Provide your email for confirmation

Tell us a bit about yourself

country *
province *

why we ask about location

Please provide your email address


To share your thoughts

Don't have an account?

Login with email

Check your inbox

We just sent a link to your inbox. Click the link to continue signing in. Can’t find it? Check your spam & junk mail.

Didn't get a link?

Sign up

Ready to get started

Already have an account?

Sign up with email

By signing up you agree to Rappler’s Terms and Conditions and Privacy

Check your inbox

We just sent a link to your inbox. Click the link to continue registering. Can’t find it? Check your spam & junk mail.

Didn't get a link?

Join Rappler+

How often would you like to pay?

Monthly Subscription

Your payment was interrupted

Exiting the registration flow at this point will mean you will loose your progress

Your payment didn’t go through

Exiting the registration flow at this point will mean you will loose your progress

Rappler Talk: Data and facts – what the PH can do to deal with coronavirus

Bookmark to watch this episode Friday night, March 27

MANILA, Philippines – Rappler CEO Maria Ressa talks to former health secretary and former dean of the Ateneo School of Medicine and Public Health Manuel Dayrit and Erika Legara, one of the country's top complex systems scientists and the Aboitiz Chair in Data Science at the Asian Institute of Management. 

Legara and her AIM colleague, Christopher Monterola, wrote this based on a transparent model: Coronavirus cases in PH could reach 26,000 by end-March if random spread not contained. Dayrit ran the Philippines' lauded reaction to contain the SARs virus between 2002-2003. (Watch: Rappler Talk: ex-DOH chief Manuel Dayrit on containing outbreaks)

In this conversation with Rappler's Maria Ressa, Dayrit and Legara talk about the data we should be monitoring and how the numbers released on a daily basis are actually a measure of cases from at least two weeks back. They also compare the actions taken with what Singapore and South Korea are doing.

Based on data, you'll learn:

For background of global numbers, read Tomas Pueyo on Rappler:

Watch Dayrit and Legara's interview on Rappler.


MR: Thank you so much for joining us today on Rappler Talk. I’m so looking forward to speaking with two experts who I hope will help us make sense of the world we’re living in today. You will see Erika Legara - she is one of the Philippines’ top complex systems scientists. And she is the Aboitiz Chair in Data Science at the Asian Institute of Management. And then of course, a familiar face - former Secretary of Health Manuel Dayrit. He was in charge of the Philippines’ reaction from 2002-2003 to the very first SARS 1, if this pandemic is SARS 2. Welcome, both of you, to Rappler Talk. Erika, I’m going to throw the first question at you, because you’ve been looking at data forever, and at how complex systems revolve around this. What kind of data are you monitoring today?

EL: Right now, we are really very limited to what the government is releasing out there. This means the number of tests conducted - which is given on a daily basis - but we don’t have historical data on that one. We also have the confirmed cases. So I just want to emphasize that this is confirmed and detected. This is not the total number of cases. And of course, the deaths and recovered cases. So these are the kinds of information available to us at the moment, and this is the data that we are currently playing with.

MR: And what meaning do you get out of all of that data? What is the nugget you are looking for? Why should we monitor these?

EL: It’s the same as what we are seeing globally. Of course, the growth trends, which is in my opinion - Secretary Dayrit, please feel free to correct me - the most important at the moment in terms of truly understanding the trajectory of this epidemic. The scale, what is the potential of this epidemic? We can see that from the growth trends of our infected cases. Unfortunately, right now, we only know the reported ones. So that is an important indicator. And right now, Dr. Chris Monterola, who is my colleague, and I are trying to develop a new way to - with this limited information - instantaneously track whether or not certain measures are working, given the current situation.

MR: Many, many other scientists around the world have talked about the exponential growth of this curve, and how globally we’re still not at a point where it has peaked. Is that correct?

EL: We’re not seeing that yet unfortunately, Maria. And most countries are not yet seeing the peak, unfortunately. But we can take a hint from China and South Korea. For example, if you look at China and South Korea, it takes about two weeks to a month before you see the peak. But one of the things that we also need to understand are the measures that they took for them to reach this peak at a much shorter timeline, and then flatten out at about 1-2 months later. But for the Philippines, right now we’re really not seeing it yet. We need more data. And when I say more data, we need more tests.

Imagine this, I’m pretty sure a lot of us have read Tomas Pueyo. Tomas gave this example. If you have one death today, and the mortality rate conservatively is at 1%, it means that this person was infected maybe 20 days or two weeks ago. And 1% mortality rate, which is conservative, means that there were probably 100 infected cases at that time, 14 days ago. So he observed that every 2-3 days, that’s the doubling time - the doubling time of number of infected cases. It means that if you have 100 cases, two days from that day, you’ll have 200. And then two days after, you’ll have 400. The thing is, we’re only testing below 200. So how will you be able to capture that growth if we’re very much limited to the number of tests we’re conducting? So no matter how good we are in tracing the trend - which is the J-curve or hockey stick that they are talking about - if we’re just going to look at the detected [cases] in the Philippines given the rate by which we are testing individuals, I think we are in big trouble.

MR: Thanks, Erika. Dr. Dayrit, former health secretary who really led the charge against SARS from 2002-2003, I’ve heard you talk about how different this is today. And you’re right that the scale - we’ve never dealt with anything like this. What data would you need in order to get a sense of where we are today?

MD: what we actually look for in order to understand how the epidemic is going, is what they call the reproductive rate. The reproductive rate is the number of cases that one infected case might infect. And I know based on talking to some of the people who were involved with inter-agency, that they actually postulated a reproductive rate of 4. Which means that they’re looking at a surge of cases in the level of about 75,000. The other reproductive rate that they considered was actually 1.4, which meant a surge of cases to 20,000. I think they used these reproductive rates - 1.4 and 4 - based on the data they got from China. So I think that’s one of the things we need to understand. And the way that you compute for reproductive rate is actually to collect the data Erika has mentioned, and then plug them into mathematical models. And try to compute what the actual reproductive rates might be. So that’s one of the things that can be done. 

Now I’ve read some of the mathematical projections that are online. And the other statistic that would be of interest is the doubling time. And for example, in one of these models that I saw - I think it was Our World in Data, that was the source - the doubling time for the Philippines was five days. Compare that to Singapore, which had a doubling time of 10 days. And compare that to Italy, which had a doubling time of two days. So these are the statistics that you would be looking at in order to understand how this epidemic is moving. Now, because we’re not doing enough testing… You see, when you do the tests, you actually try to cover as many of the population, as many of the general population who are out there who think they may have encountered somebody who is infected. So you’re just covering your population with tests. And that, in a sense, gives you what epidemiologists would call some ‘prevalence information’ of what is the level of infection in your population. By doing that, you can then have your actual data from these tests, and you can compare that to all the mathematical projections that are being created. Because these mathematical projections still rely on assumptions about your reproductive rates, your doubling times, and so on and so forth. This is the way you would try to map the progress of your epidemic.

MR: One of the things I’ve heard you talk about in the Philippines, and I’ll bring some of the data you mentioned into the real world... Five days doubling time in the Philippines, 10 days in Singapore, two days in Italy - although that also seems to have changed now, right? From Singapore to the Philippines - is that because of the measures that are put in place, or also because of the level of poverty?

MD: I would say Singapore has a doubling time of [10] days because their interventions kicked in early and they were actually able to suppress the infection. In Italy, I suspect what happened there was because it was tourist season, there were just so many people that came in infected, and it just snowballed. And therefore, with that cumulative math of infected, they then just got their epidemic going very, very fast.

MR: In terms of the response rate of getting all of this data out… I guess I’m wondering why we’re not talking about data of health and warning our people. You managed a crisis like this before - what kind of communications can we do with this kind of data? 

MD: You know, I think the government is trying to balance on one hand, giving data, and on the other, trying to prevent panic. And so they’re veering away from large numbers… of saying these are the large numbers, because of the possibility of these being misinterpreted. And so it’s piecemeal, it’s an approach of piecemeal. 

But what I can say is, if you look at the figures they announce on a daily basis, that data is the picture 7-10 days ago. So if we are now at 700 already, that was the figure 10 days ago. But really, what you also have to consider is the rate at which they’re now doing testing. Because their capacity has probably doubled. A week ago, there was only one lab doing the testing, which was RITM. Now, as I understand, there are now three or four labs doing the testing. And they’re probably clearing the backlog which was at the level of about 1,000 samples. So you are seeing now a picture of 10 days ago - plus the backlog, depending on how long that backlog actually was. And therefore, you then compute your projections and say, maybe it’s doubled or tripled. And using the factor that was recommended by Tomas - either do it times 4 or times 8. There were doubling times in that period of 7-10 days. If it was a 10-day period, then you say we doubled four times. So what are our number of deaths? How many?

EL: It’s at 45.

MD: OK, so 45 x 100 = 4,500. And that times 4, that will give you - we’re close to 20,000 cases.

EL: Yes, a little more - which was our initial estimate. Our initial estimate was at 26,000. And Dr. Dayrit is actually correct because when we first released it, if you remember, Maria, a lot of people were like, “Oh, you’re causing undue panic.” But that was what we saw in the trend. And one of the things that I found lacking was that information. We are lucky because we’re always online, we always get to read all of this stuff. But how about the rest of the Filipino people? They don’t have any idea, right? And then what they’re seeing is that the number of confirmed cases is increasing only by the hundreds. But what we know is that these are just the detected. For all we know, we’re now in the tens of thousands. Again, it’s a speculation. But if you will look at data from other countries, that is what the dynamics are telling us. It’s just that we don’t have that much data. 

So one of the suggestions of my brother - who is based in Cagayan de Oro - is to have billboards. I know it will cause panic, but I think we really have to inform the people. Because as early as March 15 or 16, I wasn’t seeing information on the Philippines that was data-driven. That’s why I’m very grateful to Rappler for releasing different models, so that people can start talking about it. I know the numbers are scary - in fact other agencies are giving much higher numbers. And some people just cannot imagine this big of a number, and then they’re like, panicking. But we should panic!

MR: Actually, that’s the reason we published your model, Erika. Because it was one of the first ones. And then a day later, WHO and DOH also published a number without any of the assumptions. From 26,000, which was your model - theirs was at 80,000. Again, as a communications person, I think transparency becomes critical in a time of crisis so that you can pull your people together. Dr. Dayrit, you talked a lot about intensive care facilities. In fact, you were the first person I heard who talked about the fact that private hospitals a week and a half ago were managing 52% of the COVID-19, and yet the government response doesn’t take that into account. And then you also talked about intensive care facilities and ventilators. Please remind us, why is that number critical once you have the other numbers we talked about?

MD: You know, I’m told that there is a figure that’s been released. That there are about 1,500 ventilators in the country at the moment. But they had a computation that the number of serious cases could actually be upwards of 3,500. So if you’re thinking just by averages, we know that we don’t have enough ventilators to treat all of these patients if they came in all at the same time. So these are some of the figures that are very useful that will drive our managers, our hospital administrators, and government to make sure that the resources are there to address these concerns. 

MR: And the basic assumption in this, in managing for the number of ventilators, is that you don’t want to overwhelm the healthcare system to the point that the people who desperately need that ventilator cannot - and that means they will die. Correct?

MD: Yes, that is correct.

MR: In the absence of that 1,500 ventilators, the government has actually talked about creating COVID-19 facilities. What measures do you think will be effective in dealing with the fact that just based on the numbers, we know our intensive care system and healthcare system will be inundated?

MD: The whole idea for creating COVID-19 dedicated hospitals is to make sure that - one, the hospitals that have to deal with a lot of cases as well that are non-COVID, are able to do that. What has been happening now in many of the hospitals that admit COVID patients, they are just getting swamped. And because of the nature of the contagion, they just have to close down - and their facilities - if they are contaminated. And therefore the strategy is to create these COVID-19 dedicated hospitals, put the COVID cases there who need critical care, and hopefully release the other hospitals to take care of non-COVID concerns, which actually happen everyday. Imagine the number of women that need to deliver, that need to go to the hospital, and so on and so forth. That might be put on hold and be jeopardized because of COVID-19. 

MR: I know I promised you both only 20 minutes. We’re at 19. So why don’t I ask you, Secretary Dayrit again, there are many different ways countries around the world have been dealing with the crisis. Can you describe the top three?

MD: Well, the top three would be - one, where they had a multi-layer approach. When I explain it, I say, they had an approach that reached people and families in their household, gave them information there, and explained what behavior they needed to do. And then they had the primary care layer. And these are the primary care physicians in the community, who were able to manage it so that they didn’t swamp their hospitals. And when their hospitals actually saw patients, they were able to treat them very well. No deaths in Singapore. 

I guess you’d put South Korea as another model for this. And essentially, what they deployed was really widespread testing in their community, really to find people who were positive, and to eventually get these people to quarantine and isolate themselves, and prevent the infection from spreading. And therefore their reproductive rate went down very low as a result of this strategy.

And I guess the third example would be China itself - that did all of these draconian measures, including shut down and really massive suppression of the population so that the virus wouldn’t spread. So these three are examples we can look up to and emulate or borrow from for the situation that we have.

MR: Our response to this now as a nation seems to be a combination of all of that. I mean, in a way, it is good to put a lockdown in place, but the chaos of the way it was executed also meant that a lot of people with the virus left Metro Manila in the two days after March 12, right? Erika, let me ask you - looking at the numbers, and in a complex system, with these kinds of exponential growth, what do you think are the possible mitigating factors that could walk in to try to find the turn of the curve?

EL: You already articulated it. It’s about delaying, to be honest, the number of infections going to the hospitals. It’s about flattening the curve. And the social distancing measures should be done in a proper way. One of the things that’s really bothering me at the moment, Maria and Dr. Dayrit, is that when it was announced on the 15th that we’re going to have a quarantine and they’re going to lift this on the 12th of April, my question was - what’s the basis of that? Would we be ready by then? I’m willing to stay longer, to be honest, but we also need to make sure that we are balancing things. Like physical health, the mental well-being of individuals, and also - not a lot of people appreciate this aspect - the economy. But what is the right date? Without data, without data-driven decision-making processes, things could get even worse for us. So, yes we should flatten the curve, but for how long? They say 1-2 weeks should be enough, but we need to see that working. And the only way for us to really see that working is if we have enough data to give a more honest assessment, to have a baseline that such measures are truly working.

MR: Erika, with that baseline, do you think you could factor in the rate of poverty in the Philippines? I mean, could we do some kind of an urban… I mean, the areas where there are people in close quarters, to be able to predict where testing should go? Or isolation?

EL: That’s a very good question, Maria. In fact, the data I was just talking about is just the infected, confirmed - just the cases. But another data that would be useful, like Dr. Dayrit said earlier, is the capacity. And you’re right, we can compare that with the urbanization of a certain locality. So that is a more data-driven way. In fact, we’re talking to one international organization at the moment, and then they’re asking us if we have a wishlist of data coming from the businesses, what would it be? And I said, mobility data. We need to have mobility data because that will really give us… Because we know that people moved out of Metro Manila. Where did they go, right? Where are they now? And how many are they across the Philippines? Which pockets of barangays, of cities? And that will give us a hint on where to deploy most of our resources. 

MR: Like the test kits?

EL: Like the test kits, like the equipment, the facilities, the manpower. If there are volunteers, we will know where to send them out, if we know where the people are.

MR: Dr. Dayrit you talked about… Singapore did have some fatalities despite their very efficient management of it. And part of the reason was because they did control the local rate of transmission in their community, but traveling came in. They had one day where they had 47 cases, and 33 were brought in. That was how efficient they were in tracing all of these. Let me ask Erika’s question to you. If we don’t have baseline data, how will our government manage this? You’re familiar with the way DOH works, the people who have lived through SARS. And it seems like… There are people in our bureaucracy who still have that knowledge. How will we decide on what measures, what data points are being used?

MD: Maria, despite the limitations of our data gathering, we can actually learn from them. And if you look at the data the DOH has released, they have actually indicated where positives have been identified. For example, in ARMM, in Bangsamoro autonomous region. And that was the result of some of our Muslim brothers there traveling to Malaysia and getting infected. I just received a Viber message from the former provincial health officer of Northern Samar. They’ve detected a case there. So we know that our provinces are already being seeded by cases. Our hope is that our local governments are already implementing the type of contact tracing that needs to be done, and quarantining of contacts so that the infection doesn’t grow there. So that’s really the way they’re managing there. But I’ve made the call to local government officials that I’ve been able to talk to, that once they see cases appearing in their hospitals, it might actually be too late. So they have to understand if the cases that they’ve seen are firstly, imported - in which case, they can do the containment measures. 

That’s what happened to Metro Manila. The first three or four cases were actually imported. But after a period of a month, infection crept in under the radar, and eventually it exploded starting March 5 and 6 - because there was already a build-up of those cases in Manila, and it hasn’t stopped. And given the backlog in our testing, every day we announce these new cases as if they’re growing and growing, but that’s actually also a result of the backlog of tests. So we know that the virus is already out there in the provinces. And therefore, that’s one message that has to go out. That the virus is there, and they have to do the proper containment so that it doesn’t spread.

Let me address that issue of - when is it going to peak? If we’re assuming that we’re now at 16,000-20,000 today, and if you’re assuming doubling times of every 5 days… we’re going to reach 75,000 in 15 days. Which is what the model stated. And that’s when our lockdown period ends.

EL: Yeah, correct.

MD: So it just tells us that this is going to go on, and government will have to figure out what we’re going to do at the end of this community quarantine period, given what the numbers are telling us. Now, we hope… Because the other question is, how do we know that our community quarantine efforts are working, right? Given that our data is late 7-10 days. Therefore, we can only… The number crunchers are going to have to figure out what were the cases that were occuring before the lockdown period started, and what are the cases after the lockdown period - those computations only being able to be made after three or four weeks after the lockdown period. I think that’s the other thing that’s interesting to actually monitor. The ability to monitor what’s happening in the lockdown period. So I would say, clear the backlog so that the cases that we’re reporting - we know are already at least 7 days late - would provide us an even more accurate picture, even if late. Make the projections as a result of that, and then from there, work out the plans. Because from what it looks like, you know, the cases are still going to go beyond maybe even that surge of 75,000 that was actually predicted by the WHO model.

MR: Wow. Thank you for that. It’s quite alarming, but very great advice. Erika, your last thoughts?

EL: Yeah. We are now living in this day and age of data. We should take advantage of it. We should not be afraid of this. If we want a more objective way of approaching things… We are dealing with a complex system wherein if we do something wrong in one entity, it’s going to have some cascading effects. So we really have to look at data, we cannot downplay this. Have more data-driven decision-making. That’s what I want to push right now.

MD: I’d like my last thought to be positive. One, we’ve moved on the lab testing issue. We now have more tests being done. I’d like to see the backlog cleared and eventually more testing done in the population. I hope that can be done within the next two weeks. Number two, I hope that in the next two weeks, we’re actually able to set up the COVID-19 dedicated hospitals - in Metro Manila as well as in the provinces. So that whatever critical cases occur, they can be treated well. And I guess my third closing remark would be my thoughts for the frontline health workers. That it is imperative that they be given all of the tools and all of the weapons - including the personal protective equipment - to be able to do their job properly so that they’re not exposed and that [they do not] eventually die from this disease. 

EL: I just want to add to what Dr. Dayrit is saying. It’s not all grim. In fact, just two days ago, we’ve started working with a team of scientists - both local and abroad - who are working closely with the DOH. I’m really keeping my fingers crossed. We worked all night last night to build some prototypes but I cannot announce yet the details because I will leave it up to the government to do that. Yeah, in 2-3 weeks, I really, really hope that we will see significant changes with how things are moving.

MR: Fantastic. Thank you both for your thoughts, your analysis, and your positivity in a very difficult time. Thanks again for joining us, and hoping for better days ahead.