In our meeting, we discussed the kinds of questions that researchers have: how to collect information, how to organize or group the data, and what should be included or excluded. For instance, we had blood-lead levels for children of different ages. Should we lump them all together or limit the age range to children under five, since they are the most developmentally vulnerable?
As we looked at the numbers, we realized we had no idea where these kids lived. They had come to our Flint clinic, but did they even get Flint water? Jenny and I searched on ZipCode.org for Flint and decided to use 48501 through 48507 on our data set. I wrote to Elin to ask if the water supply went just to Flint or beyond the boundaries of the city. Then I wrote back to IT and asked if they could run the numbers again, including zip codes in the report.
I heard back just a few minutes later. That’s the miracle of electronic medical records. So much time goes into crunching—or into fine-tuning and designing a study. But once you ask the right questions, spitting out the results happens in an instant.
We were far from finished with the fine-tuning process. We still had other things to sort out, like how to filter out repeats—patients who were in our system twice because their lead levels were measured twice in our time period. And when we saw a superhigh lead level reported, should we look individually at each of those patient charts to see what the determined cause was? If we saw that the child had eaten paint chips, should we exclude them from the study?
We knew we’d be doing the research twice, for data sets on lead levels before the water switch and after. But how were we going to break up the time periods? Should we look at a date immediately after the water switch, or should there be a lag time for the leaching of lead? How long did it take for the leaching to occur?
The residents began pulling up research on their laptops. As I spoke, their focus on me was intense, and I could see their minds were fully engaged. Most of the research that medical students and residents do is important but academic: one special report of a rare case of encephalitis, or a study of patterns of treatments that happened years ago, in other places, other hospitals. But here we were, talking about lead poisoning, lead exposure, and blood-lead levels, in numbers from the previous year and over the summer, numbers from just a couple of days ago. It was real, something that was happening all around us, the blood of our own patients, and water that flowed in the pipes of our own city, where we sat. The residents were engaged in a way I’d rarely seen before, vibrating with a weird new energy, tense but invigorated by the feeling that we were finally doing something. And our results weren’t going to be stuffed away in a digital archive and forgotten. Our results could change our world.
The residents eagerly pulled up related literature and the electronic charts of some of the kids with the highest lead levels: “Oh, this one says there was peeling paint in their house,” or “This one doesn’t mention a potential source.” Another said, “This kid was shot.” Bullets left lead in the bloodstream too. We decided to look only at kids five and younger; to look at only one level per child—the highest level; and to include all kids no matter what was in their chart as a potential source. As for time period, we agreed that we’d look at an equal length of time before and after the water switch—January 1, 2013, to April 24, 2014, versus April 25, 2014, to September 9, 2015.
Jenny took careful notes on her laptop while we talked and said she’d have something soon, even by tomorrow. I knew from our other projects together that, like me, Jenny wasn’t a nine-to-fiver. Once she got going on something, she took it home and kept working. Putting a study together was like working on a jigsaw puzzle—it was kind of addictive, hard to stop, and offered a dopamine hit of pleasure when you reached the end.
When the meeting finished, the residents filtered out of the room to see patients in the clinic. They were silent, deep in thought. I had tasked them with overnight assignments—looking up more literature on lead exposure, finding similar studies, taking a deeper dive in a number of areas we needed to know more about. They knew this work loomed large for me, for the clinic, and for all of us, but especially for our patients and our community. We would be fashioning an argument out of blood tests, numbers, and zip codes.
When the last resident left, I noticed Jenny hanging back, waiting to talk to me. She had a few questions, she said, and I could tell they were personal. She had a look on her face that wasn’t about numbers. It’s the same look I see on every parent’s face who comes to the clinic worried about her kids.
It was about Drew, her baby son. He was just a year old and still nursing. Jenny wanted to know, if she was drinking Flint tap water at the hospital, could she be passing along lead to him in her breast milk? She was using a breast pump and sending bottles of it along to his daycare in Grand Blanc, a suburb south of Flint.
I knew this was the kind of question all the parents and caregivers in Flint would have once they knew what we suspected—including all the moms who worked with us at the clinic.
“Aside from coffee and tea, or whatever I get at the cafeteria,” Jenny said, “I haven’t been drinking the water at work.” As a Flint native, she had been alarmed like everybody else when the city’s water source was switched to the river. And when she saw the early news reports of water-quality issues, she had the luxury of taking extra precautions, even bringing bottles of tap water from her home, which was outside Flint. I assured her that all those protective steps were good—and she should keep doing them.
“But what about when I was pregnant with Drew?” she asked. I could see her calculating the date of the water switch in relation to her pregnancy. Drew was born in August 2014, five months after the Flint River started flowing out of the tap. A look of worry and sadness passed across her face. “That means seventy-five percent of the time I was pregnant, we were on that water. And I was coming to work every day and drinking the water.”
“I know,” I said, nodding, and then reached out to comfort her.
Jenny studied my face as much as I was studying hers.
“Do you think Drew’s okay?” she asked. “Is there anything else I should be doing? He’s been sick a lot lately.” Drew was due for a bunch of vaccines that week, she told me, but was still recovering from his first ear infection. Jenny, being an obsessive researcher, had become paranoid about antibiotics, vaccines, the germy environment of daycare, and all the other worries that can plague a parent, especially a parent let loose on the Internet. So much of what we do as pediatricians is listening, reaffirming, and reassuring, especially first-time moms and dads.
“He’ll be fine. The ear infection is a common virus. Don’t worry. Daycare is good for him. All the germs will make him stronger,” I told her. “He does not need antibiotics. And vaccines are important.”
After a parade of reassurances and a tight hug, we said goodbye, and I sat alone in my office chair for a few minutes, trying to clear my head. I felt bad for Jenny and knew how terrible it was to have mom-worries. I remembered those days. But imagine how the moms of Flint were going to feel when they did the math for themselves. And think of the ones who were pregnant, like Jenny, and were not just having an occasional cup of coffee at work but were drinking the water all day long. How worried and angry they’d be. Who would answer to them?
* * *
—
EARLY THE NEXT MORNING, on September 10, Jenny texted me. True to her hardworking (and somewhat obsessive) style, she had already run the numbers—and had completed a preliminary analysis of blood-lead levels from Hurley Children’s Clinic.
JENNY: I got something.
ME: Email it.
JENNY: Okay. I’m coming to your office.
I was sitting at my office desk when her PDF came in with the actual statistical analysis. It was just one page. Easy to read and comprehend.
Comparing children under five years of age with elevated lead levels (greater than or equal to 5 μg/dl) f
or about the same duration of time pre–water switch (January 1, 2013, to April 24, 2014) to post–water switch (April 25, 2014, to September 9, 2015), the percentage increased from 1.5 percent to 8.5 percent.
The p value was 0.007.
In statistics, the “p value” shows statistical significance. And the lower the p value, the stronger the significance.
Anything less than 0.05 is significant. There it was.
Lead. It was real.
The impact was there in the numbers.
It was exactly what Elin had feared, and what Del Toral had suspected.
My clinic results showed more kids with higher lead levels since the water switch in 2014. The increase in the percentage was there, and it was big and statistically significant. But the sample was so small—4 out of 270 kids in the pre-period who tested high, and 6 out of 71 kids in the post-period.
Jenny appeared in my office, out of breath. She’d sprinted down three flights of stairs in a hurry. Even though her cheeks were flushed, her skin looked unusually pale, and there were dark circles under her eyes. She was tired and probably hadn’t slept much, having run numbers all night.
We didn’t even say hi or greet each other. We jumped in immediately, talking about holes in the analysis—or potential holes. One way to produce a perfect, unassailable study is to be a scientific devil’s advocate. It’s the way clinical reasoning works. Also, part of me was desperately hoping to find an alternative explanation for the results we were getting. I didn’t want them to be true.
For instance, the percentage of kids with elevated lead levels was higher and was significant. But there could be a practical explanation. What if the six kids with high lead levels were siblings in a family that lived in one home that was infested with lead paint?
And we should have more test results, or data points. Where were they? Why were so few kids tested after the water switch compared to before the water switch? Before the switch, 270 screenings had been done. After the switch, only 71.
What was going on?
What I needed was a study that left no room for doubt. I needed a larger sample—and that was going to require getting records from the county or state. After I sent more persistent emails, the county health department finally said they could retrieve lead levels only one at a time—by downloading a PDF for each child. Was this how they did basic public health surveillance? (It was hard to imagine how these public servants slept at night.) Of course, Kildee’s office might come through, but it could take time. There could be delays.
“We need more data, Jenny. We need more—”
“Wait a second. Hurley pretty much processes the labs for the entire county, not just for our clinic. Couldn’t you just pull the numbers pretty quickly with our EMR?”
“I will email IT right now.”
“Hold on,” Jenny said. “I think we’ll need to get an IRB approval for that.”
She was right, we did. IRB stands for institutional review board. Since I was the director of the residency program at Hurley, I had the authority to review data from my own patients and patients at the Hurley clinic, but I didn’t have the power to look at data from patients outside our practice. To get access to a much bigger sample, Jenny and I would have to have a review board at Hurley give us permission.
All research done on humans is protected by ethical guidelines that were established in 1979 in reaction to terrible historic incidents in which people were experimented on in cruel and exploitative ways, especially poor, disadvantaged, or mentally ill people. The infamous Tuskegee experiment that I teach my residents about—when government researchers withheld treatment from black subjects with syphilis—is a prime example of what an IRB is meant to keep from ever happening again. So while the process of getting an IRB approval could be bureaucratic and time-consuming, the principle behind it—that people are more important than any institution or system or research project—was so good, solid, and right, I couldn’t complain.
At a big university, it could take months for researchers to get approval for a study; those big institutional review boards sometimes met only four times a year. One of the blessings of working at a small teaching hospital was how quickly we could get things done. At least we hoped.
“I’ll give a heads-up to the IRB folks,” Jenny said. She worked just a couple of doors from their office.
And if the stars were aligned and we busted our butts—and I really pushed it—we might get an IRB approval in a week or two.
* * *
—
MEANWHILE BRAD WURFEL, the rabid pit bull of MDEQ, was going after Marc Edwards. After the findings of the Virginia Tech study were posted, the local media picked up the story, which prompted MDEQ’s vicious pushback. It was all over the morning news. Wurfel criticized Edwards for “fanning political flames irresponsibly” and described his Virginia Tech team as a group that “specializes in looking for high-lead problems.”
Trash-talking the leading water expert in the country is a stupid strategy, if you think about it. Wurfel told The Flint Journal, “They pull that rabbit out of that hat everywhere they go. Nobody should be surprised when the rabbit comes out of the hat, even if they can’t figure out how it is done.”
How long had it taken Wurfel to polish these cute quips? If only MDEQ had spent as much time listening to the people’s concerns and carefully looking at Edwards’s findings as they had minimizing and dismissing them.
But now that I was lying awake all night, eyes open, mind buzzing, I used the time to read all the books I’d ordered on the public health history of lead. I was beginning to see that over the last century, lead seemed to bring out the very best and the very worst in people. And I knew already which side of history Wurfel and his MDEQ bosses would be on.
IN FLINT, EVERYBODY KNOWS THE NAME Kettering.
His inspiring quotes can be found on plaques and signs around the city: “Believe and act as if it were impossible to fail,” or, my personal favorite, “My interest is in the future, because I am going to spend the rest of my life there.”
Our world-renowned engineering school in Flint, once called the General Motors Institute, is now named after Kettering—as are a number of honors and prizes given out in the city. There is a city near Dayton, Ohio, population 56,000, that was named after him too.
It makes me a little sick now to remember how proud we were, and how much we celebrated in 1992, when my dad received a “Boss” Kettering Award, GM’s highest engineering award.
Good ol’ Boss Kettering. GM has tried to keep his flame alive. He was an engineer, an inventor, and the head of GM’s research department for twenty-seven years, from 1920 to 1947. He was on the cover of Time magazine in 1933. By most accounts, he was jocular and adventurous, loved by his family and friends—and his breakthroughs in automotive technology, including the invention of the modern car starter and the two-stroke diesel engine, made a lot of money for GM. Along with Alfred Sloan, the president of GM at the time, he helped to found Memorial Sloan Kettering Cancer Center in New York—and personally promised to oversee the organization of a cancer research program employing the most cutting-edge industrial techniques.
But sometimes you have to look away from fanfare and familial love, corporate achievements, and even remarkable philanthropic work and look at a person’s entire legacy. Even I didn’t know much about the man until I started doing more reading. I wasn’t sleeping anyway.
To my mind, there is no greater public health villain.
* * *
—
LEAD IS PROBABLY THE most widely studied neurotoxin, a metal that humans—including my ancient Assyrian ancestors—have been using for thousands of years. Even the word “plumbing” is derived from the Latin word for lead, plumbum. On the periodic table of elements, the symbol is Pb. Its wonder was its remarkable malleability. The Romans used it to coat their famous aqueducts two
thousand years ago. It was also employed in almost every aspect of wine-making, something else the Romans were famous for. A food additive, it was sprinkled on everything, like salt. All this lead in the Romans’ food, wine, and water has provoked theories that lead poisoning contributed to the fall of their empire.
Since the first humans used lead, millions of tons of it have been dug from the earth and dispersed into the environment—far more than mercury or arsenic, the other two toxic trace elements. Aside from plumbing, its most widespread preindustrial use was in paint—to make it washable, brighter, and much more durable. Art historians suspect lead pigments may have contributed to the illness and demise of many great painters, including Correggio, Raphael, and Goya. There’s even a story that Vincent van Gogh was fond of the flavor of one pigment in particular and liked to suck on his paintbrushes. That wouldn’t surprise a pediatrician. White lead is noted for its sweetness, which is why lead paint tastes good to kids.
Although lead was known to be poisonous and unhealthy even in ancient times, its convenience—particularly the durability of lead paint—created a conflict between public health advocates and industry that pretty much continued into the twentieth century, as Christian Warren describes so well in Brush with Death.
Everybody knew lead was toxic, but what it did to the human body was insidious and invisible, while its benefit to industry was tangible and quantifiable in dollars. The story of lead in the United States, where industry has the upper hand, is a little different than in Europe. After childhood poisoning was linked to lead paint in 1904, several European countries banned its indoor use in 1909. But when the League of Nations banned lead paint in 1922, the United States declined to go along. At that point, the regulation of lead paint in this country was almost nonexistent. Lead was big business in America.
What the Eyes Don't See Page 14
