Science without sense…double nonsense

Píldoras sobre medicina basada en pruebas

The Ten Commandments

This post is also available in: Spanish

Print Friendly, PDF & Email

Do not panic, we are not going to talk about Bible stories today. Neither was I thinking about the film, but remembering the famous movie it came to my mind the image of what a typical movie’s scientific guy looks like. He’s easily recognized: tall, handsome, sturdy, and extremely intelligent, wearing a white extremely white coat. One last detail, he’s always flirting with the hunkiest girl of the cast. And do not forget scientist girls: it is the same for them.

However, real life is much gloomy. Anyone can be a scientist (although nowadays you might have to go abroad to be a scientist in Spain). Just to give you an idea, I myself was a scientist for a period of my life, a longer than I like to remember ago. And if you think about it, most of us have done some kind of research, because for doing research it’s not essential to enclose yourself in a lab or to be as handsome as scientist in the movies are. What it’s indeed convenient is to have very clear ideas about what do you intend to do and to take into consideration a number of aspects before starting to spend time and money (what is, apparently, the essence of research in our times).

The first issue is the identification of the doubt or problem we have, which serve to define the question we intend to answer with the study. This will be our working hypothesis, the study goal. It may be useful to formulate a clinically structured question (do you remember PICO?), that help us to define the main outcome of the study and, besides, to serve to design the search that allows us to get and review the state of knowledge on the subject, so we focus the work properly and justify their relevance and viability.diez mandamientos

Second, we have to decide what type of design will be best suited to achieve our goal. This is very important because it will condition many of the following steps. Furthermore, if we choose the wrong design, the study can be unviable. For example, if we want to show the detrimental effects of tobacco smoking we won’t be able to set a clinical trial, because we can´t allocate participants to smoke two packets a day. We’d better set a cohort study and check what happens to those who smoke and to those who do not.

The third aspect is to select the appropriate study population. We will have to think how we will select the participants and how we will assess their suitability for entering the study and the technique we will use to form the different groups. Also, fourth step, we’ll previously calculate how many we will need. Calculating sample size allowed us to gauge the power of the study to answer the target question without spending more resources than necessary.

Fifth, we must pause and think what will be the primary outcome variable and how we are going to measure it, plus if we will need another type of secondary outcome variables. The main outcome variable must be significant from the clinical standpoint, both for the investigator and the patient who will benefit from the study, but it also must be able to discriminate whether participants benefit (or harm) from the intervention or exposure factor.

Sixth, it should be carefully planned how we will collect the data and what will be the strategy for analysis. We will think about the way we’ll describe our results and what measures of central tendency, dispersion, association and clinical impact we will use. Obviously, all these parameters depend on the type of study we undertake.

The seventh step is to organize the implementation of the study according to all that we have previously considered to finally carry it out with rigor and thoroughness (eighth step).

After completing the study, we have yet to meet two points of this Decalogue. It is very important to interpret the results with caution. We will not settle with statistically significant differences, but we always complete the study with measures of association and appropriate clinical impact. We will have to evaluate the beneficial effects for the patient that may arise from the study, but before making conclusions or recommendations we always have to consider aspects about adverse or annoying effects, cost effects and, do not forget, patient preferences, because he’s the one we want to benefit.

Anyway, although we do a masterful job, if applied only to our practice the effort will surely not worth it. The real utility of research is that its results are applied to clinical practice of so widely as possible. Hence the importance of the tenth step of the research process: the dissemination of results. It is essential that results are made known, usually through publications and conferences. In cases of clinically important results the ideal is to include them in international databases that are very consulted, as they may be located, evaluated and used by most professionals.

If any of you are still reading at this point but has no plans to do any research work in the foreseeable future, do not worry: everything said so far can also be helpful. Because it is that a similar pattern of thinking can be very useful for assessing research done by others, which we practically do (or should do) a lot. Moreover, these steps are part of the requirements of many checklists that publishers use to evaluate scientific papers and many tools used in critical appraisal. But that’s another story…

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: