professional literature. what i am hoping to do ensure that probably with the public. we are going to talk about the fair test are not all created alike. how there can be structural design flaws in the way that we run trial so that they no longer our tests about what treatment is best. we are talking about marketing. although that i find marketing a little bit worrying.

we have good quality evidence from academic journals and so on. first and more importantly, we are talking about the problem of missing child data. the unspoken dirty secret, if you like, half of all the clinical trials, from her best possible estimate have not been published. so this has a huge impact on the information that we used to make informed decisions as doctors and patients about which treatment is best. this is a phrase used by young hipsters. but a very long time ago in the dawn of greediness, these are

magazines that you would have for the punk scene or punk music, music arts and fashion. this is the founding for phase one clinical trial participants. first clinical trials, when you develop a new molecule, you first of all trial it in maybe six healthy young men to make sure that nothing horrible happens to them. they are often -- they are often volunteers with low wages. so i think that it is fantastic because doctors like to think of this stuff as being very precise and clinical. but there is a real practical culture for the stuff as well. with guinea pig zero, they talk

about the reports, they talk about uniting as guinea pigs get better conditions and they talk about the history of phase one trial participants and human guinea pigs and self experimentation. some of it is very interesting. oh, wow, i did that in a welsh accent. [laughter] okay, so this is the world of phase one clinical trials. this was a participant in the uk called [inaudible] it was very well known because it went try virtually wrong. these men were very dissimilar,

sometimes things rapidly began to go downhill and they ended up in intensive care, they were incubated, they were ventilated, their whole blood was replaced and then their circulation began to shut down and then they went black and began to fall off. these pictures are from the news coverage at the time. in the uk government, inquiring with what has happened here, there are two observations and recommendations made. the first observation and recommendation have once been adopted.

okay, i'm not participating in these phase one clinical trials and i'm not sure how anyone can have that kind of relationship with red. in some ways it is foreseeable. this had been trying for at least one person. the research is working and he had no idea that this is what has happened. that this trial haven't been shared with the research community. the recommendations include that the results of the phase one trial should be disseminated more widely. only one in 10 trials is published within a year. only one in five within five

years is published. the problem of withheld trial data goes just beyond one trial were posing a risk to phase one trial participants. this is a molecule. it prevents abnormal heart rhythm. in the 1980s there was an indication of a much broader situation. so there were some people who just had heart attacks and they were likely to be given this kind of drug.

so it turns out that this abnormal heart rhythm, it turns out that something is wrong. so the cost trial, which we reported the findings in 1992, we discovered that in fact, patients were given these drugs for these particular numbers of abnormal heart rhythms were actually increased in the risk of dying. firstly, no one was being bad or malicious or evil here. the real-world outcome is death. but we were wrong. and because it was described in practice is so widespread, it has been handled so frequently. it turned out that some people

have estimated that because they are necessary or avoidable, there can be 100 30,000 u.s. citizens tested before we gather the real-world outcome. sometimes there are tons of academic papers that talk about these 100 30,000 people that died as a result of this, that it's actually only two or three on the american side. [laughter] okay, now, what accounts for this, well, it was an important wake-up call for doctors to make sure that we don't rely on surrogates outcome. interestingly, there has been

this definition. he says when we carry this out, we thought that this was an effective charm. there was one used in this effect in the 80s. so he was a little bit more optimistic. there were 100 people in this trial, nine deaths in the 50. the trials were abandoned. and the development was abandoned for commercial reasons. this is very common. the study was never published and no one ever knew that this had happened.

trials with unflattering results can go missing. they can be raised with the academic record. so the result can provide an early warning. just remember, this understatement is trouble ahead. this includes 100 30,000 u.s. citizens. now, to understand why missing clinical trial data goes on, we must understand the basics of how we use evidence as doctors in order to make decisions.

let's say the treatment of depression for those who are in the hospital. you might go, okay, do you remember a few trials and i have read about in the past? whenever i come across the trial from nine months ago and it looks relevant, i will put that in the box, ready for when i write my chapter. maybe your friends would then

produce a textbook article that would become a definitive story about this. even though it's workable for all kinds of unconscious things. this includes a piece of science about how you the -- how you examine the evidence. you say, okay, what did you use to identify these randomized trials, and this might be part of these controlled trials. that he would get the most systematic complete summary of all of the evidence that could possibly be drawn together.

i was producing systematic reviews and there were doctors worldwide seeking these treatments. now, would you like me to explain how this works? it includes the significant benefit. as you see, someone has this big trial that is left vulnerable to

random error. that gives you benefit of treatment. so there is a mother that is about to deliver a premature baby. we have to keep the baby from dying. this is really important stuff her in we can't call this the big dead baby graph. i'm genuine about this. this is a great trial of evidence and teaching epidemiology you have to really

feel that there is a connection between that and the very real world of suffering and pain and flesh and blood and death. so, when you add them all up together, what you get at the bottom is the summary effect, which is overall the beneficial treatment. all this existed for many years. even though, if you have them all up together, this stuff was affected, there were many doctors who didn't use steroids when mothers were about to deliver premature babies because some doctors would say, there are no reason to chooses.

others would say, why should i expose people to unnecessary side effects by giving them steroids. so babies died, right? the babies died not because we didn't have the evidence, because we had all of these trials. they died because doctors at that stage weren't having enough to realize the importance of emphasizing the information that we have to get the right answer.

now, let's talk about the importance of this with the babies. so, this whole project, right? of bringing together all of the evidence and one place, to get the best, closest and most accurate estimate that we can of the treatment. this break, it breaks if we don't have the evidence. we know that this has not happened. the negative studies have been withheld in two different ways. you can do with stories, which is a fictional way of relating

science or with statistics. so, when there is a trial that is conducted in 30 different hospitals and 10 different countries, employing 100 and 80 people with a cost of $20 million, the way it is written, right? okay, so each of these dots is one trial. okay, so you can probably

already be up at the top, these big trials give you a more accurate estimate. so all of these trials, we should recognize that there are smaller results with these. then we have some here in the middle and then we have him on the other side. now, these publications show that the studies will be missing from this picture. but it's not the big ones or the smaller ones.

will they be missing here? known. will they be missing here? well, that's interesting. so this is the plot and you can see here. here's the true answers come over here you can see if you go down, smaller studies and over here, you expect other studies to be and they are not there. this aggression line is not a statistically significant deviation from the funnel shape you would expect. okay, but you would also demonstrate the president and show that this is an issue. i describe it in patients.

so this is most likely the best response. so now, here is an antidepressant. i had a quick look at the literature, there are public trials that show that it is better than a sugar pill. a couple of trial that says it is just as good as any other antidepressant. so i find my name and give it to a patient. in 2010 in trial was published in the british medical journal by researchers any governmental agency. they discovered the extraordinary battle that they had were they said we will not look at this until you give us all the trials that you have done. and they said, well, we're not going to have anything to do with it. finally, they got a hold of them and there were three studies comparing what is beneficial and they were all published.

there were five more studies showing that they have not been published. as a doctor, i feel misled. [inaudible] i concluded that the trials were well designed, yet, i was still over misled. so by some people's estimate, it has no benefit. there were a lot of side effects. at best, you get something that is marginally beneficial. that is a really important thing that we have to grasp here. i'm not saying that things are always actively harmful. but the real problem is when

half of the evidence is from the medical literature, we don't know which treatments are best. even if the treatment is something, you are being deprived of a truly effective treatment. that is genuine harm. okay? so we are talking about a middle-aged man, there aren't any major breakthroughs, but there is a gradual accumulation of these things.

you give people what is an effective treatment. but there are 800. so you're down two lives out of 100. so we kind of have the failure of ambition. the flaws of our such that we cannot discern what is necessarily the best. okay, so we have this the best that we can. and it turns out that the

available evidence comes from a systematic regime. this is a systematic summary of all the studies being done so if you want to find out if the cows have simply disappeared, without posting negative results, what you need is ideas to complete it. for example, you can go to a local ethics committee and get things approved, to be conducted in a hospital in the area of the university and then publish things. you can also use regulatory documents.

remember, this is not all of the trial. it is before it was approved on the market, it was used to get approval from the fda. so it is a real pain to have this freedom of information act, everything is kind of a waiting game time, you have to wait after you apply. i talked to the man who did this paper yesterday.

they have published these results, they're all kinds of these techniques. now, it seems to me that overall, where you go from these results, this is research misconduct. who here has done some kind of research project okay, now, you have half of the data points, you do we have the data to make the line go the way that you wanted to go. so this is correct.

so what i find fascinating is that we have this strange cultural blind spot in academia. where for some reason there is a whole study of results. the overall impact in the overall effects are what we use with a real-world effect. you know, this is an a story -- you know, an idol of people trying to make money.

this is a trial. so we have people that are lured into a false sense of security. so we must be very adamant about coming out and talking about these different things. so we will never again, they said, publish a clinical trial unless it is being posted on a public website. and this was really interesting for the moment. because they said, okay, everybody does a trial, if they want to publish it, they want to

make glamorous academic finds. and so therefore, if we have that terror, we can tell people that you are not allowed to publish in the journal unless you have registered. so we can see where this has been part of these cases and then we publish half of the information. this includes spotting the missing data. so we had the uk head of merck pharmacy. and he said, oh, this is rubbish. this medical journal went public. ..

a. >> that is just the papers that were published. and it isn't surprising that it has been ignored because they're not just financial but cultural and professional for example, example,, to take a big poster finding from industry

trial that would get cited by other academic journals so that one part the impact factor is derived from the number of citations that it gets. remember academic journalists are part time it is not a full-time job and lastly they make money from the reprint of a but when it posted trial or review is published in the academic journal then the drug companies by lots of reprints at $20 a pop then they pushed those and use those marketing materials to give those to doctors to encourage them to use that treatment. that means academic journals have eight conflict of interest. so then there is a conflict of interest section at the bottom of the paper they're not asking you your conflicts but what kind?

financial above all like how much money you had for giving talks, a free conference travel, if they funded research but here is what i find interesting last year by publishing the british medical journal we went to the top five and said could you tell us for these trials how many reprint orders you got? they said no. that is confidential information. [laughter] new england journal commented gemma but i will tell leo -- you they are perfectionists to make sure you give a clear declaration of your conflict of interest but yet when we ask a simple direct question when this know how much money you got for selling reprints we're interested in the whole process with dissemination of evidence which is important now it is confidential.

interesting. anyway. [laughter] so second with these amendments act as a you have to post the results of any type of trial within one year of completion. to a website and if you do it is a $10,000 per day per find that would be a struggle for most of you i am not judging by and just say. [laughter] the 3.$5 million per year that is a parking ticket if you have of blockbuster drug making a couple of billion dollars. no routine audit when one was finally conducted in the british medical journal last year the rate of compliance was genuinely astonishing the everybody said the we don't know why you talk about publication bias

because everybody knows you have to posted one year the rate of compliance was 22%. for out of five trials ignored the act of 2007 which is the single thing that is most commonly cited for those who want to poo-poo those concerns but no fine has ever been levied in the history of the fda amendment act. that is bizarre. with a combination of it didn't work and everybody is pretending that it did? but even more preposterous, even if it was uniformity it still would have done absolutely nothing with the evidence that you practice medicine today because 85 percent of the treatments prescribed of medicine today came on the

market more than 10 years ago. most of the research is done when it is approved of than fixing everything for the trials might have an impact of medicine in the year 2029, but it will do anything for us now. we need access retrospectively to all clinical study reports today and that is out there. and reconditioned and bunkers and whatever. [laughter] so this is a long and detailed story of the european medical agency which i will not tell you. except. [laughter] only because sometimes you

can get so embedded in a problem that you start to lose your orientation of what is normal and i kept having to say i am bad and my wife and my friends say yes. so this was sent by the european madison agency in response saying we are worried about these drugs we would like to get a hold of the stuff can you tell us what you told? they sent us this. just for every day people from all walks of life the moment they see that they spontaneously laugh. what was going through their there had when they put it in the envelope? were there laughing? [laughter] did anyone say my job has gotten funny lately? it is a strange business.

because now they will now give thoughtful clinical study reports that they hold upon request and they invite you to as if they give it free and fabulous add of their own generosity. the reality is they did in response to one of the most damning and findings the european medicine agency that failed to give a consistent account of why they were withholding -- was holding this information it was extraordinary. we have to stop that for according to this lady. [laughter] but finally the final nail in the coffin but upn medicine agency only hold

clinical studies and reports that they approve in they have only been approving things over the last few years. when it comes alive over the last four years and we have 2002, 1998 and so one. finally did drug tamiflu we stockpiled lettuces -- reduces complications of the flute you don't want to talk about death so they use the word like complications. astonishingly wheel have the evidence of a does that but there was a gold standard summary bellyaching decisions which treatment was worked best they were asked to update their review and they got in touch, and now we really don't have time for that destruction. [laughter]

no. really. we're only one-fifth of the way through. so half of the trial has not been published they got in trouble -- and after the company we want to see not just the results but the fourth clinical study reports that are important is it interesting new thing in medicine because we just heard realize why it is important but even misleading so we can see how was done and measured and so one. they promised they would hand this over on tamiflu now still three years and three months later they have failed to do so.

this is extraordinary of a drug that governments have spent billions of dollars of 5% of the u.k. annual drug budget with the nhl's. gone on this stuff and we don't even know if it is any better than aspirin. is mind bending. that is the first one-third of my talk. [laughter] i know that there is such use your own venture component because i cannot see what time it is. 20 after eight. that is 8:20 p.m. in english money. [laughter] if i talk extremely fast would you like that or shut up and have questions? are you sure? genuine? are you sure? okay. i will try to speak faster.

[laughter] if i do in the american accent is that easier? [laughter] i hope you understand welch. [laughter] right. okay. all randomize controls they're not all perfect and immaculate but in reality they could be flawed which means that they are no longer having the actual estimate of the recent benefits of treatment so i will tell you about the simple ones in before we get there, it is not up coffee ought but the pork and background is the business of missing clinical trial data but there is an anecdote the single story i gave you were illustrations of a wider phenomenon and that weather phenomenon is evidenced by systematic reviews of dozens of studies. if anybody said that is jerry picking it is a major

lier with pants on fire. i am not saying that these design flaws are hugely prevalent some have been more and some have been less but i think they're interesting and clever. first of all, you make it look better than it is by comparing against something that is rubbish. the new treatment against the old treatment against the dosage or a high dose so the other has the worst side effect. were people on the old generation got high doses so with the new ones it looks like it has fewer side effects. also you to make it look better by a comparing it against a placebo, nothing. you can get the drug on the market by only having evidence showing it is better than nothing. that doesn't mean that they

use it but it is important to show the fantasy that some people have the every drug is approved for use is worth using because as a doctor and patient we're not interested in the abstract scientific question is a better than nothing, you're interested is it better than the treatment we already have right now? that is a decision you're trying to make and yet for every single drug approved by the fda 30 percent of those early compared against a placebo when they were approved on the market even though there was the currently available effective treatment for those conditions. it is widespread getting on the market with ising but the placebo. so you roughly get the game and you have to guess where the ball is. you put down $1 if you guess you get $3 back.

how many cups are you allowed to look under in this game? one. if you look under to then you are achieved in at $93 but yet we tolerate this with clinical trial design. for example. if i run a trial of a new anti-depressant i will not just measure do you get better in one way but anxiety in this upscale and how can i do those scales as well and that's upscale i would also use the activity daily living scale also of back to work scale and the general health questionnaire and before you know, what i have 12 different things i am measuring it by measures will different things in a steady, and i allow myself to say if any one of them shows a statistically significant benefit at the end then that means the

treatment caused it and it was a success then i am an idiot. the measure 12 things and you have the 5050 chance even if it is just statistical noise to find a benefit even with nothing happening after all this is preposterous lee common. also of the academic journals have the gate keeper of wisdom but they, in the past trials through with the protocol done before the trial began but it is completely different to the primary outcome in a published paper. if anyone has ever published from the "journal" you know, it is often tedious stuff about throwaway comic in the discussion section and you are the it because it doesn't take too seriously anyway. but people obsess over that stuff they fail to catch the most basic thing did use which the primary all come? next.

you can run up at trial now this is interesting because it shows it shows the vulnerability is of the fabric of how we run medicine makes is vulnerable. we made trials inappropriately expensive. this exception in medicine and we cover them with red tape and then it has the cost per patient you are from a saugh population so they do what they can to wrap up the chances to put the drug into the patient population those who are younger, healthier, and much more likely to get better better, no other drug or medical problems in a much greater chance to show a benefit. but then once you have done the test and the perfect population, as a doctor i say are the results are they

applicable to my population consisting of a lady in her '70's with renal failure on four other medications? satellite know this drug works in the perfect patient population is working here? it feels passive aggressive it is so preposterous that the scale of the disparity between the world of the trial and the reality. there are 5.5 million people in the u.k.. and what they did they took 179 patients from family doctors in the u.k. whose treatment was being managed in accordance with the current best practice guidelines. those are informed by clinical trials and they say we have 179 real world patients treated in accordance with the results. so let's see what proportion would be eligible to

participate in a couple of these trials. down a 6%. 6%. it is not difficult to find people to do trials but yet it is the impossible population. it is on page 172 of the book. [laughter] next. so as we go into marketing feeding trials are absolutely fantastic you have to salute the pain business are haney of city. [laughter] it is your word. so if you have your migraine drug approved anyone to do would trial in the real world population now make grain is the common or rare it is common. soaked with those patients how many migraine cases will

they see a year? >> this is like bread and butter you will breakers -- a record 2,000 patients every time you have an extra location to give everybody a to speed on codes of conduct and the rational way to design the trial if you want to get 2,000 patients is each recruiting 500 patients or somewhere around there. so what do we think when we see a trial that recruits to patients from 1,000 sites? this is says ceding trial because you dare not say just because you see a peculiar pattern you'd be a brave person to say that is a seeding trial they will say no know we want to

spread the risk but it is probably nonsense but you know, is seeding trial if you have access to the court documents. [laughter] that is what we have here and this is why companies are desperate to settle out of court because the court case may be a headline problem but once you get access to the extra documents then people say this is interesting. you wrote about the trial when you got recruited to patients from a dozen clinics and the marketing department designed to. the idea of the seeding trial is not primarily to find out what it does but to get large numbers of doctors to prescribe it for the child. the thing that i love so much is that the matt to the doctors boasting, yes, you

should try this drug because we're having a lot of good results in using you are the embarrassing human being. [laughter] so leslie marketing a classic example a recurring theme of the interlocking ecosystem of problems to reinforce each other not malicious but the vulnerabilities of the system and these incentives and marketing should not matter. is on the issue because of the evidence that we have with the decision makers to ensure it was a rapid accelerated uptake that is like the biggest treatment of pharma also rational disinvestment of things that don't work and we are incredibly bad added.

so if you put me in charge of the one world government and the medical research budget for one year and i can reassure you it will never have been, i can guarantee i will counsel all primary research for one year and we can live without another clinical trial just for one year. i would counsel all transfers and spend every single penny we have to create a better infrastructure for the synthesis of the evidence already have to get it out to the decision makers and putting into practice. currently this is a low status occupation we spend hundreds of millions on a blind child attends of millions but yet on the important business of bringing it together to get it out we dropped the ball. you get prescribing adviser adviser, project what happens if you do know reviews on geriatric patients and it is nonsense.

it is hopeless. because we don't do that stuff we are vulnerable to a marketing. with those decision makers to be informed of what works but we should be above it. i have not place this with the american equivalent but she has of love lee facing gets her breasts done and who'd you used here? a celebrity? i am crowds were saying. you have no equivalent. yours are opera singers and novelist. [laughter] array of sending it comes as no surprise that it probably wasn't written by them. that is no great surprise they basically understand

that and you prefer that. [laughter] now the funny thing is the prevalence is hard to spot and there are shades of gray but with reasonable frequency you find commercial license employed by the public company writes the article submit to the academic journal then they take it to the doctor to say would you like to put your name on this and we will handle the submission and resend it to the "journal" and you get your name in the academic journal and sometimes they have cash to sweeten its. now sometimes the commercial medical licensing system is declared and not very flamboyantly or publicly it is not exactly the lead author on the paper it is a little footnote and mention somewhere and sometimes it

is completely hidden. i think the heinousity of this distorts the contents of the academic literature because there is a bunch of people with money to make with resources to spend a lot of time and offer it -- effort to get their message with the overall discourse. but there is a second concern it is this a bigger concern over the medical work force because who works in academia here? how do you get ahead in academia? you publish. you get academic publications and you become head of the department and so on. what is interesting is where people of need lots of help getting lots of papers out

those people are being selectively propelled up the hierarchy in academic medicine and so we are selectively promoting the people who are most willing to participate and we may regard this as the slightly moral gray area on the edge. [laughter] we're selectively promoting these people but specifically into positions that they accumulate be a parent's of independence it is absolutely bizarre and peculiar like publishing a paper. is that easy or difficult to do? it is difficult. it is a pain. it is difficult business somebody is giving that to you. you should be grateful it is better than getting $3,000.

maybe i should employ a commercial medical writing company. [laughter] i don't know but interestingly with the whole journal, a few journals just appeared in austria a few years ago. someone out to new general practitioners of tens of thousands of doctors and it turns out they were full of reprints of articles about how fantastic merck drugs were. [laughter] the only reason i put this picture up is that they said they represented as academic journals. this is what it looks like. anyway. ridiculous. next. drug companies sales representative store around they talk to doctors and convince them and monitor what they prescribe and they give free samples, a cherry pick information and of course, they are regulated but now it a totally unbiased dissemination of

evidence. i am not impressed by as a doctor and i feel like i am not very clever i can remember the fact that i am not sure it is four years after i returned it to this silly you remember the providence i'm not sure i remember or trust myself with those benefits of treatment whispered in my year by an attractive representative in the revealing talk. [laughter] but this comes of a systematic review from 2011 over 58 studies that the drug reps overall were the same as they have hired prescribing a more likely to prescribe them but why would people spend billions of dollars on this? it goes without saying but of course, it works.

we should be able to talk about this and we should be st. jude doctors to see the drug reps are you sure? also key opinion leaders of fantastic illustration of how this isn't about bad people but the limits of what is acceptable and small tweaks of the corners to end up in the overall process to create biased dissemination of information. he qualified the doctor in the 20's now you are a specialized specialists in your 30's the you practice independently over the next three or four decades now medicine changes around you one doctor that was qualified in the '70s now using medicines invented after they were trained almost universally. they have been self-taught.

doctors are keen to pay for unbiased academic education but the state should be so industry steps in and spends more on doctor/patient education than anybody else. they are taught about which treatments work best by industry and those who make them. i think it should be irrelevant and okay to talk about it. but of course, it is regulated and people point* at the big books but we see whether they actually work but of course, the evidence shows when you have mr. shoppers along of course, it is biased a little bit and it is not corrupt people that is the interesting thing but to

understand the more framework of which this is happened so if there are three doctors, there are hospitals, of the pill hill? up on till hill. [laughter] there could be one rheumatology clinic there are three doctors and two of them for one particular problem there is generic drugs fetter cheap and just as good one will have a genuine belief on completely rational and reasonable reading of literature will say i think this one single brand of drug is better. a genuine belief. not correct but he genuinely thinks. the local drug company wraps

will identify a that doctor who likes that drug and literally give that dr. it platform and a microphone and their message will be amplified for the medical community so they would get a few hundred dollars to teach other doctors some will get thousands to teach specialist where maybe they go to conferences and denies golfing resort or teaching internationally with these programs. now that dr. hasn't changed their view for many and the rap just says it is a person who's genuinely loves our drug we think he is a charismatic person so we will shepard him around to help him share his message. actually they do make the slides themselves. [laughter] but they tried to be nice. so it is not necessary that

people are bad and evil. but with very few exceptions , i draw the line when people deny the existence of problems. i john the line when people say the color fall denunciations it is a systematic preview and they say you're the cherry picking conspiracy theory. i draw the line when people say these real problems, a just shut up you're not allowed to talk about them but the incentives are misaligned and the information architecture has several interesting technical flaws, and we should fix them. we should fix the vulnerabilities in the system that leaves it open and force people for all clinical trials it is

strange that we don't. and i genuinely believe the people of the future, the doctors of the future will look back on this era of medicine how we tolerate half of all clinical trials in the same way we look back at medieval bloodletting and you spent hundreds of millions of dollars on producing these carefully designed clinical trials to detect very modest differences over when treatment and the other than after words several times over you completely throw the baby out by letting half of them go unpublished and you knew it was a problem for decades and do nothing about their. they're unable to believe how stupid we are. thank you very much. [laughter] [applause]

>> if you have a question come to this microphone and form a line. we have 20 minutes so please keep your comments in the form of a question. >> thank you for your talk. one of the observations is that i am in biochemistry and pharmacology it takes six to seven years to provide day manuscript others to half a dozen per year plus half a dozen reviews and could do speeto that also the enormous proliferation of medical journals and science journals every week i received two or three invitations to be the editor

of eastern africa in general mec. [laughter] so there is the exclusion of thousands that are trying to get you to become part of the editorial board but it is the virtual journals online that are receiving profit from that model. >> the issue of those journals is different but the thing that i find interesting is to the extent architecture of evidence based medicine and is the ad hoc ecosystem from legacy systems demonstrably outdated. the volume of paper is being published today in tens of thousands or millions of manuscripts every year clearly requires a systematic approach but yet this is at the incredibly

low priority. those clinical trials that are identical experiments that are all pretty much the same and the outcome to measure is different but basically every trial is the same and instead of forcing people to report to them in a structured data format we let them write the six page essay and it is utterly bizarre so it is a part of the microcosm of how we haven't thought carefully enough about the infrastructure of the scientific and data. we need to be building the framework instead of piling up in a room. it is bizarre.

>> stake you for coming sid nobody out there can see your eyebrows. nevermind. carry on. [laughter] >> this is the jump off but in the mid 2000's they could bring that and a ah as the actual administrator for all clinical trials and have all funding and approval and eventually all results coming out of the national institutes of health rather than this clinical trial any feedback on actual implementation of results we could achieve in the next five years to have more transparency? >> the problem of incomplete dissemination from the publication bias that have

been sent academia to in research frequently. i would not hold them up as the hope but people have this very different fantastic notion of how to fix this like nationalizing the industry or make it all centrally administered to put the money into a big pot than a central body does it. i am not sure that is necessary because i am not sure we tried the basic stuff yet. with the missing clinical trial results we have not really tried asking even very carefully. we have clinical trials but nobody has been called off for not having over. and we're not asking for very much. when you look at the trials with the extraordinary or partially overlapping

patchwork around the world of the incomplete list of trials no way is anyone saying we need a list of all trials and have been conducted with what we have been using right now. so it is used as part of the marketing approval and it was held entirely in secret from march 2011 with the unique transparency with the entire contents of the european clinical trials register a assault from europe in 2004 but it should be running the european register as a list of tiles from europe and that is what the public would expect us to be doing. but before we get to centralize the entire

process of billions and billions of dollars with the massive political upheaval upheaval, we could try by forcing people to report appropriately and the agency should say to the company's and to say it is a list of files that has ever been done in here are the forms could you jot down the trial simi will put them up on the site and somehow we think it is okay even though they know that is the market over 30 years. don't worry about that but i try asking and fixing the current model before i try

to completely turn everything upside down. other than barricades' and stun guns. [laughter] >> i wonder if anyone reviews the reviewer's? are you from the u.s. dr. dr. tom jefferson that has the anti-flu vaccine as a hazard went on the carry knowles radio show that is similar to coast-to-coast. i have seen the stuff he has written he compares the analysis of the flu vaccine effectiveness but then he compares it to the side effects that were not in the analysis but numbers that people's out and i have seen things where he has said the date that it shows a mild effect but in his summary it shows no defect. that was the wording and i just wonder if you are familiar are how do we keep

reviewers in check? >> i do know tom jefferson but don't know anything about the flu vaccine story but i think he is a pretty good guy he did the tamiflu review and any reviewer you can comment on actually more easily said to get a letter in the academic journal but the problem with tamiflu there is the spotted after the british and austrian governments and then the japanese doctor came along to post a comment on the cochrane site saying i am worried because most of the trials in this tamiflu review come from one analysis that describes one dozen trials where all but to but have never been

published. that is weird because normally you would expect people to carefully and critically appraise the literature instead of relying on a brief summary that was done by somebody else and then put the numbers into your own review. immediately as a fantastic example of the group said you are right and that is sloppy. we will have the look and that is how these tamiflu saga began because somebody reviewing cochrane and said i think that is rubbish. you have not done it properly and those of evidence based medicines at oxford said we will do it properly now. so if you have concerns the systematic reviews.

>> but then post them online. suppose then there and criticize. suppose then there and criticize. and the answer is you. and it that is exactly how science should be. say hello after words and sounds like you have more to say. >> but there were very specific things. >> i am not the ombudsman's especially on a talk radio -- review. >> my question is from the perspective of the health insurers but we do have access to data over time and different affirmation coming together.

if you we're doing research on behalf of the house in scheerer, what kind of questions would you be asking of your work to steer the ship to a better outcome and/or what changes to the way of payment policy or pharmacy benefits that would try to effect? >> the epa ought to do it without massive cultural shift and i would use my power as the payer to address market failure in answering really important relevant questions about which treatment is best that they have no financial interest in addressing. said you have the potential to hold administrative data how patients are doing and in the u.k. to give you a parallel example, i work on a project involving

electronic health records we have 3 million people's records in the database as part of the nationalist utopia and bulleting and disease but which of the two treatments is best so they are running a trial against each other to see which is the best and we run for that half a million klay it that the traditional way is tens of millions this is a really long answer but apparently you can get an extension until 5:00. [laughter] the project is all of those on the market are against the placebo and since they have been shown to reduce

your chances of dying. they have been compared against each other but not for which is best to shown to be equivalent for how much they lower your cholesterol. now to do a big trial compare one against the other to prevent you from dying would be a big expensive than long project you have to get follow-up data in one of the annoying things of medical research is people take ages to die. [laughter] see need to find a way to follow people love so our project is basically you go to the doctor if they decide if you need a statin then as they prescribe it the screen says way we don't know which of these is the best press the button to randomly assign a universe speak about it ever again this addresses the issue doctors don't get involved because

and they have to come back for the blood test and also they are anxious over some new experimental thank you compare the two that are shown to be safe than tens of millions of people around the world then you get the follow-up data so you can see if they have of heart attack from the data if they die, and cause of death, you can pick up things of side effects and so one. the potential here by harnessing the data we already have is that they could turn the health system into machines that are into machines that are burning and adapting so i am not proposing this new drug or reid don't know what works best as situations where nobody could possibly care what statin they got because nobody has an idea which is better but they are

so widely used in heart attack is a common cause of death it could only be one for 2 percent better it would be massively workers but it is not worth the companies while to find that out but it is our interest in the government we just have to persuade the populace that it is okay to do. by the way having doctors to have the humility to say i don't know which of these two is the best that requires we hold the line because of one doctor says i know what is best for you look at my cheesy smiles then that dr. blows the game but when i say to you it is a broken egg dishes have been the title of the book. [laughter] but another example of how we fail i have a whole chapter of this in the book

the we're not constantly in identifying and resolving it see you have the data like we could say any new treatment has been approved for use that is expensive not currently in the data to show you can only get it you could do that and save you want to get the new treatment? we genuinely don't know if it works so the right thing to do is to minimize the amount of time where we continue to not know during which rarely some people will be given a treatment that will subsequently turned out to be less effective. let's find that out as soon as possible. could you do that? >> no. [laughter] so we can count on having a member for a few years may

be and making health care less expensive doesn't make us more profitable but only if we could make more money relatively? i know how to solve that. >> so you could compete on other aspects but you could collaborate on gathering data resolving the uncertainty. >> what i was asking was are their report cards for pharmaceutical manufacturers as far as how dodgy they are? [laughter] or do those payments scales based on that? >> yes. yes. yes. yes. i should have told you this i and a bad person. >> we do have to move onto the signing at this moment

unfortunately. i want to remind everybody he can't sign books to the table to the right after that talk. >> the book came out in the u.k. and lots of nice things happened so now this is what happens if you went to visit the health minister and he said we agree it is a problem and have editorials and academic journals and in they have a special inquiry into this problem. [laughter] and in the european peace. [laughter] and then they decided we needed campaign and the whole trial stop there. it is now just 30,000 signatures and now with the medical research and the council but also over 80 patient groups if you are a member of any professional body actually make them sign off. and last week it glaxo

smithkline signed up one of the biggest companies in the world. [cheers and applause] and in answer to question it is interesting because we're now in a situation where you may say they're too treatments we don't know which is best but the apparent benefits of the two are equivalent but one of them is made by a company whose as we give you everything one is made by a company not willing to say that i think actually once you have a situation with the imbalance of how much transparency there is, doctors and payers might find themselves starting to say do you know, what? if they are equivalent of one of these is not made a commitment to share the evidence they have got that i will use the one because that gives money for transparency and that is the fix is that. thank you very much. [applause]

>> not mean to put you on the spot but representative, what is on your summer reading list? >>. [laughter] i am reading the wonderful book the hopkins touch halfway through that now that is the legendary aid to fdr and we probably don't have the same politics but it is a compelling life. probably next i have not had a chance to read the kaiser book but the reviews have been compelling and that will be an interesting case study. when you read a book when you know, the characters senator dodd, barney franken the legislators it is interesting to get that perspective and some of the staffers. there is a book i have just

ordered on james byrne legendaries south carolina from a political reporter said he will love this book and a guy who's very nearly vice president instead of truman in 1944 and continued to play an extraordinary role in politics and became one of the architects of nixon's success in the south '68 and '72 and it just popped up working with harry hopkins with this book i am reading on the 1940 nomination fdr of the third term which is neat political work. i like to read about the process and to study history and i should read more policy unless history but i learned better from history.

>> in excellent health setting the record straight on america's health care that is the name of the book. the author is dr. scott atlas, also a senior fellow here at the hoover institute on the campus of stanford. dr. alice what works about the american health care system when you look at it

was a larger sense? given that could take up the entire interview. what works really is in essence of what a lot of what the book is about the actual medical care both the availability or access as well as implementation or introduction of diagnostic and therapeutic treatment of diseases. in fact, a lot of the impetus was declared by the background information that people don't really have despite what it has been said about the health care system. the bottom line it is both superb access and quality of care. . .