To Gather Drug Data, a Health Start-Up Turns to Consumers
SAN FRANCISCO — For years, Thomas Goetz had been a spirited armchair advocate of the use of digital technology and data to improve health care.
At Wired magazine, where he was executive editor, Mr. Goetz assigned and wrote articles on the subject. He organized conferences, lectured and wrote a book in 2010, “The Decision Tree,” which hailed a technology-led path toward personalized health care and better treatment decisions.
In early 2013, just as he was leaving Wired, Mr. Goetz met Matt Mohebbi, a Google engineer who shared his interest in technology and health. Their conversations continued for months, and prompted an epiphany.
“It struck me that I could help make it happen, not just write about using data to personalize and improve health care,” said Mr. Goetz, who has a master’s in public health from the University of California, Berkeley.
And so the two men founded a company, Iodine, in July 2013. Its offering, an online service for tailored drug-taking information and advice, is being introduced on Wednesday at the Health 2.0 conference here.
Iodine reflects sharply increasing interest and investment in digital health companies. Venture investment in digital health start-ups in the first half of 2014 surged to $2.3 billion, surpassing the total for all of last year, according to Rock Health, which conducts research and provides seed funding for start-ups.
The investor enthusiasm for health ventures, said Halle Tecco, managing director of Rock Health, is fueled by the belief that the health care industry is both huge and technologically backward, and thus ripe for an assault with clever software and data-driven decision-making.
Iodine joins a growing collection of health start-ups using data analysis in innovative ways, including Omada Health, Ginger.io and Propeller Health, sometimes employing sensors and smartphones to provide early warning signals about conditions like diabetes, depression and asthma.
Iodine is charting new territory with its ambitious use of Google Consumer Surveys as a research tool. With 100,000 surveys completed and more added daily, Mr. Goetz said that Iodine was building the largest survey ever taken of Americans’ drug experience, intended to help consumers and perhaps guide policy. Clinical trials to determine drug approval, by contrast, routinely involve a couple of hundred people.
Google Consumer Surveys, begun in 2012, is best known for its performance in the presidential campaign that year, being the most accurate of the Internet polls and second-most-accurate of the polls tracked by Nate Silver, editor in chief of the website FiveThirtyEight.
Today, the Google survey service is used by more than 300 websites, including those of The New York Daily News, The Boston Globe and Bloomberg. Typically, readers are asked to fill out a survey for access to one or a few articles. For each completed survey, the publishers are paid a few cents. Iodine’s use of the service, Google said, is on a scale matched only by major corporations for marketing research.
Iodine combines the survey results with other data sets it taps, including published clinical research, physicians’ reports to the Food and Drug Administration and surveys of pharmacists.
For hundreds of drugs, Iodine’s website provides not only basic information about each drug — its purpose, how it works, side effects and warnings — but also a section with the upsides and downsides of taking a drug, price and co-pays, alternative drugs and interactive tools.
The interactive graphics distill and present the results from the Google surveys. They enable a user, for example, to see how people rate their experience with a drug based on survey questions: Was taking the drug worth it over all? How well did it work for you? Is taking the drug a “hassle?” The hassle factor, survey takers are told, is their experience of side effects and inconvenience caused.
Then, users can see in greater detail how the drug rated for people like themselves, by clicking by gender and age group. The results are presented on a grid with nine squares, with the upper right-hand square being ideal — reflecting the highest ratings for “works well” and “hassle free.”
The handful of questions on the Iodine surveys certainly lacks the depth of clinical trials, but they cast a broad net. Outside experts who have seen Iodine’s early work say they are impressed. And Iodine’s applied research, they say, is certainly addressing a notable information gap in health care, for consumers and providers.
Seventy percent of Americans take at least one prescription drug and 50 percent take two or more. Yet studies show that half or more of people with chronic diseases do not take their drugs as prescribed. Lack of adherence to drug regimens is estimated to cause 125,000 deaths a year and cost between $100 billion and $289 billion annually.
“So the intent of the therapy is lost, and the killer is we don’t know why,” said Dr. David J. Brailer, the national coordinator for health information technology in the Bush administration.
The potential of the research Iodine is doing, Dr. Brailer said, is that it could provide “a window into behavioral models” to improve drug-taking compliance.
Iodine’s early survey results already raise questions and challenge assumptions. High rates of nonadherence are often attributed to older people resisting or forgetting to take their drugs. But people over 55 years old report significantly less “hassle” with their drugs than younger people.
In another analysis, the Iodine team compared the survey reports on five of the top-selling drugs in the United States with those of five off-patent alternatives. The satisfaction scores for the generics were about the same as for brand-name drugs. If the generics were prescribed instead of the branded varieties, the estimated savings would be $27.5 billion a year.
“Food for thought, at least,” Mr. Goetz noted.
Iodine faces a host of hurdles. There are several drug reference guides online, like those of WebMD, Drugs.com and the Mayo Clinic website, but they lack Iodine’s wide range of material. “Our approach is data-driven, while the sites with drug information now are mainly content-driven,” said Mr. Mohebbi, explaining the difference.
But whether that difference will attract a sizable audience remains to be seen, and Iodine has no business model yet. Mr. Goetz says he is leery of advertising, with its potential for visual clutter and conflicts of interest. Still, just as medical clinics and laboratories are valuable, Mr. Goetz said, there should be “experience value” in the aggregated, anonymous data of the drug experience of millions of people.
“But our first goal,” Mr. Goetz said, “is to create a consumer experience that is so compelling it seals that deal.”