Quality problems of medical research
When I first read that my genetic variant is supposedly benign, I was shocked. “That’s an unfortunate mistake”, I thought. But since, I have read many more medical scientific papers, and it’s not just a mistake, it’s one of many signs of the poor state of medical scientific writing. Before Fabry hit, my field of interest was historical linguistics. In that field of research, no new data can be generated, and new insights are mainly gained from combining the same sources in ever new ways. From this background, I have identified a number problems in medical scientific research. Two of them pose a systematic challenge.
1. The extrapolation trap. Sometimes, no distinction is made between solid conclusions and tentative conclusions that are partially based on assumptions. Regarding my GLA-variant D313Y, there’s the following chain of extrapolations. First, the association between Gb3 deposits and Fabry disease is turned into a causal association. Next, that causal association becomes a singular exclusive pathomechanism. Only then can a variant that does not trigger that pathomechanism be considered benign. The final extrapolation is that, assuming that that specific genetic variant is benign, the Fabry symptoms in patients with that variant must have different a cause. Every time an extrapolation is presented as a fact, that moves the interpretation of data further away from the actual data. In the end, it might lead to a complete disconnect between the two.
2. A miraculous multiplication of sources. Conclusions, including tentative conclusions that are treated as facts, are repeated in review papers. Then, those review papers are cited in other review papers. In the end, there appears to be a large body of literature supporting certain views, without a correspondingly large amount of data to back them up. This effect distorts the relative reliability of specific statements. Certain positions are echoed over an over again (like the assumption that my genetic variant is benign), and other positions, that might be better supported by available data (like the fact the ER stress is a pathomechanism for Fabry disease) disappear from scientific discourse again and again. Which positions are repeated and which vanish is likely partially due to reseacher’s bias. Data which fits exisiting models is repeated, other data ignored. This introduces an element of subjectivity, which has no place in science. Apart from that, it’s the data that doesn’t fit existing models that can really broaden our understanding of the subject matter at hand, so that’s what we should be focusing on.
Other problems include nonsense citations. In this paper, the quote “However, it was clear that such a high dosage of galactose is not practicable in patients with Fabry disease” is in no way backed up by the source that is mentioned. Sometimes, the problem is simple nonsense. In this paper, the quote “AGALopathy is thus not FD, which is based on an enzyme deficiency resulting in intracellular Gb3 accumulation. Thus, AGALopathy cannot be successfully treated with ERT.” is also not backed up by the source mentioned, but even without consulting the original source, the statement is obvious nonsense. Another problem is that research papers are sometimes created to fulfill certain quota, without adding any new insights. The sheer amount of available publications can sometimes make it difficult to assess the robustness of the data underlying given statements, especially in view of the second problem mentioned above.
The peer review process doesn’t work as intended. To prevent each of the above points from happening, there’s a peer review process. Even if most academic researchers write well, all it needs is one or two reviewers who themselves adhere to subpar standards to publish a paper.
By and large, medical research reminds me of archaeology around 1900. Everyone was racing to find something grand spectacular, sometimes dragging of parts of temples or even entire temples from faraway places to Europe. Modern archaeology, on the other hand, involves sifting through endless piles of dust. Even small finds can rewrite history. Similarly, in medicine, there seem to be people who strive for a breakthrough through original research. Claiming that the most common GLA variant is benign sure is a spectacular headline, but in the end, it’s an overinterpretation of data that harms the very people medical research set out to save.