Biostat Core Throughput
A research centre with 250 investigators and 6 or 7 biostatisticians queues every request. It takes 1 to 2 weeks to get the first meeting, then rounds of email to settle the question, the variables and the method, and the report comes back weeks later with scripts the investigator never sees.
Teams need the intake, the plan and the clarifications handled before the statistician's time is spent, the analysis run and checked without a second person re-deriving it by hand, and the report written from the run.
Done by hand, the core's capacity is its headcount, every study restarts the same conversations, and the backlog decides which questions get answered.
3× study capacity with the same team
Intake, cleaning, plan extraction, clarifications, execution, independent validation and the report draft run in Sutrix; the statistician's time goes to approvals and judgement.
Plan approved before anything runs
The estimand, the plan, its deviations and the objection ledger are hashed and signed; execution refuses an unsigned plan.
Investigator questions asked once
Each open question arrives as one question with two to four options and a recommendation; the answer is recorded and the plan updates.
How it works
Take in the study: dataset, codebook, protocol, scoring key and prior paper in their own slots, screened and hashed on upload.
Extract and review the plan: outcomes, covariates, model, missing data and the multiplicity correction, with the objection ledger written before the review gate.
Run and validate: the analysis executes from the signed plan and the independent validation re-derives every number.
Approve the report: 7 sections drafted from the run, every number checked against the results, exported as an APA Word file with figures.