Do you need a biostatistician for a survey or PRO study?

What a biostatistician does on a survey study

The role is often imagined as running the tests at the end. In practice the highest-value contributions come earlier. Before data collection: choosing validated instruments, estimating the sample size for the primary comparison, and writing an analysis plan that names the outcome, the comparison, and the model. During collection: checking early data for scoring and coding problems while there is still time to fix the survey. After collection: preparing and scoring the data, fitting the planned models, handling missing responses defensibly, and writing methods and results text a reviewer can follow.

Most of the delay and cost in survey studies sits in the middle step. Cleaning exports, mapping items, reverse coding, and documenting every decision routinely takes longer than the modeling itself.

When early input pays for itself

Three situations justify a conversation before the first participant is enrolled. A primary comparison that will be underpowered if the sample is smaller than hoped. Repeated measurements over time, where the analysis model shapes how many visits and participants you need. And any plan to modify a validated instrument, since a modified instrument loses its published scoring and norms.

Retrofitting a plan after collection is possible, and it is how most reanalysis projects begin, but the options narrow. A missing baseline variable cannot be recovered, and a sample that is too small stays too small.

Institutional core, consultant, or service

An institutional biostatistics core, usually housed in a clinical and translational science institute, is the default route at academic medical centers. It is often free or subsidized for a first consultation, staffed by trained biostatisticians, and familiar with your IRB. The constraint is capacity: intake queues of several weeks are common, hourly rates apply beyond the initial consultation, and preparation work is frequently handed back to the study team.

An independent consultant offers faster access and deeper engagement on a single study, at hourly or project rates. Quality varies, so ask for a written analysis plan and examples of methods sections they have produced.

A managed analysis service takes on the preparation and analysis as a project with a fixed scope and price, delivers tables, figures, and methods text, and explains the decisions. It suits teams whose bottleneck is bandwidth rather than access to expertise, and who want the work documented so it can be inspected. Sutrix operates this way for questionnaire, PRO, and clinical assessment studies.

• Core: lowest cost, longest wait, preparation often stays with you.
• Consultant: faster, hourly, quality depends on the individual.
• Managed service: fixed scope and price, preparation included, decisions documented.

What to bring to any of them

The research question in one sentence. The instruments used, with versions. How the data were collected and exported. The primary comparison and any secondary ones. The deadline that actually matters, such as an abstract submission or a grant renewal. A statistician who receives these five things can scope the work in a single meeting. One who receives a dataset and "see what you find" cannot.