WEBVTT - The Roadblocks to Mass Testing 0:00:15.396 --> 0:00:22.036 Pushkin from Pushkin Industries. This is Deep Background the show 0:00:22.076 --> 0:00:25.196 where we explore the stories behind the stories in the news. 0:00:25.716 --> 0:00:28.996 I'm Noah Feldman. A few weeks ago on the show, 0:00:29.276 --> 0:00:33.076 I spoke to Nobel Prize me economist Paul Romer, and 0:00:33.196 --> 0:00:36.596 Paul had what he considered a simple plan to reopen 0:00:36.636 --> 0:00:40.716 the economy without risking people's health. All we need to 0:00:40.796 --> 0:00:44.556 do is switch to a strategy where we're testing everybody 0:00:44.716 --> 0:00:48.156 with regularity. As soon as we find somebody who's positive, 0:00:48.476 --> 0:00:52.596 we need to isolate them without isolating lots of people 0:00:52.636 --> 0:00:55.676 who could otherwise just go back to daily life and work. 0:00:55.956 --> 0:00:59.076 Paul is not alone. Public health experts have been almost 0:00:59.196 --> 0:01:02.436 unanimous in saying that we need a lot of tests 0:01:02.596 --> 0:01:06.036 very soon in order to protect health and eventually reopen 0:01:06.076 --> 0:01:10.556 the economy. And yet we are behind Germany, Canada, and 0:01:10.636 --> 0:01:14.116 even Italy when it comes to per capita testing. And 0:01:14.276 --> 0:01:16.956 as of now, less than one percent of the US 0:01:17.076 --> 0:01:21.596 population has been tested for coronavirus. So what is the 0:01:21.636 --> 0:01:25.156 hold up? What are the bottlenecks that stand between us 0:01:25.476 --> 0:01:29.756 an effective testing? To learn more about coronavirus testing both 0:01:29.796 --> 0:01:32.836 for diagnosis and then for antibody testing to see if 0:01:32.836 --> 0:01:36.156 people have already had the disease. I'm joined by doctor 0:01:36.196 --> 0:01:40.796 Ohmi Garner. He's the director of Clinical Microbiology Testing for 0:01:40.996 --> 0:01:44.316 UCLA Health. His lab does more than one point five 0:01:44.556 --> 0:01:47.996 million tests a year of all kinds, and so he 0:01:48.036 --> 0:01:51.596 finds himself right in the thick of the question of 0:01:51.756 --> 0:01:57.196 how we test for COVID. Oh my, I'm so grateful 0:01:57.236 --> 0:01:59.676 to you for agreeing to talk to me. I want 0:01:59.716 --> 0:02:03.756 to ask you to help guide us through the process 0:02:03.756 --> 0:02:08.796 of just why testing takes so long to generate and 0:02:08.836 --> 0:02:12.916 to cause to function. So start wherever you want. Maybe 0:02:12.956 --> 0:02:16.276 start with the diagnostic tests first. What are the challenges 0:02:16.596 --> 0:02:19.956 to having millions of diagnostic tests up and running on 0:02:19.956 --> 0:02:22.876 a daily basis? You know, that's the million dollar question. 0:02:23.116 --> 0:02:26.956 I think that there are lots of challenges for many ends, 0:02:27.476 --> 0:02:29.996 and so just I think it's used to a little 0:02:29.996 --> 0:02:33.316 bit walk through the process of getting a diagnostic test 0:02:33.796 --> 0:02:37.236 for COVID nineteen by PRECR because it starts to then 0:02:37.316 --> 0:02:41.276 outlay where the challenges are in the system. And so 0:02:41.476 --> 0:02:43.756 the first part of the diagnostic test is the collection 0:02:43.796 --> 0:02:45.996 of the sample. This is the part I think people 0:02:45.996 --> 0:02:48.916 are most familiar with. You either go to your doctor's office, 0:02:48.996 --> 0:02:52.036 or a hospital or a drive through location, and a 0:02:52.156 --> 0:02:56.476 swab is inserted deep into nasal cavity and then that 0:02:56.676 --> 0:02:59.836 sample is sent to a centralized testing lab. I think 0:02:59.836 --> 0:03:01.916 there's a bit a little bit of confusion around this 0:03:02.396 --> 0:03:06.596 because they call those locations testing locations, but no actual 0:03:06.636 --> 0:03:09.876 testing happens there, and so I think people think, why 0:03:09.876 --> 0:03:12.076 would it takes so long if the test happens right 0:03:12.076 --> 0:03:14.396 where you collected it. So the sample needs to be 0:03:14.436 --> 0:03:17.876 sent to a centralized laboratory like my laboratory where we 0:03:17.956 --> 0:03:21.876 perform the PCR test, and the PCR test, depending on 0:03:21.916 --> 0:03:25.956 which platform is used, can take anywhere from two to 0:03:25.996 --> 0:03:28.916 six hours. And I think this is also something that's 0:03:28.916 --> 0:03:31.636 not well known because if the test only takes two 0:03:31.636 --> 0:03:34.396 to six hours, why are people waiting seven to fourteen 0:03:34.476 --> 0:03:36.876 days to be able to get a test result right? 0:03:36.916 --> 0:03:41.596 And I think one of the infrastructure challenges around this 0:03:41.676 --> 0:03:45.276 particular outbreak and testing itself, is that there are just 0:03:45.436 --> 0:03:49.076 not enough centralized laboratories that are able to do this 0:03:49.156 --> 0:03:51.996 testing right now. And this starts to key in on 0:03:52.076 --> 0:03:55.436 why we don't have testing across the country and a 0:03:55.436 --> 0:03:57.396 lot of and we have labs, so it's not to 0:03:57.436 --> 0:04:01.556 say that there aren't centralized clinical laboratories. There are centralized 0:04:01.556 --> 0:04:04.716 clinical laboratories in almost every single city across this country. 0:04:05.156 --> 0:04:08.276 The challenges that each one of those laboratories does not 0:04:08.436 --> 0:04:12.156 have the equipment to be able to test for COVID nineteen, 0:04:12.596 --> 0:04:15.476 and the labs that do have the equipment there is 0:04:15.516 --> 0:04:19.476 such a shortage on they test themselves that many labs 0:04:19.476 --> 0:04:22.756 that have the equipment still can't run the capacity of 0:04:22.796 --> 0:04:25.916 tests they could run within a day. One of the 0:04:25.956 --> 0:04:29.796 fascinating questions to me is is there some alternative technology 0:04:29.876 --> 0:04:33.316 potentially in the pipeline they would make it easier to 0:04:33.356 --> 0:04:37.716 collect samples by means other than the long swab that 0:04:37.756 --> 0:04:40.636 goes deep into the nasal cavity. You know, Donald Trump 0:04:40.676 --> 0:04:42.596 himself said it was a miserable experience for him to 0:04:42.596 --> 0:04:44.396 have that done, And though I'm not super worried about 0:04:44.476 --> 0:04:47.676 his own experience, it does mark the fact that we 0:04:47.716 --> 0:04:50.116 need not only the swabs, that we need medical professionals 0:04:50.116 --> 0:04:52.036 to do the swabbing, Whereas it might be a lot 0:04:52.036 --> 0:04:53.556 faster and more efficient if people could do it at 0:04:53.596 --> 0:04:55.876 home or there was just alive a test right now, 0:04:55.876 --> 0:04:57.596 why is it the case that we can't do that? 0:04:58.116 --> 0:05:00.236 I think that a lot of this is about clinical 0:05:00.276 --> 0:05:03.636 sensitivity of what you're collecting. And when I say that, 0:05:03.716 --> 0:05:06.916 I mean how likely is it for a false negative 0:05:06.956 --> 0:05:09.436 to be given to a patient. So in the vironment 0:05:09.476 --> 0:05:11.996 of COVID nineteen, we want to try our best to 0:05:12.036 --> 0:05:14.956 avoid a false negative, and in order to do that, 0:05:15.436 --> 0:05:19.876 you want to take the best possible specimen to increase 0:05:19.916 --> 0:05:24.036 your likelihood of actually collecting virus. Now this is really 0:05:24.076 --> 0:05:26.716 then the question is where is the virus? Right, Is 0:05:26.716 --> 0:05:28.956 the virus out at the end of the nasal cavity, 0:05:29.116 --> 0:05:31.516 Is the virus very very deep in the nasal cavity. 0:05:31.916 --> 0:05:34.436 Is the virus in the throat? Is the virus in 0:05:34.516 --> 0:05:39.156 different various oral fluid compartments. And the reason why nasal 0:05:39.156 --> 0:05:42.556 pharyngeal collection was the first thing that was used is 0:05:42.596 --> 0:05:46.956 because that's where we know other respiratory viruses live. And 0:05:46.996 --> 0:05:50.316 so whether you're doing a PCR test for influenza or 0:05:50.356 --> 0:05:54.116 a PCR test for something like respiratory sinciitio virus, the 0:05:54.316 --> 0:05:57.956 best possible specimen, meaning the specimen that gives you the 0:05:58.036 --> 0:06:02.116 highest likelihood for recovery of virus, is actually that really 0:06:02.236 --> 0:06:07.156 deep nasal pharyngeal specimen. So then I think ultimately the 0:06:07.276 --> 0:06:09.676 question is do you need to take the best possible 0:06:09.756 --> 0:06:14.236 specimen or in this particular case, with this particular virus, 0:06:14.596 --> 0:06:17.276 can you find an equal amount of virus in some 0:06:17.356 --> 0:06:21.356 of these other specimens? And those studies are ongoing. I 0:06:21.396 --> 0:06:23.556 agree with you if we could just use oral fluid, 0:06:23.596 --> 0:06:26.116 I would change all of my test to oral fluid tomorrow. 0:06:26.356 --> 0:06:28.476 But I won't do it if it's going to mean 0:06:28.556 --> 0:06:32.916 we produce more false negatives. Now let's turn to the lab. 0:06:33.036 --> 0:06:35.276 So you were saying the sample comes to you, it 0:06:35.316 --> 0:06:36.876 reaches you in the lab, and you're going to perform 0:06:36.916 --> 0:06:39.876 a PCR test. What is a PCR test. The test 0:06:39.996 --> 0:06:43.436 is actually in two components, with PCR being the second component. 0:06:43.836 --> 0:06:47.716 So the first component is really it's called RNA extraction. 0:06:48.196 --> 0:06:50.476 And what happens is that when the sample comes in, 0:06:50.956 --> 0:06:54.396 this first step actually takes out all the nucleic acid 0:06:54.476 --> 0:06:57.396 from that sample. And so now what you have is 0:06:57.396 --> 0:06:59.396 instead of the full sample from the patient, you just 0:06:59.436 --> 0:07:02.796 have a pool of RNA and then you run the 0:07:02.836 --> 0:07:06.276 PCR test. And so the PCR test is called an 0:07:06.396 --> 0:07:11.676 rtPCR test. It's a reversed trend description polymerase chain reaction 0:07:12.076 --> 0:07:15.356 that's that rt PCR, and what it does is because 0:07:15.356 --> 0:07:19.996 we're looking for RNA, PCR is a technology that examines DNA. 0:07:20.116 --> 0:07:21.596 So the first step is you have to turn the 0:07:21.716 --> 0:07:25.676 RNA into DNA, and that's that reverse transcription. The second 0:07:25.716 --> 0:07:28.476 test then is the polymerase chain reaction, and the polymerase 0:07:28.556 --> 0:07:33.396 chain reaction is really a way to amplify a specific 0:07:33.556 --> 0:07:37.356 target on DNA to see whether or not that target 0:07:37.476 --> 0:07:39.716 is there. And of course the target we're looking for 0:07:40.316 --> 0:07:43.436 is COVID nineteen, and so if some of that viral 0:07:43.596 --> 0:07:46.556 RNA is there, it's been converted to DNA, and then 0:07:46.716 --> 0:07:49.876 PCR can target to tell you whether or not that 0:07:49.996 --> 0:07:54.516 viral RNA was originally in that specimen. And it's exquisitely sensitive. 0:07:54.516 --> 0:07:57.236 The test that we're using in my laboratory get down 0:07:57.276 --> 0:08:01.916 to about five hundred copies of virus per mill of fluid, 0:08:02.396 --> 0:08:04.276 and so what I'd like to tell people is that 0:08:04.356 --> 0:08:07.116 if the virus is in the sample, the test will 0:08:07.156 --> 0:08:10.556 find it. That's super clarify and helpful. Now, you said 0:08:10.596 --> 0:08:14.756 that there are enough laboratories in the United States to 0:08:14.796 --> 0:08:20.556 handle even a substantial volume of testing, and that the 0:08:20.636 --> 0:08:23.836 problem is that they don't have the necessary equipment in 0:08:23.916 --> 0:08:26.796 a concrete sense, what is missing in these labs, because 0:08:26.796 --> 0:08:28.236 if we could figure out what that is, maybe we 0:08:28.236 --> 0:08:31.316 could talk about how we provide it. Sure So, there 0:08:31.356 --> 0:08:35.876 are now I don't know somewhere around twelve FDA Emergency 0:08:35.996 --> 0:08:41.836 Used Authorization approved PCR tests for COVID nineteen. The challenge 0:08:41.876 --> 0:08:45.236 is that the manufactures of those tests need to get 0:08:45.316 --> 0:08:49.196 those tests to those laboratories to be able to provide testing. So, 0:08:49.676 --> 0:08:52.876 giving an example from my own laboratory, I actually run 0:08:53.116 --> 0:08:57.276 four different FDA approved tests for COVID nineteen. And the 0:08:57.356 --> 0:08:59.596 reason why I do that is because I can't get 0:08:59.756 --> 0:09:04.236 one manufacturer to give me enough volume of test kits 0:09:04.916 --> 0:09:07.156 to meet the need, so I actually have to bring 0:09:07.196 --> 0:09:09.276 in four different tests to be able to do that. 0:09:09.556 --> 0:09:11.676 You get a couple one hundred a day from one place, 0:09:11.716 --> 0:09:13.956 you get a couple one hundred a day from another place, 0:09:14.196 --> 0:09:17.516 and you combine all of that volume together and I 0:09:17.516 --> 0:09:20.396 can get up around the thousand tests a day or 0:09:20.396 --> 0:09:23.356 so that I can offer in my laboratory. This is 0:09:23.756 --> 0:09:28.316 a significant challenge and the shortage by the manufacturers of 0:09:28.396 --> 0:09:32.356 the diagnostic tests really is contributing to our inability to 0:09:32.476 --> 0:09:35.676 have widespread testing. And I don't want to put the 0:09:35.716 --> 0:09:38.636 blame on the manufacturers for this. What they've been asked 0:09:38.796 --> 0:09:41.916 is very very difficult, which is to basically pivot their 0:09:42.076 --> 0:09:45.956 entire operation. And these are diagnostic manufacturers that make all 0:09:46.036 --> 0:09:51.116 kinds of diagnostic tests, HIV viral load tests, goneria, clamydia 0:09:51.236 --> 0:09:54.196 PCR tests, and they're saying to them, Okay, we need 0:09:54.236 --> 0:09:57.716 you to make a COVID nineteen test and ramp it 0:09:57.836 --> 0:10:01.276 up a hundred times more the volume you would normally 0:10:01.356 --> 0:10:04.956 make for testing. What do you think is slowing them 0:10:04.996 --> 0:10:07.556 down in doing that? I mean, just at the most 0:10:07.636 --> 0:10:11.156 basic level, picturing the factory where they make the tests, 0:10:11.476 --> 0:10:15.196 and I'm picturing them shifting over the functionality they usually 0:10:15.196 --> 0:10:18.676 have to produce other kinds of tests onto producing COVID 0:10:18.756 --> 0:10:23.076 nineteen tests. What's the hold up for them that stops 0:10:23.076 --> 0:10:25.916 them from doubling or tripling, to say nothing of going 0:10:25.996 --> 0:10:28.396 up to one hundred times. So I think one of 0:10:28.396 --> 0:10:32.436 the hold ups is just manufacturing capacity, because you can't 0:10:32.556 --> 0:10:35.636 just stop making the other tests because people still have 0:10:35.716 --> 0:10:39.076 those other diseases, and so it's almost like you need 0:10:39.116 --> 0:10:42.996 to on top of what you were making before. Now 0:10:43.076 --> 0:10:47.396 go ahead and produce one hundredfold times COVID nineteen testing, 0:10:47.756 --> 0:10:50.636 and so you know, I really think it's a manufacturing bottleneck. 0:10:50.676 --> 0:10:53.756 You can only do so many runs at a time 0:10:54.036 --> 0:10:58.276 because you only have so much manufacturing capacity in that setting. 0:10:58.876 --> 0:11:02.316 In speaking with the manufacturers, this is what I've been told. 0:11:02.676 --> 0:11:05.516 The other part of the problem that we've discovered kind 0:11:05.516 --> 0:11:08.596 of over time is that a lot of these manufacturers 0:11:08.716 --> 0:11:12.396 rely on some of the same chemicals to make their 0:11:12.396 --> 0:11:17.236 own proprietary test kits. And if one chemical, let's say 0:11:17.276 --> 0:11:22.316 that multiple manufacturers use, is in shortage, that then stops 0:11:22.476 --> 0:11:25.876 all of them from making more tests. So there's a 0:11:25.876 --> 0:11:29.156 whole supply chain here of the chemicals that are necessary 0:11:29.196 --> 0:11:32.316 to make this work. And so one has to go 0:11:32.356 --> 0:11:34.756 back a further step and go to the chemical companies 0:11:34.756 --> 0:11:37.556 and make sure the chemical companies are manufacturing more of 0:11:37.596 --> 0:11:40.756 this in terms of stopping though the production of their 0:11:40.836 --> 0:11:44.996 other tests can't. An argument be made that given the tremendous, 0:11:44.996 --> 0:11:48.716 almost unimaginable cost of keeping our economy shut down, and 0:11:48.796 --> 0:11:51.916 given the testing is so crucial to reopening, that they 0:11:51.916 --> 0:11:55.476 actually should stop manufacturing the other tests rely on whatever 0:11:55.556 --> 0:12:00.356 backlog they have and just prioritize COVID nineteen tests ahead 0:12:00.356 --> 0:12:03.996 of everything else. I can agree with you there. You know, 0:12:04.036 --> 0:12:06.356 I don't work for the companies. I can only speak 0:12:06.476 --> 0:12:09.836 as an end user that's in an academic medical and 0:12:09.956 --> 0:12:13.596 running these tests. That yes, I think that the pivot 0:12:13.636 --> 0:12:15.636 needs to be fast. I think part of the challenge 0:12:15.636 --> 0:12:18.636 as well, though, is that these companies aren't made to 0:12:18.756 --> 0:12:22.036 pivot like that, and so you know, getting the either 0:12:22.156 --> 0:12:26.276 federal support or whatever would be necessary to encourage them 0:12:26.316 --> 0:12:29.316 to be able to do this is one of the challenges. 0:12:29.356 --> 0:12:32.196 But yes, I will agree with you. Overall, to meet 0:12:32.276 --> 0:12:35.876 this immediate need, there needs to be a shifting, and 0:12:35.916 --> 0:12:37.916 I know a lot of companies are shifting. It's just 0:12:38.316 --> 0:12:40.916 the scale of the shifting. You know, when we think 0:12:40.916 --> 0:12:44.076 of clinical lab testing. You know, there has never been 0:12:44.116 --> 0:12:46.676 a test in my laboratory where I need to do 0:12:46.756 --> 0:12:51.396 two thousand tests a day. So the scale of this 0:12:51.556 --> 0:12:56.876 is just it's mind boggling from an overall diagnostics perspective, 0:12:56.876 --> 0:12:59.636 because I don't want to represent it as something like, well, 0:12:59.676 --> 0:13:02.236 the manufacturers just should have shifted and this would have 0:13:02.276 --> 0:13:05.036 been relatively easy for them to do, and they're negligent 0:13:05.076 --> 0:13:07.676 to not do that. That's not what's going on. This 0:13:07.876 --> 0:13:11.556 is really an unpress sedented shift that's being asked for, 0:13:11.876 --> 0:13:13.876 Yet at the same time we have to be able 0:13:13.916 --> 0:13:17.276 to do it in your lab. The PCR test, if 0:13:17.316 --> 0:13:20.196 you have the equipment, can be run in two to 0:13:20.276 --> 0:13:23.956 six hours. So what is driving the backlog that causes 0:13:23.996 --> 0:13:25.716 people to have to wait seven to ten days for 0:13:25.756 --> 0:13:28.156 a result. A lot of that is that every single 0:13:28.276 --> 0:13:32.596 state doesn't even have large scale testing that's available. One 0:13:32.596 --> 0:13:37.556 of our reference laboratories in California, Quest Diagnostics was receiving 0:13:37.676 --> 0:13:41.196 samples from New York and New Jersey. Well, if you're 0:13:41.316 --> 0:13:44.356 shipping samples across the country to be able to have 0:13:44.476 --> 0:13:46.636 them tested at a facility, and then once they get 0:13:46.676 --> 0:13:49.956 to that facility, if the queue or line is two 0:13:50.076 --> 0:13:53.636 hundred thousand tests long, you can see that it just 0:13:53.836 --> 0:13:58.796 increases exponentially the turnaround time, which is the expression that 0:13:58.796 --> 0:14:01.156 we use in lab diagnostics for the amount of time 0:14:01.156 --> 0:14:04.076 it takes from the sample to be collected all the 0:14:04.076 --> 0:14:07.116 way to the person getting the result back. Is there 0:14:07.156 --> 0:14:11.796 any in centralized national planning for where tests should go. 0:14:11.956 --> 0:14:15.196 I mean, it seems very crazy that someone could be 0:14:15.236 --> 0:14:17.636 tested in New York and then have their samples sent 0:14:17.716 --> 0:14:20.596 to LA so that a lab there can do the work. 0:14:20.796 --> 0:14:22.676 But I get that you have to send the test 0:14:22.716 --> 0:14:24.676 where there's access. It sounds like the kind of thing 0:14:24.716 --> 0:14:28.396 which would benefit from a centralized model. I agree with you. 0:14:28.476 --> 0:14:31.436 I think a centralized model would be helpful. But you know, 0:14:31.476 --> 0:14:34.116 I think that this is tied into our healthcare system, 0:14:34.356 --> 0:14:37.796 and our healthcare system is not built around a centralized model. 0:14:37.876 --> 0:14:39.796 And so I think this is why you see some 0:14:39.876 --> 0:14:43.356 of the great disparity across the country is because there 0:14:43.476 --> 0:14:47.796 is no centralized model, and thus individual areas some can 0:14:47.836 --> 0:14:50.716 be very very fast, and some may not have access 0:14:50.756 --> 0:14:54.276 to testing at all. What realistically is going to happen 0:14:54.476 --> 0:14:58.236 in your view, if in the next few weeks, you know, 0:14:58.356 --> 0:15:01.196 sort of like end of April, first few weeks of May, 0:15:01.796 --> 0:15:04.876 we see an effort to get people back to work, 0:15:05.356 --> 0:15:08.516 coupled with big companies trying to get tests done for 0:15:08.516 --> 0:15:13.236 their employee eventually over the course of the summer, campuses 0:15:13.276 --> 0:15:16.236 like university campuses trying to get people tested in a 0:15:16.236 --> 0:15:20.876 systematic way is not at all realistically doable from your perspective, 0:15:21.156 --> 0:15:23.636 Given where we stand, I mean, you're perfectly placed to 0:15:23.676 --> 0:15:26.036 give a credible answer to this question because most people 0:15:26.076 --> 0:15:32.556 are just speaking theoretically. Yeah, I'm optimistic. So within that optimism, 0:15:32.556 --> 0:15:35.836 and again I talk to multiple manufacturers of these tests 0:15:35.836 --> 0:15:39.796 on a daily basis. Everybody is ramping up. And as 0:15:39.836 --> 0:15:42.956 everybody ramps up, I see more and more hospitals, even 0:15:42.996 --> 0:15:46.996 just in the Los Angeles area, being able to provide testing. 0:15:47.436 --> 0:15:50.076 And so I think we're moving in the right direction. 0:15:50.236 --> 0:15:52.756 The question is are we moving in a direction fast 0:15:52.916 --> 0:15:55.796 enough to match what we're going to do with changes 0:15:55.836 --> 0:16:00.636 in social distancing policies? And that's a really difficult thing 0:16:00.756 --> 0:16:03.516 to be able to predict because I think the two 0:16:03.596 --> 0:16:07.636 need to be tethered together. As we increase our testing capacity, 0:16:07.676 --> 0:16:10.916 I think that's one way for us to start in 0:16:10.956 --> 0:16:16.156 a responsible way, opening back up, getting people back to work. 0:16:16.436 --> 0:16:17.956 But you know, I think if you just flip the 0:16:17.956 --> 0:16:22.396 switch on May fifteenth or whatever arbitrary day that said, 0:16:23.276 --> 0:16:27.596 it's going to so outpace our pace of testing that 0:16:27.596 --> 0:16:30.236 we're going to end up right back where we were. So, 0:16:30.316 --> 0:16:31.916 you know, I think that states that are going to 0:16:31.956 --> 0:16:35.516 be able to do this in a scientific and educated 0:16:35.596 --> 0:16:40.476 way to meet the testing needs while you open up, 0:16:40.836 --> 0:16:43.516 I think that we can be successful at it. We'll 0:16:43.556 --> 0:16:54.676 be back in just a moment. So far, we've been 0:16:54.716 --> 0:16:57.956 talking about PCR testing, which is to see if a 0:16:57.996 --> 0:17:01.036 person from whom example has been taken has the virus now, 0:17:01.676 --> 0:17:05.956 but part of opening up will also be extensive antibody testing. 0:17:06.556 --> 0:17:09.996 Do you guys do that in your lab? So not yet. 0:17:10.116 --> 0:17:13.796 I'm going through the validation process now, and so my 0:17:13.916 --> 0:17:16.756 expectation is that we should go live at UCLA with 0:17:16.796 --> 0:17:20.796 antibody testing, hopefully sometime early to mid next week. I 0:17:20.796 --> 0:17:23.356 do want to make kind of a general comment about 0:17:23.356 --> 0:17:27.356 the antibody testing compared to the PCR though, so I 0:17:27.436 --> 0:17:29.796 said they were like nine to twelve manufacturer of the 0:17:29.836 --> 0:17:34.116 PCR test. Ultimately they're using the exact same technology, even 0:17:34.156 --> 0:17:36.756 though there are small differences between the tests, and so 0:17:36.796 --> 0:17:38.956 you can really see those tests as the same tests. 0:17:39.636 --> 0:17:42.716 So whether I offer a PCR test or hospital down 0:17:42.756 --> 0:17:44.796 the street offers a PCR test, they're all going to 0:17:44.876 --> 0:17:48.996 be kind of the equivalent sensitivity and high quality. Unfortunately, 0:17:49.076 --> 0:17:52.396 that is not true for antibody testing at all, and 0:17:52.436 --> 0:17:54.596 so the challenge of antibody testing is that it was 0:17:54.676 --> 0:17:57.716 unregulated by the FDA to begin with, So there were 0:17:57.716 --> 0:17:59.916 a lot of tests that flooded the market that were 0:18:00.276 --> 0:18:05.396 very very low quality, and every single antibody test manufacturer, 0:18:05.436 --> 0:18:07.236 because I'm now talked to five or six of them, 0:18:07.596 --> 0:18:10.596 are using a different target on the virus to be 0:18:10.636 --> 0:18:14.596 able to look for antibodies, and that distinctly affects whether 0:18:14.716 --> 0:18:17.476 or not these tests are cross reactive, and they produce 0:18:17.556 --> 0:18:20.676 false positives. And so unfortunately, what you're going to see 0:18:20.676 --> 0:18:24.076 with antibody testing is that depending on the platform each 0:18:24.116 --> 0:18:28.396 testing area uses, they could have vastly different results. And 0:18:28.556 --> 0:18:31.276 this makes antibody testing then, in determining kind of what 0:18:31.316 --> 0:18:35.396 it means, very very challenging. So how did you go 0:18:35.436 --> 0:18:37.916 about the process of choosing the approach that you're going 0:18:37.956 --> 0:18:39.996 to use in your lab? So I was looking for 0:18:40.036 --> 0:18:42.756 the best test possible, right, I think the challenge with 0:18:42.756 --> 0:18:44.996 antibody testing is that it needs to be of the 0:18:45.076 --> 0:18:48.396 highest quality if we're going to have any chance of 0:18:48.436 --> 0:18:52.196 trying to establish some level of immunity or even affect 0:18:52.316 --> 0:18:56.036 behavior by having something like a positive test result. And 0:18:56.116 --> 0:18:58.596 so in speaking with the companies, what I really wanted 0:18:58.596 --> 0:19:01.676 to look at was the size of their validation data, 0:19:01.836 --> 0:19:05.636 whether or not they proved non cross reactivity with something 0:19:05.676 --> 0:19:11.076 like seasonal coronavirus, influenza, some other viruses that we all 0:19:11.156 --> 0:19:13.716 have IgG four that you would not want to be 0:19:13.836 --> 0:19:18.236 cross reactive in a COVID nineteen antibody test, And so 0:19:18.356 --> 0:19:21.476 we did a long process of evaluation and then now 0:19:21.916 --> 0:19:24.676 once we've chosen our company, we want to move forward 0:19:24.716 --> 0:19:28.556 with I'm going to do an extensive in lab validation 0:19:28.636 --> 0:19:32.316 with serum that I already have from COVID nineteen positive 0:19:32.316 --> 0:19:35.956 patients and from patients with other viruses to prove that 0:19:35.996 --> 0:19:38.276 it works before I go forward with a test for 0:19:38.356 --> 0:19:42.276 my patients. So you're actually doubling up. First, you're choosing 0:19:42.276 --> 0:19:44.076 what you think is the best test, and then you're 0:19:44.076 --> 0:19:46.556 going to test it yourself in the lab to make 0:19:46.596 --> 0:19:48.756 sure that you have confidence in it before you start 0:19:48.916 --> 0:19:51.236 using it. Absolutely, I think it's the only way in 0:19:51.556 --> 0:19:53.876 the current environment with the amount of anibody tests that 0:19:53.916 --> 0:19:56.316 are out there. It's just I have to be sure 0:19:56.356 --> 0:19:59.196 because clinical decisions are going to be made based on 0:19:59.276 --> 0:20:03.076 these results. Personal decisions are going to be made based 0:20:03.116 --> 0:20:05.076 on these results, and so you know, that's really the 0:20:05.156 --> 0:20:08.876 role of a clinical lab director in choosing the best possible. 0:20:10.636 --> 0:20:13.756 One of the things that has fascinated me is to 0:20:13.796 --> 0:20:17.076 hear about scientists around the country trying to come up 0:20:17.076 --> 0:20:22.156 with outside the box solutions to massively increase testing capacity. 0:20:22.556 --> 0:20:24.876 And one of the most intriguing ones that I read 0:20:24.916 --> 0:20:28.196 about was produced by a group at the Broad Institute 0:20:28.236 --> 0:20:30.276 here in Cambridge, Massachusetts, which is where I'm based, not 0:20:30.316 --> 0:20:33.356 at the Broad but in Cambridge, that was proposing, at 0:20:33.436 --> 0:20:38.236 least theoretically, a massive throughput approach where they would bar code, 0:20:38.836 --> 0:20:45.036 using a crisper like DNA technique, bar code samples so 0:20:45.076 --> 0:20:48.636 that they could then run through sequencers hundreds of thousands 0:20:48.636 --> 0:20:51.276 of tests at a single run. That's obviously a very 0:20:51.276 --> 0:20:54.676 different technology, and it's using very different kinds of methods 0:20:54.676 --> 0:20:58.036 than are used in most clinical labs. As someone who 0:20:58.196 --> 0:21:00.556 does it the old fashioned way, as it were, what 0:21:00.636 --> 0:21:02.716 do you think are the odds of success here? I mean, 0:21:02.836 --> 0:21:04.756 it might be worth trying it even if the probability 0:21:04.756 --> 0:21:07.036 of success is very low, because it will be great 0:21:07.036 --> 0:21:09.196 to test so many people so quickly. What are the 0:21:09.196 --> 0:21:11.716 big challenges that that approach faces in your view? Yeah, 0:21:11.716 --> 0:21:15.236 So the scientists that run these approaches are very talented scientists, 0:21:15.276 --> 0:21:18.756 you know, their research based scientists, and I think that 0:21:18.796 --> 0:21:23.316 they will be able to successfully put together a system 0:21:23.356 --> 0:21:27.676 that could theoretically do one hundred thousand samples. Part of 0:21:27.716 --> 0:21:31.156 what differentiates a research test from a clinical test, something 0:21:31.156 --> 0:21:33.796 that's allowed to be used on patients is that you 0:21:33.876 --> 0:21:38.036 have to prove that it works before you can actually 0:21:38.356 --> 0:21:40.876 use it clinically. And part of the challenge of a 0:21:40.876 --> 0:21:43.516 system like this I lives literally having this exact same 0:21:43.556 --> 0:21:47.236 conversation with some researchers at UCLA, is that in order 0:21:47.276 --> 0:21:49.956 to prove that it works on scale, you need to 0:21:49.996 --> 0:21:54.956 test fifty thousand samples one hundred thousand samples. And the 0:21:55.036 --> 0:21:57.356 challenge of doing that is how do you get those 0:21:57.396 --> 0:22:01.876 samples within a research setting under IRB approval and not 0:22:02.036 --> 0:22:05.196 have it take nine months, a year, a year and 0:22:05.196 --> 0:22:07.876 a half, which is typically what these sorts of things 0:22:07.876 --> 0:22:11.036 would take before or you even began to have enough 0:22:11.156 --> 0:22:15.476 data to submit to the FDA to get approval of 0:22:15.556 --> 0:22:18.636 your new technique. And so, while I do think that 0:22:18.676 --> 0:22:21.196 this could work, I don't know if it's going to 0:22:21.276 --> 0:22:25.396 move in a time frame that's going to make it feasible. 0:22:26.076 --> 0:22:28.596 There are other concerns that I have kind of wrapped 0:22:28.676 --> 0:22:31.356 up in this. A lot of the challenges of let's say, 0:22:31.796 --> 0:22:34.316 the paper that you had talked about, it doesn't use 0:22:34.716 --> 0:22:39.876 rtPCR uses LAMP, which is a different nucleic acid amplification 0:22:39.956 --> 0:22:44.516 technology that can have sensitivity issues. This is also part 0:22:44.516 --> 0:22:47.436 of some other challenges when you pool large things together, 0:22:47.596 --> 0:22:51.836 is that on an individual sample basis, sometimes you're just 0:22:51.956 --> 0:22:54.716 not as good as the gold standard. And these would 0:22:54.716 --> 0:22:56.916 be the things that these places would have to prove 0:22:56.956 --> 0:23:01.196 to the FGA before they got approval to do massive testing. 0:23:01.836 --> 0:23:04.236 In addition, you know, while I respect the fact that 0:23:04.276 --> 0:23:06.956 you can run one hundred thousand samples at one time, 0:23:07.556 --> 0:23:11.236