and rare events that trials may not detect^[8].

Because real-world data is not randomised, comparing outcomes between treatment groups is more prone to bias from differences in patients' underlying prognosis. For this reason, treatment effects observed only in routine practice, without prior confirmation in a randomised trial, should be treated with particular caution. The two forms of evidence are best understood as answering different questions. A randomised Phase III trial tells you whether a treatment can work under controlled conditions. Real-world data tells you how it tends to perform once it is used broadly, by a wider range of patients, under everyday circumstances that a trial protocol cannot fully replicate^[8].

How likely is a treatment to succeed once it reaches Phase III?

Not every treatment that enters Phase III testing is ultimately approved. Approximately 25 to 30% of drugs that enter clinical testing pass through Phase I, II and III combined^[2]. Roughly 70% of drugs that enter Phase I move on to Phase II, and around 33% of those that complete Phase I and II proceed to Phase III^[3].

These figures mean that entering a Phase III trial does not guarantee the treatment will reach approval. Some trials are stopped early or fail to show a benefit large enough, or safe enough, compared with existing standard treatments^[2]. This is worth keeping in mind if you are considering joining a Phase III trial. Participation offers access to a treatment that has already shown promise in earlier phases. It does not, however, amount to a guarantee that the treatment will ultimately prove better than, or even as good as, what is already available^[3].

What should you report to the study team, and when?

During a Phase III trial, you will typically be monitored closely for any change in your health, including symptoms that feel unrelated to your condition or the treatment. Because larger numbers of people are being followed in Phase III than in earlier phases, there is a greater chance for the research team to detect a pattern in side effects across many participants. This only works, however, if participants report changes promptly^[3].

If you are taking part in a trial and notice a new symptom, a worsening of an existing one, or anything unusual, report it to the study team as instructed in your trial's information – do not wait for a scheduled visit if you are told to report sooner. The study team, not the sponsor's administrative systems, is your point of contact for anything affecting your health during participation. Adverse events recorded promptly and consistently across a large trial are what eventually allow researchers to calculate reliable rates, such as the 44.3% to 55.1% figures seen in the reSURFACE trials, rather than relying on isolated or incomplete reports^[4].

When should you speak with your doctor or the study team about joining a Phase III trial?

Speak with your treating doctor or the trial's study team before joining if you are uncertain whether you meet the eligibility criteria, if you are taking other medicines that might interact with the investigational treatment, or if you have questions about what a placebo or crossover design would mean for your own care during the study^[3]. If your disease has a specific molecular or genetic subtype, ask whether the trial has looked at outcomes for that subtype specifically. Benefits can vary meaningfully between subgroups within the same broader condition^[7]. It can also help to ask how closely the trial's overall participant population matches your own age group, sex, ethnicity or disease severity. Trial populations have not necessarily kept pace with the diversity of the patients who will eventually receive the treatment in routine practice^[6].

Summary

Phase III trials exist to answer the questions that smaller, earlier studies cannot. Does a treatment offer a real benefit over existing options, and what happens to people who take it over months or years, in numbers large enough to catch rarer risks?^[2] The size, duration and design of these trials – randomisation, blinding, multi-centre recruitment – are all built around making the answer as reliable as possible for people in your situation.

Diversity of participants, and understanding your own disease subtype where relevant, both affect how confidently a trial's results can be applied to you personally. If you are asked to join a Phase III trial, understanding what will be monitored, what reporting is expected of you, and how the trial's population compares with your own circumstances can help you make a more informed decision alongside your doctor.

This article is for general information and does not replace advice from your doctor, pharmacist or the clinical trial team responsible for your care.

1. [1] Four Phases of Clinical Trials (accessed 2 June 2026) – https://www.clinicaltrials.astellas.com/four-phases-of-clinical-trials
2. [2] Step 3: Clinical Research (accessed 2 June 2026) – https://www.fda.gov/patients/drug-development-process/step-3-clinical-research
3. [3] Trial Phases 1, 2 & 3 Defined (accessed 2 June 2026) – https://med.uc.edu/depart/psychiatry/research/clinical-research/crm/trial-phases-1-2-3-defined
4. [4] Key clinical-trial evidence for tildrakizumab (accessed 2 June 2026) – https://dermnetnz.org/topics/key-clinical-trial-evidence-for-tildrakizumab
5. [5] EORTC: A Leader In Clinical Cancer Research (accessed 2 June 2026) – https://www.eortc.org/blog/2018/05/11/eortc-a-leader-in-clinical-cancer-research/
6. [6] https://pmc.ncbi.nlm.nih.gov/articles/PMC9306578/
7. [7] Research promotion and good partnerships have brought rare gynaecological malignancies into the light (accessed 2 June 2026) – https://dailyreporter.esmo.org/esmo-congress-2023/2023-esmo-society-awards/research-promotion-and-good-partnerships-have-brought-rare-gynaecological-malignancies-into-the-light
8. [8] Real-world research does not need to replace randomised clinical trials to be recognised as valuable for treatment evaluation (accessed 2 June 2026) – https://dailyreporter.esmo.org/spotlight/real-world-research-does-not-need-to-replace-randomised-clinical-trials-to-be-recognised-as-valuable-for-treatment-evaluation