In early March, Manu Prakash, a professor of bioengineering at Stanford University, returned from the south of France, where he’d been visiting collaborators at a marine station. He went straight to his bedroom and into quarantine—which was a good thing, because he promptly came down with the symptoms of Covid-19. He spent the next three weeks there, restless, barely able to see his wife and kids. But for Prakash, the isolation turned out to be useful—a space to start thinking about how to gear up his lab full of tinkerers for Covid-19 response.
Prakash’s lab works in the field of “frugal science,” which is devoted to creating low-cost scientific devices and scaling them up to mass production, mostly for use in economically developing countries. Prakash is best known for the Foldscope, a durable origami microscope that costs about 50 cents to manufacture. It’s become a staple of science classrooms in countries such as India, where Prakash grew up—a window into the microscopic word of fly antennae and plankton. But it’s not simply a curiosity; the device was first conceived as a tool to diagnose parasitic diseases, like malaria, in remote areas where hauling out expensive lab equipment is impractical. Other projects from the lab include a hand-powered centrifuge for processing infectious samples, and methods to safely dispose of those samples.
The lab’s first Covid-19 project was an N95 mask refashioned from a full-face scuba mask. (The concept was fresh in his mind, he says, as he’d been diving in France.) Since March, the mask has been approved as personal protective equipment, or PPE, in France, and about 40,000 have been distributed so far. (In the US, Food and Drug Administration’s authorization to use the masks as an N95 equivalent is pending, though it has gotten the green light as a reusable face shield for medical workers.) Now the team is focused on testing methods, as cases of the disease quietly surge in regions with little access to testing kits, as well as to the machines and chemical reagents required to process them. Even in the US, diagnostic testing has been slow since the start and hindered by bottlenecks at lab testing companies, although a few companies, including Cepheid and Abbott Laboratories, have received emergency approvals from the FDA for faster, point-of-care tests.
This week, WIRED spoke with Prakash by phone about low-cost tests, diagnostic black boxes, and how to make the most of final exams during a pandemic.
This conversation has been condensed and edited for clarity.
WIRED: What was on your mind when you were in quarantine?
Manu Prakash: I had just come back from France, near the border with Italy, so I had seen the international context and was just starting to see what it meant for the US and the rest of the world. I would say that before I left the US I did not anticipate the disease would truly be unchecked. But then starting to see numbers from France and hearing every day that people were positive, and it was becoming clear that this was going to be unchecked.
Then I experienced what it was like to be in the emergency room, and the number of times the doctors came into my room and had to remove their PPE and throw it out and come back in with fresh PPE. It was the early days, so reusing PPE wasn’t common. I was baffled and surprised by how much PPE is actually used.
I underwent a nasal swab, which was very unpleasant, and you start thinking about how you start implementing that elsewhere. You can’t have a health care worker do that in a low-income setting, where there aren’t enough swabs and PPE. People cough and sneeze, and even though everybody [in the ER who] is doing these tests are in full PPE gear, they’re still at risk of infection.
All better now?
Totally recovered. It was scary for a bit in the emergency room, but ultimately good that they sent me back home to recover.
Tell me about the testing strategies you’ve been developing for low-resource settings.
We’ve been asking the question of whether we can reduce the dependence on the supply chain for both PPE and reagents. The intention is to make tests simpler and make them easier for community health workers to use them. We want to make sure all aspects of an assay can be done without the need of a lab facility.
One of the challenges in terms of diagnostics is that all pieces have to be in play before a solution gets implemented. So we have to start with: What does it mean for samples to be collected and processed in a low-resource setting? That means avoiding the swab situation by using saliva samples, which can be collected safely. The processing relates to our prior work with handheld centrifuges that don’t require electricity.
The second thing we’re working on is how do you really make sure that infection control is in place? Once it’s been in the centrifuge, how do you dispose of the potentially infectious sample?
We’ve been hearing about competition for reagents here in the US. What’s the situation like in other countries?
If you’re talking about a small number of tests you can still get access, but if we’re talking about millions of tests per day, that’s a big problem. The supply chain of reagents and how it’s locked in has been really challenging. There are several places where we work, including many countries in Africa, where even if they have the funds, they don’t have access. There’s a lot of politics involved.
Is there a solution to that?
You really have to think about, OK, can these tests be made locally? This is an issue both in the context of doing tests and research for developing the test assays themselves. Even with all the lockdowns, we’ve been able to get a lot of antibodies and reagents from commercial entities for research. But there are a large number of entities in developing countries that are trying to develop assays as well.
It might be possible for you to literally make a test anywhere in the world, but usually it hasn’t been optimized. Most reagents just off the shelf are easy enough to find and they get you so far, but the enzymes and chemicals that make a test really error-proof are behind closed doors. Often, in the act of making a diagnostic test simpler and easy to use, there is an approach that you make it one button and everything else is “magical” inside the box.
It’s a black box, in other words?
Exactly. The diagnostic world operates on the printer cartridge model, where you buy a machine—the printer—and then the reagent cartridges make the cost too high. That’s been a bottleneck in the reality of why molecular assays haven’t scaled up more broadly for all diseases around the world. The way we deliver these assays is a huge part of the challenge. It’s not sustainable on a larger scale.
A cartridge [test] doesn’t make sense right now to test widely in some places, given the cost per patient. It doesn’t make sense for a second-tier city in Bangladesh. The numbers don’t pan out. This is not going to scale unless we actively do something new.
One of the analogies is to go through those types of processes in a short period of time and then open source those reagents completely. I haven’t seen this so far. There is a tremendous amount of work in open-source efforts in ventilators and PPE in the hardware world. I would like to see that translate to the biochemistry world as well. That’s the uphill battle we have to fight.
We’re in this unusual situation where the entire world is competing for resources and tests. Even the rich countries can’t get everything they need. What is that showing us about how the supply chain works?
Covid-19 is a spotlight into the current state of our health care around the world. It seems like a very special situation with the disparity of access to health care and diagnostics. And I mean, yes, the numbers are staggering. But it also highlights on a day-to-day basis the shortage of technologies for diagnosis and the lack of access. It just so happens that usually it’s for other things. You could talk about malaria, you could talk about dengue, or many other infectious diseases where the diagnosis bottleneck has always existed. Now we’re seeing that in the limelight. What an individual can afford for a valuable diagnosis is so different around the world. You really have to adopt these in the context of the market that you’re trying to deliver it to. The Cepheids and Abbotts of the world might not have the appetite for the global market because there’s nobody there to pay.
It’s obviously one thing to design a cheap solution in a Stanford lab, and another to put it out in the world. Did you learn any lessons from getting the Foldscope out as a diagnostic tool?
That’s a beautiful analogy, because one of the goals of Foldscope was for us to first learn how to scale a tool. The way that we applied it is we could test a Foldscope in its full capacity for a few diseases, but our capacity to optimize it and finalize it for all the breadth of diseases it might or might not be applicable to is fairly limited, because we can only do one thing at a time.
We set up a program where any researcher or individual who had ideas about a capability could apply. At this point about 1,000 people have applied for that program and published 400 papers on a broad range of applications. I don’t think of Foldscope as a solution to a problem. It is a means to a solution.
This is where molecular diagnostics needs to go. It’s to have a standardized generic platform where people have the capacity to tune them for their local needs, and that becomes a solution that’s locally provided. That’s a very different business model to what we’ve done in molecular diagnostics.
But what about unanticipated uses, or other problems you might encounter with inexpensive solutions?
That’s always in the back of the mind in terms of rigor and testing and regulatory context. We can propose an implementation, but we can’t dictate the implementation. I can tell you in Madagascar I’ve been looking at this situation and worrying about these cheap rapid assays that come out of China and have questionable validity. The challenge is that people need actionable information. People will implement what they can find. Right now the government can’t wait for better technologies because they have a tsunami, in some sense.
I agree that this is not just about everybody trying to build and perfect their diagnostic test. I think these things have to be in the same framework of regulatory bodies. Usually we try to partner with a local academic organization so that they can take over the regulatory aspect.
In some places they don’t have their own regulatory framework, and that’s worrisome. I was just talking to someone in Kenya—a collaborator on a ventilator project, where we were sending them ventilator designs and they would build replicas of our machines. They said never before in the country has a ventilator been developed locally, so they didn’t have the support from regulators.
It’s a chicken and egg problem. Unless we see growth and development in local contexts, we don’t get the regulatory framework to build things.
How does doing frugal science change in the middle of a pandemic? There are risks to doing things cheaply, especially now.
Everybody—not just us—is operating in a crisis mode in some sense, because things are not under control by any means. That is universal to any approach in science right now. We’re being extremely plastic about our methods, and preprints are coming on board rapidly. But because it’s frugal, it allows us to scale in the hands of lots of researchers around the world. Reproducibility is everything in a response effort like this. You have to build trust.
The other issue is scale. I was teaching a class and gave a hypothetical—but not so hypothetical—Covid-19 diagnostic assay on the final exam. The assignment was to strategize what it would look like to deliver 1 billion tests at a price point at less than $1. If you put a dollar number to something even before you begin to work on an idea, it changes your approach. Most of the time a solution is drawn up without the context of whether you can scale it. And then it’s hard to label it a solution. I think starting from costs has been personally quite valuable because we come up with lots of ideas that go out the window quickly. They’d be fantastic research methods, but they wouldn’t work.
This is not an academic exercise. In these types of times we have to ask what’s the target point—for tests, for PPE. The same goes for vaccines, in a sense. Will they be available to the larger population around the world? Or are we talking about an endemic disease where X percent gets a vaccine and the rest do not?
Did any good ideas come out of the final exam?
I haven’t read through the submissions yet—I’ve been too busy! But I’m genuinely excited to. Honestly, the reason I assigned it is that everybody is talking about how there’s a massive disruption in education, and to a certain extent, that’s true. But the way I look at it is we’re seeing science unfold on the news on a minute-by-minute basis, and there’s a tremendous need now for scientists to engage.
For the people who can—and I understand it’s not true for everybody, because it puts people in economic hardship and challenges with health—but for those who can, it’s a front-row seat into how science unfolds. What if it were dependent on them to come up with the solution? Here’s a scenario where one single assignment could lead to an implementation at a global scale.
WIRED is providing free access to stories about public health and how to protect yourself during the coronavirus pandemic. Sign up for our Coronavirus Update newsletter for the latest updates, and subscribe to support our journalism.
More From WIRED on Covid-19