Malaria vs. Mobile Phones
December 2, 2015 2:54 PM   Subscribe

 
There's a picture of the lens and a description of it in the second link:

The add-on device, known as a mobile-optical-polarisation imaging device, makes use of a smartphone's camera features to produce high-resolution images of objects 10 times smaller than the thickness of a human hair.

The device images a blood sample using polarised light that can detect Hemozoin crystals, a malaria parasite by-product which appears as very bright dots in the image and is an accurate indicator of infection.

posted by jacquilynne at 3:15 PM on December 2, 2015


Last time I had malaria, I went to a small health clinic in the largest town near where I work. They did a quick blood draw and a tech microscoped the slide. It took about 10 minutes to confirm that I was full of plasmodia. They didn't have microscopes or trained techs at the health clinics in the smaller villages, so people had to walk at least 5 km, as many as 30, to get to this town to be tested and treated. Having an easy to use tool which connects to a smart phone (which I would say about 2/3 of the adults had) and doesn't require skill with a microscope would certainly make it easier and quicker for people to get tested and start treatment quickly. Malaria is the worst.
posted by ChuraChura at 3:26 PM on December 2, 2015 [5 favorites]


A lens that bolts onto a smartphone? Computer image processing and detection? I would be shocked if the false negative/positive rates are not terrible.
posted by Mitrovarr at 3:49 PM on December 2, 2015


Malaria is the worst.

My dad had malaria. Supposedly, that was how he went bald in his twenties: 106 degree fever due to malaria.

So, yeah.
posted by Michele in California at 4:11 PM on December 2, 2015


Malaria symptoms are really obvious, especially if you've had it before as many people in the developing world where this would be beneficial have. I would imagine that in the case of a negative test but positive symptoms (very high fever, general malaise, body aches, chills, throbbing headaches, etc.), they'd do another test, and/or send someone to the nearest town with a tech.
posted by ChuraChura at 4:12 PM on December 2, 2015 [1 favorite]


In addition to all this neat stuff, regular old-fashioned clinical and epidemiological studies have really benefited from participants and patients having cell phone access. Our follow-up is much higher than it used to be, and it makes it possible to find people for whom there is no address or reasonable set of directions.

Where I work in rural Malawi, trying to find households even within 1-2 km of the district hospital - an area which is significantly more built up than areas further away - was fraught. We had 35% attrition from initial study enrollment just because we couldn't find people 3-6 weeks after they had visited the clinic. A study in the same area a year later, we got the cell phone number of each village chief we were working with, and had < 5% attrition. In a study I'm currently working on, it is feasible to get a cell phone number for almost every participant. The rapidity of the change is astonishing.

A lens that bolts onto a smartphone? Computer image processing and detection? I would be shocked if the false negative/positive rates are not terrible.

The quality of diagnostics is a difficult issue to get away from in these contexts. The sensitivity and specificity of microscopy by lab techs, who require training, remains fairly poor. Rapid diagnostic tests require almost no training to use, but are expensive and have poor specificity (ie, they tend to give false positives for infections that have already cleared). PCR is the gold standard, but is expensive, time-consuming, and requires a fair amount of lab training. Here is an open-access paper comparing diagnostic methods in Angola and another from Malawi. The latter gets into issues of who is treated or not based on test results, which, because of the ubiquity of malaria in this part of the world, is still often treated presumptively based on symptoms (ie, irrespective of what the test says).
posted by palindromic at 4:25 PM on December 2, 2015 [4 favorites]


palindromic: PCR is the gold standard, but is expensive, time-consuming, and requires a fair amount of lab training.

I'm curious why it's considered expensive. I do PCR assays, and I design PCR assays, and I know it isn't expensive to do. I mean, it's not super cheap, it's hard to get it down to less than $1/reaction (although I can do it, and if pressed, I can sometimes get below $0.50 depending on the template), but it's not even close to what places charge for it, particularly for diagnostic purposes. Is it all to pay for the machines and development time? Is it just markup? Is it for all of the liability and reliability safeguards built into human testing? I don't know.
posted by Mitrovarr at 4:34 PM on December 2, 2015


Well, OK, but you've got a PCR, ready access to electricity, you've been trained in the techniques. Source every village's clinic a PCR. Get them connected to a reliable energy source. It's not feasible to drive samples to the nearest PCR (so maybe a 15 hour drive, if you're coming from one of the villages around where I work), figure out if people have malaria, then find them and start treatment. Rapid testing really is a more efficient way to do it.

In places where access to a microscope is prohibitive, a PCR-based diagnostic test really isn't a viable option. The clinic I was tested at didn't have electricity the majority of the time. Sinking money into developing that sort of diagnostic when it will really only be beneficial for a tiny proportion of the people who have malaria isn't a very good use of resources.
posted by ChuraChura at 7:10 PM on December 2, 2015 [2 favorites]


In the setting where I work, PCR diagnosis is almost entirely limited to research applications. Microscopy is the most common form of clinical diagnosis because it is fast and relatively inexpensive. The expense of PCR is not about markup, as such. Medical treatment through public clinics and hospitals is free in Malawi, for example, so it's not a matter of the patient being unable to afford it. PCR diagnosis of malaria is expensive/inappropriate in Malawi because:

1) Need to acquire machines, reagents, etc.

Reagents in particular are a problem, especially if a cold chain needs to be maintained. Most equipment like this would either be donated or come from research grants.

2) Need to train and pay lab techs

In regions with less than 50% literacy, this is often difficult.

3) Unreliable transport/electricity access

Whoops! Power went out for an hour when you were 5 minutes into the cycle.

4) Difficulty of servicing broken machines

Most equipment is donated by companies who may provide service occasionally, but do not have a regular presence in the country. This is a problem with almost every piece of donated 'high-tech' equipment that makes its way into a poor country. Hospitals have sheds full of broken devices that they cannot fix.

5) Time

This doesn't add to the expense as such, but it does make it less appropriate as a diagnostic device in some settings. Waiting 1-2 hours for a test result before being treated is okay in a place where you can phone in a prescription to a pharmacist, or where possibly infected people are treated on an in-patient basis. In practice, most malaria patients would be seeking to catch a bus or bicycle back home before early evening, and will leave without medication if they have to wait overnight.

Related to this is that in order for PCR to be cost-effective, the clinic would need to include multiple samples for each round of PCR. Morning patients, then, could be waiting a long time for results, and a lot of patients would just be presumptively treated so that they could have medication in the event they leave before lab results are in.
posted by palindromic at 7:13 PM on December 2, 2015 [3 favorites]


Oh, there are solutions for most of those problems (use LAMP to reduce the need for a difficult/expensive machine, lyophilize your reagents so they can keep at room temperature, machine analysis of a qPCR curve is way more viable than machine analysis of a blood smear). And the traditional methods have problems too; I would say it's way more difficult to train someone to do microscopic blood analysis than to cookbook a PCR reaction, particularly one that's been set up as PCR beads or an canned analysis. Nonetheless, I can understand why they are difficult to do in the third world.

What I can't understand is why the method is not used more for diagnostics here.
posted by Mitrovarr at 7:54 PM on December 2, 2015


Malaria symptoms are really obvious, especially if you've had it before as many people in the developing world where this would be beneficial have.

It also has a lot of symptoms that are common to many other diseases though. In Burkina Faso, the protocol in village clinics is pretty much to always treat for malaria if there's a high fever, in addition to treating for whatever else they think the patient might actually have (happened to me... The nurse told me I almost certainly didn't have malaria as he handed me the malaria meds). The first time someone told me they were enrhumé , I had to go home and look it up because for years whenever anyone I knew was under the weather, they just said they had the palu.

Later I actually got malaria. And yep, it is the worst. My initial screening came back negative, too. I checked myself in to the clinic despite that, because I figured auditory hallucinations were probably a better indicator than an assay kit of unknown expiration.

Anyway, I guess my point is that a more accurate test is definitely a good thing. And also that I agree that malaria is the worst.
posted by solotoro at 3:46 AM on December 3, 2015 [2 favorites]


« Older The Peculiar Ascent of Bill Murray to Secular...   |   Some Mothers Do Ave Em Returns Newer »


This thread has been archived and is closed to new comments