Maryland young maker Jack Andraka isn’t old enough to drive yet, but he’s just pioneered a new, improved test for diagnosing pancreatic cancer that is 90% accurate, 400 times more sensitive, and 26,000 times less expensive than existing methods. Andraka had gotten interested in pancreatic cancer, and knew that early detection is a challenge. He gleaned information on the topic from his “good friend Google,” and began his research. Yes, he even got in trouble in his science class for reading articles on carbon nanotubes instead of doing his classwork. When Andraka had solidified ideas for his novel paper sensor, he wrote out his procedure, timeline, and budget, and emailed 200 professors at research institutes. He got 199 rejections and one acceptance from Johns Hopkins: “If you send out enough emails, someone’s going to say yes.” Andraka was recently awarded the grand prize at the Intel International Science and Engineering Fair for his groundbreaking discoveries. [via Fast Company]

Watch Andraka talk about his improved test:

138 thoughts on “15-Year-Old Maker Astronomically Improves Pancreatic Cancer Test

  1. One hundred ninety nine of allegedly highly educated and prestigious professors didn’t recognize the potential of what Mark had?

    This is yet another marker for serious problems in higher education.

    1. FredB, You make an assumption that the 199 rejections equals a problem in higher ed. Yes, it would have been great to see more positive responses but there is no correlation between that and problems in education. If anything, you presented only your opinion. Please do your research. Opinions, like manure, usually stink.

      1. “there is no correlation between that and problems in education” <- Have you made your research? If we could only throw research results out of our mouth, the sharing of ideas would cover much more precise matters, but it will be as well mostly non-existant. Please keep your epistemological orthodoxy for the proper environment, and also please apply it to yourself, as you are trying to force it on others.

      2. “Opinions, like manure, usually stink.”– unless it’s your opinion, right Scott?

        Higher education does just that- squash opinions of individuals. You seem very closed minded if you think opinions are manure. Can’t wait for your generation to leave power.

        1. Wrong. Higher education is not longer the bastion of fact over opinion. Big pharma sponsors so many university studies it’s ridiculous. Just read this NYT piece on how they have infiltrated to med schools:

          Special interests (corporations) will do whatever it takes to protect their bottom line, and if this kids discovery will dent some pharmaceuticals profits from a pancreatic cancer blood test, then they will have no problem trying to stifle it. Nothing is sacred anymore.

      3. Scott, you must be self interested to respond so foolishly. I think it is pretentious to use the word correlation about one instance. Did you do the calculation? You seem very certain. My guess is that most of the professors did not read it because they had better things to do than read it. We all have to select which E-mail we will read and which we will delete. I would also guess that many of the professors thought it was a question the young man was asking when they scanned the contents. I think it is pretentious to use the word correlation about one instance. Did you do the calculation? You seem very certain.

      4. 199 rejections is nothing new. This has been a problem for ages now.

        “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die and a new generation grows up that is familiar with it.”
        – M. Planck

    2. There is indeed a problem in the way higher education systems recognize research of this kind. Am I the only one in the room that thinks that most of the instituions who received the email with corresponding documents did not bother to read the article? Not only that but have we forgotten the more obvious fact, namely that the money spent on cancer research did not produce such testing and research.

    3. I cant find anything peer reviewed or in depth on the topic, but his explanation includes visualization by scanning electron microscopy and construction of interaction specific nanotubes. I dont know if my institution recieved an e-mail from this kid or not, but even if I had believed him I would have had to say no. I do not have the facilities or expertise necessary to facilitate nanotube work nor do i have reasonable access to SEM technology. Undertaking development and the long course of trials involved to ensure no cross reaction with other protesins etc. is a monumental task which my lab simply doesnt have the funds or ability to do. I would be willing to bet many of the rejections were not based on a belief that this kid was incorrect, but more likely that the labs just didnt have the space or time. Imagine if someone with a revolutionary idea came into your workplace, said I am a full time student with no experience in the field. Take my idea and make it happen on your dollar. would you be alble to take him in? I’m glad johns hopkins could, but i’m not surprised others couldnt.

    4. I’m sad to see that many people commenting on here don’t even have a brain to share amongst themselves. The kid is 15, he’s smart, he’s well intentioned, he had a unique idea superior to anything any of you (us) have ever or will ever have in our lifetime. What are all these negative comments?? Jealousy? He’s not that smart…just look at how i fixed my shelves. I’m so smart.

      Look, if the test is dirt cheap and uses 1/6th a drop of blood, line up 100 of em and i’ll volunteer my 16 drops of blood. If all 100 test positive, i’d like to see the math on that one shoot it down. Ya got cancer…its that simple.

      Kudos for the kid, wishing your teenager was as smart as him rather than maybe playing World of Warcraft.

  2. You GO Jack! There will always be naysayers – most of them jealous because they didn’t think of it or too insecure to believe in someone who doesn’t know what they can’t do. (I placed 4th in my category at the 1984 ISEF)

  3. Incredible. All power to him, and as a side note I don’t think at 15 I would have been able to express any of my ideas that eloquently with a camera stuck in my face.

  4. The problem is that a test that is only 90% accurate is going to be worse than useless.

    According to Wikipedia, tens of thousands of people have pancreatic cancer in the US. So, estimating an upper limit of that as around 50,000 in 300,000,000 American citizens means that a random person has about a 1/6000 chance of having the cancer. So, if you’re given the test, the odds that it gives you a true positive result are around 1/6000, but the odds of a false positive result are 1/10, so a positive result from this test means a likelihood of only about 1/600 that you actually have pancreatic cancer.

    What makes it worse than useless, as opposed to just not very useful, is the fact that any time you get a positive result, you’re going to have to go have procedures (such as a biopsy) to determine if you actually have cancer, and those procedures are going to have side effects, everything from scarring to death from infection. That means that if you give a lot of people this test, vastly more people are going to have those procedures even though they don’t need them. It’s why giving everybody the PSA test was such a bad idea. It resulted in a lot of diminished health and really didn’t catch very many prostate cancers.

    Cancer tests need a much lower error rate in order to be useful. Like a false positive rate around 1/10000. I don’t know if it’s possible to reduce the error rate of this particular test to that level, which would make it a starting point for further research or not, but it’s kind of important that more people understand the implications of error rate in light of how Bayesian statistics work.

      1. No medical test has 100% accuracy. There’s always some source of error, because there are always exceptions to rules and examples that don’t fit patterns in medicine. But that’s ok; it need not be perfect to be useful.

        But to the parent post, kbolino raises the point which I was going to make: 90% accuracy does not mean 90% specificity (i.e. a 10% false positive rate). It could mean 90% sensitivity (1 false negative in every 10 positive cases) or 90% positive predictive value (if the test is positive, there’s a 90% chance you have cancer). It’s just too vague a term to understand what’s going on, and I suspect that a statistician looking over the winner’s raw results would be able to clarify them a great deal.

        EVEN IF the test has a 10% false positive rate, that simply makes it less useful for screening without compromising its use in diagnosis. It would then be administered only to patients who are already suspected of having cancer – people who WOULD HAVE HAD BIOPSIES UNCONDITIONALLY IF IT DIDN’T EXIST. A negative result would clear them and avoid the biopsy. Therefore it would cause fewer procedures.

        If overly skeptical people don’t gang up to shoot it down before it even sees the light of day, like they did with something very similar that I had been working on for breast cancer.

        1. If no test is 100% correct, it’s better to have a 90% accurate test than have nothing, right? Would you rather go through a deathly cancer than a biopsy that might cause a scar on your pristine body?

    1. “What makes it worse than useless, as opposed to just not very useful, is the fact that any time you get a positive result, you’re going to have to go have procedures (such as a biopsy)”

      So… you would rather everyone just wait like they are now, and then much later in the stage of cancer (if they have it) go out and get a biopsy which you yourself have defined as having side effects from “scarring to death from infection”, when their systems are already under stress from a cancer?

      90% isn’t 100%, but it’s not exactly 50% either.

      1. Yes that’s what he said. And he explained how acting based upon imperfect information in many cases results in more death than cure. E.G. you may cure 10 sick people and kill 20 healthy people. Here’s an example, I have a friend who had a false positive that resulted in his skull being cut open with a medical saw. Oops.

      2. To answer your question: Yes. I think that not doing the test as it is described in the article is a net benefit to society.

        Those biopsies are going to be done in hospitals. If the odds of picking up a fatal infection in a hospital during a biopsy are about 1/1000, then you would be condemning six perfectly healthy people to die for every person you save by detecting their cancer early. Does that choice make any sense to you? Of course, that 1/1000 is a number I pulled out of my nether regions because I’m too lazy to look it up, but it’s probably somewhere within the ballpark.

        My point remains that merely getting a positive result on a test can have a significant negative health impact, so tests for relatively rare conditions need to be quite accurate before they are useful.

        Of course I AM saying “sure sucks to be you” to the people who have the cancer, but I think that’s a better choice than saying “sure sucks to be you” to the perfectly healthy people who stop being so because of a chain of events triggered by a random false positive. “First, do no harm” and all that.

    2. You assume that “90% accuracy” means “the Type I error rate is 10% and it is independent of all factors–including the number of times the test is administered” in an article that throws out such gems as “26,000 times less expensive” and “400 times more sensitive”?

      Methinks you doth read too much into a fluff piece.

      1. Another point worth mentioning is that they can be selective about who they administer the test to. It’s not as though the base rate fallacy is new to medical research, however many people have been tripped up by it. If the prior probability that a random member of the population tested is high enough, 90% accuracy isn’t so bad.

    3. Mr Jonathon, one small problem with your opinion and calculations, on a forum like this, few of us are statisticians, or researchers. When you start throwing numbers like that out in a forum, few people would really know if you’re correct or just “made it all up”. I doubt that the people at Johns Hopkins would have been interested in this new discovery if it came with all the problems you describe. Maybe you should contact John Hopkins and enlighten them with your information. Cancer researchers don’t usually check Make magazine articles when evaluating new research. Thanks so much for helping all of us.
      By the way we would like to see some of the research you have been working on.

    4. Nice job mutilating statistics. I’m sure that if it were 99% accurate, you could construe and twist numbers to make it appear dangerous.

      What 90% really means is that “Sir, we are 90% convinced that you do, or 90% convinced that don’t have pancreatic cancer”. Physicians don’t just rely on one test to conclude the issue. Instead, they compile evidence. There are many tests that are not conclusive on their own, but they serve as guides on how to proceed.

      1. No, accuracy rate says ABSOLUTELY NOTHING about how sure you are that someone does not have it. Accuracy rate tells you how many people WITH IT will be detected. It does not tell you how many people WITHOUT IT will be incorrectly detected, that’s what a false-positive rate is. Also I’m absolutely certain that doctors would think this is useful when it isn’t, a huge percentage of doctors make the base-rate fallacy, see:

    5. It is very unreasonable to rip on a prototype for its lack of performance. The point is the proof of concept, as incremental refinements belong to later stages. This is not a commercialized product.

  5. I am reminded of the Steve Jobs quote, A class people hire other A class people.
    199 rejections, and the Professor that decided to let a 15 year old try a wild idea is at … Johns Hopkins.
    America may be doing a lot of things foolishly, but mentoring great talent for the future seems taken care of

  6. “Yes, he even got in trouble in his science class for reading articles on carbon nanotubes instead of doing his classwork.” Pure example of how our educational system does not assess nor encourage talent.

    1. Exactly! The winners of these huge science fairs are often like this: projects technically done by a kid, but with extensive help from (and access to) established world-class laboratories. His results are promising, but very preliminary, and he certainly isn’t the lone maverick these stories are selling him as. This is hopefully the start of a good career in a vibrant, highly collaborative research environment. But he isn’t Doogie Howser.

  7. Amateur engineers and scientists have made strides in many fields. To be recognized and your contribution seeing the light of day can be pretty low likely hood of happening. The why’s for this can be simple ego bruising to economic havoc to big business.

  8. Problems with higher education notwithstanding, I’m of the opinion that such science contests are complete rubbish.

    Guy hasn’t shaved once, but saves the world.

    Come on….

    1. To you and all the other haters. This kid has done more in his 15 year life than anyone on this comment section will probably fulfill in their lives put together. How about a congratulations for potentially saving thousands of lives. Also to the guy spouting off numbers and statistics about why this test would “harm more people than save” is complete nonsense, sounds like someone just got out of their first statistics class and wanted to sound smart on the internet. The kid himself states 100% accuracy and the media states 90%, it’s likely that the percentage is greater than 90, and with further research will be even more accurate. It’s not like everyone is going to go to the doctor and get screened and if they show up positive it’s straight to the biopsy. Most of the patients receiving this test will probably be of older age and complaining of early stage symptoms related to the cancer. Older people receiving the test have a higher chance of infection (>65 years old) and hospitalization due to biopsies. Having a very cheap, fast, and effective way to detect the cancer would REDUCE the number of biopsies needed by weeding out the majority of those not affected by the cancer. Biopsies always have a risk of infection, and one major problem now is the rate at which infections happen is increasing due to the amount of antibiotics being prescribed across the world, which bacteria are adapting to. More simple tests like this are EXACTLY what we need.

  9. My Mum died in September of pancreatic cancer. I sure would welcome it if there were a simple test so my siblings and I can be tested as we get older.

  10. It is truly refreshing and surprising when a young person can produce high value results in a critical field like oncology research. Not only that, but without financial backing from big Pharma or NIH, etc. Millions of dollars in research over decades has in general, produced far more Porsche’s than cures.

  11. The very fact that this invention could be easily acquired for a song and buried by one of the big Pharma companies shows the immoraliity of allowing profit driven corporations to make decisions about matters of public health. Research and the development of things such as new Antibiotics and Vaccines should be done by not-for-profit corporations that would be mandated with the responsibility of public health and public health only. There would still be room for profit driven free enterprise but only where it would be to produce a better product for the public good.

    1. Sorry, not in this mix of atmosphere. Pham is a very big business, however it does rely on innovation. If this kid has something, I only hope he has the business forsight to do more than spout percentages.

  12. how can I make an appointment with him to have this test done? My family has lost 6 members to pancreatic cancer, so far.

  13. Here’s the thing….. Its a step forward… Not perfected, but something to build on in the future… Good job, John Hopkins for taking a potential leap to the future of pancreatic cancer research!

  14. Of course they rejected him, if everyone is able to do research for free they will never get grants for tons of money to put in their pockets, odd that no one is recognizing that only a few years ago this wouldn’t have been possible. Excellent independent work from a bright young man, I’m glad he’s getting recognition for it.

    1. Lab equipment is expensive, carbon nanotube equipment is extra expensive. Lab space is expensive. This kid has great ideas but he does not have the hands on experience to physically cary out many aspects of this project. So yes, it would be great for everyone to do free research, but its just not feasible. Unless you want to pay some more taxes for publicly funded research, but that makes everyone angry.

      1. @voidwarranties: This is a warm experiment. While he does not have the tools, he understands the priniples and the workings in his head.

        I say, give him enough rope to hang himself with his findings. *Shrug* it could lead to nothing. Then you win. When it leads to something, you might still win. Give the guy a break. (Maybe a yummy breakfast…you might owe him one).

        However, maybe his immediate discovery leads to low results. Maybe, another scientiest will become inspired and work off of Andraka’s findings. This is how discovery has always worked. This is pretty much how science works.

  15. What I want to know is where a 15 year old kid got his hands on carbon nanotube synthesizing equipment and ready access (or the money to access) a scanning electron micrograph to check his results. Not to mention the molecular libraries and manipulating tools required to build tubes with such protein specificity? Or did he win off of a proposal, and these tests are yet to be done?

  16. Hey, I just wanted to thank you for this post! In all honesty, pancreatic cancer gets very little exposure compared to say breast cancer or even prostate cancer, and yet the five-year survival rate for pancreatic cancer is only 6%! I watched the video on this guy, and was blown away by the incredible method he had devised for screening pancreatic cancer! I think that since it is pancreatic cancer awareness month, it is important to spread the word about new advancements in pancreatic cancer treatments and diagnosis. That is why, in honor of this month, I wanted to share a brand new infographic on pancreatic cancer:

Comments are closed.


I'm a word nerd who loves to geek out on how emerging technology affects the lexicon. I was an editor on the first 40 volumes of MAKE, and I love shining light on the incredible makers in our community. In particular, covering art is my passion — after all, art is the first thing most of us ever made. When not fawning over perfect word choices, I can be found on the nearest mountain, looking for untouched powder fields and ideal alpine lakes.

Contact me at or via @snowgoli.

View more articles by Goli Mohammadi