Feature: Why did Google Health fail?
This article first appeared in the April 2012 edition of Pulse+IT Magazine.
When Google announced in June last year that it was dropping its Google Health online personal health record (PHR) platform, there was much debate as to exactly why the offering failed so publicly. Many reasons were given for the failure, including consumer apathy, concern over security and privacy and the fact that no one really knew what to do with it.
In a much-discussed article in US magazine InformationWeek Healthcare earlier this year, reporter Paul Cerrato rejected the analysis of Colin Evans, the former CEO of personal health record provider Dossia, that the problem was the unwillingness of healthcare providers to give consumers control over their medical data.
Instead, Cerrato wrote, the main reason the public doesn't sign up for online PHRs en masse is they don't really care that much about their health. “Yes, concerns about security and privacy and the reluctance of providers to share patient information slow things down, but at its core this is about apathy,” he wrote. “As I've said before, most Americans care more about their cars than their health.”
The current CEO of Dossia, Mike Critelli, told Information Week that Google Health failed because it required consumers to do too much of the work in getting data downloaded. Jessica Ryan Ohlin, an analyst with Frost & Sullivan, described Google Health leaving the market as “sobering”.
“Google didn't fully commit to this, they didn't build enough relationships with health IT vendors, payers, providers and other healthcare stakeholders that would leverage the value chain across healthcare, and they relied on consumers – many of whom don't see the need for a PHR,” Ms Ohlin said.
Not everyone agrees with this way of thinking, however. Kenneth Mandl, the co-founder of the open-source PHR project Indivo, which is said to have inspired the creation of Google Health, recently told US National Public Radio (NPR) that there were a number of reasons the service didn't take off. It didn't have enough apps and it didn't open itself up to the developer community, something that companies like Apple do so well, he said.
Google also didn’t put much effort into establishing a trust model, Dr Mandl said. “What would Google do with your health data once they had it? And do I trust a company to manage my health that does not even have a customer service number? Is search and advertising the revenue model I am looking for in my health records company?”
David Dembo, a clinician and Asia Pacific business development executive for Microsoft's Health Solutions Group here in Australia, has a different theory. He believes that Google dropped the service too early and should have waited for critical mass to build.
“I actually don't think that Google Health was a failure,” Dr Dembo says. “I think the only failure was their early withdrawal. They didn't wait for the body of evidence or for the patients to be convinced. What they appeared to do was make their judgement all around the test base and scale and they thought that would come.”
Microsoft is convinced that scale will come, but it will take time, Dr Dembo says. “It is a real pity that Google Health did withdraw because that would have been another investment by a big, talented multinational to help build on that body of evidence.”
A body of evidence is critical to the success of any PHR venture, he believes, and technology companies must realise that both clinicians and consumers alike must be convinced something will work and will be of benefit before they take it up wholeheartedly.
“This is an interesting conversation to be had in that it ties in with the case for the PCEHR,” he says. “As a collective, we get distracted by technology and really this is a change management issue. This is a human issue and is about changing the way we behave, changing the way clinicians prescribe and consumers consume.
“The pivot here remains the doctor. That is still the first person that the consumer goes to for advice – that is their gatekeeper, so to speak. And there's very few industries that have done a good job of clinician behaviour change management.”
Dr Dembo points to the pharmaceutical industry as one sector that has been “ masterful” at changing clinician behaviour. He says the pharmaceutical industry understands that to get clinicians to change what they are currently doing to what it would like them to do, it is essential to study the factors that motivate them in the first place.
“The first thing is [clinicians] have to be convinced that changes are in the best interest of the patient,” he says. “Second of all, because we are scientifically trained, we must have robust data to support the change that is being made, and a subtle but critical component of getting clinicians to change is it has to be positioned as mainstream and not too far from the fringe. The first principle of medicine is do no harm, so you tend to practice medicine that is well accepted.
“And that is exactly what the pharmaceutical companies do – they gather a robust data set that allows them to make claims about what they'd like you to do rather than what you are currently doing. The body of evidence for the use of online tools is only beginning to emerge and unless the evidence becomes robust enough to convince clinicians to change the way they practice, there will be sluggish adoption by patients.
“We still need to convince our medical colleagues that not using technology to augment the doctor-patient relationship leaves both the doctor and the patient exposed to poor health outcomes. Intuitively we all seem to know that consumer technology has massive potential in health, but change in health relies on evidence and that is what all stakeholders should be focusing on now.”
Despite the relative success of Microsoft’s PHR offering HealthVault and of Dossia, with the demise of Google Health it is still pertinent to ask whether other consumer-managed systems will go the distance. The UK, as always, provides a salutary lesson in what not to do.
Prominent researcher Trisha Greenhalgh, head of the healthcare innovation and policy unit within the Centre for Primary Care and Public Health at Barts and the London Medical School, has extensively researched and written about eHealth in the primary care setting. She published an article in the British Medical Journal (BMJ in 2010 about the uptake – or lack thereof – of HealthSpace, an internet accessible personal electronic health record that was introduced by the UK's National Health Service in 2007.
HealthSpace was inspired by the success of Kaiser Permanente's My Health Manager in the US, which offers a customisable portal through which members may access parts of their centrally held record. “By mid-2008, 2.4 million of Kaiser’s 8.7 million members had registered for My Health Manager, most commonly for repeat prescriptions,” Professor Greenhalgh and colleagues wrote.
“Use of the system was associated with up to 10 per cent fewer visits to the physician and a significant reduction in telephone calls. A survey of members who were actively using this technology showed that most perceived it as useful and easy to use.”
Not so HealthSpace. As Professor Greenhalgh reported, between 2007 and October 2010, 172,950 people opened a basic HealthSpace account, with less than one per cent opening an advanced account, which offered increased functionality and was aimed at the chronically ill.
“Overall, patients perceived HealthSpace as neither useful nor easy to use and its functionality aligned poorly with their expectations and self management practices,” she wrote. “Those who used email-style messaging were positive about its benefits, but enthusiasm beyond three early adopter clinicians was low, and fewer than 100 of 30,000 patients expressed interest [in messaging].
“Policy makers’ hopes that “deploying” HealthSpace would lead to empowered patients, personalised care, lower NHS costs, better data quality, and improved health literacy were not realised over the three-year evaluation period.”
Perhaps one reason why Kaiser's My Health Manager has succeeded where HealthSpace and Google Health have not is that in the latter two cases, there seems to have been little effort put into developing relationships with consumers and other vendors. And, to repeat Dr Dembo's point, the body of evidence is not yet there.
So where does that leave projects such as Healthbook, the consumer portal being developed by IBM and private insurer Medibank as one of the Wave 2 PCEHR projects in Australia?
Healthbook is planned to be more like a health diary than an full service like HealthVault, but one reason why Medibank believes it will be successful is that it will be offered initially only to Medibank customers, and it will come with the resources of 100 nurses to provide advice and guidance.
According to Isabel Frederick, general manager for eHealth at Medibank, Healthbook will be different from other consumer health portals in that the initial users of the product will already have established relationships with Medibank, “so it will be an extension of that healthcare relationship”, Ms Frederick says.
An existing relationship might be the reason why other offerings such as Dossia are successful. Dossia was launched in 2006 by a consortium of employers to offer a PHR to employees as a way of helping to control spiralling healthcare costs. Members include Intel, AT&T, Wal-Mart and BP, who between them employ over two million people.
Dossia has now launched a Health Manager extension, which integrates game and social networking dynamics and messaging to improve user engagement and behaviour change.
In its new version, Dossia is certainly trying to build relationships with vendors and consumers, and will now be able to aggregate data from multiple sources, including from users' health plans, healthcare providers and wellness and medical devices.
This is something that Microsoft's HealthVault already does, according to Dr Dembo. “We built HealthVault to be that bridge between the consumer world and the health system world,” he says. “And that is what has happened in the US. It is available to organisations that are convinced that consumers need to be a participant in their care, to have a much stronger relationship with consumers, that face to face is augmented with online. They can use HealthVault to do that.
“We have over 200 organisations that are doing it; there are around 300 applications that are live on HealthVault and over 80 consumer devices. We are already beginning to see the evidence of its impact, the kind of evidence that gives us confidence that a greater body of evidence will come. It garners confidence in enabling consumers to participate more in their care. That kind of cascade happens and people become bullish about consumer control.”
HealthVault is not likely to be introduced into Australia in the short term for lack of a sustainable local market to fuel its growth. Also, if Australia continues down the path towards data sovereignty, in which legislation would predicate that all health data resides here and is not to leave our borders, the Australian health system will not be able to access HealthVault in other countries.
“For there to be a business case for HealthVault in Australia, the role that the consumer plays in eHealth needs to go further than the role that it has been given in the PCEHR,” he says.
“The way I would differentiate between what HealthVault offers and what the PCEHR is going to offer, is that PCEHR enables Australian consumers to have transparency on their data and it provides them with the ability to consent for their information to be shared within the health system. It is kind of a portal in to the health system, rather than the other way around.
“HealthVault is more about a portal out of the health system that enables a health system to see what is happening in the community and it provides the consumer with much more empowerment to be participants in their care rather than simply recipients of care. With the PCEHR, the consumer still remains a recipient of care, while HealthVault enables people to become a participant. When we see the market move towards genuine consumer empowerment, then there will be a strong case for HealthVault in Australia.”
So, progress in consumer-managed online health records will be slow. However, Dr Dembo certainly believes it is the way of the future, with a few provisos.
“There is a hell of a lot of risk in the approach that I see to eHealth projects, in that we focus too much on the technology and not the human element. The other risk that I see is that we tend to – by we I mean healthcare stakeholders – is that we do get distracted overly by the stereotypical chronic disease management patient that may have many pathologies, that may be a little recalcitrant or perhaps not well educated, and that is a bias away from the majority of society of healthcare consumers.
“In the main, we have a well educated population. They are concerned about their health and are motivated to take a more active role in remaining healthy. It is in this area of wellness that significant momentum already exists for consumer devices and online tools. A gap however exists between what the services consumers seek and what the health system can provide, so consumers have to go outside the system to find help.
“There are some great articles I have read around consumer devices and the gamification of health and how a large chunk of society is already moving towards the online self-help world and generating a large dataset of health and wellness data. There is recent research showing that there are over 200 million mobile health apps that are in circulation – about a third of Americans track their health data online but don't share it with anyone else.
“The community surrounded by smart devices and smart portals is digitising in advance of the health system and if the health system wants to leverage such a potentially powerful data set, it will have start building a stronger connection between the consumer and itself. Consumers, whether from frustration or aspiration seem convinced that technology is a pillar of wellness and they are embracing it in droves. It is now up to us healthcare stakeholders to prove that traditional health services can extend beyond bricks and mortar into the online world.”
Posted in Australian eHealth