Conveying Information...

...or "Welcome to My Caveat Emptor World"!

          Imagine someone knocks on your door and when you open up, it is your neighbor and they start telling you all sorts of things in an intense tone:

“I noticed you have a big crack in your sidewalk – someone might trip on that.  Also I see that downspout is loose and it might fall and hit someone.  I saw a couple of termites coming out from that post there holding up the porch – probably should get that checked out.  I found this nail in your driveway.  I noticed that there are flames shooting out of your upstairs window.  There are some frayed cable wires by the side of your house – some animal is chewing on them.  Oh, and part of the fencing around the base of your deck is broken – animals love to get underneath and take up residence there…”

…and while he is still talking, another neighbor rushes up and says:

“There are flames shooting out of your upstairs window!”

          Now I ask you – which neighbor gave you the most information?

          Both neighbors told you that there were flames shooting out of your upstairs window.  But the first neighbor told you a lot more than that.  So isn’t it obvious that the first neighbor gave you the most information?

          Well, not to me.  To me it is obvious that the second neighbor gave the most information.  Even assuming that everything the first neighbor said was true, I say that the additional information clouds the important issue and therefore it borders on dis-information.  More information is not always more information.

          What’s the point?  The point is that our society is operating under the full assumption and belief that the first neighbor is clearly the better neighbor and is clearly providing more information.  And I don’t buy it one bit.

          I’ll just pick on one example, but there are many many other examples.  Specifically:  commercials for a new drug.  The FDA requires those commercials to list all of the possible side-effects of the drug.  Great idea.  Make sure the public knows about those.  For example, how about this one[1]:

“…Invokana can cause important side effects, including dehydration, which may cause you to feel dizzy, faint, lightheaded, or weak, especially when you stand up. Other side effects may include kidney problems, genital yeast infections, urinary tract infections, changes in urination, high potassium in the blood, or increases in cholesterol. Do not take Invokana if you have severe kidney problems, or are on dialysis. Stop taking and call your doctor right away, if you experience symptoms, such as rash, swelling, or difficulty breathing or swallowing, Tell your doctor about any medical conditions, medications you are taking, and if you have kidney or liver problems. Using Invokana with a sulfonylurea or insulin may increase risk of low blood sugar. ...”

          I know that many people consider drug companies evil and just out to get your money and that they would (and do) try to hide all of the side-effects of their drugs in order to make a profit.  Of course they do – they are a for-profit company.  Drugs have side-effects.  You should expect that.  As a result, you probably do have to require drug companies to present all of their side-effects or they will have a tendency not to do it.  I am not opposed to that.  What I am opposed to is being required to list every side-effect, essentially regardless of severity and frequency, in a big monologue laundry list. 

          What I would rather see is that the commercials must list the three (maximum!) most common or most severe side-effects.  Probably, in general, this would end up being an agreed-upon combination of frequency and severity.  As a default, I would say that in most cases we would want to know the most frequent side-effect (probably dizziness in the case of Invokana[2]) and the two most severe side-effects (I’m not sure what that would be in the case of Invokana – maybe the fact that you shouldn’t take it if you have kidney problems or are on dialysis).  In most cases, you would need to parse out the data a bit first to make a determination of what should be listed.  For example, if a study subject dies after taking a drug, that is obviously a severe side-effect.  But if a single death occurred in a study of 1000 subjects, and the death had an unknown relation to the drug, then that one might fall off the list in favor of something more common but somewhat less severe.  For example, if that same drug resulted in osteoporosis such that 5% of the study subjects had fractures within one year of being on the drug, that would certainly rise to the top of the list!  What I am suggesting is not simple to implement because it does take a judgement call.  I believe, however, that it would provide more information than an unranked laundry list.  Even on the package inserts, I would want to see those top three items in bold, larger font, and all the rest of the laundry list in smaller font (or just direct the super curious to a website where they can read the whole study results). 

          I think this approach would allow consumers to make a better comparison about the severity of side-effects of different drugs.  If drug A has a top three of:  1) light-headedness, 2) urinary tract infections, and 3) don’t take it if you have kidney problems; and drug B has a top three of:  1) tachycardia, 2) heart attack, 3) death…well, doesn’t that tell you something very important right away?

          I know that there are those who argue that such an approach is hiding a bunch of other side-effects and “what if I get one that isn’t listed?”  Well, that is certain to happen – rarely.  If you take a drug, any drug, you just have to know that you might be the one in one thousand who gets the odd, unusual, and possibly severe side-effect.  Or maybe you are the one in one thousand who gets a side-effect that no one encountered in the clinical trial and so it was unknown anyway.  In my opinion, you just have to know there is an underlying risk going into it, and accept that risk (or live in a bubble).

          Going back to my original illustration, let’s just say that one in ten-thousand side-walk cracks indicate an impending sinkhole, of which 1% of those sinkholes will be big enough to swallow your whole house.  Thus it may be – just possibly – that your sidewalk crack indicates a bigger impending doom than the fact that there are flames coming out of your upstairs window.  To me, the trade-off is worth the risk in this case.  If, by giving a laundry list of all the risks, you miss the one key risk that is most critical, most common, and most immediate, then that is a bad outcome.  For the one person who is about to be swallowed up in a sinkhole while his upstairs burns, well…he is having one really really bad day and there is nothing we can do about it.  Instead, I say, be like neighbor #2 and convey the critical information.  Better yet, bring your hose with you!

---

[1] I don’t know anything about Invokana – I just picked this one randomly because I could find the text of their commercial online easily.

[2] I didn’t go back and check the data here about Invokana – I’m just guessing for the sake of example here.

Drugs and Devices

          This entry is about my thoughts regarding medical treatment using drugs compared to medical treatment using devices (generally speaking, implanted devices).  First, some personal background so that you can see my biases.  I work in the implantable medical device field, so naturally I am biased towards them.  I have seen firsthand the impact that they can have – not only for the implant recipients I see at work, but also in my own family.  However, I personally do not have a device implanted inside me (yet!) – unless you count the pencil lead lodged in my thigh from an accident in second grade; or the small hunk of metal in my thumb from changing a tire in my rusted out Galaxy 500.  I have had electrodes implanted in my forearm and hand for various testing purposes, but they were just temporary.  I don’t like surgery – but who does?  Also, I’ve never been on any regular medication for anything other than the occasional antibiotic and various NSAIDS and cold medications.  So - that's my background.

          In general, devices are the option of last resort in medical treatment today.  By that I mean that the practice of medicine is designed to try to treat any disease with medications first, generally proceeding from the “least intense” to the “most intense” medication.  Along with treatment via medication, there will also be treatment via non-invasive therapies.  For example, if you have back pain, it will first be treated with some therapy (exercises and so on) and NSAIDS.  As it gets more intense, you will progress to more intense painkillers.  Eventually you may be prescribed an opiate.  If you fail all of those, then you might get back surgery.  And if you fail that, you might get an implanted device called a “spinal cord stimulator” (SCS) device for pain relief.  Of course the treatment path is not always linear, and it varies from specialty to specialty, but in essentially any case, the device is the last resort.

          Because of the field of research I am in, I frequently have the opportunity to talk to disabled individuals about their interest in having a device implanted in them to improve their function and independence.  Occasionally I come across someone who says that they never want to have a device implanted inside them.  At first I thought that it was because they didn’t want to go through surgery again, which is fully understandable.  But no, they are often willing to undergo surgery if it would help, just not have a device inside of them.  I found that to be rather interesting.  I hope they never need a pacemaker!

          To me, it comes down to this:  we are more afraid of what we can see than what we can’t see.  If you take an x-ray of someone who has a device implanted in their body, you’ll see the device plain as day.  It is obviously “unnatural” and that is rather unnerving to all of us.  If we see an x-ray somewhere that shows a lot of “hardware” inside of a person, we’re likely to be thankful that is not our x-ray.

          By contrast, if you take an x-ray of someone who is fully addicted to pain-killers, you won’t see anything different.  Even if you take an MRI, you won’t see anything different.  In fact, even if you take their brain, dissect it apart, and examine the individual neurons, you still might not see the addiction itself.  It is, for all intents and purposes, invisible to us.  For most of us, the effect of most drugs on our body is never immediately visible.  And so…it seems safe.  Even if we drink a bit of a poison like arsenic, there may be no immediate effect.  But it has effectively invaded every cell of our body.  We can’t get rid of it.

          The reality is that drugs are frequently more dangerous to us than any device.  Are there any drugs that do not have side-effects?  Often, a second drug is added to reverse the side-effects of the first drug, and so on.  We accept the side-effects because it seems like an acceptable trade-off if we get the relief we are seeking from our primary symptoms.  The problem is that some of the most serious side-effects are very slow developing.  Drugs that increase the likelihood of getting cancer, or damage our liver, are usually very slow acting and those more serious side-effects do not become apparent until years, even decades, later.  By then it is too late.  In those cases, you can’t reverse or even treat the effects.  If you stop taking the drug, you may still be left with its side-effects – sometimes for the rest of your life.

          To be sure, there are some potential “side-effects” of devices.  The most common disadvantage with any implanted device is the risk of infection.  On rare occasions devices move inside the body, or some part breaks and it no longer works.  But, the point is, these are fixable problems because, in the worst case, the device can be removed to resolve the problem.  The thing that seems the scariest to us – the fact that we can see it on an x-ray – also makes it inherently safer.  If we can see it, we can remove it and get rid of it.  In most cases a device can be removed without any further consequences.

          Another apparent disadvantage of devices is their initial expense.  Because they involve surgery, and they are often very expensive themselves, the whole cost of getting a device can be quite high.  Insurance companies are reluctant to pay that up-front cost.  They would rather pay monthly for drugs you have to take daily than to pay for the one-time cost of implanting a device.  There is an aspect of human nature that would rather spend $5/day every day for the rest of their life than to pay $5,000 now and be done with it.  Often the math clearly comes out in the favor of the device, particularly if it is more effective than the drug, but that is usually ignored.  In that sense, we seem to prefer the “death by a thousand cuts”, though maybe that is a poor idiom to use in this case!

          The point is that since devices are nearly always relegated to the “last resort”, it means we all have to first live through the side effects of drugs before a device will be considered for treatment.  However, there are certain situations where that leads to two major disadvantages.  First, we are stuck with the side-effects of the drug, and some of those side-effects are difficult, if not impossible, to reverse.  Second, by the time a device is utilized for treatment, the disease has progressed to the point where it is much more difficult to treat by any means.  Thus a device, which might have provided relief at an earlier stage, can no longer provide any relief.  There is even some evidence that if devices were used at an earlier stage, they might halt or even reverse the progression of the disease.

          Medical devices do have to be made safer, smaller, less-invasive and more effective.  They are certainly not always the answer.  But there are cases where progression to treatment with a device is better for the patient and is more cost-effective, than continued treatment with drugs.  For example, our current health care system is set up to keep treating chronic pain with opiates until the person is potentially addicted, before considering treatment with a device like a spinal cord stimulator.  The device is not always the solution, but there is evidence that the device would be more effective if it were used earlier, and it does not have near the negative personal, social, and financial consequences that opiate addiction has. 

My point is that medical devices should be given a fair consideration in the treatment of various diseases and disabilities, and not always relegated to the “last resort.”  This is a significant change in thinking.  First, the practice of medicine will have to be convinced of the potential of devices to be more than a last resort in certain situations.  Second, the reimbursement structure will have to change to reflect this difference in thinking.  And, finally, the perception of devices by the general public as being more invasive than drugs will have to be adjusted.  Each one of these steps is a major undertaking.  It will certainly not happen overnight and probably not in my lifetime.  But I do think that eventually the general perception will change on this issue.

Writing Grants for Federal Research Funding – Some Tips!

          When I finally successfully defended my Ph.D. in Biomedical Engineering, one of the members of my committee wanted to meet with me separately after it was over.  When I met with him, he gave me some “advice” that has been seared into my brain ever since.  He said: “your writing is so bad that you’ll never get any papers published and you’ll never get any grants funded.”  A truly optimistic way to start a career in research!

          Well, he was right in one sense: my writing was terrible.  But that was over 25 years ago, and since then I have managed to get many papers published and get many grants funded.  I don’t have any teaching duties, so I and my staff are essentially fully supported by the grants I write.  My funding is mostly from National Institutes of Health (NIH) and the Department of Veterans Affairs (VA), but I’ve also had funding from the Department of Defense (DoD), the Food and Drug Administration (FDA), and a variety of different private foundations. These days it is very hard to maintain continuity in a research program.  The funding rates at NIH are typically around 10-15%, meaning that only one in ten grants is funded in any one cycle.  And you’re competing with a lot of really smart and creative people trying to solve equally valuable problems.  So I thought I would write some of my thoughts about the process of grant writing and give you my personal tips on how to write successful grants.  Hopefully it will be of some help to any of you who are trying to make a career in research.

Most of my thoughts are specifically related to medical research grants submitted to federal agencies where there is a scientific peer-review process, such NIH, VA, FDA, and DoD.  I personally haven't had funding from the National Science Foundation, but I presume the process would be fairly similar there as well.  Also, my area of research is implantable medical devices and rehabilitation, so that will totally color my comments.  For example, if you are doing basic cell research in a neuroscience field, my comments may not all be relevant or even correct.  In my experience, the different disciplines even within NIH have different “personalities” with respect to the types of projects they like to fund.  But, with that background in mind, here are some suggestions.

          Let’s start off by trying to put into perspective the difficulty that researchers face in getting funding.  I tell people that getting a grant funded is like bowling a perfect score. Bowling a perfect score is very hard to do – 13 strikes in a row – yet there are lots of bowlers who bowl a perfect score.  How do they do it?  Well, I would say there are two things that you need in order to bowl a perfect score: one is you have to be a very good bowler (obviously), and the second thing is that you have to bowl a lot.  Even the best professional bowlers don't bowl a perfect score every time - it's still a rare event. 

          The analogy with respect to grant writing is this:  you have to be a good researcher, and you have to submit a lot of high quality proposals.  By that I mean that you have to have good ideas, you have to do good research, you have to work as hard as you can to improve your skills and knowledge, and you have to be good at presenting your proposed work in written form.  But in this day and age you also have to write a lot of high quality grants.  This is, of course, hard to do because it takes time to write even one good, high quality grant.  Efficiency is therefore important – you have to keep evaluating what you are spending your time on.  I often ask myself as I’m working on a grant:  “will spending more time on this particular aspect of the proposal improve the grant’s chances of getting funded?”  If not, then don’t spend too much time on it!

Given the difficulties in getting funding, you have to accept that sometimes you will have an excellent idea and write an excellent grant, and it just won’t get a good enough score to be funded.  The process takes a lot of perseverance.  There will be ups and downs.  Because of that, I highly recommend being part of a larger collaborative group of investigators.  Collaborative research itself is an important issue that may be a topic for future discussion, but from the standpoint of grant support, it almost seems necessary.  It's extremely difficult for a single isolated investigator to maintain continual grant funding.  By having a collaborative group, it is possible to “ride out” some lean times. 

I’m not going to talk here about the importance of your scientific ideas.  Obviously you have to have good ideas.  You have to conduct research on issues of importance.  You’ll have to listen to your colleagues and reviewers; and sometimes you’re going to have to face the hard truth:  you need to change your focus.  That can be very difficult.  About fifteen years ago I spent three or four years trying to get funding for an idea I had that involved restoring function in cases of peripheral nerve damage.  After a number of attempts I had to face the reality that it just wasn’t going to happen.  I still think it was a good idea, but I just was not going to get it funded, so I had to leave it behind.

You’re going to have to hone your writing skills.  Grants are written documents and, for the most part, science is conducted through the written word (grants, journal articles, books, etc.).  Grants themselves are never scored explicitly on grammar and clarity, but it absolutely plays a role in whether a grant gets funded or not.  For most of us in the hard sciences and engineering world, our college education didn’t include a lot of specific training in good writing skills.  Kind of odd, really.  However, there are certainly ways to improve your writing skills.  Early in my career, I read a couple of books on scientific writing that helped.  There are classes that you can take that help as well.  But I would say that the thing that helped me the most, by far, was reading and reviewing other people’s grants.  I found it especially instructive to read poorly written grants and papers and try to figure out what made them poor and how they could be improved.  It’s hard work.  You should also keep asking yourself “do I make the same mistakes in my own writing?”  If you are a young investigator, take advantage of opportunities to help review grants for others.  Do internal reviews of papers and grants.  Write out your comments and try to figure out what could be done to improve a grant to make it fundable.  This kind of review and introspection will help you immensely.

          The other major thing is that you absolutely have to understand the audience you are writing for.  If possible, find out the backgrounds of the type of people who are likely to be reviewing your grants.  If you are submitting to NIH, the general make-up of any review panel (Study Section) is publically available [here].  You should go through every name on the study section roster and see what department they are in and, if possible, find out what their area of study is.  Is the Study Section mostly composed of a group of basic neuroscientists?  M.D.s?  Engineers?  You need to know this because every sentence of your proposal needs to be written with that group of reviewers in mind.  I really mean every sentence – from the opening lines of the Specific Aims to the grant conclusion.  If I’m writing a grant that will be reviewed by a group of clinicians, then my whole grant is going to be couched around the disease state and clinical application that I am pursuing.  If I’m writing to engineers, then I need to catch their attention at the beginning with the innovation of my approach and I will certainly need to include more technical details. 

          As important as knowing the make-up of the Study Section, this next issue is even more important:  you have to understand and appreciate the general mental state of the individual who will be reviewing your grant.  By that I mean that you need to understand the personal conditions under which your grant is going to get reviewed.  Reviewers are generally successful researchers, which means they are busy people who have to write their own grants.  They participate in Study Sections because it is good scientific community citizenship.  They want to do a good job reading your grant, but when can they fit it in to their day?  Reading your grant will get pushed off until the late evening.  And it will get pushed off until close to the due date for the review.  It’s just human nature.  Ultimately they will not have quite as much time as they had hoped to read your grant in detail, but they will put time into it.  This puts a premium on the clarity of your writing.  Use of clear figures is mandatory.  Write your grant so that it can be read for someone who is tired and bleary-eyed and still has three more grants to read after they finish yours!

Also, in most cases, the topic of the grant will not be directly in the reviewer’s area of research.  In fact, your grant may cover some areas of scientific exploration that the reviewer is pretty fuzzy on.  Not that they don’t understand the basic science, but they surely have not been reading the same literature you have been reading.  For example, my research is in electrical stimulation for restoration of motor function, primarily in spinal cord injury.  It would be quite possible that, for example, I would be asked to review a grant on something like diagnostic ultrasonic imaging.  I know the basic principle of ultrasonic imaging, and I “know of” people who use it, but that’s about it.  I certainly haven’t read a paper about the advancements in that field in the past 25 years.  The odds are high that there are things in that grant that I really will not understand very well.  However, if the grant includes a brief “tutorial” on ultrasonic imaging and a clear introductory figure, that will help significantly in understanding the proposal.  In my opinion, it is worth the space, even though space is at a high premium in a grant.

          Along the same lines, I always encourage people to try to minimize the use of acronyms.  It is extremely difficult to read a grant if you have to keep going back to the front of the grant to find out what each acronym stands for.  I have a personal rule that I try to introduce no more than three new acronyms in any grant.  This means, again, that you have to know the background of your reviewers.  Will they all know what CNS stands for? (yes)  MRI? (yes)  EMG? (highly likely)  FES? (maybe)  NNP? (definitely not)  So, I try to use, at most, three acronyms that the reviewers are unlikely to know.  That means that I have to spell out a lot of other terms where an acronym would take up less space.  But it’s worth it.  Also, if it has been a few pages since I’ve used a particular acronym, I’ll spell it out again to remind them.  You wouldn’t be allowed to do that according to the rules of writing for a journal article, but a grant is not a journal article.  Your goal is to make it as easy to read as possible for the reviewers. 

          I’ll stop here for now.  There is no advice that can guarantee that you’ll get a grant funded – it just doesn’t work that way.  But if you keep working at the craft of grant-writing, you will greatly improve your chances of getting one funded.