"Our objections, as I have pointed out, were really on theoretical grounds, largely based on the fact that there was no evidence that it was safe. Until we had such evidence we had to question the safety."
When, in 1960, US Food and Drug Administration drug reviewer Frances Kelsey (1914-2015) had an assignment to review thalidomide, a painkiller for pregnant women, she stubbornly refused to certify it, despite the fact that women taking it reported no significant ill effects. Her approach was vindicated when this drug, in other countries, was revealed to cause the birth of severely deformed children. The opnening quote is the whole story in two sentences. What Kelsey had was a procedural objection that could not be satisfied: the claims in the application were not backed by adequate evidence, and an unproven thing cannot be certified as safe merely because it has not yet been proven dangerous. She held the line not because she saw what was coming but because the paperwork was incomplete and the rules said incomplete applications do not pass. That is an LSI's "inspector" relationship to a standard: the structural criteria are not met, and no amount of pressure changes the criterion.
Her autobiography is a four-decade record of a person who trusts only what has been measured. The frogs: "For a year I sat surrounded by frogs in little cages set in water. I would lift them up, dry them, weigh them, inject them, put them back in the water, and then weigh them at 15 to 20 minute intervals for four hours." The rats in the sulfanilamide investigation, where the lethal culprit revealed itself by direct observation: "one could see the urine gradually turn red, and then decrease in volume, and then finally stop." The fish entrails on the deck, "down like soothsayers poring over the entrails looking for these tiny little nubbins of tissue." She remembers her life as a sequence of careful procedures performed on physical material until a clean result emerged. This is a deeply sensory-logical mind — it works by handling, weighing, measuring, and checking, and it reaches conclusions only when the measurement forces them, without any "eureka" moments of insight.
At the American Medical Association, she and the other reviewers "kept a sort of informal list," jotting down the names of authors whose papers were poor: "Oh, it's Dr. So-and-so again." At the FDA, decades later, she enjoyed the same classificatory and categorizing impulse: "A bunch of us would jot down a name when we saw it and compare notes. It was the same thing I had done when I worked at the AMA." And then she built it into the institution: her data-retrieval system matching physicians to drugs, the formal list of disqualified investigators ("From the first one in 1964 up to December of 1990, there are seventy-one listed", boasts she). This is a LSI tracking of who is reliable and who is not, with the conviction that conduct should be monitored and all deviation noted. She is not vindictive about it, but she cannot not keep them. Quality control is, for her, a moral activity expressed as bookkeeping.
After thalidomide she chose, and effectively built, the function of making sure investigators were actually doing what they signed agreements to do — going out to verify records, catching "fudging in all areas, clinical and animal," disqualifying the crooks, checking that consent forms were where they were supposed to be. She designed and ran a surveillance apparatus. An LSI who has identified a domain of rule-violation will construct the system that detects and punishes it, patiently, for years. "Every time a regulation is made insisting people do something some way, there has to be an inspectional program, to make sure they are doing it that way." She even used the "inspector" root here.
Throughout, her emotional register is very level. The two volunteers who died in the South Dakota morphine study, the addicted dean, the deformed infants — all are reported in the same precise, slightly detached tone as the LD50 procedure or the armadillo supply chain. "I was — I admit it — very surprised" is, at the climactic moment of her professional life, the upper limit of her expressed emotion. Her Fe may not have been unvalued, but it was certainly not strong.
She definitely had some "peripheral" accent — her scientific instinct points to some LII traits not being alien to her, and a few notes in her auto-biography lean more comfort-and-autonomy oriented SLI-style. And there's a dry, understated humor running throughout (the Texas vet who tried his remedy "on his secretary without ill effects, and he planned next to try it on cattle") that is lighter than the LSI norm, but not outside the boundaries, either, and a certain humility in her self-presentation even fits Talanov-style Di (a supposedly strongvalued LSI function).
Overall, she was a typical "benign" LSI, and a useful corrective to the type's reputation. The list skewing toward dictators and security-service men (Park Chung Hee, Isser Harel) is a selection effect: the historical record over-samples the LSI pointed at people. But Kelsey's inspector role is clear: she literally built and ran an inspectorate, went out to verify that investigators did what they'd signed agreements to do, kept the disqualification list, caught the fudging. The deformed-infants catastrophe she prevented is what an LSI "strict inspector" instinct looks like when it's guarding the public against industry. For a LSI's societal effects, it quite matters what counts as a violation and who wrote the rulebook.
Comments
Post a Comment