Last Friday Google confirmed the rumor that it is shutting down its personal health record (PHR) service, Google Health. The service will officially end January 1, 2012 but users will have until January 1, 2013 to retrieve data stored in their PHRs.

The reason given for terminating Google Health as well as the home energy conservation service Google PowerMeter was that “they didn’t scale as we had hoped.” That seems a roundabout way of saying that not enough consumers were interested in creating and maintaining their own PHRs. However, Google is the largest data mining operation in history and I doubt it is simply walking away from health records. Google probably concluded there are better ways to get at the data.

One option would be to work with major hospitals migrating to electronic medical records (EMRs). The federal government has been pressuring the health care industry to implement EMRs for some time. Perhaps Google’s lobbyists know something the rest of us don’t?

Another option for Google is to continue gathering the data in other ways. Google Health Product Manager Aaron Brown alluded to this when he said, “…we’re always working to improve our search quality for the millions of users who come to Google every day to get answers to their health and wellness queries.” He also wrote, “There has been adoption among certain groups of users like tech-savvy patients and their caregivers, and more recently fitness and wellness enthusiasts.” Perhaps Google plans to introduce free fitness and wellness applications.

I suspect the underlying problem is that most health care providers are reluctant to use health records controlled by patients—and most patients sense this. Some, such as BJC Healthcare in St. Louis, offer a compromise solution: Web-based health records set up by patients but owned by the health care provider. Unfortunately, health care provider-owned Web records have very limited features.

UPDATE: June 30, 2011

MobiHealthNews discusses What Google’s exit means for Microsoft’s HealthVault.

CollabRX chairman, CEO, and chief scientist Marty Tenenbaum woke up one day and discovered a lump that turned out to be metastatic melanoma, a particularly virulent cancer. He went to see several local oncologists and was troubled to learn that each had a different treatment recommendation. As his company puts it, patients like Tenenbaum were expected “to make life and death decisions based on trial and error, medical guesswork, or even their zip code – what was available locally.”

Information is power. Tenenbaum researched the options and bet his life on a vaccine trial that ultimately failed. But he was one of the participants who responded well. It’s been more than a decade since he was first diagnosed.

There are thousands of papers on cancer treatment published each year. Oncologists can’t possibly study them all. Plus, how well a patient responds to a given treatment is based in part on their cancer’s “genomic subtype.” Put another way, “cancer is actually hundreds or thousands of rare diseases, and ...every patient's tumor is, to some extent, unique.”

CollabRx developed the Cancer Commons platform for collaboration between patients, physicians, and research scientists based on a molecular disease model (MDM) and how the patient responds to specific treatments. This approach not only benefits the current patient, it builds a knowledge base for subsequent patients.

Cancer Commons is today focused on melanoma. The site plans to add other forms of cancer over time.

The problem with our health care system—in the plainest and simplest terms—is that each of us is expected to choose a doctor, buy insurance, and then do what we are told.

Everyone agrees that has to change. Many organizations are working on changing how we pay for health care. Few are addressing how we choose doctors.

Most people find a doctor by either asking someone they trust for a recommendation or by contacting a referral service (usually operated by a local hospital). In my experience, these methods are hit or miss. People often recommend their own doctors just to reassure themselves. I’ve had doctors I trust refer me to other doctors I could never trust. Referral services direct you to doctors who are either just starting out (i.e., they are inexperienced) or, worse, are still looking for patients.

The Internet is slowly changing the way we choose doctors. If you are looking for a doctor with extensive knowledge and experience treating a specific problem and are able to travel, then the Web is a very powerful tool. Most of the top physicians have online CVs describing their current and past positions, publications, and awards. It’s never been easier for consumers to track down leading specialists.

However, if you are looking for a general practitioner with whom you can establish a rapport, then things haven’t changed much. You can search for physicians at local hospital and “find a doctor” websites. These sites typically provide office addresses, types of insurance accepted, and specialties—and little else.

I can think of two ways that the Internet could be used to improve the process of searching for local doctors. First, we need at least one site with detailed profiles and reviews of most local doctors—a site not run by a local hospital or medical school. Second, there should be an affordable way to meet two or three doctors before making a commitment. Imagine a site that lets you reserve a ten minute slot by making a non-refundable $25 online payment. Perhaps the site should be modeled after eHarmony.

A site called HealthInReach is creating a transparent marketplace for health care. According to the company, “HealthInReach members can see what specific procedures a doctor offers, the cash costs of each, and receive pre-negotiated group discount rates.” That sounds like a step in the right direction. Unfortunately, the service is currently only available in the Los Angeles area.

The rise of sites that make choosing doctors easy and affordable seems inevitable. For a variety of reasons—some of them good—many doctors are reluctant to make themselves available on a trial basis. But there should at least be a place where you can learn more about a local doctor’s experience, competence, and style.

UPDATE: June 10, 2011

Another enhanced "find a doc" service is ZocDoc. The website enables users to make appointments with doctors by clicking on a time from a list.

Three weeks ago I signed up to use HealthVault, Microsoft’s personal health record (PHR) solution. HealthVault is designed to let you create, manage, and share your own medical records. So far I’ve found HealthVault to be not as consumer friendly and polished as I would have expected three years after it was launched. However, I’ve only begun to explore HealthVault’s dizzying array of third-party tools and services.

What do most doctors do when you see them for the first time? They ask a series of questions about your medical history, your family’s medical history, and your current complaints or concerns. I’m surprised that Microsoft doesn’t have a software wizard to help you set up your HealthVault personal health record. Instead, you are left on your own to add health information (conditions, family history, drugs, and so on), authorize others to access your information (family members and health care providers), and browse dozens of online health tools.

The descriptions of many of the online health tools are too general—you really have to test drive them to determine which ones are right for you. Unfortunately, you have to authorize each tool individually to access your HealthVault information, and most of the tools have their own terms of use and privacy policies. So you must either read all of these legal documents or ignore them. Too bad there isn’t a standard agreement for all HealthVault partners' products.

HealthVault and several partners’ tools have the ability to collect, share, chart, and analyze data from devices such as pedometers and blood pressure monitors. Unfortunately, things don’t always work as they should. For example, Microsoft offers an application called HealthVault Connection Center that automatically opens when I connect my pedometer to my computer. It usually takes three attempts to upload the new data. Plus, my pedometer’s archived data could only be uploaded as a PDF document. The new data can be used by applications to create graphs; the archived data is simply being stored.

It seems like Microsoft put more effort into enlisting partners than creating a great user interface. I wish it was the other way around. But I’m still exploring HealthVault’s online health tools. Today I tried two that look promising: Mayo Clinic Health Manager and Surveyor Health’s Medication Risk Maps.

Companies such as 23andMe, DeCode Genetics, Navigenics, and Pathway Genomics will analyze your personal genome for a fee. They quantify your risk for certain medical problems, determine whether you are a carrier of select genetic conditions, and indicate whether you are likely to respond well or poorly to specific drugs. Some also tell you what your genes say about your ancestry and basic inherited traits.

It sounds wonderful. After all, there are specific genetic tests that have proved tremendously valuable. And there’s little doubt that personal genetic testing will become increasingly useful and accurate in the years ahead. However, personal genetic testing is just getting started and it’s important to understand its limitations. If you are an early adopter, you might enjoy having your genome tested and analyzed.

I say “enjoy” because some of the information promised appears to be mainly for entertainment value. For example, 23andMe will tell you if you are genetically predisposed to curly hair. But first they make the disclaimer that a single curly hair gene has not been found and probably doesn’t exist. The interpretation is based on “several genes affecting different aspects of hair development.”

Determining whether you are a carrier of a specific genetic disorder could be useful if you are planning to have children. Though these disorders are relatively rare, they can be devastating. However, 23andMe’s website indicates that 90% of people tested do not have genes associated with any of the 24 conditions covered.

The value of many drug response tests is debated. If you could be told that due to your genetic makeup Plavix will not provide you with adequate protection against blood clots, that information could save your life. However, all that can be said today is that certain genetic variations may indicate that you will receive less protection. There is no genetic test that can say with certainty that you should not rely on Plavix. What you get today is merely a sense of how personal genetic testing will help people in the future.

Personal genetic tests also promise to tell you about your risk for common medical problems such as abdominal aortic aneurysm, Alzheimer’s disease, breast cancer, gallstones, and so forth. However, your risk of developing a specific problem is due to a combination of genetic, environmental and lifestyle factors. Genetic tests would be extremely helpful if they could tell you, for example, that due to your genetic makeup your odds of developing kidney stones are extremely high but you can offset that risk by eating three avocados per week. Unfortunately, you are more likely to be told that your odds of developing kidney stones are slightly greater due to your genetic makeup but that you can reduce the odds by eating healthy foods. (Duh! Eating healthy foods will reduce your risk for a multitude of medical problems.)

I agree with 23andMe: people should be free to decide for themselves whether to undergo personal genetic testing. Even if only a tiny percentage of consumers derive significant health benefits from such tests, everyone has the right to pursue the benefits. (Though I also believe the potential for harm should be re-examined as the technology progresses.)

But I have some concerns. At first glance, it looks like there aren’t any major privacy risks to personal genetic testing. All of the companies that offer the tests promise not to share your results with your employer or insurance company. But I’m not convinced those are the only privacy risks; I'll talk about additional risks in another post. In the meantime, I’d like to see the companies agree to independent audits of their privacy policies and practices.

Web-based personal health records (PHRs) empower you to control your own medical records. They also let you interact online with health care providers (example: your doctor), applications (example: a weight loss program), and devices (example: a blood pressure monitor). PHRs are an alternative to the hospital- and physician-centered electronic medical records often touted by politicians.

Unfortunately, research shows that consumer adoption of PHRs has been very slow. In fact, there are reports that Google plans to scale back its PHR initiative, Google Health.

However, I wouldn’t give up on PHRs just yet. Consumers won’t sign up in droves until pharmacies, doctors, hospitals, and others offer PHR users compelling features and benefits. It takes time to build an ecosystem. But progress is being made.



To kick off our PHR coverage, I’m going to sign up for Microsoft HealthVault. I’ll report on my experiences as I proceed. Later, I’ll try other PHRs such as Google Health, AHIMA (MyPHR), and WebMD, and I’ll let you know how they compare.