espite earlier denials, the US National Institutes of Health has now
admitted that NIAID funded several grants allowing gain-of-function
virology research. Some of these funds made their way to the Wuhan
Institute of Virology, from which it is widely believed that SARS-CoV-2
may have escaped. This has led to a flurry of commentary on the Internet,
some of it strongly worded, that reveals an long-standing undercurrent
of fear about whether science is out of control.
Here's an example. One blogger blames science for the suppression of discussion about the virus and says it is the same institutional arrogance that led to Chernobyl:
An "age of science" is thus an age of authoritarianism. Not necessarily of peace and plenty, but of self-assured lab coats pressing expensive panaceas, and government lab rats using violence to back them. . . .
We're in the beginning stages of scientific and social tyranny. Others have already passed through the late stages — and the scientific outcome they got was Chernobyl.
Another article says:
‘[G]ain of function’ means bio-war against humans like you. It's not about your health. It's about the health of our predators . . . which might just include the celebrated Dr. Fauci.
The control of scientific institutions by the government makes politicization, and hence fear and mistrust of science itself, almost inevitable.
Of course, most of the blame for the virus belongs to the institution from which it likely escaped and, some speculate, might have been engineered. The heavy-handed actions of local and federal governments in forcibly imposing unwanted vaccinations and lockdowns on the public contributed to the feelings of resentment. But the NIH also contributed to this fear by its lack of transparency on its funding of this research.
As much of an embarrassment as Dr Fauci's conflicting statements and his institute's support for this research may be, many people seem not to realize that it's not the job of the head honcho or even the vast army of NIH bureaucrats, but the grant peer review panels to decide whether a grant gets funded.
How grant sausage is made
Grant review panels are meetings held by the NIH to decide whether a proposal has sufficient scientific value to be worthy of funding. These days, to prevent massacres, they're held over the Internet using a specially written secure version of Zoom.
The panel consists of a large group of academic scientists, an SRO (Scientific Review Officer), an assistant whose job is to try to prevent the NIH's software from crashing, and a moderator. Half of the grants are immediately tossed out because of bad writing skill, unfashionable approach, or flawed design. The remainder are discussed by the group with a strict time limit and the grant is then given a score which determines whether it is worthy to be funded.
Before the meeting, the NIH gives reviewers a lecture about scoring bias. Apparently, nobody at NIH ever noticed that the panels were handing out sizzling scores to those whom they already admired and triaging those from people who had productivity gaps caused by becoming pregnant, getting COVID, or being screwed over by the bureaucrats at their university. The lecture seems to have had little effect, and much of the discussions consist of praise for applicants who are famous in their field.
Because the number of grants is far too many for the reviewers to read, few read any other than the ten or so they have been assigned. You can't just read the grant itself; you must also read the applicants' earlier papers and three or four other papers in the field to understand whether the experiments make any sense. This may reveal that they had already obtained evidence that disproved their own theory, but had ignored it and plowed forward anyway. If the reviewers don't do their homework—and many don't—they will be fooled into thinking the project could work. Thus, despite repeated warnings from the SRO, the reviewers have to describe the aims and experiments in the grant and explain why it is (or is not) important. Being academics, the main goal of the reviewers is to avoid insulting each other, the grant applicants, and above all the SRO and the moderator.
Funding criteria
The sole task of reviewers is to determine the proposal's scientific merit. Even if they believe the proposal is too dangerous, they're only allowed to say so if it affects the scientific value.
They also check a box indicating whether protections for animals and human subjects look adequate—based on what the applicants have written—and whether equal numbers of males and females and, for clinical trials, representative numbers of all racial groups are represented. This doesn't materially affect the score; NIH bureaucrats just tell applicants these are deficiencies that must be corrected.
What needs to change
First, we need two more boxes to check as to whether the research is inherently risky for public health and whether the proposal is ethical. These need to be included as part of the score.
Second, I've been on several of these panels and the discussion has always been inadequate. The best way to sort out thinking would be for the three reviewers of each grant to meet (or email each other) several times and consider each other's arguments for and against the grant. That doesn't happen: you're allowed to read what they wrote, but mostly they write pablum, so we're often blindsided by irrational and irrelevant arguments that spring up.
Third, moderators aren't required to remain neutral. They have great authority and use it to influence the panel. If the applicant investigator is a rock star or if the method is “cool,” everyone makes excuses: oh, they just forgot to mention controls; oh, even though they published garbage in their last five papers, they'll certainly do it right this time. In one grant, the applicant had never once detected what they were looking for because their method was fatally flawed. No one cared; it got a good score.
Fourth, panels get fixated on some small detail and trash every grant that misses it. They may decide that omitting one particular bit of information is a fatal flaw and nail a dozen grants for it.
It's exhausting and discouraging to spend two weeks of preparation and then watch bad grants getting funded and good grants getting trashed. Here is what I learned from it:
- Most people, even many scientists, are unable or unwilling to use incisive reasoning and use social status, i.e. power, as a substitute for making decisions.
- To get funded in science, you need to do something fashionable regardless of its scientific value. This is one reason we get so many dead-end fads and groupthink.
- Universities that don't provide enough resources for their faculty to do cutting-edge research will not bring in much research funding. So the bigger get bigger and the smaller get smaller.
- Separate pre-panel meetings are needed so reviewers have time to think about each other's critiques and avoid being blindsided by nonsensical arguments during the meeting.
- Although grant review panels are responsible for scoring a grant, they need to be allowed to consider ethical and public health risks. The NIH should require a statement of possible dangers in the project. This needs to factor into the final score.
- The US government cannot cannot program a computer to save its life.
In the past, the NIH has always been responsive to public concerns. No doubt even now these failings, and possible reforms, are being discussed in some windowless office in Rockville or Bethesda. But if people demand that the director of an institute be given veto power over research grants, they're giving more power to someone who will, sooner or later, abuse it.
oct 30 2021, 5:45 am. updated oct 31 2021, 4:20 am
