1
00:00:04,170 --> 00:00:07,690
Okay. Hi. Welcome, everybody. Welcome.

2
00:00:07,710 --> 00:00:14,250
Some of you started the must've from stragglers from the department of wandered in to see how you might contribute to IPM.

3
00:00:14,730 --> 00:00:17,400
For those of you I know, most of you. My name's Carl Hennigan.

4
00:00:17,400 --> 00:00:23,430
I'm professor of evidence based medicine and director of the Centre for Evidence based Medicine, and I'm also a GP.

5
00:00:25,050 --> 00:00:29,400
This talk, basically, I figured for those in the math as it applies to everybody,

6
00:00:29,400 --> 00:00:38,970
if evidence based medicine is often seen to people like something that's really difficult to get involved in and actually actually can be quite scary.

7
00:00:39,300 --> 00:00:46,200
So after a busy day in the masses, this is a talk about sort of some of the things I've learned over the last 20 years that have really

8
00:00:46,200 --> 00:00:53,520
profoundly influenced me in the way I think and how I go about my contribution to evidence based practice.

9
00:00:54,120 --> 00:00:57,480
And some of you might start to think about how you may contribute.

10
00:00:57,930 --> 00:01:02,129
So I'm just going to ramble around from 50 years worth and feel free to interject.

11
00:01:02,130 --> 00:01:06,600
Stop me any time you don't quite understand what I'm saying or you want to stop.

12
00:01:06,600 --> 00:01:09,810
And maybe at the end we'll have a little discussion about how you might contribute.

13
00:01:10,440 --> 00:01:17,999
So I think this is an important photo and this is an important landmark in how we

14
00:01:18,000 --> 00:01:24,030
think because thalidomide with a really important issue in terms of regulation.

15
00:01:24,480 --> 00:01:30,150
And this is John F Kennedy giving the gold medal or bronze medal at the time to Francis health.

16
00:01:31,080 --> 00:01:38,820
Because in America, what they said is we're not going to allow you to have thalidomide unless it's in the context of a clinical trial.

17
00:01:40,650 --> 00:01:44,460
And that was not for the pregnancy issue. That was because of the neuropathy problems.

18
00:01:45,180 --> 00:01:51,660
That meant you were about 100 times more likely to have a child with fluid amide in the UK than America,

19
00:01:52,170 --> 00:01:59,910
despite the fact they outnumber 6 to 1 in population. And at the time what happened is drugs were just basically brought about in the market,

20
00:02:00,390 --> 00:02:03,540
released on the market, and people then said, let's see what they do.

21
00:02:03,870 --> 00:02:09,329
And with this drug, they got spontaneous reports that actually it's really good in nausea and then said,

22
00:02:09,330 --> 00:02:14,700
oh, well, there's a really good group to come to the fore. There's a really interesting group of people that this could apply to.

23
00:02:15,390 --> 00:02:16,560
And it's pregnant women.

24
00:02:18,090 --> 00:02:25,800
And at the time, John F Kennedy said, I'm going to put £2 million into the US FDA so that this problem would never occur again.

25
00:02:26,730 --> 00:02:30,630
So that really is the birth of regulation around about the fifties today.

26
00:02:30,650 --> 00:02:40,379
You've touched on this today. It's a really important interest in arena that you talk about the idea of antiarrhythmic and that's a real problem.

27
00:02:40,380 --> 00:02:44,190
And it was the birth of in the eighties of evidence based practice, if you like.

28
00:02:45,000 --> 00:02:51,120
This is an arena and area where people were pretty clear about clinical practice and

29
00:02:51,120 --> 00:02:55,920
then along came this clinical trial and really turned over what we think we fall.

30
00:02:57,270 --> 00:03:04,349
And often you should read stuff you've got to read if you're going to be influenced and think about the bigger, wider issue.

31
00:03:04,350 --> 00:03:10,860
And this is a great book, actually. Thomas Moore, why tens of Thousands of Heart Patients Died in America's Worst Drug Disaster.

32
00:03:12,180 --> 00:03:15,870
I would like to make my job participatory as well. He feels good about reading.

33
00:03:16,260 --> 00:03:19,860
Feel good about reading. Can you read? Okay, good. We get the variety.

34
00:03:19,860 --> 00:03:26,640
Students are not thrilled if they just read out. And I said loud and clear and slow and just read the first sort of ten lines for everybody.

35
00:03:28,650 --> 00:03:34,110
This book tells the story of America's worst medical drug disaster over just a few years.

36
00:03:34,110 --> 00:03:38,760
An estimated 50,000 people died from taking drugs intended to prevent cardiac arrest.

37
00:03:39,540 --> 00:03:42,660
After hundreds of thousands of patients routinely took these drugs,

38
00:03:43,140 --> 00:03:47,850
a definitive medical experiment proved that they did not prevent cardiac arrest, as doctors had believed.

39
00:03:48,780 --> 00:03:50,940
Instead, the drugs caused cardiac arrest.

40
00:03:51,810 --> 00:03:57,210
Often the effect was so sudden and unexpected that people literally dropped dead while going about their normal lives.

41
00:03:58,050 --> 00:04:02,000
The result of this single medical misjudgement about the properties of these drugs produced

42
00:04:02,010 --> 00:04:07,170
a death toll larger than the United States combat losses in wars such as Korea and Vietnam.

43
00:04:09,030 --> 00:04:12,810
Okay. Thanks. Quite interesting, isn't it?

44
00:04:12,990 --> 00:04:16,000
Quite powerful when you think about what's really going on,

45
00:04:16,040 --> 00:04:22,530
the capacity of medicine to do benefit but great harm and that we all start in equipoise is really interesting.

46
00:04:23,040 --> 00:04:34,020
So superb evidence. Both practices born all disappear, all the problems disappear around about just early late 1999, 2000, I become a House officer.

47
00:04:34,290 --> 00:04:41,999
In fact, at the back there's Jeffrey Aronson, who now works in the department it with my consultant at the time and I oh yeah we practised.

48
00:04:42,000 --> 00:04:49,559
It's all disappear but I want you to talk go forward to about 2000 and I'm going to talk to you about a particular drug in a minute.

49
00:04:49,560 --> 00:04:52,410
But this is where the birth of evidence based practice comes from,

50
00:04:53,040 --> 00:04:57,180
is the integration of breath research, evidence with clinical expertise and patient values,

51
00:04:57,630 --> 00:05:04,650
trying to bring evidence to decisions, which seems like any time I ever speak to anybody who's not in health care,

52
00:05:04,650 --> 00:05:07,889
they say, Well, that's obvious, isn't it? Everything should be based on evidence.

53
00:05:07,890 --> 00:05:13,560
Why do we need a professor evidence based medicine? Surely you're out of a job, you're redundant.

54
00:05:13,830 --> 00:05:22,230
And actually it seems nonsensical to patients and the public that medicine would be practice without evidence.

55
00:05:23,220 --> 00:05:28,890
Yet it continues to be and continues to be in all sorts of ways influenced by a lack of evidence.

56
00:05:29,910 --> 00:05:32,910
But I want you to take forward to 2000. So I'm a House officer.

57
00:05:33,810 --> 00:05:35,700
I work in a team called Firm Fee,

58
00:05:36,450 --> 00:05:46,709
and there's a particular drug called via a Cox two inhibitor that is being marketed as a pain reduction medication that's advantageous over current

59
00:05:46,710 --> 00:05:58,200
non-steroidal because of its lack of gastric adverse events and is widely being used by people who I'm around because there are lunches going on,

60
00:05:58,200 --> 00:06:06,150
there are people being taken out to dinner and I reckon about a thousand people within our care within a two year period were put on this medication.

61
00:06:06,180 --> 00:06:10,710
So I want you to hold a thousand people. Many, Jeff, in the team would say about 5000 people.

62
00:06:11,130 --> 00:06:16,980
It could be 30. Say we take a thousand. Now, what's interesting, when you come to this Vioxx, this is actually, again,

63
00:06:16,980 --> 00:06:24,980
really shocking is results from the clinical trial in 2000 show five times the number of heart attacks a month,

64
00:06:24,990 --> 00:06:28,530
patients taking Vioxx compared with those taking naproxen.

65
00:06:29,160 --> 00:06:33,360
Merck say the reason was naproxen, strong protective effect on the heart.

66
00:06:34,140 --> 00:06:39,690
So they argued that way. 2001, the company's internal intention to treat analysis,

67
00:06:39,930 --> 00:06:44,639
which are all going to start to learn about that means the people, everybody, not the ones just on the drug.

68
00:06:44,640 --> 00:06:52,170
If you dropped out, identified a significant increase in total mortality nearly five times four and a half times.

69
00:06:53,070 --> 00:06:58,560
Okay. In one study. And the second study was 2.55, so nearly two and a half times more likely.

70
00:06:59,820 --> 00:07:09,690
Overall, that meant there were more 54 death around amongst the 1069 compared to 12 deaths among the placebo.

71
00:07:10,050 --> 00:07:16,350
So if you round that to about a thousand, that's around about 20 patients, give or take a few.

72
00:07:17,070 --> 00:07:22,680
Do you see what we were doing? Anybody see the significance of of putting a thousand people on average on this treatment?

73
00:07:23,070 --> 00:07:28,890
We're actually assigning people in the intervention to increase mortality thinking we're doing the right thing.

74
00:07:30,130 --> 00:07:36,510
Carl, actually, what happened in the jump was that John Reynolds, who was chair,

75
00:07:36,930 --> 00:07:42,610
said we should not have this struggle to formulate in 2000 yet all the people

76
00:07:42,630 --> 00:07:48,870
we saw we're getting it outside were coming in the hospital in huge pressure.

77
00:07:48,870 --> 00:07:54,900
At that time, pharmacy reps would go to the street to general practice, talk to people about this.

78
00:07:55,440 --> 00:08:01,649
And that's an incredible just hold that point. Jeff said, let's think about that as a signal 2001.

79
00:08:01,650 --> 00:08:04,229
They told them there was no increase in risk of death. Why?

80
00:08:04,230 --> 00:08:11,969
They knew it was three times the FDA said, could you just send a letter to everybody saying, oh, it's it's a bit unsafe.

81
00:08:11,970 --> 00:08:22,500
Think about it. Now, when you go to when you look at evidence generally is you don't need statisticians to interpret what's going on.

82
00:08:23,250 --> 00:08:29,190
To be honest, when you look at a graph like this and you look at this is a cumulative death rate and say that the intervention in the placebo,

83
00:08:29,370 --> 00:08:34,680
you don't need to be a statistician to work out what's going on. You can see there's an increased mortality.

84
00:08:36,030 --> 00:08:42,300
Yeah. What happened with this is that this was systematically withheld from clinicians for around about three years.

85
00:08:43,700 --> 00:08:53,090
And it only became in 2003 for apparent that this three times mortality effect was happening and they consistently withheld evidence.

86
00:08:53,750 --> 00:09:00,540
And what was interesting at the time is this is from Catherine DeAngelis, who was editor of one of the big journals at the time.

87
00:09:01,550 --> 00:09:06,800
This is her and this is her quote. This is occurred because the physicians allowed it to happen and it's time to stop.

88
00:09:08,250 --> 00:09:14,549
And that's because sometimes commissions and everybody's complicit in not using evidence based medicine principles.

89
00:09:14,550 --> 00:09:20,760
And what Jeff said to you with John Reynolds, who is a very smart physician, looked at the evidence and said, there's a problem here.

90
00:09:21,150 --> 00:09:25,139
We should follow the evidence base as opposed to not using the evidence and saying, well,

91
00:09:25,140 --> 00:09:29,280
there are all these pressures, all these guidelines, all these commercial interests telling you to do something.

92
00:09:30,750 --> 00:09:37,230
And that's very profound to me in my thinking to say, gosh, you can really get it wrong very rapidly if you're not careful.

93
00:09:39,480 --> 00:09:46,799
And then we come to the steps of EPM, which is about critical appraisal validity and thinking about these issues, about applicability.

94
00:09:46,800 --> 00:09:51,840
And then this morning before this particular trial, which is rosiglitazone the dream trial.

95
00:09:52,020 --> 00:09:54,540
Great name is the dream trial, the Lancet,

96
00:09:55,020 --> 00:10:01,410
one of two journals in the world that have a profound impact on medications in practice, surgical interventions in practice.

97
00:10:01,680 --> 00:10:05,940
Then two journals really do have a monopoly on these types of studies.

98
00:10:06,720 --> 00:10:10,440
But what's interesting is when you do read this and you read the bottom two lines and I

99
00:10:10,440 --> 00:10:14,610
can take people who've never practised before now nothing about evidence based medicine,

100
00:10:14,910 --> 00:10:21,030
and I get them to read. And at the bottom they read this cardiovascular event rates were much the same in both groups,

101
00:10:21,360 --> 00:10:28,500
although 14.5% participants in the glitches are group and 2.1% in the placebo group developed heart failure.

102
00:10:29,970 --> 00:10:31,860
Well, when you think about that, that is nonsense.

103
00:10:31,860 --> 00:10:40,019
Even as a GP, I know the heart is part the cardiovascular system and even my kids can tell you that from gxp biology.

104
00:10:40,020 --> 00:10:48,510
So that statement can't be right. And if you look at what it is saying, is it saying it's five times more likely, 5 to 1 simple math?

105
00:10:50,540 --> 00:10:54,530
And the P value is .01.

106
00:10:55,100 --> 00:10:59,800
So 99 times out of 100, this is the effect you get. And one time in a hundred this could be my chancellor.

107
00:11:00,920 --> 00:11:08,720
You likely if you expert. So if you just think about it, you repeat this experiment 100 times, you're going to get this effect 99 times out of 100.

108
00:11:10,100 --> 00:11:17,180
That's not rocket time when I said that. So an oh my God. So when you see signals, you should start to take them seriously.

109
00:11:18,020 --> 00:11:19,700
And that's been profound in the way we think.

110
00:11:20,390 --> 00:11:24,260
And when I wrote this, I wrote this to The Lancet and they said, Well, we're not going to publish the editorial.

111
00:11:24,260 --> 00:11:28,460
We don't want editorial from you. Hennigan So I rang up Fiona Godlee and said, This is a disgrace.

112
00:11:29,240 --> 00:11:35,750
And we publish that immediately that the study came out. The number 82 harm free years is 250.

113
00:11:35,750 --> 00:11:39,979
So for every 250 patients you put on there, you're going to have one case of heart failure.

114
00:11:39,980 --> 00:11:43,850
And this is in really low risk people. This is not in high risk people.

115
00:11:45,830 --> 00:11:50,240
So you say, well, that's pretty low, but think about it, this was a $3 billion drug at the time.

116
00:11:51,760 --> 00:11:55,180
It's been hundreds of thousands. We're back to the corks to argument.

117
00:11:55,180 --> 00:11:58,060
We're back to the Antiarrhythmic story again.

118
00:11:58,780 --> 00:12:05,860
And what's interesting is when you look at the timeline and go back and look in the regulatory timeline about Rosiglitazone, you can take it back.

119
00:12:06,310 --> 00:12:12,290
It was approved by the FDA in May 1999 and they asked for a doctor, the phase four trial.

120
00:12:12,310 --> 00:12:16,150
It was 2000 by the European Commission record as a phase four.

121
00:12:16,150 --> 00:12:17,620
So they wanted some further trial.

122
00:12:17,650 --> 00:12:27,940
They said we thought they had all this final document on violated on the cardiac disease signal start to appear here from the firm's working group.

123
00:12:27,940 --> 00:12:34,959
An Analysis of CV Events 2004 Summary First Meta Analysis Meta Analysis by Steve

124
00:12:34,960 --> 00:12:40,210
Nissen And to have said there's a 43% increased risk again of heart problems in

125
00:12:40,210 --> 00:12:46,600
this group and here 2010 we go down box warning here FDA gave a box warning in

126
00:12:46,600 --> 00:12:52,299
2010 suspend sales and again if you start to dig around the information you find,

127
00:12:52,300 --> 00:12:55,330
this is in the original 1999 medical review.

128
00:12:55,960 --> 00:13:00,100
The increase in body weight and undesirable effects on serum lipid is a cause for concern.

129
00:13:00,490 --> 00:13:07,360
Heart disease is a major cause of morbidity and mortality. It cannot be assumed that Rosiglitazone will decrease the CVD risk.

130
00:13:07,690 --> 00:13:11,620
My concern about delivery theory with long term effects on the heart should be addressed,

131
00:13:12,190 --> 00:13:19,000
but with the 11 years before it withdrew off, the market was the biggest selling diabetes drug around about 206.

132
00:13:19,540 --> 00:13:26,169
So very interesting again about this idea that treatments have benefits and harms and out there is lots of information

133
00:13:26,170 --> 00:13:32,079
around treatment and your job in evidence based practice is if you're smart enough is we'll get there in the end,

134
00:13:32,080 --> 00:13:38,200
but you'll get there about four or five years before anybody else because like what you just said about John Reynolds is you go,

135
00:13:38,200 --> 00:13:42,370
hey, I don't understand this. And it's the same for benefit and for harms.

136
00:13:42,580 --> 00:13:46,540
It works both ways. People who are doing stuff beneficial get there, really.

137
00:13:46,630 --> 00:13:55,940
They stop doing the harmful. I'll move on again then. So that's just so interesting and this is probably again really being quite enjoyed.

138
00:13:55,960 --> 00:13:59,560
This is Margaret Chan in 2009, you'll remember the swine flu pandemic.

139
00:14:00,850 --> 00:14:06,070
Anybody have swine flu? Yeah, I caught it when I came to my first class here.

140
00:14:06,250 --> 00:14:10,150
Okay. And he's still alive to tell the tale.

141
00:14:12,040 --> 00:14:13,899
That's the definition of a pandemic, which,

142
00:14:13,900 --> 00:14:21,610
with change for this means it should cause serious an epidemiological infection that has widespread and causes serious morbidity.

143
00:14:22,240 --> 00:14:26,709
The definition with change to suit swine flu at the time and at the time.

144
00:14:26,710 --> 00:14:33,010
This is my girl from 2009. They're a bit bigger now. And tell me to shut up, dad, and go away because they're teenagers now.

145
00:14:33,010 --> 00:14:38,409
But at the time they used to listen to me. Then the question is not this was my girlfriend.

146
00:14:38,410 --> 00:14:43,000
And what happened is my sister also rang me. She said, I just rang up the doctors, we've got a fever.

147
00:14:43,180 --> 00:14:47,410
And they said, you need to go to Northampton to pick up a prescription of Tamiflu.

148
00:14:48,100 --> 00:14:53,680
And she said, What do we do? Call? And I thought, I've no idea at this time.

149
00:14:54,520 --> 00:14:59,020
It's a systematic review that I thought had been done in adults. Oh, this is not that interesting.

150
00:14:59,020 --> 00:15:03,339
So what we did is we'd been involved with a group who'd been doing the Cochrane Review.

151
00:15:03,340 --> 00:15:08,080
We said, Let's prioritise doing the systematic review in children, and that's exactly what we did.

152
00:15:10,560 --> 00:15:20,910
And that was the major headline that came out. The research is I was astonished that the lack of evidence base for trials in the world for children,

153
00:15:21,330 --> 00:15:24,659
less than 1500 children had been in randomised trials for this treatment,

154
00:15:24,660 --> 00:15:30,730
despite the fact influenza is hundreds of thousands, millions of people infected at any one time.

155
00:15:31,050 --> 00:15:39,540
You could do this trial in a week or two. And what it showed was that basically it has a small reduction in symptoms,

156
00:15:39,540 --> 00:15:43,469
no effect on complications, and it makes children have diarrhoea and vomit.

157
00:15:43,470 --> 00:15:46,440
About one in 20 children have serious vomiting and diarrhoea with it.

158
00:15:46,740 --> 00:15:51,510
And I was a bit shocked and we published that with major national news with all over the TV,

159
00:15:51,750 --> 00:15:55,410
and it had no impact whatsoever on the prescribing of Tamiflu.

160
00:15:56,190 --> 00:16:01,950
In terms of guidelines, policy, in terms of GP's, I started to get messages from individual clinicians going, Yeah,

161
00:16:01,950 --> 00:16:06,600
this is really good, we think you're right and this is what we don't want to be pressured into prescribing this.

162
00:16:07,350 --> 00:16:12,299
But there's been an interesting change in how the world is now pushing us with technologies.

163
00:16:12,300 --> 00:16:16,920
And if you think drugs, devices, surgical, they're all technologies.

164
00:16:18,480 --> 00:16:22,469
And why is this once it's come about in effect, why do we slightly think differently?

165
00:16:22,470 --> 00:16:27,780
And I think there's been a profound change in about the last ten years around the business of medicine.

166
00:16:29,040 --> 00:16:35,429
And this is 2005, and this is an article in The Independent about good old Donald Rumsfeld.

167
00:16:35,430 --> 00:16:38,489
What was that? There's no known knowns or unknown unknowns.

168
00:16:38,490 --> 00:16:43,649
But he may he was actually one of the shareholders in Gilead who had the licence for the drug.

169
00:16:43,650 --> 00:16:49,350
That's why he made a lot of money. But one of the things is he has for a UN report on how many predicted death would bird

170
00:16:49,350 --> 00:16:54,990
flu remember bird flu that's before swine flu court So if I asked you worldwide,

171
00:16:55,650 --> 00:17:01,440
how many predicted death would you think in this model that the UN to put together the worldwide

172
00:17:01,710 --> 00:17:06,270
number of deaths were to be predicted at that point and anybody would like to hazard a guess.

173
00:17:07,140 --> 00:17:12,960
You can phone a friend 1000 thousand, 10,000.

174
00:17:13,440 --> 00:17:17,190
You're going to go for 10,000. You're going to go for 10,000.

175
00:17:17,190 --> 00:17:21,900
What about the third? How many predicted death? They looked at it. They said bird flu.

176
00:17:22,440 --> 00:17:25,470
It's got a few cases. The been about 9000 deaths so far.

177
00:17:26,840 --> 00:17:30,560
You've got ten. Hmm. You're going to go for 100,000.

178
00:17:30,950 --> 00:17:39,680
So it was 150 million deaths is what the UN predicted and the number of deaths with 700,000 that might be expected in the UK alone.

179
00:17:41,210 --> 00:17:51,290
That's when we pull our first order of 14.6 million courses of this drug started stockpiling the drug because we got particularly worried,

180
00:17:52,550 --> 00:18:01,070
despite the fact that only 90 deaths so far. So the real commercial operation around some of these pushing you into and I could show you the

181
00:18:01,070 --> 00:18:05,990
really interesting 205 the marketing material said flu is a disease that everybody stays at home.

182
00:18:06,140 --> 00:18:10,640
We're going to make you all for your doctor. And when you see doctor, what happened?

183
00:18:10,640 --> 00:18:15,770
You're going to get intervened. In fact, you're going to spread it more quickly and we're going to intervene.

184
00:18:15,770 --> 00:18:20,010
We medications. And that's what the marketing material said. But I particularly like this.

185
00:18:20,050 --> 00:18:26,690
See, this is the way you get in in in problematic thinking because of the way the news media now comes to the public.

186
00:18:26,690 --> 00:18:28,700
And you and this is Hans Rosling,

187
00:18:28,700 --> 00:18:34,520
who if you go and Ted has done some remarkably good talks about this whole rather and it's called the news to death ratio.

188
00:18:35,210 --> 00:18:43,400
And so you can look at the number of death in the first 15 days from 24th of April, when we first learned about swine flu to the sixth May to oh nine.

189
00:18:43,580 --> 00:18:47,480
And you can compare that to TB and you might want to think about that 13 days

190
00:18:47,960 --> 00:18:51,710
in terms of now what's happened in Ebola and think about it in context of TB,

191
00:18:51,710 --> 00:18:52,800
you could do the same for Ebola.

192
00:18:52,820 --> 00:19:03,320
In fact, the number of deaths in that first 13 days with 31, the swine flu in the same period, the number of deaths for TB 63,066.

193
00:19:03,890 --> 00:19:11,300
The number of news articles with 254,442, the swine flu and 6501 for TB.

194
00:19:12,260 --> 00:19:20,330
That means if you see one, you have to see 8176 news articles for one death when it comes to flu.

195
00:19:21,350 --> 00:19:25,640
However, if you see one article on TB, ten people have died.

196
00:19:27,050 --> 00:19:32,600
When was the last time you ever thought to myself, Oh my gosh, people are really talking about TB as though it's an important issue.

197
00:19:33,380 --> 00:19:38,450
We should really be solving this. So this is almost what we always see is the inverse care law,

198
00:19:38,450 --> 00:19:42,989
which is if you want to read something interesting going back is inverse verse was a chap called

199
00:19:42,990 --> 00:19:48,080
Choo to Heart who said health care is often disproportionate to the needs of the patient.

200
00:19:48,320 --> 00:19:54,890
And he was a man in a mining village in Wales, in South Wales, and he said, look at all these people with chronic airway disease,

201
00:19:54,950 --> 00:19:59,989
high blood pressure, they get zero care and all the healthy people get all the attention and the care.

202
00:19:59,990 --> 00:20:03,330
And it's really interesting if you look at that, it's not sexy TV.

203
00:20:03,440 --> 00:20:09,320
It seems to be in less deprived populations, lower socioeconomic, classic people are not interested in it.

204
00:20:09,620 --> 00:20:17,440
But boy, this is a big problem globally. And if you think of that number, 63,000 in nine, in 13 days.

205
00:20:18,670 --> 00:20:23,590
Ebola is really starting with that 4000 since about Christmas.

206
00:20:23,830 --> 00:20:27,130
Important, yes, but nobody is worried about this.

207
00:20:27,680 --> 00:20:35,649
So what happened is I was so amazed at the lack of evidence that I managed to meet with a call from Jefferson.

208
00:20:35,650 --> 00:20:38,290
And this is Deborah Cohen, who we do a lot of work with.

209
00:20:38,290 --> 00:20:42,960
And Jeff knows Deborah Cohen, I mean, started in the adult what's going on here in the adults.

210
00:20:42,970 --> 00:20:47,890
And he said, well, we think that's a bit dodgy here and we're not happy.

211
00:20:48,430 --> 00:20:54,309
And I'm just me. This is a whole group of us who ended up meeting, and this is it in Oxford.

212
00:20:54,310 --> 00:20:59,889
There's Tom Jefferson there. There's a few of the people, Peter Doshi and Chris Del Mar from Australia, a few of the people.

213
00:20:59,890 --> 00:21:05,890
But we ended up combining our efforts together because the Japanese shepherd sent

214
00:21:05,890 --> 00:21:09,940
an email to Cochrane and if you send an email to Cochrane reviews to the authors,

215
00:21:09,940 --> 00:21:14,260
you have to respond within six months and now we could go off into that, he said.

216
00:21:15,070 --> 00:21:23,140
Of the ten trials you did, you published in the Adults and used in your Cochrane review, if you actually look at it, only eight.

217
00:21:24,320 --> 00:21:26,660
Only two of them have ever been published.

218
00:21:27,590 --> 00:21:36,559
The other eight were included in a meta analysis by a pharmaceutical employee who was not actually employed but was actually paid for with a doctor,

219
00:21:36,560 --> 00:21:40,670
had been had his lab paid for, had included that data, but it never got published.

220
00:21:41,480 --> 00:21:45,410
And he said the unpublished data has a far more greater effect than the published.

221
00:21:45,800 --> 00:21:50,360
And it could go and startling. And I just thought, well, that's quite interesting, but.

222
00:21:51,490 --> 00:22:01,719
This is what when Tom first said to me, we got our first ever clinical study report, which is the stuff that's not published the most to cite cited.

223
00:22:01,720 --> 00:22:10,000
So these are the two sided trials. Nicholson 2000 and Trinity found that the only two published one FAI did not mention

224
00:22:10,000 --> 00:22:14,260
any serious adverse events or stated they were no drug related serious adverse events.

225
00:22:15,250 --> 00:22:18,550
And these findings were repeated by bodies such as the UK NHS.

226
00:22:19,930 --> 00:22:24,970
No theory events were noted in the major trials and no significant changes were noted in laboratory parameters.

227
00:22:26,350 --> 00:22:30,400
However, the first clinical study report we got, which is of that study,

228
00:22:30,790 --> 00:22:35,830
describes ten theory adverse events, not in nine participants in the two trials.

229
00:22:36,400 --> 00:22:44,170
Now, if I when I film everything I know, when you see that I cannot reconcile what's going on here with this.

230
00:22:45,940 --> 00:22:52,030
And we have the clinical study reports, clinical study reports go into thousands of pages and they detail all the adverse events,

231
00:22:52,180 --> 00:22:57,160
the serious adverse events, serious, unexpected, serious thing that adverse reactions.

232
00:22:57,490 --> 00:23:03,280
And that's what it said. And at that point, I was shocked, to be honest about this.

233
00:23:03,880 --> 00:23:07,420
Again, I'm in a position where we go back to 2000, all the other aspects.

234
00:23:10,020 --> 00:23:15,930
And then it emerged that 60% of patient data from randomised trials had never been published,

235
00:23:16,770 --> 00:23:23,610
including the largest ever treatment trial undertaken with just over 1400 people of all ages.

236
00:23:24,450 --> 00:23:30,540
We've now already spent £450 million on this in the UK and this is the situation of what we spent.

237
00:23:32,830 --> 00:23:38,760
I mean, and it's interesting, these things don't go away. We make a complete nuisance of myself and the team do still.

238
00:23:39,060 --> 00:23:44,280
We've just published in The Lancet today a letter, opinion letter, because the Lancet,

239
00:23:44,280 --> 00:23:48,299
the people who were all funded by the drug industry offered actually all the observational data,

240
00:23:48,300 --> 00:23:52,260
say that this treatment is fantastic and you should all studies.

241
00:23:52,260 --> 00:23:56,760
We started to say, well, if you basing treatment effects on observational data, you start to get worried.

242
00:23:57,540 --> 00:24:03,750
It's on the W.H.O. Essential Medicines List on the basis of three observational trials that show a 70% reduction in mortality.

243
00:24:04,440 --> 00:24:07,590
But when you think about it, well, people get Tamiflu, unwell people.

244
00:24:07,680 --> 00:24:11,340
I will send you to a hospital without the Tamiflu. That's what happened.

245
00:24:12,180 --> 00:24:15,840
If you really want. Well, I'm going to send you in and we're going to treat you appropriately.

246
00:24:16,740 --> 00:24:24,000
And on Monday, next Monday, I met the has the parliament on Monday at the public accounts committee because they're

247
00:24:24,000 --> 00:24:29,490
really pissed off because they're saying we've spent over £500 million on this treatment.

248
00:24:30,360 --> 00:24:32,819
And if you look at what's happening currently,

249
00:24:32,820 --> 00:24:42,389
a GP who says I prescribe this readily and we stockpile it and I'm not quite sure where and we still spend another £50 million a couple of

250
00:24:42,390 --> 00:24:48,870
months ago because we're in a binding contract and then everybody says no money in health care can't quite understand and reconcile that.

251
00:24:48,870 --> 00:24:54,719
But again, if you prepared to use the evidence, the evidence, it allows you to look and dig and think.

252
00:24:54,720 --> 00:24:57,420
You need to have all the evidence around the benefits and harms.

253
00:24:58,230 --> 00:25:01,680
I'm just going to finish that about well and then to talk about things that we may do differently.

254
00:25:02,550 --> 00:25:06,240
This is some stuff, you know, evidence based practice.

255
00:25:06,240 --> 00:25:12,240
If you prepared to be a bit of a nerd for me thing where you can go and this is about five years ago,

256
00:25:12,240 --> 00:25:18,570
I got backed into a corner and said, we think there's a major problem with metal on metal hips.

257
00:25:19,920 --> 00:25:22,530
And this is people at Channel four with Deborah Cohen at the BMJ.

258
00:25:22,770 --> 00:25:30,090
But we've got this evidence and we can't really find any orthopaedic surgeons who are willing to look at it and give us a sort of impartial viewpoint.

259
00:25:31,320 --> 00:25:34,620
And I looked at it and I said, again, it was a bit like the Vioxx.

260
00:25:35,760 --> 00:25:42,360
If you looked at ceramic hips, the failure rate with about 3% in ten years, superb operation, had some problems.

261
00:25:42,720 --> 00:25:47,100
A leg could dislocate it, put people on the bed, they get better and they go back to work.

262
00:25:47,100 --> 00:25:48,900
But there were a few issues with it but not make it.

263
00:25:49,170 --> 00:25:55,230
Metal hips came along and people said You can get back to work quicker, you can have a better range of movement and it doesn't dislocate.

264
00:25:56,160 --> 00:25:59,430
And by about five years, the failure rate, 15, 20%.

265
00:26:00,150 --> 00:26:03,570
I'm like, you don't need many statisticians to know we're in a really bad way.

266
00:26:04,110 --> 00:26:12,309
And it turns out in devices it's like the Wild West compared to drugs because often a device can get on the market.

267
00:26:12,310 --> 00:26:14,850
And we looked at it, we did this look to all the medical data.

268
00:26:14,850 --> 00:26:23,489
I was shocked to learn about how devices are regulated, even in terms of implantable devices that you can get away with.

269
00:26:23,490 --> 00:26:26,670
Just equivalence. My device is similar to somebody else's on the market.

270
00:26:26,670 --> 00:26:34,320
I don't need a clinical trial. That's okay. You'll get CE marking and us FDA marking on what's called a 510 K license.

271
00:26:35,730 --> 00:26:42,150
Well, I'm just alluding to that is I've ended up as a world expert in device regulation

272
00:26:43,320 --> 00:26:49,200
almost by default on the basis I don't think anybody else read this stuff or Cath.

273
00:26:50,850 --> 00:26:55,950
And if you go and read this stuff and say I'm really interested in it, you'll be amazed at what goes on in devices.

274
00:26:56,580 --> 00:26:58,290
I can tell you it's unbelievable.

275
00:26:58,500 --> 00:27:06,600
But actually understanding the evidence that's required to get drugs or devices onto the market puts you in a space where

276
00:27:06,600 --> 00:27:12,480
actually you can be really influential in terms of thinking of how you might contribute to how health goes forward.

277
00:27:13,200 --> 00:27:19,530
So that's a bit about some of the profound ideas and issues that have impacted on me.

278
00:27:19,800 --> 00:27:27,150
And I went to the DEA who fund all of our research in January, and what they wanted to know is what should we think about differently?

279
00:27:28,140 --> 00:27:31,320
And so I put these ideas, seven ideas to them.

280
00:27:31,440 --> 00:27:35,100
I said, This is what's going wrong or what we might do differently.

281
00:27:35,610 --> 00:27:40,680
I think there is a constant need to re-evaluate interventions that are already on the market.

282
00:27:42,000 --> 00:27:45,480
And here's a classic example aspirin for everyone older than 50.

283
00:27:45,780 --> 00:27:54,330
This is an editorial in the BMJ. In 2005, everybody at 50 was going to be like Smarties would be on aspirin by 209.

284
00:27:54,630 --> 00:27:56,580
We're actually saying this is a bad idea, actually.

285
00:27:56,940 --> 00:28:04,120
And when you look at it and think by 209, actually the benefits of aspirin in primary prevention are outweighed by the harm.

286
00:28:04,590 --> 00:28:07,950
So by four years later, we're like, no, this is. And when you look at.

287
00:28:08,330 --> 00:28:14,540
It's really interesting. If you go back to here about 1758, if you go up to Chipping Norton up the road,

288
00:28:14,960 --> 00:28:20,540
you can make the Reverend Stone, who published here is the account of the success of the bulk of the Willow.

289
00:28:21,020 --> 00:28:26,210
And he was a fellow. It was what him actually was a fellow at Wadham.

290
00:28:26,870 --> 00:28:35,000
But when you look at this, it takes until 1974 for the first randomised trial of aspirin and my reported.

291
00:28:36,720 --> 00:28:45,310
I mean only 40 years ago it seems ridiculous that only 40 years ago actually people didn't really have trial evidence to support aspirin in America.

292
00:28:45,620 --> 00:28:48,590
So most of what we know is really young and early.

293
00:28:49,310 --> 00:28:57,450
What's a puff of physiology is is hundreds of years old and all basic sciences, teaching pathology and physiology are coming from two 300.

294
00:28:57,470 --> 00:29:01,520
But we're coming from a period where actually the evidence base is not that amazing.

295
00:29:01,520 --> 00:29:06,200
And if you look at 1996, the FDA approved it for suspected MRI.

296
00:29:06,980 --> 00:29:10,490
That's not that long ago. And I put this slide up, you'll see this slide.

297
00:29:10,490 --> 00:29:14,770
Maybe people talk about this idea that cumulative meta analysis here,

298
00:29:14,780 --> 00:29:19,040
the meta analysis and this is probability is 12 and that the overall effect favour treatment.

299
00:29:19,370 --> 00:29:25,309
And what they do and I've done this for about ten years and then if I can't be right, this they go about here and they say,

300
00:29:25,310 --> 00:29:29,660
well, everybody could have stopped doing the evidence here and it should have been implementing.

301
00:29:30,470 --> 00:29:34,880
And that's correct, in a way. But actually, what's going on here?

302
00:29:35,450 --> 00:29:39,050
What are all these trials doing right up to 1988?

303
00:29:40,370 --> 00:29:46,160
They are confirming the finding in lots of different populations, generalising the results.

304
00:29:46,380 --> 00:29:54,800
They're consistent. What happens now with Thrombolysis has become widespread and it's not something that people go.

305
00:29:54,980 --> 00:30:00,380
I'm not quite sure about the evidence. What happens today is much of the evidence, the roundabout here.

306
00:30:01,190 --> 00:30:06,050
And then we move into all sorts of other areas called knowledge translation, implementation, science.

307
00:30:06,350 --> 00:30:09,350
But actually the problem is we're in 13 now.

308
00:30:09,350 --> 00:30:12,480
When I went to the DEA and said, we don't we don't fund confirmatory trials.

309
00:30:13,580 --> 00:30:19,790
And I said, that's why there's no other trial of Tamiflu outside of the manufacturers, because nobody said,

310
00:30:20,150 --> 00:30:27,170
let's just confirm these results in a different population is sensible as opposed to spending 550 million.

311
00:30:27,410 --> 00:30:33,649
I said to them, you could have spent in the telephone line 5000 people received the treatment in the first week.

312
00:30:33,650 --> 00:30:41,120
You could have randomised them done the trial within 28 days and it cost you about £500,000 as opposed to 5 million.

313
00:30:41,450 --> 00:30:42,140
Not interested.

314
00:30:43,100 --> 00:30:51,079
Number two, insufficient evidence exists around most current intervention to determine if they are effective, which follows on from one.

315
00:30:51,080 --> 00:30:53,690
And this is the classic example and we've been discussing this today.

316
00:30:54,110 --> 00:30:59,750
I am amazed that in primary care around about 3 million people are on antidepressant.

317
00:31:00,770 --> 00:31:05,000
And I'm even more amazed when you look at the amount of evidence that informs that decision.

318
00:31:05,420 --> 00:31:12,420
This is in primary care, the systematic review 14 studies of which ten examined 58 to assess rise.

319
00:31:12,440 --> 00:31:22,790
But forget that the number of participants again is only 1364 in the intervention and 919 in the placebo, and they're typically 6 to 8 weeks.

320
00:31:23,660 --> 00:31:25,940
The discussion I was having with Kemal today said, Well,

321
00:31:25,940 --> 00:31:30,170
how come the guidelines or the guidance is you've got to stay on the treatment for 18 months before you come off it,

322
00:31:31,130 --> 00:31:37,490
if that's what the evidence says, no dose information and the offers were in Britain and the appropriate duration a three year treatment.

323
00:31:38,450 --> 00:31:42,430
So it's a bit like the Wild West, but we've ended up with 3 million people on the treatment.

324
00:31:42,650 --> 00:31:51,580
I can't get them off. I think that's very thought provoking. Third is this is actually now becoming a big problem for us is very difficult.

325
00:31:51,580 --> 00:31:56,090
It's becoming very expensive to develop the evidence base. It's almost like an outcome to the old trials.

326
00:31:56,480 --> 00:32:01,760
The front end regulation is so great that it takes about a year just to get off the ground.

327
00:32:02,060 --> 00:32:08,420
And these trials now run rate of 20 like 2.5 million and you're like, we need more money just to get going.

328
00:32:09,590 --> 00:32:17,030
If I came tomorrow and I said, I want to do a new intervention in pharmacy and we've discussed this for anticoagulation,

329
00:32:18,140 --> 00:32:22,190
I want to do some educational well, the commissioners can give it to one pharmacist.

330
00:32:22,190 --> 00:32:24,860
We can give it to every they can give it to half. You can give it to a third.

331
00:32:24,860 --> 00:32:29,630
But if they decide to randomise, it, then inflated the cost by half a million pound.

332
00:32:30,080 --> 00:32:35,209
So I think this is a real issue that has to be in there's a new thing called the Health Research Authority that's

333
00:32:35,210 --> 00:32:43,230
trying to improve how we do clinical trials for this came up recently in a Twitter debate and got picked up.

334
00:32:43,250 --> 00:32:49,770
Most interventions don't apply to the people we see in practice, and this was something I did with Jamie Oliver.

335
00:32:49,770 --> 00:32:53,400
It's TV production said to many made of thin.

336
00:32:53,720 --> 00:32:56,450
You're not the man who made it fat, the man who made it thin to deceive.

337
00:32:56,510 --> 00:32:59,839
And he said, Well, look, this evidence, they all say this evidence for Weight Watchers works.

338
00:32:59,840 --> 00:33:04,640
And I said, well, it does work, but it doesn't work to anybody we see in practice.

339
00:33:05,450 --> 00:33:09,170
So if you take the evidence, this is based on the trial evidence if. And look at the evidence here.

340
00:33:09,750 --> 00:33:14,360
You take a thousand people and invite them, 885 won't come.

341
00:33:14,960 --> 00:33:18,830
So that's 88% a population are not interested in intervention.

342
00:33:19,130 --> 00:33:24,020
So this is a public health. It's to both solve the problem of the hundred and 15 who do come take up the intervention,

343
00:33:24,770 --> 00:33:28,250
fix well, attend the first class and 53 do not attend all the classes.

344
00:33:29,150 --> 00:33:34,130
However, of the 62 attend all the classes, they will lose about 5.4 kilos at the end of them classes.

345
00:33:35,000 --> 00:33:40,280
If you follow them up with two years, you down to 13 people who maintain the goal weight three and up by five you

346
00:33:40,280 --> 00:33:43,940
that's about ten of which nine are women and one's a man maintain that go away.

347
00:33:44,240 --> 00:33:48,680
So he started out with an intervention 4000 people and he managed to get to ten people.

348
00:33:49,340 --> 00:33:52,880
Superb, fantastic public health intervention.

349
00:33:53,120 --> 00:34:01,520
Do you think it's actually useless? It's useless if we think this is going to be our strategy for attempting to lose weight.

350
00:34:02,120 --> 00:34:07,620
And if if you are one of the people who likes classes and is in this and you like going to Weight Watchers,

351
00:34:07,620 --> 00:34:11,299
then you can keep going for five years and you've got a good friend and great it's a great thing to do.

352
00:34:11,300 --> 00:34:15,530
But the health service is this what we should be spending our money on commissioning?

353
00:34:16,520 --> 00:34:19,580
I think this is deeply and this is about thinking about these.

354
00:34:19,880 --> 00:34:22,880
Who does this research apply to? Five.

355
00:34:22,880 --> 00:34:25,520
I'm not going to march about this, but I did touch on this.

356
00:34:25,520 --> 00:34:33,500
One of the HCA and John Breast is going to talk about this on Thursday that the current need to develop a rapid but also robust evidence base.

357
00:34:33,920 --> 00:34:36,440
We need much more efficient, quicker information.

358
00:34:37,250 --> 00:34:42,860
And when we decide that that information is important, we then need to think about doing a robust job.

359
00:34:43,610 --> 00:34:48,679
The problem is, is if you seen today, if if I go to PubMed and I'm not wrong here,

360
00:34:48,680 --> 00:34:55,280
but last time I checked it with 24 million articles and climbing and it's almost exponential.

361
00:34:56,000 --> 00:35:00,110
PubMed counts for about 45% of the day's publications out there.

362
00:35:00,110 --> 00:35:07,490
So the number of 50 million plus of which will bottom out about 50,000 clinical trials in about three or four years time.

363
00:35:08,120 --> 00:35:11,929
It's massive. The explosion in evidence base and how do we get through efficiently?

364
00:35:11,930 --> 00:35:18,410
What we want to know is the real problem. This you say, and this probably fits with the weight.

365
00:35:19,310 --> 00:35:25,910
Here's my two important question for any intervention down here is the critical appraisal that you're all going to learn down here.

366
00:35:26,690 --> 00:35:36,590
Yeah, but my two is the top. If you take a questionnaire, my two important questions are who does this apply to and does it make a difference?

367
00:35:37,100 --> 00:35:40,669
And if you can't answer it, yeah. Applies to the people I see.

368
00:35:40,670 --> 00:35:45,899
And it does make a difference. Why move on? Why even bother doing any appraisal?

369
00:35:45,900 --> 00:35:50,690
And now I've been thinking about this and I might write this, that actually the real way we do it is the other way around.

370
00:35:50,700 --> 00:35:54,200
You tend to ask a question and do the appraisal and say high quality evidence,

371
00:35:54,200 --> 00:35:57,980
and then at the end of it run out of steam and go, Oh, who does it apply to?

372
00:35:58,430 --> 00:36:01,880
But actually, sometimes the other way to what you find it, does it make a difference?

373
00:36:01,890 --> 00:36:07,130
Are all these other issues like surrogates, composite end points, end points that don't mean anything.

374
00:36:07,310 --> 00:36:15,110
6 to 8 week studies. Yeah, not long enough to even sufficiently say this is a useful end point and when you do it that way.

375
00:36:15,590 --> 00:36:23,280
The problem on coming to in the way I'm thinking of this is we've got this mantra of cost effectiveness is what it's all about.

376
00:36:23,280 --> 00:36:26,720
And people are not understanding that actually, where's the clinical effect?

377
00:36:27,680 --> 00:36:34,160
What do they mean? So any time you look at an intervention, you should be able to say, what is the actual benefit here?

378
00:36:35,240 --> 00:36:40,010
And when you do something like that in Tamiflu, you actually start to see that.

379
00:36:40,010 --> 00:36:43,909
Actually, if you answer that question and anybody goes to me, well, there is a bit.

380
00:36:43,910 --> 00:36:50,240
So I say, Well, okay, just imagine you are a GP and you're going to say to somebody, inform them, give them an informed decision.

381
00:36:50,510 --> 00:36:54,860
What would you say about the benefit? What would you say at home and who does it apply to?

382
00:36:55,340 --> 00:36:57,290
And when you do that, it breaks down the arguments.

383
00:36:57,290 --> 00:37:06,440
Breakdown seven and then seven is this interesting issue is that there's a real problem supporting our regulatory bodies.

384
00:37:06,830 --> 00:37:13,760
And this is just one example. This is an example of a drug called Rivaroxaban for acute coronary syndrome that came up at the FDA.

385
00:37:14,300 --> 00:37:19,160
And this chap, Toby Matheny, is a real sceptic and does a really good job.

386
00:37:20,150 --> 00:37:22,010
But he said we're in deep trouble.

387
00:37:23,390 --> 00:37:31,340
He said the end point is about 1 to 1 and a half percent difference in the endpoint rates between river oxybate and placebo, however.

388
00:37:32,470 --> 00:37:35,920
About 12% of the patients had an incomplete follow up.

389
00:37:37,600 --> 00:37:43,150
And we would say, well, actually, if the loss of the follow up exceeds the endpoint, you've got a problem.

390
00:37:44,050 --> 00:37:47,520
You back at the the Vioxx argument that you could have this intent.

391
00:37:47,530 --> 00:37:54,700
What's happening to them patient. And when he did a simple plot of the year to the patients with missing data.

392
00:37:55,060 --> 00:37:58,420
This is what it starting to look like over the last ten, 12 years.

393
00:37:59,560 --> 00:38:07,090
In fact, what should be happening and this should be going this way because the end points a smaller here, the difference is smaller.

394
00:38:07,090 --> 00:38:14,870
Three trials need to be larger, but they need to maximise following up patients because the really important what goes on in these study and

395
00:38:14,950 --> 00:38:21,280
we're really interested in that stuff because this does make a difference to how you interpret the endpoint.

396
00:38:21,820 --> 00:38:28,060
If we use this treatment, we could be right back at where we are with Vioxx because what are these 12% patients telling us?

397
00:38:28,420 --> 00:38:34,459
So coming back to where you are, what you might think about these are the three ways when I think about anybody you worked with,

398
00:38:34,460 --> 00:38:37,630
those contribute to the development of the evidence base,

399
00:38:38,260 --> 00:38:46,620
contribute to the methods of EPM and contribute to the teaching dissemination and communication of VBA and then the free aspect.

400
00:38:46,640 --> 00:38:50,469
So anybody work through this, I'm always pushing them a bit to say, Well, where are you?

401
00:38:50,470 --> 00:38:56,650
So we, we do lots of here. So this is you could be the evidence base in terms of doing systematic reviews,

402
00:38:57,490 --> 00:39:01,360
could be in doing primary studies, could be in implementation studies in practice.

403
00:39:01,600 --> 00:39:12,160
So that's one aspect. They're interesting. They will make a difference in summary, but the other two areas contribute to the methods of EPM.

404
00:39:13,600 --> 00:39:16,290
I told you that the first ever trial of aspirin was only for two years.

405
00:39:16,300 --> 00:39:21,430
It's some information we know what we're doing yet we're only into the into some of these issues.

406
00:39:21,430 --> 00:39:25,570
There's lots to be learned and thought about. There's so many areas you can look at.

407
00:39:25,900 --> 00:39:29,139
It's almost it's almost exponential.

408
00:39:29,140 --> 00:39:32,350
But you should be thinking, oh, I'm really interested in losses to follow up.

409
00:39:33,040 --> 00:39:38,710
Just really interested in adverse events. Jeff, have we got a sensible system for measuring adverse events, would you say?

410
00:39:38,730 --> 00:39:42,850
You know, Annette's really interested in diagnostics.

411
00:39:42,860 --> 00:39:47,020
Would you know, do you think we've got a sensible way of assessing diagnostics in primary care?

412
00:39:47,140 --> 00:39:53,530
No. So that's our contribution. But then the third is contribute to the teaching dissemination communication of VBA.

413
00:39:55,480 --> 00:40:00,670
That's what I'm doing now. After I'm communicating, I'm trying to disseminate some way of teaching.

414
00:40:00,670 --> 00:40:04,090
That's why we were on teaching courses this year.

415
00:40:04,090 --> 00:40:10,510
We had a meeting in summer. We're going to start teaching IBM to schools because by the time you're at school, it's all too late.

416
00:40:10,510 --> 00:40:14,260
And I can't believe that. And in fact, schools are starting to do a good job.

417
00:40:14,260 --> 00:40:20,440
Actually, they are talking about drug development and clinical trials is on the collect curriculum, so it's already in there.

418
00:40:20,440 --> 00:40:24,790
So we're going to work with schools a bit because we think there'll be interesting dissemination and communication.

419
00:40:25,630 --> 00:40:28,930
There are many ways and I'll just finish it so you can think about them.

420
00:40:28,930 --> 00:40:34,750
Here's here's a group that I did. And so thinking about communication and dissemination, you can participate in all fourth way.

421
00:40:35,410 --> 00:40:41,170
This is a group of people who came to a three day workshop, practice workshop for three days.

422
00:40:41,740 --> 00:40:48,670
And at data we looked at a systematic a paper that was published in the BMJ about the Association

423
00:40:48,670 --> 00:40:55,210
of White RIF and Type two Diabetes and this group with so infuriated by how crap that paper was,

424
00:40:56,050 --> 00:41:04,120
it said We have to do something about this. So I said, Well, why don't you get together and write a letter to the BMJ and when you write it,

425
00:41:04,360 --> 00:41:08,079
so you really so we in the mini appraisal that we did it with like you're going to do tomorrow,

426
00:41:08,080 --> 00:41:11,180
they spent 90 minutes to appraise the paper and they said, Look at all these problems.

427
00:41:11,700 --> 00:41:15,790
I said, Well, why don't you go and write it? So they got together, wrote it up and communicated it.

428
00:41:15,790 --> 00:41:21,250
That's me at the end, at the end and sent it to BMJ and it got published.

429
00:41:22,900 --> 00:41:29,080
The communicating, they're participating. So if you develop these skills, you certainly should be using these skills in some way.

430
00:41:29,080 --> 00:41:30,310
You shouldn't just put them in the closet.

431
00:41:30,640 --> 00:41:37,660
That annoys me in a way, because I think if you've got something that you think to say in health care and you've seen something in a signal,

432
00:41:38,170 --> 00:41:44,469
you should participate. Nobody's going to tell you head off. Because if you stick to your evidence and not your opinion, stick to the evidence.

433
00:41:44,470 --> 00:41:49,270
This is what the evidence is saying. You'll be all right. Here's another. And we put all these and I just put here the left.

434
00:41:49,270 --> 00:41:54,729
If you go to the CBM, you can put all the projects and here's your thoughts in all sorts of different ways.

435
00:41:54,730 --> 00:41:58,450
This is a meeting that led to a publication about real birth is rubbish.

436
00:41:58,450 --> 00:42:03,370
Even this is a load of people who are challenging us to think differently.

437
00:42:03,910 --> 00:42:04,989
And some people are saying.

438
00:42:04,990 --> 00:42:10,819
Trish Greenhalgh who's saying, look, there are loads of problems with the evidence base and some of it's rubbish and some of it,

439
00:42:10,820 --> 00:42:13,570
and Jeremy's going, This is really bad and this is what the difference is.

440
00:42:13,570 --> 00:42:20,710
And they published an article in the BMJ a couple of months ago that the biggest downloaded paper they've had this year,

441
00:42:21,850 --> 00:42:25,129
because they were just saying, look, let's participate in thinking ideas.

442
00:42:25,130 --> 00:42:31,630
There's no space where you can't contribute. I could spend hours talking about all the different ways you can contribute to different.

443
00:42:31,670 --> 00:42:38,270
Project. Have a look at the website. Some of the different stuff we do and I wanted to finish and maybe this is a good point to finish.

444
00:42:39,680 --> 00:42:44,410
So out of this, again, a contribution. What did we do after the Tamiflu story?

445
00:42:44,420 --> 00:42:51,590
I was contacted by a couple of people, one with Ben Goldacre and the other with Fiona Godlee at the BMJ.

446
00:42:52,970 --> 00:42:58,400
And what happened is they said, Oh, there's an MP in and I was amazed that people actually do this.

447
00:42:58,910 --> 00:43:02,450
They actually really put their neck out and bingo, they could put your neck out.

448
00:43:02,450 --> 00:43:04,820
And when that MP called Fair Wollaston,

449
00:43:05,300 --> 00:43:10,310
who is prepared to talk to us about this issue because she'd read about the Tamiflu and they couldn't believe it.

450
00:43:10,700 --> 00:43:14,659
Could you come down to London and talk to us about all this publication?

451
00:43:14,660 --> 00:43:19,940
So we went down to London and basically out of that was born an initiative called All 12.

452
00:43:21,500 --> 00:43:26,719
And what's been interesting and working with people like this is, is centred sense about things.

453
00:43:26,720 --> 00:43:35,420
BMJ that Ben Goldacre James Lynn Library is most of the great projects I'm involved.

454
00:43:35,420 --> 00:43:42,260
He'd never have a single penny of funding at all and all the stuff we get involved in,

455
00:43:42,260 --> 00:43:50,090
often the question or the clinical issue is compelling because you've skilled yourself in some of these issues in evidence based medicine,

456
00:43:50,420 --> 00:43:54,650
and often you don't need lots of funding to go and dig around, be involved.

457
00:43:55,970 --> 00:43:59,690
520 organisations, 80,000 people have signed this position.

458
00:43:59,690 --> 00:44:03,620
If you haven't signed it, could you go to trial? Because that's me telling you to sign it.

459
00:44:03,620 --> 00:44:09,710
There'll be another 20 people help and it's a classic example of using the skills, the VBM and what we're trying to do.

460
00:44:10,010 --> 00:44:16,640
We solve the problem that half of the clinical trials are never published or have not been published,

461
00:44:17,360 --> 00:44:22,520
which prevents massive problems if we can understand the benefits and harms of treatment,

462
00:44:23,210 --> 00:44:26,300
and that's a good example of how you might contribute or get involved.

463
00:44:26,930 --> 00:44:30,200
Thank you very much. Does anybody want a couple of questions for you all?

464
00:44:32,930 --> 00:44:38,960
How did you manage to get the Tamiflu people to tell you about the unpublished trials?

465
00:44:39,350 --> 00:44:45,980
Because I tried to find out about trials and it's like, yeah, correct.

466
00:44:46,010 --> 00:44:51,350
Totally correct. In fact, what happened is we ended up started with involving the BMJ,

467
00:44:52,070 --> 00:44:59,210
then started involving the media and actually got channel forward involved and they did a TV programme.

468
00:45:00,230 --> 00:45:08,480
Then we had this ridiculous where we'd write these emails to people like Roche and copy in 50 different media people, and they're all available to go.

469
00:45:08,480 --> 00:45:11,510
And then we got to the position where we just kept ramping it and ramping up.

470
00:45:11,510 --> 00:45:14,419
And then if you go on there, there's a BMJ Open Data campaign.

471
00:45:14,420 --> 00:45:18,620
If you look at it, you can if you're a real nerd, you can see all the emails are published.

472
00:45:18,620 --> 00:45:25,440
We even published all the emails. Instead, what we're going to focus on is because you keep saying you're going to give it and then you don't.

473
00:45:25,590 --> 00:45:28,680
And that was about an 18 month movement.

474
00:45:29,130 --> 00:45:34,220
And probably the bit that made a difference was to think the publication of Ben Goldacre is.

475
00:45:34,230 --> 00:45:39,360
But, but, but pharma had a slight shift in it just resonated in some way.

476
00:45:39,360 --> 00:45:44,519
And what it did really well with communicate. So we all I think I do all the work but not me.

477
00:45:44,520 --> 00:45:50,010
But I know all the people who do all the work in this area. What he did is really communicate it to the wider public really well.

478
00:45:50,460 --> 00:45:57,400
And that's what was exceptional. And then the old trials came and about four months after all trials.

479
00:45:57,420 --> 00:46:01,950
Roche, GSK made a commitment that we're going to make this stuff available and they signed all trials.

480
00:46:02,970 --> 00:46:06,390
And then it looked a bit odd because Roche were on the other side going, Well, we're not giving it here.

481
00:46:07,110 --> 00:46:16,340
And then about three months in, they committed and retrospectively, well, they sent us everything on a they talked about privacy and all this.

482
00:46:16,350 --> 00:46:22,739
And I was like and we had a discussion and they just posted me the CDs in a package to the department.

483
00:46:22,740 --> 00:46:23,850
I was like, This is ridiculous.

484
00:46:23,850 --> 00:46:30,030
After all this privacy and all these issues you just sent in the mail and I was just so I've got all these all this data.

485
00:46:30,030 --> 00:46:35,069
We've now got it constituted for Tamiflu,

486
00:46:35,070 --> 00:46:43,290
over 150,000 pages of data compared to about 12 pages that had been published in The Lancet and JAMA to that point,

487
00:46:44,340 --> 00:46:49,770
to the bulk of the trials and not published. What are the excuses they have for not forcing trials?

488
00:46:51,060 --> 00:46:54,990
Well, it's allowed if it's allowed, it's permitted. So you can get away with it.

489
00:46:55,890 --> 00:47:00,090
I mean. Well, let's be clear. You have a drug, you're selling a drug.

490
00:47:00,420 --> 00:47:05,400
You have a science department and a marketing department. A marketing department saying we've got $44 billion of sales.

491
00:47:05,640 --> 00:47:09,570
And you're like, Yeah, but I'm fat on half these trials, which we want to publish that show with no effect.

492
00:47:09,570 --> 00:47:10,740
That's publication bias.

493
00:47:11,610 --> 00:47:17,430
Why would you then incentivise everybody to make that available if there's no regulations or laws that make you have to do that?

494
00:47:19,200 --> 00:47:25,589
The FDA in 2007 brought in an FDA regulation that says for all drug trials they've come through,

495
00:47:25,590 --> 00:47:31,320
FDA regulation is supposed to publish a summary of results within one year of the trial closing,

496
00:47:32,550 --> 00:47:37,320
and people continually audit that and say nobody does it and not a single person has yet been fined.

497
00:47:38,740 --> 00:47:42,850
So there's no regulation. Basically, there's no regulation, and that's part of it.

498
00:47:42,940 --> 00:47:49,060
Last two weeks ago, the European Medicines Agency ruled that all out of 2016, in general,

499
00:47:49,660 --> 00:47:55,900
any drug that comes to them, they are going to make the full clinical study report fully available on their website.

500
00:47:56,290 --> 00:47:59,560
That's just all we do.

501
00:47:59,800 --> 00:48:05,560
I mean, simple project I'm involved with with Alex Thomson in our department at the moment we're auditing the University of Oxford,

502
00:48:05,560 --> 00:48:10,090
be our season, be all youth and saying because why a patient asked us to do it.

503
00:48:10,090 --> 00:48:13,540
The patient representative said, I go to all trials and I'm really interested.

504
00:48:13,570 --> 00:48:17,139
I'm on the board of the BRC and I quite like to know whether they're publishing their child or not.

505
00:48:17,140 --> 00:48:21,100
And we thank you very much. Can't get the clinicians to say it. They all now have to do it.

506
00:48:21,280 --> 00:48:24,879
It's been fantastic. We've now identified 115 trials.

507
00:48:24,880 --> 00:48:28,990
We're going to audit them. Anybody can do that. This takes time, doesn't it?

508
00:48:29,710 --> 00:48:33,630
But it's been great that a patient came because they force the issue, whereas everybody else is.

509
00:48:33,820 --> 00:48:36,820
We don't quite don't want you to know what's going on, to be honest.

510
00:48:38,380 --> 00:48:42,970
It's in the public good, isn't it, to make this stuff available? It's not there yet.

511
00:48:43,750 --> 00:48:46,150
One more form and I'll let you go for a drink.

512
00:48:48,130 --> 00:48:58,430
You think with the older drugs you know where normal practice is based on some it's one of the we've moved into midfielder I think that's available.

513
00:48:59,660 --> 00:49:04,459
Well there should be because the aspirins are very sobering story because the aspirin,

514
00:49:04,460 --> 00:49:10,250
if you make the individual data available which had been made available, if you go beyond 2009,

515
00:49:10,550 --> 00:49:16,370
what happened is that the chap here called Peter O'Farrell is a neurologist who saw that individual data and was looking at it and said,

516
00:49:16,370 --> 00:49:17,450
well, that missed something here.

517
00:49:18,710 --> 00:49:25,640
And what that missed is the fact that aspirin might have a massive role in the prevention of cancer, particularly metastases.

518
00:49:26,240 --> 00:49:30,380
And if aspirin came along now, they probably looking at it in cancer and not in heart attack.

519
00:49:31,220 --> 00:49:36,710
And that actually it may be a really important drug in prolonging life in lots of cancers.

520
00:49:37,460 --> 00:49:44,720
And that was because the data was made available. So there's probably lots of drugs out there that people could start to look at.

521
00:49:45,020 --> 00:49:50,059
And we've got one antibiotics is one indication we're running out of antibiotics.

522
00:49:50,060 --> 00:49:53,600
What if you made all the data available? You might say, well, we relooked at some of these antibiotics.

523
00:49:54,090 --> 00:49:56,150
Actually, some of these may be useful in over here.

524
00:49:57,260 --> 00:50:02,990
So, in fact, even from a beneficial point of view, we have to do it more so than the harms in some ways.

525
00:50:03,950 --> 00:50:05,300
And if we don't do it, yeah,

526
00:50:05,840 --> 00:50:13,639
my concern is we could spend 20 years here and actually clinical medicine has bottomed out and actually could start going the other way.

527
00:50:13,640 --> 00:50:21,860
We'll get actually increase morbidity, reduce reduce mortality, life expectancy or shorten we're getting fatter, we exercise less.

528
00:50:22,670 --> 00:50:29,899
Our children may not live along with the people in this room. And then the great thing is though, like the weight watch is perfect.

529
00:50:29,900 --> 00:50:32,900
You could have a drug and an eye.

530
00:50:32,900 --> 00:50:36,650
When we had a few weeks ago the overdiagnosis conference something else we got involved

531
00:50:36,650 --> 00:50:43,670
in interesting the idea of overdiagnosis that actually you can find a disease

532
00:50:43,670 --> 00:50:48,229
and diagnosed with the disease and diagnose more people with the disease but have no

533
00:50:48,230 --> 00:50:51,740
effect whatsoever forever on the mortality or the morbidity is really interesting.

534
00:50:51,740 --> 00:51:01,460
And there are also the TV that appearing like pre-diabetes which is superb for if you in the manufacturing pharmaceutical.

535
00:51:04,320 --> 00:51:12,150
So yeah that's next on the agenda that we said is part of the fight is to say right you've made everything from 2016 available.

536
00:51:12,360 --> 00:51:17,880
We now need to go back so that many of you people could say, I'm interested in I might look at this.

537
00:51:19,110 --> 00:51:23,160
And then we need lots more people who are prepared to do the work. Okay.

538
00:51:23,730 --> 00:51:26,730
All right. That's 7:00. You can have a drink. Thank you very much.