Riesgo de sesgo de publicación en intervenciones terapéuticas para la COVID-19

This article describes publication bias, its most frequent causes, its characteristics, the regulatory tools to avoid it, and some statistical techniques to analyze it. These techniques are explained and applied to three therapeutic interventions related to the 2019 coronavirus disease (COVID-19): corticosteroids, ivermectin, and tocilizumab. Risk of publication bias was detected for ivermectin and tocilizumab. Systematic reviews and meta-analyses are secondary research designs that provide a reference to guide decision-making. They are prone to different types of bias, i.e., a systematic deviation in the results. Even if carried out with methodological rigor, their validity can be threatened by publication bias. This is defined as the act of concealing or delaying publication, withholding data arising from research studies, or both. Up to half of controlled trials remain unpublished. During the H1N1 virus pandemic, publication bias from industry-funded studies led to the recommendation and large-scale procurement of oseltamivir, a drug that later proved to have no relevant beneficial effects. Two-thirds of clinical trial funding for COVID-19 comes from the pharmaceutical industry. In the context of the COVID-19 pandemic, studies are published at an accelerated pace, making it very important to understand and identify publication bias. To reduce publication bias it is necessary to regulate the registration and publication of clinical trials, but this requires coordination among countries and international bodies. It is important to suspect and attempt to identify publication bias for decision making.