The research spectrum

Different questions require different methods

Most people are trained to think of human research as a single tier with the RCT as the gold standard In practice, human research is a spectrum of designs, each answering different questions at different costs, with different types of evidence and different regulatory implications.

Selecting the right design for the question is as important, if not more, than executing it correctly. An RCT answers whether a compound causes an effect under controlled conditions. An epidemiological study answers what happens in real populations over time. A meta-analysis synthesizes what the entire body of literature actually shows. Registry studies capture real-world outcomes at scale. None of these is inherently better. They answer different questions.

Case Studies Individual observations, hypothesis generation
Epidemiology Population patterns, real-world outcomes
Registry Studies Longitudinal real-world data collection
Meta-Analysis Evidence synthesis across the literature
RCT Controlled efficacy under defined conditions
Scientific services

A full methodological toolkit for botanical research

Primary service

Clinical Trial Design and Execution

Randomized controlled trials for natural products are the gold standard if the goal is efficacy evidence and claim substantiation. NRI designs, conducts, and analyzes human clinical trials with methodology built specifically for botanical and nutraceutical interventions, not pharmaceutical assumptions applied to supplements.

Phase 2 through Phase 4 designs. Placebo-controlled, active-controlled, crossover, and parallel group. PI-led from protocol to publication.

Pilot and feasibility studies Pre-efficacy work to confirm methodology, refine endpoints, and estimate effect sizes before a full-powered trial.
Full efficacy trials Properly powered RCTs designed for peer-reviewed publication and regulatory claim substantiation.
Epidemiology and observational research

Real-world outcomes research

Not every research question is best answered by an RCT. Epidemiological and observational designs answer questions about what happens in real populations under real conditions, at costs and timescales that RCTs cannot match.

For safety questions, long-term outcome questions, and population-level pattern questions, observational designs often provide stronger, more generalizable evidence than a controlled trial in an artificial setting.

NRI example: The lavender and tea tree endocrine series

NRI's three-paper epidemiological series on lavender and tea tree essential oils investigated the proposed link between topical essential oil exposure and prepubertal gynecomastia in children. The research question required a population-level epidemiological approach, not an RCT. No controlled trial could ethically or practically answer whether long-term low-level topical exposure at population scale produces endocrine effects in children. The epidemiological series, combined with measurement instrument development, produced the most comprehensive evidence available on this question and closed a significant regulatory concern for the industry.

Cross-sectional studies Population snapshots examining the relationship between exposure and outcome at a single point in time.
Cohort studies Prospective or retrospective follow-up of defined groups over time to examine outcome patterns.
Case-control studies Comparison of cases with specific outcomes against matched controls to identify contributing factors.
Registry studies Systematic longitudinal data collection on defined populations to capture real-world outcomes at scale.
Evidence synthesis

Systematic reviews and meta-analyses

A single trial answers a specific question under specific conditions. A systematic review with meta-analysis answers what the entire body of evidence shows, with what consistency, and at what effect size. For ingredients with multiple existing studies, this is often the most powerful and most credible evidence deliverable available.

Meta-analyses also identify where the existing evidence is weak, inconsistent, or methodologically flawed. For brands planning a trial, a systematic review of existing literature is the essential first step: it defines what is already known, what the effect size estimates suggest for power calculations, and what methodological approach is most likely to produce publishable results.

NRI example: Pediatric aromatherapy meta-analysis

NRI's meta-analysis on pediatric aromatherapy for procedural distress synthesized the existing evidence across multiple trial designs and populations to produce the first comprehensive quantitative estimate of effect size for this indication. The meta-analysis established NRI as the leading methodological voice in pediatric essential oil research and provided the evidence foundation for subsequent trial design in this space.

Systematic review Comprehensive, reproducible literature search with quality assessment using Downs and Black or equivalent checklist.
Meta-analysis Quantitative synthesis of effect sizes across studies using appropriate statistical models in Stata.
Scoping review Rapid evidence mapping to characterize the volume and nature of research on a topic before a full systematic review.
PROSPERO registration Pre-registration of systematic review protocols for peer-reviewed publication credibility where timelines allow.
Statistical services

Biostatistics

Simplistic, elementary statistical models lead to false negatives and failed studies. The statistical analysis plan determines whether a well-designed trial produces publishable results or an inconclusive dataset. NRI uses Stata v17 for all primary analyses with full pre-specified statistical analysis plans.

NRI example: REDRESS wound healing trial

The REDRESS study on a novel medical device wound healing required statistical modeling for blood clotting and tissue repair outcomes, outcomes with non-normal distributions, time-varying measurements, and multiple correlated endpoints. Elementary parametric models would have missed real effects or inflated false positive risk. The biostatistical approach was a core methodological contribution of the study, not a secondary consideration.

Power analysis Sample size determination based on realistic effect size estimates for botanical interventions, not pharmaceutical assumptions.
Statistical analysis plan Pre-specified primary and secondary analyses, handling of missing data, and sensitivity analyses.
Primary analysis Full statistical analysis in Stata v17 with appropriate model selection for the data structure and outcome type.
Manuscript support Statistical results section preparation and methods reporting to peer-review standards.
Publication services

Manuscript preparation and publication support

Conducting a good study and publishing a good study are different skills. NRI provides full manuscript preparation services for studies conducted by NRI and for studies conducted elsewhere that need publication support. The overwhelming majority of NRI studies are published in peer-reviewed journals or presented at scientific conferences.

Full manuscript preparation IMRAD structure, methods reporting to CONSORT or PRISMA standards, results tables and figures.
Journal selection and submission Target journal identification, formatting to journal requirements, and submission management.
Peer review response Reviewer response letters and manuscript revisions through acceptance.
Conference presentation Abstract preparation, poster development, and oral presentation support for scientific conferences.

Need research beyond the RCT?

Tell us your question and your timeline. We will recommend the right design.

Get in touch