Nonrandomized comparative interventional study
How do outcomes compare when investigators assign interventions without randomization?
When to use this project
Compare prospectively assigned interventions using a stated nonrandom rule, with an explicit plan for baseline differences, time trends and selection.
How it differs
Investigators assign the intervention, distinguishing this from routine-care observational comparisons. Assignment is not randomized, so measured adjustment alone cannot establish exchangeability.
Decisions before you start
- State the assignment mechanism and whether comparators are concurrent or historical.
- Align eligibility, assignment and follow-up; identify baseline confounders without automatically adjusting for post-treatment variables.
- Use reporting guidance suited to the intervention. TREND specifically concerns behavioral and public health evaluations; it is not a universal replacement for all trial guidance.
How the flows work together
Each flow below links to its diagram, steps and tables. Flow order organizes the project; it does not imply that all work is strictly sequential. Data assessment, analysis planning and protocol refinement may inform one another.
1. Comparative intervention overview
Overview · Question through nonrandomized intervention reporting
Use this overview to state the purpose and link to detailed decisions. Keep a versioned protocol and decision register, and add evidence references for the rationale.
2. Recruitment and assignment
Selection · Document enrollment and assignment mechanisms
Investigators assign the intervention, but without a random sequence. Document the selection mechanism and assignment timing. Historical controls additionally require attention to calendar changes and ascertainment differences.
3. Endpoints and estimands
Design · Define outcomes and comparison targets
An estimand specifies the population, treatment conditions, outcome variable, handling of intercurrent events and population-level summary. An intercurrent event changes interpretation or existence of an outcome; it is not synonymous with a missing measurement.
4. Sample size and confounding
Statistics · Plan precision adjustment and overlap
Information planning must reflect the proposed estimator, imbalance, clustering and attrition. Increasing sample size narrows random error but does not eliminate selection bias or residual confounding.
5. Follow-up and safety
Conduct · Track treatment assessments and disposition
Stopping treatment need not end outcome follow-up. Distinguish adverse-event seriousness from severity, define ascertainment windows and keep clinical safety review separate from statistical efficacy monitoring.
6. Data and analysis plan
Statistics · Specify populations estimation and sensitivity
Select baseline confounders using clinical and temporal knowledge. Assess overlap and plausible unmeasured confounding. Avoid automatic adjustment for post-treatment variables, and account for the assignment level in uncertainty estimates.
Equations and statistical interpretation
These explanations describe notation and assumptions. Flow does not calculate these values.
Standardized outcome contrast
m̂(a,Xᵢ) predicts the outcome under intervention a at baseline covariates Xᵢ; the empirical averaging population defines the target contrast.
A causal interpretation needs consistency, positivity, sufficient confounding control and a suitable model. This expression alone does not solve selection, time trends or cluster dependence.
Published example
This intervention evaluation used a before–after comparison, not randomized concurrent assignment. Examine how calendar changes and differences between periods could affect interpretation.
This publication illustrates a study design, not an endorsement of every method or a reproduction of its results. Follow its protocol, supplement and update history when studying the example.
Hypothetical scenario
A service introduces a discharge support program in selected wards and compares it with wards continuing usual care.
- Comparative intervention overview: Define the intervention question, assignment rule, approvals and prospective registration requirements.
- Recruitment and assignment: Document ward selection, participant entry and possible contamination.
- Endpoints and estimands: Define comparable outcomes, intercurrent events and a clear estimand.
- Sample size and confounding: Plan information needs allowing for clustering, imbalance and uncertain confounding control.
- Follow-up and safety: Apply consistent follow-up and safety procedures in both settings.
- Data and analysis plan: Adjust using justified baseline variables, assess overlap and examine selection, calendar trends and residual confounding.
Deliver a populated protocol, decision log, data specification, and planned outputs. Enter numerical findings only after analysis and verification. Keep decisions made before seeing results distinguishable from later amendments.
Sources
- TREND: reporting nonrandomized evaluations of behavioral and public health interventions
- ICH E9(R1): estimands and sensitivity analysis (2019)
- WHO: clinical trials and prospective registration
- Hernán and Robins: Causal Inference—What If
- ISRCTN: what the clinical study registry records
- ClinicalTrials.gov: about the registry
Source check: 2026-09-17. Record the version of each reporting or appraisal tool used in your actual protocol.
Compare all Clinical trials project types · Open editor in a new tab