It’s hard not to view Google as an 800-pound gorilla, beating competitors at every turn thanks to its vast mountains of cash and engineering talent. But there’s one field where the Mountain View-based search giant has frequently stumbled: repeated attempts to build a foothold in the biomedical realm have either failed or not borne fruit yet. Now it’s trying again.
Earlier today, Bloomberg Business reported that Google is developing what it calls a “health-tracking wristband”—in other words, a fitness band. But this one isn’t going to be marketed at the general public. Nope, Google’s aiming for the academic crowd and wants the device to be used in clinical research. Based on Google’s history, should we expect it to catch on?
The history starts back in 2008 with Google Health. Aimed squarely at Microsoft’s HealthVault, Google Health was a patient-level electronic medical record. It was designed to be a one-stop shop in the cloud (before we called it that) where users could store their medical records, prescription records, and other important health data. We were skeptical of Google Health’s prospects for longevity back then, and so it proved three years later when Google Health went the way of Google Reader—albeit without anything like the same public outcry.
Then there’s Calico, a biotech focused on “aging and associated diseases.” Founded in 2013 by Google and Art Levinson (a former head of biotech giant Genentech and an Apple board member), Calico seems like a natural move from a group of billionaires hitting middle age and concerned about their mortality. It’s too soon in Calico’s brief life so far to judge it. It’s main action came last year when it signed a deal with the pharma company AbbVie, each investing $250 million to build a SF Bay Area research lab.
Last year we got news of more Google biomedical activities. A ‘smart contact lens’ is under development in conjunction with pharmaceutical giant Novartis. Different functions have been suggested for this device, including monitoring glucose levels (important for diabetics) as well as working like an aftermarket ‘autofocus’ for people who have difficulty switching from near to far vision. Then there’s Baseline Study, Google’s answer to the quantified self. Baseline Study is an attempt to create a database that would, if successful, give one a granular and quantitative look at human health.
Baseline Study is a perfectly reasonable approach to a pressing problem in biomedical research. Although DNA sequencing is faster and cheaper than ever, genomic information on its own can be of questionable value without good phenotypic (i.e. expressed characteristics) and biometric data. For example, just looking at genome sequences would suggest that quite a few of us ought not to be alive, since we’re walking around with loss-of-function mutations in genes that we’d thought were essential for life.
Getting good quality phenotypic data and linking it to each individual’s genome is a solid approach, and Google is not alone in studying what some have dubbed the narcissome (the less snarky call it personal genomics). The problem with the Baseline Study may be the small sample size. Google announced it would start with 175 volunteers, which sounds like a lot until you consider the vast number of variables it will be tracking (Google has told is this is just the first phase and it will increase the study size in due course).
Google Genomics is an entirely separate product, this one focused on hosting biological data. You can view it as Mountain View’s answer to the popularity of Amazon’s S3 and EC2 cloud services among biologists, who use it for storing and analyzing large genomic data sets. (If each genome is 6 billion base pairs and you have 100,000 genomes in your study you can imagine how much data is there.) It’s not an exaggeration to say that Google Genomics has barely registered among that research community.
(It should also be noted that many have linked Google to genomics through 23andMe, the personal genomics company co-founded and currently run by Anne Wojcicki, the spouse of Google’s Sergey Brin, but there’s no sign of Google doing anything more than investing in the company.)
With that history, what can we make of health tracker? We reported last year that the company planned to take on Apple’s HealthKit developer framework with Google Fit. This new product would take on Apple’s ResearchKit, as well as wearables that monitor biometrics like heart beat, skin temperature, and so on. It’s a potentially lucrative market. In his February State of the Union address, President Obama announced the Precision Medicine Initiative, a multi-agency research program that wants to enroll a million Americans into a longitudinal health study, and mobile health and biometric trackers have figured prominently in planning discussions.
The difficulty about playing in this space is that it’s extremely regulated compared to the consumer tech sector. Medical devices (including lab tests) are the domain of the FDA, whether you want to commercially market your device or simply use it in clinical research on human subjects. Speaking to Bloomberg Business, Google’s head of life sciences, Andy Conrad, said the company was planning on seeking regulatory clearance in the US and Europe for the health band. Conrad also mentioned the need to engage with researchers, presumably to make up for ground lost to Apple since it showed off ResearchKit earlier this year.
With so many swings, one might have thought Google would have, if not a biomedical home run, then at least put some players on the bases. But there are vast differences between the tech and biomedical industries. Sure, both are home to lots of very smart people, many of whom have PhDs. But Google isn’t the first bunch of engineers to take a look at a cell and think they could do better than the people already in the field, something Derek Lowe explains well as the “Andy Grove fallacy.” Lowe also did a good job dissecting the problems facing tech companies in this area: unlike a microchip, our biology is the product of evolution, which results in all sorts of weird stuff we still don’t fully understand. Compared to that, optimizing search result ranking or delivering targeted advertising seems like child’s play.
Date: June 24, 2015