Retrospective Phase II Clinical Trial
Also known as: retrospective Phase II study, historical Phase II analysis, retrospective efficacy study, Phase II retrospective analysis
A retrospective Phase II clinical trial evaluates a treatment's preliminary efficacy and safety signals using existing archival data — medical records, registries, or electronic health records — rather than prospectively enrolling new patients. It mirrors the objectives of a standard Phase II trial (estimating response rate, tolerability, and early efficacy) but does so by looking backward at patients who have already received the intervention, making it faster and less costly than a prospective design.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use a retrospective Phase II design when a treatment has already been used in clinical practice and sufficient patient numbers exist in institutional or registry databases to estimate response rates or survival endpoints with acceptable precision, and when a prospective study would take too long or would be ethically or logistically infeasible. It is particularly appropriate for rare diseases, orphan indications, or urgent post-approval safety questions. Do not use this design when the historical records are too incomplete or inconsistently documented to reliably extract key variables, when outcome assessment requires prospective monitoring (e.g., patient-reported outcomes, frequent lab assessments), when the comparison requires an unexposed control arm subject to contemporary confounding, or when regulatory-grade evidence is required — in that case a prospective controlled trial is necessary.
Strengths & limitations
- Substantially faster and less expensive than a prospective Phase II trial because data collection has already occurred.
- Enables investigation of rare conditions or historical regimens where prospective enrollment would be impractical.
- Can leverage large institutional datasets or national registries to achieve sample sizes exceeding what a single prospective trial could enroll.
- Useful for generating hypotheses and activity estimates to power and design a subsequent prospective Phase III study.
- Subject to selection bias because patients included in historical records may differ systematically from those who would be enrolled in a prospective trial.
- Information bias is common: key variables (response assessment timing, toxicity grading, reason for dose modification) may be missing, inconsistently recorded, or not collected at all.
- Cannot establish causality; confounding by indication — the tendency to give certain treatments to patients with better or worse prognoses — is difficult to fully adjust for.
- Findings are not regulatory-grade and typically cannot serve as pivotal evidence for drug approval without confirmatory prospective data.
Frequently asked
Is a retrospective Phase II trial the same as a retrospective cohort study?
They share the retrospective observational design, but differ in framing. A retrospective Phase II trial explicitly adopts Phase II trial objectives — pre-specified primary efficacy endpoints, activity thresholds, and statistical decision rules analogous to a prospective Phase II trial. A retrospective cohort study more broadly compares outcomes between exposed and unexposed groups without necessarily invoking the Phase II decision framework.
Can a retrospective Phase II study support a regulatory submission?
Generally not as pivotal evidence. Regulatory agencies (FDA, EMA) require prospective, controlled, and often randomized evidence for marketing approval. However, retrospective Phase II data can contribute to compassionate-use applications, support expanded access programs, or provide historical context and effect-size estimates cited in a new drug application dossier.
How do I handle missing response assessments in the analysis?
Missing response data should not be assumed to represent treatment failure without justification. A pre-specified missing-data strategy is essential: report the proportion of patients with evaluable versus missing assessments, use a primary analysis based on evaluable patients with appropriate confidence intervals, and conduct sensitivity analyses under different assumptions (e.g., treating all missing as non-responders) to bound the plausible range of the true response rate.
What sample size is typical for a retrospective Phase II study?
Sample size is usually determined by available records meeting eligibility criteria rather than a prospective power calculation. As a reference, Simon's two-stage design for prospective Phase II trials targets 20–100 patients depending on the hypothesized response rates. Retrospective analyses should ideally match or exceed this range and report the exact two-sided 95% confidence interval for the primary endpoint regardless of sample size.
Which reporting guideline should I follow?
STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) is the most widely applied guideline for retrospective observational studies and covers cohort designs. Some journals in oncology additionally request CONSORT-style flow diagrams showing patient identification, exclusion, and analysis. Consulting the target journal's instructions and using both checklists in parallel is good practice.
Sources
- Simon, R. (1989). Optimal two-stage designs for phase II clinical trials. Controlled Clinical Trials, 10(1), 1–10. DOI: 10.1016/0197-2456(89)90015-9 ↗
- Clinical trial. Wikipedia. link ↗
How to cite this page
ScholarGate. (2026, June 3). Retrospective Phase II Clinical Trial. ScholarGate. https://scholargate.app/en/epidemiology/retrospective-phase-ii-clinical-trial
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Diagnostic Accuracy Study DesignClinical Research↔ compare
- Phase I Clinical TrialEpidemiology↔ compare
- Phase II clinical trialEpidemiology↔ compare
- Phase III clinical trialEpidemiology↔ compare
- Retrospective Cohort StudyEpidemiology↔ compare