Imagine a future where you can see the future. Where doctors can look at people and know that something terrible is going to happen and see what they have to do to get out of danger.
That future is now, according to Prof. Ran Balicer, Chief Innovation Officer and Founding Director of the Clalit Research Institute. Speaking Thursday evening at the Health 2040 event in Tel Aviv arranged by Lema’anchem, Balicer said big data, artificial intelligence and proper use of technology and modeling could shift the medical paradigm.
“What would happen if instead of going to your doctor with pain, your doctor would call you and say, ‘The data shows that in the next week you will have a heart attack. You should come in before evening so we can stop that from happening.’
“It sounds crazy,” Balicer said, “but it’s not.”
“What would happen if instead of going to your doctor with pain, your doctor would call you and say, ‘The data shows that in the next week you will have a heart attack. You should come in before evening so we can stop that from happening.’ It sounds crazy, but it’s not.”
Prof. Ran Balicer
Today, 30% of health care is “futile,” Balicer said – meaning there is no value added from the care. Another 45% of necessary interventions are missed because medical teams are too busy to get to them. Moreover, the third highest cause of death in the United States is medical errors – just under heart disease and cancer.
That amounts to more than 250,000 people dying every year from medical errors, the equivalent of eight jumbo jets of people per day.
Balicer said these deaths are not because of negligence, but because “doctors are human.”
He believes that effectively collaborating with computers could help reduce these errors and allow healthcare to shift from supplying patient-level, reactive therapeutics to predictive, proactive and preventative care.
The Clalit health fund
The Clalit health fund, Israel’s largest insurance agency with 4.5 million members (60% of the Israeli population), has been capturing digital patient data for a quarter century. Today, the fund collates patient-level clinical data from community clinics, specialty clinics, hospitals, prescriptions, laboratory and imaging results and demographic data.
Its research institute then leverages these electronic health records and analytical tools to support decision-making and improve care delivery – at the organizational and institutional level, as well as for the individual patient.
“I can look back 10 years,” Balicer said. “I can use that data to look five years ahead. And if we know that people will get worse, we can change the way of giving them care.”
For example, several years ago, Clalit decided to shift the way it contacts members to request they come in for a flu vaccine from being based on only age and alphabetical order to centering on how high-risk an individual is to developing severe disease. A Clalit population-specific predictive model was developed that included examining sociodemographic variables, comorbidities and more.
In another example Balicer had shared previously, the Clalit research institute conducted an evaluation of trends in diabetes prevalence and incidence, intending to prevent the onset of new diabetes cases by targeting prevention at those who are at the highest risk. Instead of using the limited set of internationally defined criteria, Clalit built and tested an internal prediabetes prediction score. The results led Clalit to tell its clinics to call patients in a different order, preventing hospitalizations and even death.
In another example, the Clalit research institute conducted an evaluation of trends in diabetes prevalence and incidence, intending to prevent the onset of new diabetes cases by targeting prevention at those who are at the highest risk. Instead of using the limited set of internationally defined criteria, Clalit built and tested an internal prediabetes prediction score.
The results were incorporated into the fund’s electronic healthcare records system so that primary care physicians could see a high-risk flag for specific patients and provide preventative care.
“WHAT DOES all this mean for doctors who have only minutes to spend with their patients?” Balicer asked. “If a doctor is handed more than 70 points of data, how can a doctor manage it?”
He said that Clalit is now completing an interface that pulls those data points together inside the electronic health record and provides doctors with actionable insights on one screen for each patient who comes in for care.
Over time, the plan is also to incorporate “digital twin” data, for example, to enable doctors to see what treatments were used and effective on patients with similar characteristics to influence care.
Genetic data becoming part of doctors’ tool boxes
Prof. Lina Basel of Beilinson Hospital and Schneider Children’s Medical Center spoke about how genetic data becoming part of doctors’ tool boxes could have a similar effect, helping doctors identify populations that should be given the highest priority.
Another speaker, Prof. Ronit Sacthi-Fainaro of Tel Aviv University, shared her belief that using 3D-printed tumor models can help decide which treatment best suits specific patient tumors. These models, she said, would replace standard research techniques, where cancer cells grow on 2D plastic dishes outside of their adjacent microenvironment, making it harder to determine drug responsiveness.
Potential risks by hackers
However, as healthcare becomes more digitized, it also opens itself up to potential risks by hackers, Gil Shwed, the founder of Check Point, told listeners Thursday evening.
Israel’s Check Point is a world leader in cybersecurity solutions.
The healthcare sector was the most targeted industry for cyberattacks, Shwed said; on average the healthcare sector experienced 1,500 weekly attacks, a 60% increase over the previous year.
Shwed said one of the reasons for this is that the potential to get ransom is easier in the healthcare sector, where not acting fast could be a matter of life or death. Moreover, he said that hospitals, for example, are among the most accessible places to attack because there are so many concurrent electronic systems – from closed to open-loop – running simultaneously.
Shwed said these attacks could come in many forms.
Take the data from an MRI or CT scan, which is generally stored on a hospital’s server. If suddenly there is no access to the server, and a doctor needs to treat someone based on the outcome of his or her MRI or CT scan, the doctor cannot give the patient the correct treatment.
A malware attack that prevented clinicians from accessing medical records could result in canceling appointments, tests and surgeries or force hospitals to divert ambulances to other facilities. In emergency medicine, time often means life.
Last year, Israel’s Hillel Yaffe hospital was hit with a ransomware attack that targeted its computer systems and succeeded in penetrating its IT infrastructure. The hospital was able to take in and care for critical patients but turned away any non-urgent care patients and asked them to seek treatment elsewhere.
Shwed said that while the majority of attacks are driven by criminal causes, there is an increasing reality that state actors and terror groups could engage in cyberwarfare against enemy countries.
In Israel, this could mean not only Russia or China – countries Shwed said are perpetrating attacks worldwide – but also increased targeting by Iran or Israel’s Arab neighbors. He said that most Middle East countries are far behind Israel in their cyber sophistication but that technology is moving fast, and Israel should not assume it will take much longer for them to catch up.
Attacks today are what he calls “fifth generation,” meaning that, like a virus, they can quickly mutate making them hard to identify.
“Today’s challenge is prevention, detection, remediation – prevention versus treatment,” as in medical care, Shwed said.
