WHEN someone goes into cardiac arrest, survival depends on how quickly the heart can be restarted. Enter Amazon’s Echo, a voice-driven computer that answers to the name of Alexa, which can recite life-saving instructions about cardiopulmonary resuscitation, a skill taught to it by the American Heart Association. Alexa is accumulating other health-care skills, too, including acting as a companion for the elderly and answering questions about children’s illnesses. In the near future she will probably help doctors with grubby hands to take notes and to request scans, as well as remind patients to take their pills.

Alexa is one manifestation of a drive to disrupt an industry that has so far largely failed to deliver on the potential of digital information. Health care is over-regulated and expensive to innovate in, and has a history of failing to implement ambitious IT projects. But the momentum towards a digital future is gathering pace. Investment into digital health care has soared (see chart).

One reason for that is the scale of potential cost-savings. Last year Americans spent an amount equivalent to about 18% of GDP on health care. That is an extreme, but other countries face rising cost pressures from health spending as populations age. Much of this expenditure is inefficient. Spending on administration varies sevenfold between rich countries. There are huge differences in the cost of medical procedures. In rich countries about one-fifth of spending on health care goes to waste, for example on wrong or unnecessary treatments. Eliminating a fraction of this sum is a huge opportunity.

Consumers seem readier to accept digital products than just a few years ago. The field includes mobile apps, telemedicine—health care provided using electronic communications—and predictive analytics (using statistical methods to sift data on outcomes for patients). Other areas are automated diagnoses and wearable sensors to measure things like blood pressure.

If there is to be a health-care revolution, it will create winners and losers. Andy Richards, an investor in digital health, argues that three groups are fighting a war for control of the “health-care value chain”.

One group comprises “traditional innovators”—pharmaceutical firms, hospitals and medical-technology companies such as GE Healthcare, Siemens, Medtronic and Philips. A second category is made up of “incumbent players”, which include health insurers, pharmacy-benefit managers (which buy drugs in bulk), and as single-payer health-care systems such as Britain’s NHS. The third group are the technology “insurgents”, including Google, Apple, Amazon and a host of hungry entrepreneurs that are creating apps, predictive-diagnostics systems and new devices. These firms may well profit most handsomely from the shift to digital.

The threat to the traditional innovators is that as medical records are digitised and new kinds of patient data arrive from genomic sequencing, sensors and even from social media, insurers and governments can get much better insight into which treatments work. These buyers are increasingly demanding “value-based” reimbursement—meaning that if a drug or device doesn’t function well, it will not be bought.

The big question is whether drug companies will be big losers, says Marc Sluijs, an adviser on investment in digital health. More data will not only identify those drugs that do not work. Digital health care will also give rise to new services that might involve taking no drugs at all.

Lunches eaten

Diabetes is an obvious problem for the pharma business in this regard, says Dan Mahony, a partner at Polar Capital, an investment firm. Since evidence shows that exercise gives diabetics better control of their disease (and helps most pre-diabetics not to get sick at all), there is an opening for new services. UnitedHealthcare, a big American insurer, for example, has a prevention programme that connects pre-diabetics with special coaches at gyms.

An app or a wearable device that persuades people to walk a certain distance every day would be far cheaper for insurers and governments to provide than years of visits to doctors, hospitals and drugs. Although Fitbits are frequently derided for ending up in the back of a drawer, people can be motivated to get off the sofa. Players of Pokémon Go have collectively walked nearly 9bn kilometres since the smartphone game was released last year.

That is the backdrop to a new firm called Onduo, a joint venture that Google’s health-care venture, Verily Life Sciences, and Sanofi, a French drug firm, set up last year. Onduo will start by developing ways to help diabetics make better decisions about their use of drugs and their lifestyle habits. Later on, Onduo wants to help those who are at risk of diabetes not to develop it. The startup is a good hedge for Sanofi, which faces a slowdown in sales of its blockbuster insulin medication, Lantus, which lost patent protection in 2015.

This kind of thinking does not come easily to drug firms. Switzerland’s Novartis is one of the few to have acknowledged that digital innovation will mean selling products based on patient outcomes. But if pharma firms do not design solutions that put the patient, rather than drug sales, at the centre of their strategy, they risk losing relevance, says Mr Sluijs.

Large hospitals, some of which count as both incumbents and traditional innovators, will also be affected. The rise of telemedicine, predictive analytics and earlier diagnoses of illnesses are expected to reduce admissions, particularly of the emergency kind that are most lucrative in commercial systems. The sickest patients can be targeted by specialist services, such as Evolution Health, a firm in Texas that cares for 2m of the most-ill patients across 15 states. It claims to be able to reduce the use of emergency rooms by a fifth, and inpatient stays in hospitals by two-fifths.

Rapid medical and diagnostic innovation will disrupt all businesses that rely heavily on physical facilities and staff. A mobile ultrasound scanner made by Philips, called Lumify, means that a far larger number of patients can be seen by their own doctors. As for data-based diagnostics, one potential example of its power to change business models is Guardant Health, a startup that is analysing large quantities of medical data in order to develop a way of diagnosing cancer from blood tests. If the firm can devise an early test for breast cancer, demand for mammograms and the machines that take them would fall, along with the need for expensive drugs and spells in hospital.

From ER to AI

There is also good news for hospitals, however. Increasingly, machine-learning programs are able to make diagnoses from scans and from test results. An intriguing recent project has been to stream and analyse live health data and deliver alerts on an app that is carried around by doctors and nurses at the Royal Free Hospital in London. The app, which is the work of DeepMind, a British artificial-intelligence (AI) research firm owned by Google, identifies the patients at greatest risk of a sudden and fatal loss of kidney function. The Royal Free says that the app is already saving nurses’ time.

Naturally enough, the health-care entrepreneurs have the boldest visions. The point of care will move rapidly into the home, they say. People will monitor their heart conditions, detect concussions, monitor the progress of diseases and check up on moles or ear infections using apps, mobile phones and sensors. Last year the FDA approved 36 connected health apps and devices. A new app, called Natural Cycles, was recently approved in Europe for use as a contraceptive. Its failure rate for typical use was equivalent to that of popular contraceptive pills. A smartphone may eventually be able to predict the onset of Alzheimer’s, Parkinson’s or even the menopause (if the information is wanted).

In emerging economies, where regulations on health data are less onerous and where people often already expect to pay to see a doctor, there is faster growth and innovation. China, which is building 400 hospitals a year, saw its two largest VC investments in digital health care last year. One went into a Chinese medical-service app, Ping An Good Doctor, which raised $500m; a video-consultations app called Chunyu Yisheng raised $183m. India is another innovator. To take one example, LiveHealth, based in Pune, is an app that lets patients assemble all their health records in one place, see test results and communicate with doctors.

In the short term, the greatest disruption will come from a growing array of apps in many countries around the world that give consumers direct access to qualified GPS on their mobile phones. Overall, telemedicine is expected to grow rapidly. In America, GPs will conduct 5.4m video consultations a year by 2020, says IHS Markit, a research firm. Britain’s NHS is testing a medical AI from a London-based startup called Babylon which can field patients’ questions about their health. A paid service called Push Doctor offers an online appointment almost immediately for £20 ($24). The firm maximises the efficiency of its doctors by reducing the time they spend on administrative duties. They spend 93% of their time with patients compared with only 61% in Britain’s public sector. Babylon reckons that 85% of consultations do not need to be in person.

In the longer term, the biggest upheaval may come from the large technology firms. Amazon and Google are not the only giants to be stalking health care. Apple has expressed a strong interest in it, though it is taking time to decide exactly what it wants to do. For several years it has provided a way of bringing together health data on its iPhone, and tools for health researchers to build apps. As personal-health records accumulate on its platform, from sensors such as Fitbits to medical-grade devices, it will encourage more app development.

An app using data from an iPhone or another smartphone might be able to warn users that a sedentary lifestyle will exacerbate a heart condition or that, based on social-media patterns, they are at risk of depression, for example. Apple and other tech firms may also be able to help patients take greater control of their existing health records. For now medical records mostly remain under the guard of those who provided the care, but this is expected to change. If patients do gain proper access to their own data, Apple is in a particularly strong position. Its platform is locked and fairly secure, and the apps that run on it are all screened by the firm.

None of this will materialise quickly. Regulated health-care systems will take time to deal with concerns over accuracy, security and privacy. In Britain the Royal Free is already under scrutiny over how it shared its patients’ data. That suggests a broader worry: that technology companies are too cavalier with their users’ data. Such firms typically use long agreements on data rights that are hard for individuals to understand. The medical world places importance on informed consent, so a clash of cultures seems unavoidable.

Yet enormous change looks inevitable. Investors hope for billion-dollar health-tech “unicorns”. Payers eye equally sizeable savings. Amid such talk it is worth remembering that the biggest winners from digital health care will be the patients who receive better treatment, and those who avoid becoming patients at all.