Bits Blog: Health Sites Under Scrutiny Over Mining of Data

Millions of people rely on Web sites like WebMD and Health.com for information about depression, sexually transmitted diseases, cancer and other sensitive personal health issues. But it can be difficult for consumers to understand how health sites may capture, analyze and share information about user searches and other activities — even the small minority of people who manage to slog all the way through the privacy policies.

Lisa Madigan, the attorney general of Illinois.

In an effort to increase industry transparency, Lisa Madigan, the attorney general of Illinois, has opened an inquiry into the data-mining practices of some popular health sites.

On Tuesday, she sent letters to officials at eight sites asking for detailed information about their companies’ data collection, data storage and data sharing practices. The sites included: about.com; drugs.com; health.com; mayoclinic.com; menshealth.com; mercola.com; WebMd.com; and weightwatchers.com.

In the letters to the sites’ executives, Ms. Madigan said she was concerned about the potential dissemination of information related to people’s private health concerns.

“Health-related information, which would be protected from disclosure when said in a doctor’s office, can be captured, shared, and sold when entered into a Web site,” she wrote. “These concerns are likely overlooked by consumers, as the disclosures about capturing and sharing their information are often buried in privacy policies not found on websites’ main pages.”

WebMD’s privacy policy, for example, says that the site does not make a user’s personal information – like a name or address — available to third parties for marketing purposes.

But third parties, the privacy policy says, may use non-personal data to target WebMD users with ads related to their interests. The policy added that WebMD may combine personal and nonpersonal information about users on the site, or may collate that data with information gathered from external sources.

Risa Fisher, a spokeswoman for WebMD, said that the company had just received Ms. Madigan’s letter of inquiry and planned to provide the information she requested about its user data practices.

“Privacy is very important to WebMD and our policies are designed to fully protect the personal health information of our users,” Ms. Fisher said.

The Illinois inquiry comes after the publication a few days ago of a research letter in a medical journal reporting that some popular health portals leaked information about users’ health searches to third parties, like social networks or ad networks, operating on their Web sites.

For his research, Marco D. Huesch, a health care policy researcher at the Sol Price School of Public Policy at University of Southern California searched for content related to depression, herpes and cancer on 20 popular health-related Web sites.

In the letter about his study, published in JAMA Internal Medicine, he said that 13 of those sites used third-party tracking elements like cookies or social media plug-ins. Seven of the sites, he wrote, leaked his health searches to third-party trackers.

Although Mr. Huesch wrote that he could not determine whether the third parties misused the information, he found the leakage of the health searches worrisome in itself.

“The ramifications could span embarrassment, discrimination in the labor market,” Mr. Huesch wrote, “or the deliberate decision by marketers not to offer or advertise particular goods and services to an individual, based solely on the companies’ privately gathered knowledge.”

The online advertising industry is keenly aware of such concerns.

This year, the Network Advertising Initiative, an industry self-regulatory association for third-party digital ad companies, revised its code of conduct to require that its members obtain user permission before collecting information about certain specific health conditions.

The conditions that would require user permission include “all types of cancer, mental health related conditions, and sexually transmitted diseases,” the revised code said, but not acne, high blood pressure, heartburn, cold and flu, or cholesterol management.

The self-regulatory group has nearly 100 members, according to its site. The updated version of code of conduct is scheduled to take effect next year.

Administration Says Mining of Data Is Crucial to Fight Terror

WASHINGTON — In early September 2009, an e-mail passed through an Internet address in Peshawar, Pakistan, that was being monitored by the vast computers controlled by American intelligence analysts. It set off alarms. The address, linked to senior Qaeda operatives, had been dormant for months.

Investigators worked their way backward and traced the e-mail to an address in Aurora, Colo., outside Denver. It took them to Najibullah Zazi, a 24-year-old former coffee cart operator, who was asking a Qaeda facilitator about how to mix ingredients for a flour-based explosive, according to law enforcement officials. A later e-mail read: “The marriage is ready” — code that a major attack was planned.

What followed in the next few days was a cross-country pursuit in which the police stopped Mr. Zazi on the George Washington Bridge, let him go, and after several false starts, arrested him in New York. He eventually pleaded guilty to plotting to carry out backpack bombings in the city’s subway system.

It is that kind of success that President Obama seemed to be referring to on Friday in California when he defended the National Security Agency’s stockpiling of telephone call logs of Americans and gaining access to foreigners’ e-mail and other data from Microsoft, Google, Yahoo and other companies.

He argued that “modest encroachments on privacy” — including keeping records of phone numbers called and the length of calls that can be used to track terrorists, though not listening in to calls — were “worth us doing” to protect the country. The programs, he said, were authorized by Congress and regularly reviewed by federal courts.

But privacy advocates questioned the portrayal of the program’s intrusion on Americans’ communications as modest. When Americans communicate with a targeted person overseas, the program can vacuum up and store for later searching — without a warrant — their calls and e-mails, too.

Mr. Obama acknowledged that he had hesitations when he inherited the program from George W. Bush, but told reporters that he soon became convinced of its necessity. “You can’t have 100 percent security and also then have 100 percent privacy and zero inconvenience,” he said. “We’re going to have to make some choices as a society.”

To defenders of the N.S.A., the Zazi case underscores how the agency’s Internet surveillance system, called Prism, which was set up over the past decade to collect data from online providers of e-mail and chat services, has yielded concrete results.

“We were able to glean critical information,” said a senior intelligence official, who spoke on the condition of anonymity. “It was through an e-mail correspondence that we had access to only through Prism.”

John Miller, a former senior intelligence official who now works for CBS News, said on “CBS This Morning,” “That’s how a program like this is supposed to work.”

Veterans of the Obama intelligence agencies say the large collections of digital data are vital in the search for terrorists. “If you’re looking for a needle in the haystack, you need a haystack,” Jeremy Bash, chief of staff to Leon E. Panetta, the former C.I.A. director and defense secretary, said on MSNBC on Friday.

Under the program, intelligence officials must present Internet companies with specific requests for information on a case-by-case basis, showing that the target is a foreigner and located outside the United States, a senior law enforcement official said Friday. If the N.S.A. comes across information about an American citizen during the search, it turns over that material to the F.B.I. for an assessment, the official said.

An administration official said Friday that agencies were evaluating whether they could publicly identify particular terrorism cases that came to the government’s attention through the telephone or Internet programs.

Representative Mike Rogers, the Michigan Republican who is chairman of the House intelligence committee, said Thursday that the phone program “was used to stop a terrorist attack.” He did not identify the plot, or explain whether the call logs in the case would have been unavailable by ordinary subpoenas.

Two Democratic senators on the Intelligence Committee who have been warning about the bulk collection of records under the Patriot Act, Ron Wyden of Oregon and Mark Udall of Colorado, said Friday that their study of the calling log program has convinced them that it was not worth its cost to privacy.

Mark Mazzetti and Michael S. Schmidt contributed reporting.