The role of non-interventional studies

Staying abreast of changes in the understanding of disease biology, diagnostic workup and treatment paradigms can be daunting; however, it is essential in the drive to continually improve patient care. As new drugs come to market, their role in clinical practice clearly needs to be understood, and perhaps one of the best mechanisms for assessing the true effect of any agent is through non-interventional studies (NIS). Indeed, both observational studies and NIS provide a wealth of information contributing to as many as nine out of every 10 research papers published.

In the past, company-sponsored NIS have been criticised for having a more commercial than science-based focus and have been used to increase clinicians' practical experience with a particular agent.

Today there are an increasing number of guidelines to strengthen the methodology, conduct and reporting of NIS to ensure they achieve appropriate scientific rigour and therefore contribute to our understanding of the disease and its treatment. In an era of increasing focus on the safety of medicines, NIS play an essential part in further demonstrating the benefit–risk profile of an agent in clinical practice where it is likely that patients will exhibit a much broader range of characteristics than those defined by regulatory phase III protocols.

While NIS are out of scope of The Clinical Trials Directive, other guidelines and country-specific laws and regulations apply to ensure the safeguard of patients in observational studies. Recently the European Network of Centres for Pharmacoepidemiology and Pharmacovigilance (ENCePP), sponsored by the European Medicines Agency, issued a guidance document and checklist for methodological standards in pharmacoepidemiological studies.

The European Federation of Pharmaceutical Industries and Associations also give guidance on the conduct of company-sponsored NIS. As with any clinical trial, a scientifically valid objective and a robust protocol are essential. Methodology should be driven by the research question and expert guidance may be valuable to ensure the outcomes contribute to clinical knowledge and are not biased.

To maximise the scientific value of the study, endpoints must be valid, even though they are observational, sample sizes must be sufficient and appropriate statistical methods should be used to analyse data and test for confounders and effect modifiers. Furthermore the protocol must address the potential selection bias inherent from not having randomised treatment due to the non-interventional nature of the study.

Given the contribution that NIS make to clinical knowledge, publication of results is essential and this should be considered at the outset, ensuring adherence to guidelines of the International Committee of Medical Journal Editors and registration on appropriate databases, for example the ENCePP electronic registry. When publishing results from studies of this nature it is important that sufficient emphasis is given to the study design and analysis to prevent over-generalisation of results, without adequately addressing the study limitations.

Since NIS contribute enormously to our understanding of disease management and the 'real-life' risk–benefit evaluation of different treatment options, they remain an essential part of the medical education arena. However, to ensure NIS contribute to the furthering of medical knowledge, the design of scientifically valid, robust protocols is paramount.

The Author
Gabriel Schnetzler MD, FMH Pharm Med, MFPM
senior medical consultant, Prism Ideas

Email him at 
gabriel.schnetzler@prismideas.com