Writing the report

This branch covers producing a clinical study report, from the date it must meet to the page a regulatory reviewer reads, and its members divide that work by the task in hand: scheduling the report, drafting it from the precursor documents and the outputs, producing its patient narratives, presenting its results, reporting and tabulating its safety data, and reporting a study whose conduct an external event disrupted. It leaves the guidelines and templates a report is written to to the branch on the standards a report follows, the choice of a shorter format to the topic on full, abbreviated, synopsis and interim reports, the structure of the dossier summaries the report feeds and the pooling behind them to the branch on the dossier and its documents, the general craft of tables and figures to the branch on presenting numbers, and quality tolerance limits, including the breaches a report summarises, to the branch on good clinical practice and trial quality.

• Scheduling the clinical study report — meeting a date the sponsor fixes before drafting begins: the complexity and precursors that drive each task, frontloading the front end before the final outputs arrive, how long each analysis and reporting task takes across the industry, negotiating to meet the date, and lean methods that take waste out of the writing process
• Drafting the report from the precursor documents and the outputs — what the front end holds and adapting it from the investigator's brochure, the protocol and the analysis plan rather than lifting it, writing factual results from the final tables, figures and listings, and where post-hoc and supportive analyses go
• Producing the report's patient narratives — the clinical and pharmacovigilance databases behind the narratives, medical monitors drafting and physicians reviewing serious adverse event narratives, when the work falls, why the workload is heavy, and narratives written by hand, programmed, generated by utilities and drafted with AI
• Presenting the results for a regulatory reviewer — a report a reviewer comparing applications can navigate, summarising its own data completely for the dossier's summaries, tables, graphs and short sentences in place of paragraphs, clear key messages, and data the reviewer can inspect and reanalyse
• Reporting complex safety data in the report and the summary of clinical safety — making the important safety findings findable among hundreds of outputs: navigating unhelpful outputs, turning adverse event tables into narrative while leaving causality to a physician, reporting what matters, how much detail lean writing should leave, and the kick-off meeting that settles it
• In-text safety tables from immutable outputs — read-only, append-only safety tables generated from immutable data and linked into the report rather than retyped: how they differ from traditional tables, building and linking them, the team's roles, introducing them, and their benefits and costs
• Reporting a study disrupted by an external event — what the COVID-19 pandemic did to trials and the emergency guidance the EMA and the FDA issued, the documents and data domains the disruption reached, whether its data are integrated into the normal report or kept separate, where contingency measures, deviations and affected participants are reported, confounders, and how much to report