Big data sets assisted by artificial intelligence technologies could help doctors personalize treatments for coronavirus patients, the CEO of Johns Hopkins Medicine said.Global cases of the virus have topped 10 million, with a quarter of those in the United States, as many states struggle to contain skyrocketing rates of infection following reopenings. Nearly 500,000 people have died as a result of COVID-19. Dr. Paul B. Rothman, MD, is the Dean of the Medical Faculty at Johns Hopkins University and CEO of Johns Hopkins Medicine (JHM), one of the world’s most prestigious medical institutions.A rheumatologist and molecular immunologist by profession, in the past Rothman’s research focused on immune system molecules known as cytokines and the role they play in the development of normal and abnormal blood cells.In an interview with The Media Line, Dr. Rothman discussed the impact that the COVID-19 pandemic has had on the health sector, as well as technologies he believes will provide a revolutionary transformation for patient care in the future.THE MEDIA LINE: What about the COVID-19 pandemic has surprised you? Was there an aspect of the outbreak that caught you off guard?Dr. Paul B. Rothman, MD: The rapid spread of the virus surprised us all. It’s a fairly infectious virus and that really surprised many of us [physicians]. The societal and economic impact of requiring physical distancing – and the decrease of economic activity that surrounds the isolation that is required to slow the spread of it – was much worse than we would have thought.TML: So many aspects of life have been altered because of the virus: the concept of social distancing; the wearing of masks. How much have health services at Johns Hopkins shifted in recent months to remote means and do you believe that this trend of telemedicine is here to stay?PBR: In response to the pandemic we did a couple of things. First of all, we really tried to minimize any non-COVID-19 work that wasn’t urgent, so elective and non-urgent surgeries were delayed. Then, as much as we could, we moved patients online.We went from several dozen telemedicine visits a day to over 5,000 telemedicine visits a day. More than 70% of all visits were via telemedicine and we did that in over just a couple of weeks so it was quite rapid.Why did it take this [crisis] to switch so much of our activity to telemedicine? The truth is it’s because the payment models are part of it. Prior to this, few payers or insurance companies would pay for telemedicine, but because of the emergency most insurance companies and the government (Medicaid) will now pay for it. That really facilitated that rapid change.Telemedicine is much more convenient for patients. They don’t have the hassle of having to find a parking space and wait in a waiting room. For certain activities, it’s fine. It’s not as great for a new patient or a patient you haven’t seen before, where there’s some sort of contact that you want to make with them so that you can help build a relationship. That is much more difficult to do in telemedicine than it would be in person. The largest [problem] is that you can’t do the physical exam, and that’s such a large part of a physician’s repertoire of understanding a patient and disease.So I don’t think we’ll ever totally switch to telemedicine but it will have a role long-term. We’ll have to rethink what our waiting rooms look like. Do we need waiting rooms if a third or 25% of our activity is now telemedicine? Do we need as many exam rooms? Will we have to build new facilities for people to do telemedicine?TML: As of Monday, the US’s seven-day average of daily new COVID-19 cases increased by more than 30% compared with a week ago, according to JHM data. Where do you see this pandemic headed in the United States?PBR: What people outside the US might not understand is that the federal government has given a great deal of responsibility for COVID-19 to the individual states. So what you’re seeing is that different states have dealt with it differently, in terms of how much physical distancing they’ve required, the use of masks, or when they reopened retail.We’re seeing the result of that heterogeneity of public policy both at the state level and down to the county and city level. Now we’re seeing [the coronavirus] spread in many places that loosened their physical distancing regulations.Where is [the pandemic] going to go? We’re going to see hotspots come up and in response, we’ll have to change local regulations until it slows down. The truth is, until we have an effective therapeutic or vaccine, it’s not going to go away.Around the world, several countries have had it under control and now they are finding more cases. This pandemic has had a huge impact on the world and we’ll just have to continue to work hard to control it. [At JHM], we’re just trying to help discover the important therapeutic or vaccine that will help control the disease.TML: Do you think that the world will have some kind of viable treatment by the time there is a second wave?PBR: I’m not sure we can even call it a second wave. We may even have just a continuous movement of waves and virus hotspots. Until we get a [treatment], we’re going to see geographically dispersed increases in cases.What people are trying to do is balance the economic effect of physical distancing rules and regulations--and the harm that that causes to health care – with trying to contain the virus until we get a vaccine.TML: Speaking of social distancing and trying to balance economic interests and health concerns, Israel almost quashed COVID-19 in May and then the country reopened. Now cases are returning to the earlier peak that we had in April. Is there any way to avoid this kind of scenario? Is it just unavoidable?PBR: People don’t know. In some places around the world, they are trying to get on top of hotspots that are developing by using contact tracing and other mechanisms to try to isolate [the spread] in geographic areas. But again, until we get a vaccine or therapeutic we’re going to be battling this virus.TML: How can we improve diagnostic testing and other medical technologies to better deal with a pandemic in the future?PBR: One of the big reasons this is such a difficult virus to deal with is that there is this long pre-symptomatic phase, which is up to four or five days where the patient has minimal, if any, symptoms.One of the difficulties with the qPCR test [quantitative polymerase chain reaction] to diagnose COVID-19 is that it requires a nasal swab and still, at best, it’s probably only picking up 80% of positives. That’s at best; it could be even less. We’re still technologically trying to see if we can get saliva [samples] to be sensitive.There are people who are working on other technologies and we’re still waiting to see if something can try to detect [the virus] earlier. The real problem is that with many other viruses – like influenza – you get symptoms fairly early on so it’s easier to isolate, versus this virus which has a phase where a patient has minimal symptoms.TML: Could Artificial Intelligence (AI) and machine learning help to deal with the coronavirus pandemic eventually? How does JHM hope to use these new technologies in the future?PBR: We have a data platform and we’re using AI and machine learning for COVID-19 data. We’re also trying to collaborate with others to get data in. The biggest question is why some people get very sick and others don’t? It is still a question that no one understands. We know certainly that there are some things such as age OR underlying [conditions] like obesity which could predispose you. There are likely some genetic predispositions and some blood types [that affect severity]. I think working through that requires the analysis of big data and AI, and these will help better predict who will get very sick and who might benefit from different therapeutic interventions.TML: What technology or technologies do you believe will become major game-changers for the health care sector in the coming years?PBR: The ability to analyze big data using precision medicine, machine learning and AI. [We are] trying to analyze big data sets to better divide our patients up so that we can provide more personalized care for them. COVID-19 is a great example; it’s very different in different people. To use big data, to understand that and find sets of patients that will respond similarly to a therapeutic agent is really important. I think that's the future of medicine: the use of predictive analytics to segment patients that will respond in a homogeneous way to a therapeutic.For more stories, go to themedialine.org.