Abstract
Science has many enemies, and editors of this journal always strive to monitor each one of them, starting with strict adherence to the high Cephalalgia standards in all aspects of quality control (1). Working and publishing in science means following rigorous rules. The reason is that evidence; that is data, are only meaningful if they are true (2). But who decides what is true? A problem inherent to science is that we do not know what is true and, even more, probably cannot know it. Science lives on doubts. A good scientist is the first to concede that he/she may be wrong (3), and there is not a single construct of truth that survived the ages. One can look at science as a never-ending battle between good and better ideas; that the current up-to-date idea is tomorrow’s error of yesterday, as, for example, Thomas Kuhn thought about it (4). Alternatively, one understands science as a discourse between schools of thoughts that constantly collect data to finally triumph over all others and get nearest to “the truth.” Either way, it all hinges on data and their interpretation. As a headache scientist once stated, “you can argue with any interpretation, but you cannot argue with data”. Whether it is about a small piece of a puzzle or changing the way we think about a cause or a treatment: It all stands and falls with data and, consequently, with the collection of these data. Furthermore, here is the root for all except the worst enemy of science: The quality of such data.
Science’s enemies are indeed numerous: It starts with a missing pre-specified hypothesis, the ignoring of order and time effects, a missing control or placebo group, too small numbers, insufficient description of methods and insufficient statistics, missing power analysis, lack of control for multiple comparisons, no report of missing/negative data and, of course, the blunt forgery of data or results. Science has developed numerous strategies and tools to control for each of the above-mentioned points, such as rules for statistics, standards for reporting and the quest to follow reporting guidelines such as STROBE, ARRIVE, and MOOSE, to name a few. If one reflects these points quite plainly one wonders which of these hazards is the worst for science. The question seems silly, since each one of these issues is damaging science profoundly, and surely plain forgery of data is the worst enemy of all?
The answer is that none of those is the worst. The worst enemy of science is personal belief. Even the most rigorous and transparent collection of well-controlled data will gain no traction when facing a belief. Belief does not need data; belief needs no discussion, since someone who believes is content with their own company. The fact remains that what holds true is not just in medicine, but in all aspects of our life, and we can observe the constant discussion and battle wherever we look. (Please note that for the purposes here, I am explicitly not referring to religion and faith, but rather more globally as a kind of diffuse knowledge about a given subject). An obvious example in headache research is the question that exists since Wolff’s pivotal book (5): Are primary headaches such as migraine vascular or not? Two schools of thought exist (please note, citations only highlight certain key aspects and do not mean that the authors represent these schools): One school argues that intracranial vessel behaviour explains most facets of migraine and treatment must consequently have these vessels as a primary target (6–8). The other school argues that central (brain) responses and homeostasis explain most facets of migraine and treatment must consequently aim for the nervous system as a target (9–11). The current compromise is the “neurovascular theory”, which builds on the direct effects of the trigeminal system on intra- and extra-cerebral vessels (12,13).
Why is this a good example? A short time ago, a review was submitted which dealt with a study that offered valid data for one of the above-named viewpoints. One of the reviewers stated a few arguments and issues to be discussed by the authors and suggested rejection by concluding “I do not believe it”.
If our reviewers, our peers, and our community already take the view that data do not count because they do not believe in the argument, how can we expect that neuroscientists and doctors not deep into the matter, or indeed the society that pays for science, accept our scientific authority? How can science evolve when stopped by belief? It is a real threat to all of us and cumulates in the fact that our achievements can indeed be snatched away from us by re-labeling falsehood as “alternative facts”. The problem is fatigue, and a laziness of thinking; the problem is tiredness of a mind that does not even bother to listen to other viewpoints. We need to stand together and always fight back that data truly matters, by reasoning and several other means that the American Association of Science (AAN) has pointed out recently (14).
There is no question that we all publish too much and the so-called “salami tactics” are only a small part of this problem. There is also no question that poorly reported science, and also forgery, exist. And this is something that we need to identify and to battle. Science and the reporting of scientific content is work (15). Constant, relentless work to become even better than the excellence we achieved yesterday. The key is sobriety and reliability. Scientists and their science must continually work on their credibility to earn the trust that society puts into their work. How? One can bring it down to the simple formula: That the data must be produced and reported in a way that allows everyone to reproduce them. If data is reproduced, and reproduced again, they are more likely to represent our current way of thinking. And this is what science is about. It is the way we think and consequently behave. Let us all defend the integrity that science is based upon.
