Once all the subjects have gone home, the samples have been analysed, the data crunched and the results interpreted, the clinical study report (CSR) remains the only narrative detailing the delivery of your study. Memories fade, key personnel move on (or retire) and content of study ‘summaries’ (in documents like the Investigator’s Brochure) are constantly reducing. The CSR is not just a requirement of regulatory reporting but also a critical contributor to collective corporate memory, emphasising the importance of producing timely, accurate and well-written reports [1,2]. Quality control (QC) plays a vital role [1].
As frequent reviewers of sponsor CSRs, we at Niche Science & Technology find that the best time to perform your QC can vary depending on the nature of the project. When done with the right checks and balances, it can take place during the CSR drafting process, which has the advantage of more relaxed timelines, or wait until CSR finalisation, which guarantees that the final product is included in the QC. In any event, a formal QC step is essential under all circumstances, and it should be factored in advance when developing CSR timelines [1,3]. The QC is much more than just a read-through or a ‘review’, as it must include a 100% check of the information in the CSR against the source information [1,4]. The QC must be conducted (usually by an experienced medical writer) by someone not involved in writing the CSR. Although the CSR author is thoroughly familiar with the narrative and the source materials, they are susceptible to what writers term ‘word blindness’ [3].
Furnished with the source documents, including the study protocol, statistical analysis plan and data package (tables, figures and listings [TFLs]), most reviewers are often eager to start confirming the study details. However, I prefer to initially perform a more comprehensive review of the CSR, including areas that are frequently overlooked by authors. This includes:
- Ensuring that the CSR’s title exactly matches that of the protocol
- That the document version is stated correctly throughout
- That the page headers and footers are sized correctly to fit the page orientation and are consistent across the entire document
These characteristics of the document can indicate a great deal about the care with which a report has been prepared [1,2]. Effectively, previewing allows me to focus on the task at hand by ensuring consistency of appearance and adherence to an agreed-upon medical writing style guide [3, 4]. An important aspect to check is treatment group nomenclature, ensuring consistency throughout the document; often variations are introduced during study conduct [1]. Descriptions that remain consistent throughout a document suggest a fully reviewed and polished document. Pagination should also be considered, making sure there are no unwanted blank pages or large areas of white space caused by incorrect positioning of page breaks.
The ‘Print Preview’ function in Microsoft Word is a practical tool to review the ‘look’ of the document before focusing on the text itself. Through this function, the following can be quickly checked:
- Blank pages
- Incorrect gaps between paragraphs
- Tables split across pages
- Incorrect justification
Such previewing may seem minor, but it is a useful way to familiarise oneself with the CSR and understand the document’s overall structure, which assists with time management and delivery planning [3].
A formal QC requires a systematic approach involving appropriate checklists directing a step-by-step approach [3,4]. The QC must follow the appropriate CSR template and style guide as well as the relevant QC standard operating procedures (SOPs) [1,3]. It can be helpful to QC check the methods first, to gain a better understanding of the study design before addressing the results. This involves confirming that the information presented reflects the source materials (e.g., protocol and amendments as well as any statistical analysis plans, study reference manuals and/or file notes). The statistical methods in a CSR (contained in Section 9.7 of the CSR as per International Council for Harmonisation E3 guideline) are often sourced from the statistical analysis plan rather than the protocol [1]. The rest of the methods, including the details regarding Ethics and Investigators, the Introduction, Objectives and remaining aspects of the Investigational Plan, are principally sourced from the study protocol [1].
The principal task with regard to the results is to perform a 100% data QC, that is, verify all cited data against the source data (i.e., TFLs) [1,4]. When checking data against the TFLs, it is essential to confirm meticulously that data appearing in the in-text tables match the source documents. Additionally, each in-text display should contain a footnote featuring the data source. These QC checks must also confirm that all data (numbers/statistics) in the text of the results match the source documents. Even if there are no mistakes, it is considered best practice to keep a formal record of the review [3]. Data checking can be the most time-consuming task carried out during a QC; however, this investment ensures that the aspect of the report most prone to error, transcription of data, is correct, thereby supporting regulatory-grade quality [1,2].
Internal consistency of the data presented in the CSR is also essential [1]. After verifying the results sections, the Conclusions are checked to make sure they accurately reflect what is stated in the main body of the results. Any statements in the Discussion section that pertain to study results must also be checked against the main body of the CSR. Similarly, the synopsis must also be cross-checked against the main body to ensure consistency. The CSR synopsis is often prepared at an earlier stage and is therefore prone to error if later changes are not mirrored [1,2].
Overall, a CSR should be checked for consistency of presentation, format, grammar and spelling [2,3]. The type of English spelling followed (UK vs US) should be consistent throughout the document. When reviewing, in-text tables and figures should follow a consistent presentation style and align with the applicable style guide [3].
Numbering of in-text displays must be sequential and in keeping with the CSR template [1]. The table of contents, list of tables and list of figures should also be updated to ensure structural accuracy. Checking the correct and full listing of abbreviations is also necessary, with definitions provided at first use and in the synopsis, which often functions as a standalone regulatory document [1].
Hyperlinks and cross-references within the CSR should be verified to ensure functionality and correct navigation [3]. When providing QC comments, reviewers should be clear, precise and reference exact locations in source materials [3,5]. Any information that cannot be located in the source materials should be marked as “unable to verify” (UTV). Referring authors to relevant sections of style guides or SOPs promotes learning and consistency [3].
When finishing the review, it is good practice to record that a 100% QC check has been conducted and cite the source documents, versions and dates used [3,5]. Comprehensive QC reviews are intensive and require attention to detail, but systematic approaches and robust checklists support the production of high-quality regulatory documents [1–5]. For further tips on CSR development, see our ‘Insider’s Insight into Clinical Study Reports’.
References
- International Council for Harmonisation (ICH). E3 Structure and Content of Clinical Study Reports. Geneva: ICH; 1995.
- European Medicines Agency. Guideline on the Clinical Study Report. London: EMA; 2018.
- American Medical Writers Association; European Medical Writers Association; International Society for Medical Publication Professionals. Good Publication Practice for Communicating Company-Sponsored Medical Research (GPP3). Ann Intern Med. 2015;163(6):461
- Hardman TC. Good Publication Practice 3 and the Evolution of Ethical Industry-Sponsored Clinical Reporting. September 15, 2015
- US Food and Drug Administration. Guidance for Industry: Submission of Clinical Study Reports. Silver Spring (MD): FDA; 2013.
