Study design and research methodology
Reliable scientific evidence begins long before the first data point is collected.
A relevant research question does not automatically lead to a reliable answer. The study population, choice of comparator, definition of outcomes and method of data collection all shape the conclusions that can ultimately be drawn. Once these decisions have been made, even the most sophisticated statistical analysis cannot necessarily correct their limitations.
Heorfy provides methodological consulting to turn clinical and scientific questions into feasible studies capable of generating interpretable, decision-relevant evidence. Collaboration can start while the research hypothesis is still being developed or when a study protocol needs independent assessment before implementation.
How Heorfy can contribute
Specific scientific capabilities, tailored to each study design and project objective.
Study design
Selection and justification of the study design best suited to each research objective.
Research questions and endpoints
Definition of hypotheses, target outcomes, comparators and measurement strategies.
Sample size planning
Power, precision, simulation and feasibility scenarios.
Bias and validity
Assessment of confounding, selection, measurement quality and threats to interpretation.
Study protocols
Development and independent review of methodological and statistical sections.
Early analysis planning
Aligning the statistical strategy and future SAP with the study design.
From a research question to a design that can answer it
The first task is to establish exactly what needs to be learned and what decision the findings are intended to inform. Describing disease progression is not the same as estimating a treatment effect, identifying a prognostic factor or evaluating the performance of a diagnostic test. Each question calls for a different approach and carries its own threats to validity.
Methodological planning helps align objectives, eligibility criteria, target populations, outcome definitions and assessment schedules. It also supports the choice of relevant comparators, identification of potential confounders and an explicit assessment of whether the proposed design can support causal interpretation or only the estimation of associations.
Choosing the right design prevents problems later
Randomised trials, cohort studies, case-control studies, cross-sectional designs and studies based on real-world data serve different purposes. The strongest approach is not necessarily the most complicated: it is the one that can produce a valid answer under realistic clinical, operational and resource constraints.
Key considerations include selection bias, measurement quality, loss to follow-up, confounding and alignment between study design and planned analyses. Where relevant, a project may also require trial emulation strategies, sensitivity analyses or causal inference methods, with their assumptions made explicit rather than taken for granted.
Sample size and feasibility are inseparable
Sample size calculation should never be reduced to entering numbers into a formula. It depends on the effect that matters clinically, defensible estimates of variability, plausible recruitment and attrition rates, and the consequences of alternative scenarios.
When early assumptions are uncertain, simulation and sensitivity analysis can clarify what precision or statistical power is realistically achievable. This creates a transparent basis for decisions and helps avoid promising conclusions that the available study design cannot support.
A protocol connecting scientific, operational and statistical decisions
Good methodology becomes tangible in a clear, internally consistent study protocol. Heorfy can help develop or critically review the scientific rationale, objectives, design, outcomes, sample size, assessment schedule and statistical approach. Clinical and operational decisions should be consistent with the future statistical analysis plan, rather than forcing statistical methods to compensate for weaknesses in the design.
Depending on the project, outputs may include a full methodological section or protocol, a documented sample size justification, an independent study design review or a decision record that makes assumptions and recommendations transparent to the research team.
Better design, more useful evidence
Sound study design does not eliminate uncertainty. It makes uncertainty visible, measurable and interpretable. That is the basis of findings that withstand scientific scrutiny and support more informed clinical and healthcare decisions.