快猫短视频

Last week, Google announced its latest foray into health care 鈥� an ambitious project to collect vast amounts of health data from 10,000 volunteers. The company鈥檚 health spinout Verily describes its Project Baseline as 鈥�an effort to map human health鈥�.

People have been mapping human health by collecting data from volunteers for decades, but Google is billing its version as a step change. That鈥檚 largely down to the addition of snazzy new technologies that Verily will use to track the study participants: with more frequent and reliable data, the company says it will be able to draw conclusions no one has.

But is this simply a techno-fetishisation? And worse, could the vaunted technology even get in the way?

Along with a range of medical scans, genome sequencing and gut microbiome analysis, every participant in Project Baseline will be given a Study Watch, worn on the wrist to measure movements, skin conductance, heart rate and rhythm, as well as a separate device that will be in their bed, monitoring their sleep.

Self-reporting pitfalls

This approach is very different from how major studies have been done over the past few decades. The Framingham Heart Study 鈥� which has so far included 15,000 people across three generations and helped us understand how smoking and cholesterol, among other things, contribute to heart disease 鈥� involves medical testing every two years and interviews on lifestyle.

The 鈥� launched in 1993 to investigate health issues in postmenopausal women, recruiting 161,000 participants 鈥� required dietary changes and supplements benchmarked by medical tests and reporting on habits.

Self-reporting is notoriously fraught though and can yield unreliable data because people omit information they may deem unimportant or embarrassing, or just plain forget.

Although the new study will also include self-reporting, via an app that will send periodic questionnaires, it鈥檚 fair to say that Google is putting its faith in the tech.

听Tech teething problems

听Judging by the tech that鈥檚 already out there, that might be unwise. There are plenty of gadgets that offer to measure every step you take, every beat of your heart, and how much you sweat. Their developers have been promising that they will make huge contributions to healthcare research ever since the devices were launched.

They haven鈥檛 yet, because these trackers aren鈥檛 too reliable 鈥� consumer wearables aren鈥檛 made for clinical trials, and most for the collection of clinical data. Even clinical-grade devices can of how much exercise someone has done.

It鈥檚 not just about the devices themselves 鈥� it鈥檚 also our ability to use them consistently. A 2014 survey of some 6000 US adults who bought wearable fitness trackers found that within six months of buying them.

It鈥檚 all well and good to celebrate the transition from self-reporting.听 But it鈥檚 important to acknowledge that wearing a tracker won鈥檛 entirely bypass other concerns about human nature.

No specific goal

The use of health-tracking devices in medical research can also mean more work for study participants. Project Baseline is scheduled to run for four years 鈥� can 10,000 people really be relied on to use a smart watch and sleep tracker every day for that period? To consistently keep their device charged?

Google told 快猫短视频 that it has built the watches specifically with these in mind, giving it a battery than only needs charging once a week, and an unobtrusive profile. The company worked with consumers and researchers, and says 鈥渨e have conducted a long-term study with wearable users so we can get feedback on what impacts user compliance鈥�.

Beyond the usual privacy and security risks, will the shiny technology really yield new insights? Particularly when Verily鈥檚 effort is not concentrated around a specific research goal?

The Framingham Heart Study specifically set out to identify factors that cause cardiovascular disease, while the Women鈥檚 Health Initiative trialled the effects of hormone replacement therapy, dietary changes and vitamin supplements on the health of postmenopausal women.

The team behind the Verily project says it is looking for clues that might link specific health factors or behaviours to diseases like heart disease or cancer, potentially helping us to develop ways to predict these diseases, and treat them earlier on.

Lost billions

But this approach has gone wrong before. The US National Children鈥檚 study 鈥� meant to follow the progress of 100,000 children for 21 years 鈥� fell apart 15 years and $1.3 billion in, when it became clear that .

Getting a huge glut data, but without guiding principles could lead to false correlations. It would be really difficult to find trends if you didn鈥檛 know what you were looking for, says Lukasz Piwek at the University of Bath in the UK. Given how much data Verily will be collecting, without having a specific research question or starting point on what to look for, this could turn out to be a needle-in-a-haystack exercise. Such data-mining approaches to medical research will need sophisticated algorithms to search for patterns before they can be useful.

Human health is not the same as online browsing behaviour 鈥� it鈥檚 a lot more complicated and needs its own set of algorithms to decipher. Data is a valuable resource, but prospective studies tend to be better than retrospective ones, because you鈥檙e collecting all the info you need.

Google sidesteps this by saying 鈥�One of the key features of the work we鈥檙e doing with Project Baseline is developing a data infrastructure for multi-dimensional data sets so that we can use it as a single query source. This has not been done before.鈥�

听If this approach is to work, Google will need a much larger pools of participants to get results. 听They have preliminary plans to run it for longer if successful, but nothing concrete as yet.

Google鈥檚 latest foray into health isn鈥檛 surprising. Why wouldn鈥檛 the world鈥檚 most successful search engine 鈥� which can already use a person鈥檚 browsing history to reveal their interests, whereabouts, habits and beliefs 鈥� reach out for even more data? And with the goal of solving some of our most pressing health problems, who could fault them? This time, however, Google may not provide the answers.

Read more: Encryption wars: The privacy debate raging in your smartphone; wearables