Monday, 25 June 2018

On fallibility, humility, and responsibility in science

It has been a rough seventy two hours in the lab. Things have gone very very wrong. We have lost a lot of animals to an unknown illness that is swift and lethal. All day we have felt powerless, and the day ended in a succession of necropsies. More dispiriting, this is the second attempt we have made at this particular experiment. It started out much more promising than the last, but very suddenly, we were once again confronted with how little power we have over biology.
We work on an animal model of infant physiology, which means we work on infant animals. Few people do. It is difficult and demanding work. Infants have care needs beyond what a university animal facilities can provide (though ours have accommodated ours to an astounding degree). So we must step in to feed animals and care for them. Our bigger problem is this: there is little data on infant mammalian physiology, particularly for large farm animals like we work on. I remember a few years ago when we were having problems with our anesthesia procedure. Every vet and textbook in the country swore by tylezol for inducing our animals. It may have worked in adults, but it was useless in infants.
For the past two years we have increased the complexity of our experiments. We now raise the animals from birth, when they are most vulnerable. With full term animals, we had good luck. We got all the data we needed, and felt confident we knew what we were doing.
The grant that was funded however, called for doing the project in premature animals.
There is a very good reason for this. Human premature babies have a host of complications, in particular with regard to feeding and respiration function, which is what we study. Yet it is almost impossible to do the type of research we do on human babies, because it requires X rays. This is doubly the case for premature babies. Crucially, only babies diagnosed with problems can be enrolled in research studies for what we do. As a result, we have no idea what non pathological physiology in this system looks like for newborns, either full term or premature. The work we do addresses a critical gap in pediatric physiology.
Nor did we rush into this blind. We have years experience caring for young animals. Before this project, my PI found a team of researchers who do research in our model on animals far more premature that we were planning. They have come to visit us and us them. We discuss our problems with them regularly. And, thanks to them, our care of full term animals was a resounding success. We were cautiously confident.
The first preterm litter was a disaster. Only a single individual survived past 48 hours. It was pretty much our worst case scenario. We sat back, took stock. We bought better incubators. We developed new protocols for cleanliness and care. We settled on a slightly less premature age for our animals, scaling back our ambition in the face of how much more delicate these preterm animals were.
Again, we thought we were ready.
The delivery of the litter went amazingly. Sixteen healthy newborns in two used medical incubators. For 48 hours, we felt good.
Then they started to fall ill. By Saturday evening, two were refusing food (a very bad sign). By sunday morning, they had started dying. Today was almost like being in a hospital during an outbreak. It was utterly heartbreaking, and sobering.
We pay for our experiments in effort. I have been working well over fifty hours since Wednesday, like everyone in the lab. And we care. Our animals are helpless. And so we try to marshal all our resources to help them when they are ill.
And today I was reminded that we too are helpless.
We are working on a animal model of an incredibly vulnerable patient population. That vulnerability is baked into the very experiments we proposed to undertake.
I understand a little what working in a neonatal ICU must be like now. I understand what realising how little you know, and how little you can do, must feel like now.
As scientists, we plan experiments. We anticipate contingencies. We think about physiology, experimental conditions, confounding factors. Yet when we create animal models of disease, we create, in effect, patients. And we are suddenly faced with the maddening, terrifying, awe full power of illness. The limits of medical science, laid bare in our own research labs.
Hopefully, the remaining animals will survive. We will get data. We will feel our bargain with our animal subjects has been at least partially fulfilled.
But today, I think, I confronted the limits of what I do. Today, I think, we all felt, perhaps, we were getting close to the point where the price of the bargain is too high.