Data Is Only as Strong as the Analysis Behind It

Your trial is complete. The database is locked. Now comes the work that turns raw data into a submission package FDA can evaluate — and that reviewers will trust.

This phase looks like execution. It is. But it's also judgment: knowing which analyses to run and which to resist, how to handle data anomalies without introducing bias, how to document methods with the transparency that regulatory review demands, and how to present results that tell a clear, defensible story.

BSC has supported the analysis and reporting phases of hundreds of medical device submissions. We understand what rigorous looks like — and what it takes to produce work that holds up when reviewers look closely.

What We Do

From Data to Analysis-Ready Datasets

We begin where the data is: raw, messy, and rarely submission-ready. We conduct systematic data screening, identify protocol deviations and violations, derive analysis populations, and build ADaM-compliant datasets with complete derivation documentation and audit trails. Every dataset we produce is traceable from source to output.

Statistical Analysis

We execute the full scope of pre-specified and exploratory analyses — primary and secondary endpoints, subgroup assessments, sensitivity analyses, and post-hoc analyses where appropriate and defensible. For trials with interim analyses, we provide DSMB packages and support ongoing safety monitoring. Our methods span both frequentist and Bayesian frameworks, and we bring the same rigor to every analysis regardless of complexity.

When the Data Isn't Telling the Story You Expected

It happens more often than sponsors want to admit. The primary endpoint missed. A subgroup tells a more complicated story. The effect size is smaller than powered for. Before concluding the program is lost, experienced statistical eyes need to assess what the data actually supports.

BSC helps sponsors navigate this with honesty and precision — evaluating what the data can and cannot support, identifying defensible analytical approaches, and determining whether a credible path to submission exists. We don't manufacture outcomes. We find the strongest defensible story the data allows, and we tell it clearly.

This isn't spin. It's rigorous, experienced statistical judgment applied at the moment it matters most.

Tables, Listings, and Figures

We produce validated TLFs that are traceable to source data, aligned with the pre-specified SAP, and formatted for regulatory submission. Every output is QC'd against the analysis dataset and the programming specifications. Nothing leaves without documentation.

Clinical Study Report

We author the statistical methods and results sections of the CSR — not just the numbers, but the interpretation, the limitations, and the scientific context regulators need to evaluate your findings. We support full revision cycles and final production, and we deliver sections that are consistent, well-documented, and submission-ready.

What This Requires

The analysis phase is where flawed foundations become visible. Inconsistent TLFs, derivations that can't be traced, methods that diverged from the SAP, conclusions that outrun the data — any of these invite FDA questions that delay approval and erode reviewer confidence.

BSC operates under 21 CFR Part 11-compliant processes with full audit trails. Our work is documented, reproducible, and built to withstand the scrutiny of regulatory review. We've rebuilt submission packages that others couldn't get across the line. We'd rather be involved early enough that it doesn't come to that.

Need a statistical team that can carry your analysis across the finish line? [Talk to us.]