Clinical research technology and CRO services

Name the snapshot, the plan and the method before you quote the number.

A sponsor should be able to see which data and which program produced a result. We write analysis plans, calculate sample size from stated assumptions, and run the methods the software implements: diagnostic accuracy, agreement, and the survival and longitudinal methods that are actually present. A textbook method that is not implemented is not offered as if it were.

What you can use

Plan

Population, endpoint, missing data and multiplicity, before the confirmatory look.

Sample size

Assumptions written beside the number.

Diagnostic accuracy

Sensitivity, specificity and intervals where those methods exist.

Agreement

Methods such as Bland–Altman where the study needs them and the code exists.

Outputs

Tables and figures tied to the run.

Verification

A second qualified person can check the output. That is a service, not an automatic certificate.

How the work moves

Prespecify

The plan is versioned.

Run

On the frozen snapshot.

Review

A statistician accepts or rejects the output.

File

The report cites that output.

Related

Deliverables

Statistical analysis plan
Population, endpoint, missing data and multiplicity, before the confirmatory look.
Sample-size note
The assumptions beside the number. We will not publish a number without them.
Accuracy or agreement output
Only where that method is implemented and the plan asks for it.
Tables, listings and figures
Tied to the named snapshot and the named run.
Verification note
A second qualified person can check the output. That is a service, not a certificate.

Methods we will not pretend to have

Diagnostic accuracy, agreement, and the survival and longitudinal methods that are actually in the software can be planned. A method that is not implemented is named as unavailable, not offered as if a general statistical language were attached. There is no claim of an arbitrary R runtime. You provide the endpoint definition and the snapshot. You accept the output before a report quotes it.

Questions

Is every R package available?

No. There is no general claim of an arbitrary R runtime.

Software on this page is what the product can record today. CRO work is a professional service under a separate scope. Neither is a certification, a regulator’s approval, or proof that a study is compliant because the software was used. A qualified person still signs consequential decisions.

See the workflow on your protocol.

The conversation is about your study design and what the software can and cannot do today.