Like so many attempts before it — drkoop.com and RevolutionHealth.com to name just two — Google has found that implementing personal health records in a meaningful way is really, really hard.

So hard, in fact, that it has given up and is shuttering its Google Health service, after just 3 years being open to the public.

Now’s a good time to review Google Health and why it may have never been the roaring success Google had hoped it would be. Perhaps this can act as a warning for future companies who think that (a) there’s a huge, pent-up demand for personal health records and (b) implementing them in a way that is engaging and helpful is easy.

Let’s start with Google’s statement about Google Health’s failure to capture any significant component of the electronic health record space:

In the coming months, we’re going to retire two products that didn’t catch on the way we would have hoped, but did serve as influential models: Google Health (retiring January 1, 2012; data available for download through January 1, 2013) and Google PowerMeter (retiring September 16, 2011). Both were based on the idea that with more and better information, people can make smarter choices, whether in regard to managing personal health and wellness, or saving money and conserving energy at home. While they didn’t scale as we had hoped, we believe they did highlight the importance of access to information in areas where it’s traditionally been difficult.

Of course, when you shut down a product, you have to put as a good a face on it as possible. It’s not a failure — it was a “influential model” for the entire industry! Despite have zero data to demonstrate that Google Health actually helped “people make smarter choices [about their health],” they promote the claim to make it seem like Google Health made a difference.

Objectively, it’s hard to exactly see what kind of difference — for better or worse — Google Health may have made. Since they’ve never publicized numbers of total or active users, it’s hard to gauge whether it had any impact whatsoever. But we know how it helped move the needle in at least better understanding the complexity and difficulties of medical and health data.

That’s because of our own e-Patient Dave, who took Google and his hospital up on the promise of Google Health. As he he detailed a year after Google Health launched, he had a difficult time using it with actual real-world data from his doctor. He had good reason to believe Google Health could help him gain better access to his health records, because that’s what the hype machine suggested:

When Google Health launched last May, my hospital’s CIO blog said “we have enhanced our hospital and ambulatory systems such that a patient, with their consent and control, can upload their BIDMC records to Google Health in a few keystrokes. There is no need to manually enter this health data into Google’s personal health record, unlike earlier PHRs from Dr. Koop, HealthCentral and Revolution Health.”

Sadly, it appears Google’s QA team never actually tried to do what they and their partners said was readily do-able with Google Health — use it to view and interact with their medical and health data.

When it launched to the public in 2008 (after being in development and private testing for two years), Marissa Mayer, who took over the Google Health initiative from Adam Bosworth, was very optimistic about the product’s potential and future:

Google Health, Ms. Mayer said, represents a “large ongoing initiative” by the company, which she said she hoped would eventually include “thousands of partners and millions of users.”

She had good reason to be optimistic. Because, after all, people already use Google to search for health information. Why not help them even more?

“If you look at health care, there’s already a huge user need, people are already using Google more than any other tool on the Web to find health information,” Mayer said. “And the health care industry generates a huge amount of information every year. It’s a natural core competency fo us, to understand how to organize all that data.” […]

While the focus will be on improving health care and making records more accessible and portable for patients, Google will also improve life for physicians, Mayer noted.

“The goal for a lot of doctors is how many patients can they see in a day,” Mayer said. “That means their minutes per patient has got to go down, and the less time they have to spend finding and going over patient records the better. Ultimately we will design a product that’s useful for users, and also helps doctors do their job more quickly and more efficiently.”

Great ideas. But they never materialized into a product that either doctors or patients (or just ordinary people) could readily or easily use and master.

My own review of the initial release of Google Health demonstrated a product that left more than a little to be desired in lack of bringing anything much new to the table. No cool visualizations. No engaging game component. Nothing other than your manually-entered data (if you didn’t want the nightmare e-Patient Dave experienced) and some boring graphs. Combined with not being a HIPAA-covered entity, it may have been enough to sink expectations as the product’s updates were few and far between.

RIP Google Health; I’m not sure how “influential” you were, other than to act as a role model of how not to role out an online personal health record without significant real-world testing ahead of time. You were not a good idea ahead of its time; you were simply an idea tried many times before and coming up against the exact same problems of prior efforts — nobody is much interested in data for data’s sake.