I didn't say that fraud does not exist, or that there isn't pressure to produce publishable results that might affect accuracy or ethics (on occasion.) I said that this is a much smaller problem than the article implies. Only the retractions were analyzed; the retractions are a vanishingly small percentage of the total. If you want to argue that the retractions are the tip of an iceberg of falsified scientific data, let me know.
In context -- PubMed has more than 22 million documents and accepts 500,000 a year, according to Wikipedia.
So, to do the math: Number of fraudulent articles, total, = vanishingly small percentage of the total articles.
I wonder if that has ever happened here before?
You are in a maze of little twisting passages, all alike.