Definition: What is a single-arm study?
A single-arm study (one-arm study) is a study design in which all study participants receive the same intervention and there is no parallel control group within the same study. Efficacy and safety are therefore assessed without a randomized comparator group, often by comparison with historical data, registry data, or external controls. Single-arm studies are used when a randomized controlled trial would be difficult to implement in practice or would be ethically problematic, for example in rare diseases, very small patient cohorts, or when there is no acceptable standard therapy.
In clinical development, single-arm studies are particularly common in early phases (e.g., Phase II). However, they may also be relevant in late-stage programs close to marketing authorization, for example in orphan drugdrug development or in indications with high unmet need. It is crucial that the study objective and the strength of evidence are defined realistically in the context of the design: single-arm studies are often exploratory, but under certain conditions they may also include confirmatory elements.
Endpoints, hypotheses, and statistical planning
Because an internal control group is lacking, objective and robust endpoints are often selected, such as overall response rate, progression-free survival, or biomarker-based surrogate endpoints. Statistical planning often focuses on precision (confidence intervals) or on demonstrating that a predefined threshold is exceeded. This threshold is typically based on external evidence. It is important that the sources, assumptions, and definitions are transparent and are specified in the protocol.
For time-to-event endpoints, the handling of censoring and follow-up is critical. Informative censoring can bias results if participants systematically drop out of observation. Therefore, follow-up processes, reasons for discontinuation, and data completeness should be closely monitored. Sensitivity analyses, alternative endpoint definitions, or different censoring rules can help assess the robustness of the results.
External controls and comparability
In practice, a historical control value is often used. This can be appropriate, but it entails risks: differences in population, diagnostics, supportive care, treatment sequences, or endpoint definitions can severely limit comparability. Increasingly, external controls from real-world evidence sources are being discussed, for example from registries or electronic healthcare data. Methodologically, matching approaches, weighting, or other adjustment methods are used to reduce differences in baseline characteristics.
Despite methodological adjustments, residual bias may remain. Therefore, it is important to describe the limitations openly and not to base interpretation solely on an external comparison. With a strong treatment effect, single-arm evidence can be compelling; with moderate effects, there is a high risk that systematic differences explain the observed effect size.
Advantages, limitations, and typical sources of bias
The key advantage of a single-arm study is feasibility: less recruitment effort, no randomization, and often faster execution. This can be attractive for sponsors and study sites, particularly in time-critical development programs. At the same time, internal validity is lower than in randomized designs because confounding and selection bias are more difficult to control. Expectation effects or changes in standards of care over time can also bias comparisons with historical data.
Typical sources of bias include a selective patient population (e.g., fewer comorbidities), differences in prior treatments, differing diagnostic pathways, and different definitions of response or progression. Best practices include prospectively defining the comparison strategy, a consistent endpoint assessment process (e.g., central review), sensitivity analyses, and a clear separation between exploratory hypotheses and conclusions relevant to marketing authorization.
Regulatory classification and implications for benefit assessment
From a regulatory perspective, single-arm studies are not excluded, but they require a robust justification. In EU contexts, Scientific Advice can be used to align early on endpoints, patient selection, and comparison approaches. Requirements for study conduct under Good Clinical Practice and EU Regulation 536/2014 continue to apply. For marketing authorization, the suitability of the evidence is assessed in the benefit–risk context, where applicable within accelerated or conditional procedures.
For Germany, it is relevant that later benefit assessments and reimbursement discussions often require an appropriate comparator therapy. A purely single-arm demonstration may therefore reach its limits in the health technology assessment context despite marketing authorization. Sponsor teams should consider this perspective early and, where possible, plan complementary evidence (e.g., randomized data, registry concepts, or master protocol strategies).
Operational implementation in the CRO and sponsor environment
Single-arm studies are often not automatically “simpler,” because data quality must be particularly convincing. The protocol, monitoring, and data management should address the key risks of bias, for example through consistent endpoint definitions, clear follow-up processes, and robust source data verification for key variables. If external controls are planned, data access, data protection, data standards, and statistical methods must be clarified in good time.
A common pitfall is conflating exploratory intent with an authorization claim. If a single-arm design is primarily exploratory, this should be clearly stated in the study objective. If, however, evidence relevant to marketing authorization is being pursued, a rigorous concept for external comparator data and transparent documentation in the clinical study report are required, including limitations, sensitivity analyses, and traceable data provenance.
In addition, communication with stakeholders is important: study sites, statistics, Medical Affairs, and later Market Access teams should develop a shared understanding of the strength of evidence early on. This helps avoid unrealistic expectations and enables timely planning of follow-on studies.
FAQ: When is a single-arm study appropriate?
Typical use cases include rare diseases, small populations, or situations where randomization is ethically or practically difficult and robust external comparator data are available.
FAQ: Can single-arm studies support marketing authorization?
Under certain conditions, yes—particularly where there is high medical need and a convincing effect size. However, the requirements for justification, endpoints, and the comparison strategy are high.
FAQ: What are the biggest risks of the design?
Lack of internal comparability, confounding, selection bias, and differences versus historical or external controls. These risks must be addressed methodologically and reported transparently.
Regulatory references (selection): EU Regulation 536/2014 (Clinical Trials Regulation), ICH E6(R3) Good Clinical Practice, ICH E9 (Statistical Principles), EMA Scientific Advice (general guidance on evidence justification).