Wednesday, June 20, 2018

Data Cities: Behind Every Record is a Human Life



I am visiting New York City with Lily, my precocious 5-year old; asking a thousand questions one moment; the next sharing a dozen unsolicited facts about sea turtles.
One of my favorite places in the world is Grand Central Terminal. The majesty of the place stops me in my tracks at the top of the stairs. I look out over the golden concourse, the light slanting down from the magnificent thousand-paned windows bathing scurrying travelers in swimming pools of light. The arresting space feels holy like a cathedral, capped with a roof of stars in a cerulean blue sky.
Lily is beside me, singing a made-up song, giggling and oblivious. She can’t really see it yet. One day when she is older, we will return. There is much I want her to see.
Years ago, when I first saw a photo of this very concourse (one of my favorite books is Mark Helprin’s “A Winter’s Tale” -- but please don't ever see the movie), I became briefly obsessed and read everything I could find about Grand Central. On an average day, 750,000 travelers flow through the Terminal using a series of generous and gently sloping ramps, linking together each part of the structure while at the same time skillfully separating incoming, outgoing and suburban traffic. 
The enormous construction project was itself ingenious, replacing an overwhelmed, dangerously-congested above-ground steam train hub with newly-introduced electric trains, which were cleaner, faster and easier to repair. The Terminal was built in sections underground, resulting in the ability to sell above-ground plots to what would literally shape the heart of New York City, while paying for the cost of electrifying the railroad's operations.
Grand Central, and in fact, the entire city, strikes me as an apt metaphor for a massively parallel computer, with the Terminal as its enormous data distribution hub, carefully engineered to scale to high volumes with minimal bottlenecks. This metaphor is not a unique thought – many have surely thought likewise. 
Enough musing now, though. Lily is hungry. We hold hands and walk back to the hotel.
It is evening now. It is Lily's bedtime soon. I pull the heavy blackout fabric of the room curtains aside. We talk about the windows lit up below and around us. “Do you see all the windows? There’s a family inside each of those.” We play a little game, making up imaginary families, living lives behind each window.
Time for Lily to go to bed. I read her a bedtime book. “Tallulah listens to Maisy’s heartbeat. That Tickles!”. She falls asleep quickly tonight. I return to the window.
There are so many windows in the city. It’s hard to grasp, in a city like New York, how all of these buildings contain people, each with a story; each with a life no more or less important than your own.
Data is like that, too; each clinical dataset a city block. Each trend in a dataset is the aggregation of thousands of people. Each medical record is a shuttered window, drawn shade, closed blind.
Inside one window is a woman, barely 35, with three children under five. She felt a lump in the shower last April, no larger than a cube of sugar. Invasive ductal carcinoma. Stage 2A. 
Behind the drawn shade, a frail man, shoulders rounded, fighting the pain as he ever-so-carefully begins to sit down. Osteoporosis weakens his bones until they cannot support his weight. Everyday things, such as coughing, stepping out of a car, or bending over can fracture a bone. It’s called a pathologic fracture.
Behind the closed blind, a mother is massaging her son’s legs. He has tight muscles, so they constantly hurt. He couldn’t walk until he was three years old due to Cerebral Palsy. Because he walks with a scissor gait, a lot of his schoolmates think it is funny. It has been hard for him to make friends. His father returns with a rental car for the trip. He found an iPhone app to search clinical trials for their son. Tomorrow they will drive to Bethesda, Maryland, the nearest site, to see if their son can participate.
Looking in one window, at one family, one person, doesn’t teach us how to make a city thrive. But when we pull back, we see the buildings, the height, depth, the structure, the patterns. 
Behind every record of data there is a full human life. When you stack the records high, tease significant variables from the insignificant, you can see how quickly a disease tends to spread -- how to slow down, stop or cure a disease. How long it takes to die. This is called medical research.
When you look at the data; 100,000 clinical trial visits since April 26, 2015, having adverse events of special interest, a clinical trial, lab test results for subject visits; it’s a rounded scatterplot distribution, ascending from the insignificance. It’s a blue bar or a red slice of pie chart. It is hard to fathom, easy to miss, that each pixel is a human life. Possibly a final chapter to a human’s story.
Just as zooming out allows us to see the larger picture, zooming back in connects us to the lives thriving, suffering and dying. By looking in the windows, we connect with the individual human beings in the architecture; from Lily peacefully sleeping beside me, to the frail man with rounded shoulders, behind the drawn shade, who is finally, painfully, in his seat.

No comments:

Post a Comment