I was just listening to Tyler Cowen’s conversation with Atul Gawande and noticed this part:
COWEN: Why do surgeons sometimes leave sponges behind in the bodies of patients who are being operated on?
GAWANDE: You zeroed in on one of my very first projects in creating intervention.
COWEN: Great paper.
GAWANDE: We had done a case control study of this problem of surgeons leaving sponges inside people, and got it published in the New England Journal [of Medicine], partly because of our whole method of going about solving this problem, which was, we studied 60 people who had sponges left inside them, compared to 240 people at the same institution at the same time with the same operation who didn’t have sponges left inside them.
I don’t want to focus on the interventions Gawande developed (he is the author of The Checklist Manifesto: How to Get Things Right, the title of which gives you a pretty big clue about one of those interventions); instead I want to focus on the fact that surgeons—who go to four years of undergrad, four years of med school, at least five years (in most cases) of residency—and who are highly motivated to not screw up procedures, because if they do people die—still manage to make seemingly elementary mistakes. Like forgetting a sponge in a patient.
Those mistakes happen, even to brilliant people, because as the cognitive load on a person increases, so does the tendency for error—even simple error. The same kinds of things happen, of course, in grant writing, although our “patients” are unlikely to die as a result. Still, the grant writing process is cognitively complex, which in part explains why so few people can become good grant writers. Interactions among the staff operating the program, the person writing the proposal, the funding agency, and the RFP are complex and can lead to errors. Even the nature of RFPs themselves lend themselves to error.
For example, I was just working on an HIV testing proposal for a client in a big Midwestern city. The narrative section of the proposal is limited to ten pages, with 1.5 line spacing, or about 7 single-spaced pages. The RFP, however, is 111 single-space pages. That’s right, the RFP is about 15 times longer than the allowed response. The possibility for error in such situations is enormous—it is cognitively difficult, and maybe impossible, to hold 111 pages of sometimes contradictory instructions, background on the applicant, and project design in one’s “RAM,” while also keeping to the max page length.
Part of our job as grant writers is to minimize error and understand where and why it might happen, so that we can prevent it to the maximum extent possible. Surgeons, who face life and death issues, don’t always manage to get the sponges out of people, even when they are very highly incentivized to do so. As such, it should not be surprising that the rest of us, who are doing cognitively complex tasks, also face major challenges in getting things right.
Everything is hard. Sometimes there is no way around that. If you’re old enough, you likely remember computers from ten or fifteen years ago that were slow and unreliable by today’s standards. Today, computers are probably more than a thousand times faster (transistor density tends to double every eighteen to twenty-four months) than they were 15 years ago. Yet Firefox is still kind of slow at times, Word still crashes, and various other programs have their foibles. One would expect computers to have transformed medicine, especially now that they’re so fast, yet every doctor hates their Electronic Medical Record (EMR) system. Isaac’s primary care physician uses eClinicalWorks and routinely complains about it being slower and less efficient than hand charting. He says finding the information he needs is harder with eClinicalWorks than it was when he charted by hand. In other words, he likes a millennia-old technology better than the latest software release.
We have faster computers, but EMRs still suck. We have faster computers, but Word still crashes. We have faster computers, but we also demand more of them. As hardware capabilities expand, we demand more of software. The software gets more complex and eats the gains from hardware speed. If I only ran programs from 10 or 15 years ago and made demands like those from that time, I could have a blazing-fast computer, but without the capabilities I like (like the ultra-high resolution 5K display on my iMac). Making software is hard, so it has problems and trade-offs.
The analogy to grant writing seems too obvious to belabor. I’ve also got to get back to the 10 page opus I’m extruding from the 111 page RFP; it’s too early for a cocktail.
Oh, and that story about the sponges? Gawande did come up with a technological fix for lost sponges: bar code each sponge and make sure that each sponge is “checked in” and “checked out.” That simple intervention means that virtually no sponges are lost in patients today. But not all problems lend themselves to technological fixes. Writing doesn’t.