Reacción a "The FDA approves a new type of drug for schizophrenia"

A new drug with a novel mechanism of action is coming out, which is fine in principle, but in these cases we must avoid conveying an image of excessive efficacy or excessive hope in the face of novelty. Just because something is new does not necessarily mean it is better. Focus on the studies and the available data.
The available data focus on two studies. In the first of these, in the short term (five weeks) and with 252 people, they found a statistically significant improvement but not a very important one at the clinical level. 

They use the PANSS scale, a scale that measures different types of symptoms in schizophrenia. The drug improves a mean value of 90 by 10 points compared to placebo in a statistically significant way, but it remains to be seen whether this is clinically relevant and, above all, whether it is relevant in terms of the cost it might entail. 

The other study is a continuation of the previous one, extending up to one year, and finds an improvement that the authors describe as significant.  

The study is based on the same 252 people, but there is one striking fact, and that is that at the time of evaluation, 110 of the 252 people remain, and at the date of publication of the data only 29 had completed the 52 weeks of treatment. The data are therefore lacking as to why this number of dropouts is so high, because it may be due to poor tolerance to the drug.
Among those who finish the figures do seem more significant, as there is an improvement of more than 30 points on the PANSS scale, but the problem is that this 52-week study has no placebo arm, so we do not know how much of this improvement is due to the new drug and how much may be due to other circumstances, because there is no placebo for comparison. This efficacy data, therefore, says little.
In short, I believe that the short-term studies do not provide data of great efficacy, and in the long-term study the efficacy data are not measurable because there is no placebo arm for comparison.
I also think there are too few people to be able to properly assess the safety profile, which will be assessed in the end in clinical practice as it is used. And that is risky.
From what they say, there are no particularly striking side effects, but we must not lose sight of the fact that it has been tested in only a few people. A potentially serious side effect occurring in an incidence of one in 1000 people would not have been detected in these studies. This does not mean that the drug cannot be used, but it must be used with particular caution, because it is not a known drug and there may be rare adverse effects. 

The question of cost is also important. I don't know what the price will be, but for this improvement in the scales we will have to ensure that it is not disproportionate, as is the case with many new drugs, because the money we spend on this drug will not be able to be spent on other aspects of the treatment. And there is clearly a lack of an active comparator study, i.e. a study that compares this new treatment with an established treatment for schizophrenia. With these studies we cannot say whether it is better, the same or worse than what we already have, which is what we clinicians are interested in. 

We see, therefore, that the process is a bit like the usual way of doing things so that at a marketing level it is also marketable, but at a scientific or clinical level I think we have to have certain reservations. We will have to see.