by Tom Farley
In what looked like the final blow, City Hall press officer Stu Loeser threatened to cancel Bloomberg’s announcement itself. The federal government seemed to be stealing the narrative, he thought, after seeing the stories on the FDA. Besides, Loeser had offered the story to a reporter at The Wall Street Journal as an exclusive, and the reporter was unimpressed. Companies saying they would reduce their “sales-weighted average” sodium level for crackers to 640 mg per 100 grams would mean nothing to readers. The reporter needed to see changes to specific, recognizable products for them to feel real. “There’s no story here,” a press staffer told us.
The companies were dropping out, the federal agencies were melting away, and the event itself was evaporating. In a few days, we had gone from changing the world to looking hopelessly inept.
The salt team met in my office in despair. Should we drop the whole effort and declare failure? But Lynn Silver, Sonia Angell, and Christine Curtis had worked for years for this moment. They had created forward movement out of little but chutzpah, data, and persistence. After decades in which the federal health agencies had accomplished zero on salt, the New York City health department had actually prodded the massive food industry to budge.
We rallied to hold it together. Heinz had promised to cut the sodium in its ketchup by 15 percent by 2012. That achievement, as small as it seemed, was heartening to my chief of staff, Kelly Christ. Heinz ketchup was as big and as American as any food gets. Her five-year-old niece ate Heinz ketchup every single day, like “every toddler in America.” Cutting how much salt we were feeding those kids was damned important. And there couldn’t be a more concrete example for the Journal. With some frantic calling, the group uncovered a few examples of specific foods that would lose some salt. They weren’t thrilling—Oscar Meyer bologna down 17 percent, Goya canned beans down 25 percent—but they were just enough to satisfy the reporter.
I lobbied Stu Loeser and Linda Gibbs. Federal regulation at this point was a pipe dream; the FDA story was a distraction. Mayor Bloomberg was often accused of liking heavy-handed government regulations; this was a chance to show him cooperating with industry. He had created the salt story and he deserved to get the credit for it. And if we canceled the announcement, we’d be fighting ugly stories about why our well-publicized initiative had flopped. That got the event back on the mayor’s calendar.
On a Monday afternoon, Bloomberg led a small entourage into City Hall’s Blue Room. I stood behind him, together with American Heart Association CEO Nancy Brown, Los Angeles health director Jonathan Fielding, and representatives from Mars, Subway, and other food companies. “Now I must make an admission here,” the mayor started. “I love salt on my food. I’ll put it on my popcorn. I’ll put it on my bagels. I’ll even sprinkle it on my saltines.” Having gotten that out of the way, he went on. “Today I’m excited to announce that sixteen of those companies have formally committed to our National Salt Reduction Initiative. . . . In total, these companies have made commitments to reduce sodium in 49 of the 62 packaged food categories and 15 of the 25 restaurant food categories. . . . They are true leaders in their industry.”
Those companies that had taken the leap were promising some important changes. Boar’s Head, a sandwich meat maker, committed to meeting 2014 targets in cold cuts, sausage, bacon, and processed cheese. Starbucks promised to meet the standards for its breakfast sandwiches, Subway for its sandwiches, soup, and cookies, and Unilever for its pasta sauces. The grocery store distributor White Rose would require its private-label manufacturers to fall in line with targets for thirty-nine packaged food categories. And giant Kraft made a commitment that was frustratingly vague but potentially huge: it promised to “meet or exceed the sodium reduction targets in 50 percent of the relevant NSRI categories, which represents the large majority of the foods that we sell.”
Many companies ignored us. PepsiCo, General Mills, and ConAgra were missing, as were McDonald’s, Burger King, and Wendy’s. Campbell Soup put out a statement that could have spoken for all those on the sidelines: “we believe the targets and timing proposed by New York City can not be realistically achieved in all of our product categories and still meet consumers’ demand for great-tasting foods.” Still, for the first time ever, America now had a concrete national plan to reduce the amount of salt in food. Many food companies were signing on, and if the rest of the food industry later joined them, every year tens of thousands fewer Americans would die of heart disease or stroke.
At the health department, we didn’t know if our initiative would build momentum over time or stall. But we saw no alternative. The food companies hated mandates and wanted to be left alone on salt, but that hands-off approach had failed for decades. The Institute of Medicine committee wanted mandatory regulation, but only the FDA could do that, and it looked paralyzed. Our scheme of coordinated and verified but voluntary action was an attempt at a middle path between two extremes that each looked like dead ends.
• • •
A few days after I landed as health commissioner in 2009, Farzad Mostashari came to my office and told me he was leaving. He was headed to the federal government as a deputy to David Blumenthal, the national coordinator for health information technology. When Mostashari left, I asked Amanda Parsons, from his unit, to take over New York City’s Primary Care Information Project.
Parsons is not a typical doctor or a typical Frieden disciple. Young, blond, and effervescent, she had found her way to the health department not though the mountains of Nicaragua but through the sleek offices of the Fortune 500. Beside her M.D., she had an M.B.A. and several years’ experience advising drug companies for the consulting firm McKinsey & Company. By 2007, after Frieden’s high-profile successes, even at McKinsey “people were talking about the health department.” When a job announcement in Mostashari’s electronic medical records project popped up, Parsons had sent in a résumé. As his new quality improvement director, she would create teams that would show doctors how to use electronic health records to practice medicine better.
She was starting from zero. “None of [the doctors] spent any time thinking about quality improvement,” she said. “They had no foundation to do it.” Worse, even doctors who had completed the training on eClinicalWorks were “completely under water” with the software. They especially struggled with their lifeblood—billing, which was a key function of eClinicalWorks. For doctors in small practices, getting paid is a nightmare, with or without software. “The level of craziness is far beyond what I ever understood,” Mostashari told me. “The regulatory requirements are so heavy that the doctors are clickety-clickety-clicking all day long, having nothing to do with what’s really clinical care.”
Before Parsons could coax doctors to think about improving quality, she had to throw them a lifeline. She established a team that did nothing but help configure doctors’ software and give them booster training, and a second team that helped straighten out billing morasses.
Only after she got past that was Parsons able to send in a third team to focus on the quality of care. The quality team showed doctors how to generate lists of patients needing preventive tests or treatments, just as Mostashari had shown the doctor in Harlem which of her patients needed flu shots. Later Parsons filtered the lists based on the kind of action needed: those patients who hadn’t been tested (like for diabetes), those who had been diagnosed but were not prescribed medicines, and those who were supposed to be taking medicine but had not been to the doctor for a long time.
Parsons’ staff also calculated the doctors’ quality measures, like the percent of patients over sixty-five who had had a flu shot. At first the doctors were embarrassed that their quality indicators were bad. But after getting over their initial skepticism, like the doctor in Harlem, they wanted to improve. They ran lists, sent their patients letters, and tracked their progress with the software.
In 2009 the Obama administration’s “stimulus package” allocated money for the program Health Information Technology for Economi
c and Clinical Health (HITECH). HITECH would distribute $2 billion to help doctors adopt electronic medical records and billions more if the doctors achieved “meaningful use” of “qualified, certified” software. For the government to certify the software, it had to have certain features, and for doctors to get the incentive payments, they had to use those features. The “meaningful use” features were based on what Mostashari had learned from the New York City project, like “clinical decision support systems,” the ability to generate filtered patient lists, and built-in quality-of-care measures. Of the $2 billion, more than $600 million was designated for “regional extension centers” that would help doctors exchange their paper charts for the software. These extension centers would copy what the New York City health department had done; they had been Farzad Mostashari’s idea.
By the time Parsons began directing the program in the summer of 2010, the Primary Care Information Project was mushrooming. Unbelievably, the team had fulfilled Bloomberg’s first promise that a thousand doctors would be using the city’s software by the end of 2008, and it would soon deliver on his second promise of 2,500 doctors by the end of 2010. The health department became one of the new regional extension centers, receiving federal funds. With that designation, the project expanded to help doctors use not just eClinicalWorks but any electronic medical record software that met the new federal standards. The project soon reached into the offices of doctors treating three million people.
Later, I visited one of those offices. Dr. Frank Maselli leads Riverdale Family Practice, a group of ten family physicians in the Bronx. He has a tree trunk for a chest and a broad black beard, making him look more like a down lineman than a family doc, but he also has a warm smile and a knack for computers. Riverdale was one of the early practices to install the health department’s software and try out its quality-improvement features. “I got my first dashboard,” Dr. Maselli told me. “It was horrible. . . . Everything’s all red.” It hurt, because he and his partners took pride in their work. “It’s not like you have bad doctors. It’s just that you have so many things to do. . . . It’s so hard to keep all the balls in the air.” When he talked to his doctors about flu shots and helping smokers quit, they told him, “‘You just gave me a list of twelve more things I have to do. When am I going to do them?’ And we realized that we had to take that function away from the doctors and give it to other members of the staff.”
Dr. Maselli trained his medical assistants to routinely check the electronic records and ask patients about important actions that were easy to overlook—not just flu shots and smoking but also mammograms and colon cancer screens and cholesterol tests. The medical assistants updated the software and then prompted the doctor to write a prescription or give a vaccine. After he reassigned this routine work, Riverdale’s “dashboard” of quality measures improved “tremendously.”
When I spoke to Dr. Maselli, I was amazed at how the software had become so central to his practice. He demonstrated the features, his large hands nearly covering a tiny laptop’s keyboard. Alerts about high blood pressure popped up on the screen. A click gave the doctors suggestions for treatment. Another click sent prescriptions to a patient’s pharmacy. Laboratory test results arrived electronically, routed automatically into individual patient records. An office quality manager printed lists of patients who were overdue for appointments, tests, or services and then used the software to contact them. Every month the health department e-mailed Dr. Maselli quality reports for each of his doctors, and every month he passed them around among the doctors so everyone saw all of them, and by peer pressure all were motivated to improve. “The [health department] program was just great,” he said. “It’s been a terrific success.”
As an enthusiastic adopter, Dr. Maselli wasn’t typical. Amanda Parsons needed to know whether, across an ever-widening footprint, the project was actually helping most doctors practice medicine better. One of her staffers, Sarah Shih, conducted a study to find out. Doctors in fifty-six offices had received an early version of the health department’s software, before the reminder systems were built. Later the health department added the preventive-service reminders and other quality features to their machines. Shih measured how often the doctors in those practices did ten different tests and treatments, first before they got the reminders and then six months afterward. Because the doctors had had electronic medical records the entire time, the study would assess not whether computerized records improved care but instead whether the reminder systems—the virtual Tom Frieden standing at their shoulders—made a difference.
And they did. After they got the reminders, doctors increased by more than 10 percentage points the number of patients they screened for breast cancer, measured for obesity, or tested for diabetes. They were also more likely to ask patients about smoking, give flu shots, and control high blood pressure. In all, the doctors’ performance improved in seven of the ten measures of quality of preventive care. And it was a large majority of the doctors who were improving, not just a few head-of-the-class types.
When I saw the results, I was astounded. Big medical practices with full-time managers had shown similar improvements with the smart use of software. But the improvements in New York City were in tiny practices, often storefronts with just one or two doctors who had no organization looking over their shoulders. The software had filled in their blind spot. And with that new vision, the doctors’ practices got better. It had taken a few years, but Tom Frieden and Farzad Mostashari had figured out a way to get independent doctors—the ones who served the city’s poor—to do a better job at saving lives. The software worked. And it was touching the lives of three million New Yorkers.
• • •
We kept nudging other food companies to sign on to our salt plan. The publicity was all good, we told them. If giants like Kraft and Unilever could do it, so could they. The other food companies didn’t rush, but by the fall of 2010 we had enough latecomers for a second announcement. Butterball promised to cut salt in its deli meats and hot dogs, Snyder’s of Hanover to change its pretzels and chips, Heinz to move past ketchup to frozen pizza, and Hostess to cut sodium in its Wonder bread, snack cakes, and doughnuts. Then the supermarket group Delhaize America, which had sixteen hundred stores under names like Food Lion and Hannaford, promised that it would reduce the salt in its private-label products in twenty-two categories, including frozen pizza, cereal, and butter.
Many others still sat it out. But our conversations with two food companies showed that we were at least getting their attention. The first had come to us when we ran an ad campaign.
Whenever we talked to food companies, they complained that by adding salt, they were only meeting their customers’ tastes. If we wanted them to use less salt, the corporations told us, we must educate consumers. I was irked that companies spending millions in advertising to persuade people to eat their unhealthy food then blamed their customers for being ignorant enough to eat it. But it was true that most people didn’t know that too much salt was bad for them. In the fall of 2010 we put out ads to educate consumers. They weren’t what the food companies had in mind.
The U.K.’s Food Standards Agency had run a three-stage education campaign. The first round told viewers that too much salt was bad, the second that people should eat less than 6 grams of salt (about 2,300 mg sodium) per day, and the third that packaged food was “full of it.” They hoped to persuade people to compare labels and choose foods with less sodium. But we had no budget for salt advertising. Lynn Silver decided that if she could gather enough money for only one campaign, our message should be the FSA’s third. We produced two ads, one featuring a can of soup and another a frozen dinner, each bursting with a snowstorm of salt. “Many foods pack a lot more salt than you think,” the ads read. “TOO MUCH SALT CAN LEAD TO HEART ATTACK AND STROKE. Compare Labels. Choose Less Sodium.”
The soup ad stirred up Denise Morrison, the incoming CEO of Campbell Soup. The ad, she complained, “suggests that salt is all that
a consumer gets in a typical can of prepared soup. This is plainly untrue, and New Yorkers would not be well served by a campaign that persuades them to eat less soup.”
Soon afterward, at Morrison’s request, Andy Goodman, Sonia Angell, Lynn Silver, Christine Curtis, Kelly Christ, and I sat at the conference table in my office with Campbell’s CEO and four of her senior staff for a soup tasting. Campbell, the executives told us, sold more than 60 percent of the canned soup in America and had been working to reduce sodium in its soup for decades. In recent years it had cut the sodium by a third in its tomato soup, which they said fed 25 million households in America. Campbell had also started a line of low-sodium soups called Healthy Request.
If they hoped to show us that Campbell had made progress, they succeeded. Their tomato soup already had less sodium than the targets we had set for 2014, and I thought it was tasty. With Campbell’s tomato soup such a staple of the American diet, that achievement was helping prevent high blood pressure across the country. But if the executives hoped to prove that Campbell couldn’t do more, they failed. The Healthy Request line of low sodium soups made up only 10 percent of its sales, which confirmed what all the food companies know: labeling a food “low sodium” drives down sales because customers assume it tastes bad. Most of Campbell’s red-and-white canned soups—other than tomato—in my grocery store had nearly 900 mg of sodium per one-cup serving. That was more than our target and more than half of the sodium that most people should eat in an entire day. (Sylvia Birnbaum’s favorite, Butternut Squash Bisque, had a little less salt than that but still more than our target for 2012.) So Campbell had shown that it could cut the salt in mainstream soups and win with customers. It just needed to do that for more types. Later, Campbell joined the NSRI but only for its canned food (SpaghettiOs) line.
