Game of the little red: clinical trials or bets?

Do you bet that the cure for type 1 diabetes will be available before 2033 (or that it won't happen, to the sadness of the more than 9 million people with the disease around the world)? If you wish—and, in all honesty, if you are not in Brazil—you can make this bet. Despite the pompous name, the "prediction market" basically allows anyone to invest their money in the possibility of a future event happening, from a politician's victory in elections to the number of plays a musician will accumulate in a year. And, recently, digital platforms in this segment, such as Polymarket and Kalshi, have announced the possibility of betting on clinical trials and drug approvals. Besides the cure for type 1 diabetes, in October, Kalshi's "Medicine" tab allowed betting on when Retatrutide, a weight loss drug, would be approved by the FDA, the agency that regulates drugs and food in the US. Or if a treatment for hereditary angioedema, a rare disease, would arrive before or after July 2027. Or at what time some company would submit the clinical studies of an innovative therapy against sickle cell disease. Before getting into the scientific repercussions, I confess that I feel a certain discomfort accessing a website that looks like a stock market monitor and seeing that many contracts traded there involve, ultimately, the lives of people weakened by various diseases. Whoever bets on the failure of a clinical trial is essentially rooting for the lack of a new therapeutic option—in short, for the misfortune of others—in the name of a bloated bank account. That's where I started the conversation with Andrei Sposito, coordinator of the Clinical Research Center at the Faculty of Medical Sciences at the State University of Campinas (Unicamp). "The conventional financial market already has other mechanisms to profit from the failure of clinical research. This is not new." What worries Sposito most is, in fact, the wave of betting on everything, from sports to medicine, a phenomenon capable of leading families to suffer considerable losses. A report from Kalshi defends that the prediction market helps generate objective information about how promising certain research would be, which, among other advantages, helps patients in search of solutions to their problems. The basis of the argument? On these platforms, the odd—pardon, the contract price, or "action"—is defined by the transactions between users, which would reflect the collective expectation about a treatment being evaluated. If many people (including experts) are willing to put money on the success of a therapy, the contract price tends to rise, which would act as an indicator of potential. However, this view encounters resistance. In an opinion article in the scientific journal JAMA Health Forum, the economist Genevieve Kanter, from the University of Southern California in Los Angeles, argued that this business model could compromise recruitment for clinical research considered less promising by the market. Who would want to enter an experiment in which almost no one is betting? But, in the case of clinical research, the prediction market would have less potential for... prediction. One of the secrets to anticipating the future is to arm oneself with the largest possible amount of information available in the present, which is particularly complicated in a sector surrounded by confidentiality agreements and double-blind methodologies. Few individuals, like researchers and sponsors, have access to more data and, even so, with restrictions. So the person who gives up on entering a clinical trial because a prediction market platform does not show prior optimistic data is making a decision based on enormous uncertainty—and, potentially, closing the doors to an improvement in health itself. The article in JAMA Health Forum also expressed concern about possible scientific manipulations, especially in the absence of clear and specific rules. "Imagine [...] a clinical trial researcher modifying exclusion criteria to influence the study results to which a bet is linked," wrote Kanter. Or worse, a healthcare professional with debt in the market who interferes with the doses administered to participants to win the bet and go into the red. For Sposito, however, regulations focused on clinical research and drug approval make this type of fraud more difficult, as there are different points of control: "The researcher, the company, the insurer, the regulatory agencies... everyone would have to be deceived or participate in the scheme." In short, the greater risk would be in bets focused on early studies, where there is less oversight. The prediction market does not like to be compared to betting houses. And there really are differences. For example: instead of "betting against the bet," you actually buy and sell contracts with other users, as in a stock exchange—the platform profits from the mediation. But, for the average user, this distinction is at least blurred on the websites. This interpretation is not mine alone: in April of this year, the federal government prohibited platforms like Kalshi and Polymarket from operating in Brazil because they were not classified in the previously authorized sectors of the betting market (sports results and online casinos, like the Tigrinho game). In other words, the government saw these platforms as repackaged bets. "We started to see the growth of a sector that appeared to be something else, but what we actually observe is that it has the look, snout, and nose of a bet," said Régis Dudena, Secretary of Economic Reforms of the Ministry of Finance, to Reuters at the time. With the provisional measure that completely banned bets, it became even more complicated for the prediction market. Would you bet that this veto will stand next year? Playing cards In 2022, the science integrity specialist Leslie McIntosh revealed, along with other colleagues, the existence of a fake network of scientists, the Pharmakon Neuroscience Network (PNN), which sold authorship spaces in thousands of adulterated studies. Four years later, she and Federica Silvi recovered a database with almost 2 thousand published articles listing at least one co-author of the PNN—a clear indication of fraud—to check how many were included in patent references, manuals, and so on. According to the data they published in preprint format, 57 were cited in public policy guides, 12 in clinical guidelines, and an incredible 480 in patents. Setting aside the alarmism, this does not mean that each of these documents is completely tainted, or even that they recommend practices without scientific evidence. Many other articles—these, apparently reliable—were also listed in these pages. Furthermore, if you want to profit from fake article factories (the paper mills), a good strategy is not to draw attention and produce texts that, although false, follow the trend and generate conclusions similar to those of most published texts. Still, the penetration of these articles in official documents reinforces how debates about scientific integrity lead to real impacts. And you are thinking of skipping reading this column... Republish