Research Transparency – a guide for students

ONLINE POST in ENGLISH, commissioned | AMP Student 2017;4

As a medical student, you may choose – or may be compelled – to conduct and publish some research. If so, congratulations: you are embarking on an important journey.

What do you need to know before you start? Is it enough to understand the principles of epidemiology, medical statistics, and maybe qualitative research? Not quite: there is much more to think about if you want to do and publish a good, worthwhile study that other people can use and trust.

As the authors of an important 2014 series in the Lancet showed, far too much health research is wasted because it asks the wrong questions, is badly designed, is never published, or is poorly written up and reported. Don’t contribute to this waste. Ensure that your very first study, even if quite small and simple, is able to add something useful to knowledge. You just have to follow the right steps.

I mentioned research ‘that people can use and trust’. This is really what we mean by ‘research transparency’. It depends on having a good research question that you can answer and interpret through the best possible methods, valid analyses, and cautious discussion. You’ll also need to ensure clear and complete reporting of the question, methods, and results so that your research is potentially reproducible. And you should declare all authors’ contributions to the work, and state any conflicts of interest.

6coubedwg24-luis-llerenaWhat do I mean by a ‘research question’? Well, your hypothesis might be ‘I think there’s a link
between A and B’; your aim might be ‘I’m going to look at this link in patients attending the clinic at my hospital’; and your objective might be ‘I will examine the link by looking at the medical records of women at the clinic’. Your research question needs to be detailed and answerable, like this: ‘in 100 consecutive women aged over 60 who attended the clinic over the past year, what was the association between A and B, according to the information I will collect from their records using a detailed proforma?’ You must define A and B carefully. For example, A might be ‘newly diagnosed type II diabetes as defined by fasting glucose’, and B might be ‘smoking status (cigarettes per day) as recorded in the clinic records’.

Every minute you spend on finding and focusing your research question will be time well spent. A good research question will satisfy the FINER criteria (Feasible and answerable with available resources; Interesting to the investigators and to others; Novel enough to confirm, refute, or extend knowledge; Ethical and approved by an ethics committee; and Relevant with potential to influence practice, policy, or future studies).

You can start to establish relevance and novelty by discussing your study idea with colleagues and supervisors. Then you’ll need to search the literature for previous studies related to your idea, and see if you can identify a gap in the evidence base. It may quite a small gap, and you may be able to answer your research question with a quite simple study design. It is far better to do a simple study well than to attempt an overcomplicated study and get stuck.

Find the best study design to answer your question using the search tool at the EQUATOR Network (Enhancing the QUAlity and Transparency Of health Research). EQUATOR has a free online library of writing tools called ‘reporting guidelines’ – checklists for transparent and accurate reporting of studies and protocols. It’s an essential resource for designing and writing up research. And, thanks to the Pan-American Health Organization (PAHO) the network’s resources are available in Portuguese too.

Boa sorte!

Disclaimer: Trish is editorial lead for the BMJ Research to Publication elearning programme (rtop.bmj.com) which is available on subscription.