1
00:00:00,030 --> 00:00:00,850
I'm going to give you a talk.

2
00:00:00,870 --> 00:00:07,530
We're going to talk about overdiagnosis, explain how it affects me, how it might affect you, and how you might think a bit differently about it.

3
00:00:08,070 --> 00:00:15,660
Okay. Well, when I first started thinking about this problem, I was in New York, actually, and that fixed avenue going down New York,

4
00:00:15,660 --> 00:00:20,940
and I opened for the word the signpost here, caution with there because of the traffic.

5
00:00:21,540 --> 00:00:24,210
But actually, as I progressed on it, it became a little bit darker.

6
00:00:24,480 --> 00:00:29,550
I realised it was about this van that was on the side of the road, which was the Brain Tumour Foundation,

7
00:00:29,730 --> 00:00:36,120
where you could have your scan of your brain on Sixth Avenue for about $700 that evening.

8
00:00:36,510 --> 00:00:45,870
So this idea that the road to early detection and so this this presence of the idea, you'll be aware of C.T. scanning being around you in areas.

9
00:00:45,870 --> 00:00:55,290
And then as I went through, you realise this sort of stuff everywhere in the newspapers, £120 test to spot genetic heart failure.

10
00:00:55,650 --> 00:00:59,370
So you can have a test like that today. Everybody in this room can have it.

11
00:01:01,170 --> 00:01:10,350
And when I got home I realised you get to a certain age, some of you are not there yet and people start sending you letters through the post.

12
00:01:10,950 --> 00:01:15,870
And online letters are messages an urgent message about your examination?

13
00:01:16,680 --> 00:01:21,840
And basically the suspect thing was telling me it's now over two years since your last examination.

14
00:01:22,050 --> 00:01:30,780
And as we mentioned in our previous letter, you were due to have your eye test on 29th of August 2014, and now it's 415 for February 2016.

15
00:01:31,410 --> 00:01:38,070
I'm a bad person now just to say I'm over 40 and I have a family history of glaucoma.

16
00:01:39,090 --> 00:01:43,200
Anybody else in that situation in the room? A couple of people.

17
00:01:43,530 --> 00:01:46,470
All right. So the good thing if you get free eye tests, don't you?

18
00:01:47,160 --> 00:01:52,530
So if you want to get free eye test and you're over 40 telling me you've got family history of coma and they'll give you a free eye test,

19
00:01:52,800 --> 00:01:58,230
both the secret. The second issue, though, then, is how often do you get a free eye test?

20
00:02:01,150 --> 00:02:04,280
Whenever I feel like it. Every year. Every year.

21
00:02:05,050 --> 00:02:11,920
So the NHS pays for people like me and our friends here to go for an examination every year.

22
00:02:12,670 --> 00:02:16,300
However, I'm the professor of Ebola and I haven't been for nearly two years.

23
00:02:17,110 --> 00:02:24,940
And so we've got differences of opinion every year. Whenever I feel like it so far, you know, I'd look at the evidence before we go.

24
00:02:25,120 --> 00:02:32,410
How can that help me? And actually, the the systematic review and economic evaluation of the evidence for glaucoma screening.

25
00:02:33,430 --> 00:02:39,670
And when you do that, you see that this often phenomenon of actually there doesn't seem to be much evidence.

26
00:02:40,180 --> 00:02:48,710
There are only two randomised controlled trials and then two randomised trials have to do two important steps.

27
00:02:48,730 --> 00:02:52,240
They have to extrapolate and they have to make assumptions.

28
00:02:54,160 --> 00:02:59,230
Whenever anybody's doing that, I often think they must be making it up at that point.

29
00:02:59,770 --> 00:03:06,760
But you can see that without treatment. Meantime, to blind for at least one eye with 23 years compared to 35 years with treatment.

30
00:03:07,090 --> 00:03:11,860
That's assumption based, but immediately that tells you whenever you get a diagnosis,

31
00:03:12,100 --> 00:03:16,870
you're going to have to take somewhere about 25, 30 years of treatment to realise the benefit.

32
00:03:17,080 --> 00:03:24,219
And you might not know, but you are committing yourself. One point somebody is going to diagnose you with raised intracranial pressure in your eye,

33
00:03:24,220 --> 00:03:27,490
intraocular pressure and you need to take 25 years of treatment.

34
00:03:27,490 --> 00:03:32,590
So that's a first for process. Okay. But it appears not to be cost effective.

35
00:03:33,070 --> 00:03:38,830
Okay. And the screening interval seems to be every ten years to approach cost effectiveness.

36
00:03:38,830 --> 00:03:46,930
That's different from the benefit. That's how much we would screen you to say we generate a genuine benefit and then a go to obviously.

37
00:03:46,930 --> 00:03:54,729
Well, look what see the US Preventive Task Force, they do guidelines and they do a rigorous job and they obviously do the systematic reviews.

38
00:03:54,730 --> 00:03:58,810
Again, look at the evidence and come to a conclusion. And this was their conclusion.

39
00:03:59,900 --> 00:04:04,719
And now we've got two people in here.

40
00:04:04,720 --> 00:04:10,570
How does that really help you make a decision? Do you think do you think that makes you more likely or less likely to screen yourself?

41
00:04:11,870 --> 00:04:18,850
Okay. Well, very helpful. Is it? So you have this discussion and so it went back to NHS choices.

42
00:04:19,200 --> 00:04:22,989
NHS choices are are evidence based.

43
00:04:22,990 --> 00:04:27,370
Tell us how to communicate to the public and it tells it is important to have regular eye

44
00:04:27,370 --> 00:04:31,600
test though I problem with glaucoma can be diagnosed and treated as early as possible.

45
00:04:32,080 --> 00:04:36,040
If you have glaucoma, it can take a long time before you realise you have a problem.

46
00:04:36,370 --> 00:04:41,860
You should have an eye test at least every two years or more frequently if advised by your optometrist.

47
00:04:42,670 --> 00:04:46,360
Well, it's obvious you're optometry. They make money out of it. You're going to tell you more frequently.

48
00:04:46,600 --> 00:04:49,660
They'd like to do it every day, if possible, one day if you were expecting it.

49
00:04:50,380 --> 00:04:55,330
But how is that compatible with that? Isn't that what's going on?

50
00:04:55,360 --> 00:04:59,620
This is the NHS in complete opposite to that information.

51
00:05:00,370 --> 00:05:08,200
We don't know what we're doing. Hey, do it every two years. So at that point, I'm like, feeling totally confused.

52
00:05:09,010 --> 00:05:12,020
What do we do now? Well, I'll go and look at the evidence.

53
00:05:12,040 --> 00:05:19,240
What did the test look like? And here's a test which have different ranges of sensitivity and specificity.

54
00:05:19,570 --> 00:05:23,110
So this is the worst case scenario if I have the test. Okay.

55
00:05:23,120 --> 00:05:28,630
So if I take a thousand white people like me, my prevalence of disease with a family history of about 5%.

56
00:05:30,090 --> 00:05:33,329
That means 50 people have the disease. 950.

57
00:05:33,330 --> 00:05:36,420
Will not everybody follow that? Happy with that? Okay.

58
00:05:37,200 --> 00:05:43,620
The test ranges from 10% to 90% sensitivity and specificity of 81 to 90%.

59
00:05:44,040 --> 00:05:50,110
So if you take the worst case scenario of a sensitivity of 10%, everybody is doing the MSI knows what that means.

60
00:05:50,130 --> 00:05:57,060
Now, I'm not going to test you my overreading, but what it means is 10% of the people with the disease will test positive.

61
00:05:58,200 --> 00:06:07,770
And if you take the specificity of 81, that means of the 950 who don't have the disease, 181 will test positive.

62
00:06:07,950 --> 00:06:16,590
You had all the numbers together and you end up with this. Only five of 186 will actually have the disease if you have a positive test.

63
00:06:16,860 --> 00:06:21,000
That's the worst case scenario. There are about 3% if I get a positive test.

64
00:06:21,390 --> 00:06:24,390
But let's take the best case scenario. The systematic review evidence.

65
00:06:24,690 --> 00:06:30,540
The best case scenario is a sensitivity of 90% and a specificity of 99%.

66
00:06:30,810 --> 00:06:34,410
We're now in a really good test. Most people would go, Wow, that's a super test.

67
00:06:35,070 --> 00:06:38,100
Sensitivity, 90% and very high specificity.

68
00:06:38,580 --> 00:06:45,180
You get the same numbers here. Happy with that? If sensitivity 90 said 45 out of 50.

69
00:06:46,330 --> 00:06:51,220
Yeah. So of the 50 people with disease, 90% of them will test positive, about 45.

70
00:06:51,670 --> 00:06:58,390
And of the 950 without the disease, 99% specificity, 95% will test positive.

71
00:06:58,630 --> 00:07:05,380
That means you end up with about 30% chance of having the disease if you have a positive test and a glaucoma screening as of tomorrow.

72
00:07:05,890 --> 00:07:17,350
About one in three. So I find that I understand the evidence, but it still doesn't help me very much because now I'm stuck in alone thinking,

73
00:07:17,350 --> 00:07:21,550
well, the US Preventive Task Force says we're not quite sure whether to screen or not.

74
00:07:21,970 --> 00:07:26,770
NHS choices say do it every two years or less. The evidence is extrapolated.

75
00:07:26,950 --> 00:07:31,209
There are three options and it's only two randomised trials and actually there's only

76
00:07:31,210 --> 00:07:34,390
one in three times if you have a positive test that will actually have the disease.

77
00:07:34,960 --> 00:07:39,220
I actually have. And so it leaves me in a position of uncertainty.

78
00:07:40,060 --> 00:07:43,150
I still don't know what to do. Now, that's interesting.

79
00:07:43,150 --> 00:07:48,880
And I've looked at all the evidence, and so I find this is a recurring theme whenever I go and look at types of evidence.

80
00:07:50,290 --> 00:07:55,029
And so often you get this scenario where somebody said, Well, oh, but good news,

81
00:07:55,030 --> 00:07:59,859
all I have to do is put I prescribed, I've dropped in few, I got diagnosed.

82
00:07:59,860 --> 00:08:05,800
So you always get a case where somebody goes if without testing and screening, I wouldn't have been picked up.

83
00:08:06,750 --> 00:08:11,729
Okay. The problem with that is the few if I've got to put eyedrops in my eyes now for 25

84
00:08:11,730 --> 00:08:16,170
to 30 years and I don't really know whether it's going to make any difference.

85
00:08:17,310 --> 00:08:22,560
Now it's an important issue. Second issue is when we label people with the disease.

86
00:08:23,280 --> 00:08:29,340
I can go back to show you a particular study going back now to the 1978 in the New England Journal of Medicine.

87
00:08:29,790 --> 00:08:35,580
This is David Sackett, and Brian Haines looked at what happens if you go to hypertension.

88
00:08:35,910 --> 00:08:42,209
Diagnose hypertension in the workplace. If you go into the workplace, you go and get people like you.

89
00:08:42,210 --> 00:08:47,640
And we say some of you have hypertension, some of you don't, and some of you never knew that.

90
00:08:48,270 --> 00:08:48,839
Well, in fact,

91
00:08:48,840 --> 00:08:57,120
one of the biggest issues about this paper is that actually what it showed is that labelling resulted in increased absenteeism from work.

92
00:08:58,260 --> 00:09:06,569
So there's something we do. And it was quite significant that actually the amount of it within previously unaware subjects before screening,

93
00:09:06,570 --> 00:09:11,550
they take about 2.7 days after screening, they take about 8.4 days off of work.

94
00:09:12,630 --> 00:09:16,650
But huge implications, doesn't it, for how we think about when we give people a label.

95
00:09:17,640 --> 00:09:23,310
If not just you're going to say, well, you have a diagnosis in here, some treatment now go away and get on with it.

96
00:09:23,490 --> 00:09:28,950
Actually, we may increase their anxiety. We may increase their health seeking behaviour.

97
00:09:29,340 --> 00:09:34,680
We certainly may increase the ability to take time off work because suddenly you think you might be a new person.

98
00:09:34,980 --> 00:09:39,730
So what does it mean if you're prepared to have a label? Then you go back to the glaucoma.

99
00:09:39,740 --> 00:09:41,930
Should I have a label or not? What? The other connotations?

100
00:09:41,930 --> 00:09:48,800
What do I need to know about in terms of the quality of my life, who I am, what effect it might have on me before I make the decision?

101
00:09:49,070 --> 00:09:56,030
I think this is really important stuff. We don't know the answer. Okay, so that's one example.

102
00:09:57,110 --> 00:09:59,420
For myself. Here's a second example.

103
00:10:00,620 --> 00:10:08,000
Dear Mr. Hennigan, I am pleased to invite you to a health screening clinic we are holding at the Kassam stadium on Thursday, the 25th of February.

104
00:10:09,220 --> 00:10:14,740
Why have I been invited? When we organise the clinic, we write to a small number of people.

105
00:10:14,800 --> 00:10:19,090
They are very special in your area for whom appointments are available.

106
00:10:19,870 --> 00:10:27,370
This is a simpler system that keeps our administration costs low, meaning we can offer extensive health screening at a fee.

107
00:10:27,370 --> 00:10:34,300
That is exceptional value to me. I'm told, look, well, the fee is £129.

108
00:10:35,710 --> 00:10:41,680
And the good news is, if ever went to BUPA, it would be £424.

109
00:10:42,460 --> 00:10:49,980
So I am being sold the message. And I was somebody I threw for Harry, the centre manager I talked to.

110
00:10:49,980 --> 00:10:54,180
I said, Look, anybody here over 40 and we've one of them people.

111
00:10:54,330 --> 00:10:58,080
And she was getting these electives as well, collective health screening letters,

112
00:10:58,380 --> 00:11:02,040
the ones you hear about 40, you start to get invited to health screening.

113
00:11:03,180 --> 00:11:06,330
Okay. And in that, you have to make a choice. Do I go?

114
00:11:06,720 --> 00:11:12,540
And in this message, it talked about measurements such as age, sex, family history, height, weight and blood pressure.

115
00:11:12,990 --> 00:11:19,080
Simple blood test to measure. And following the check, you will receive personal advice about what you can do to stay healthy.

116
00:11:20,010 --> 00:11:25,350
That's what we're doing now in health care. We send out this is from a GP practice, not from private firms.

117
00:11:26,070 --> 00:11:31,470
And then you get information like this, questions you may have. What happens at the check?

118
00:11:32,460 --> 00:11:37,140
And when you read the information. It's so basic, it doesn't really help you in any way.

119
00:11:37,500 --> 00:11:41,550
Make an informed choice. Do not talk about the tests, what the accuracy is.

120
00:11:41,800 --> 00:11:45,900
Doesn't talk about the important effects it may have on you in terms of the benefits in the hand.

121
00:11:46,260 --> 00:11:52,680
If you start pretty brown statements, some people with raised blood pressure will have their kidney checked through a blood test.

122
00:11:53,820 --> 00:12:00,750
But what they do, treatment or medication, may be prescribed to help you maintain your health without really informing you.

123
00:12:01,350 --> 00:12:05,150
Is that informed choice about the benefits and harms though?

124
00:12:05,880 --> 00:12:10,500
Health checks. We actually have just done echo on a player who's not here.

125
00:12:11,010 --> 00:12:11,760
It's done in evidence.

126
00:12:11,760 --> 00:12:19,440
Sympathies with us for the show to look at health checks, cardiovascular risk screening, and that helped it prioritise where all the evidence is.

127
00:12:19,650 --> 00:12:24,330
And here's one of the first bits of evidence about helping you make your decision about should you have a health check.

128
00:12:24,780 --> 00:12:32,340
And this is the Cochrane Systematic Review, and it's done by Lafe Kroc's Bill, and it came out a couple of years ago.

129
00:12:33,370 --> 00:12:37,239
Okay. And what does it show? This is what it show.

130
00:12:37,240 --> 00:12:45,190
Again, only child about health. It did not find that health checks had any effect on hospital admissions, disability, war,

131
00:12:45,310 --> 00:12:49,030
the number of referrals to special additional visits to doctors or absence from work.

132
00:12:49,600 --> 00:12:55,090
And they were poorly studied, all the important work. So that doesn't work.

133
00:12:55,570 --> 00:13:00,640
So it looked at this review. Here's another review that we found within the message.

134
00:13:01,000 --> 00:13:05,950
And this one is looking at opportunity. Systematic reviews, opportunistic risk assessment.

135
00:13:06,290 --> 00:13:14,050
Systematic is is we systematically call you in Ruth if you're in my practice and I opportunistically say I'll make you pressure,

136
00:13:14,620 --> 00:13:17,860
or would you like to talk to me about this today? I could look at that.

137
00:13:18,410 --> 00:13:22,330
Okay. Another Cochrane review. Okay.

138
00:13:23,350 --> 00:13:28,659
So again, you'd expect to see these important effects like all cause mortality.

139
00:13:28,660 --> 00:13:33,490
If you're going to invest in health checks, you would want to see some effect on all cause mortality.

140
00:13:33,820 --> 00:13:38,350
No difference. Cardiovascular mortality. That's what you're hoping to see.

141
00:13:38,650 --> 00:13:42,370
No effect. Have no effect on these outcomes.

142
00:13:43,390 --> 00:13:46,719
All right. You can even drill down into this.

143
00:13:46,720 --> 00:13:50,740
One is one of the only population based on where they've done it actually in primary care.

144
00:13:51,040 --> 00:13:55,780
This was published in 2014 in the BMJ June 2014.

145
00:13:56,080 --> 00:14:04,480
Effective screening and lifestyle counselling on incidence of ischaemic heart disease in a general population into 99 randomised trials in Denmark.

146
00:14:04,960 --> 00:14:09,190
So this is not just screening, this is screening giving you some extra lifestyle conflict.

147
00:14:09,480 --> 00:14:11,740
Okay, look at what they did though.

148
00:14:13,200 --> 00:14:20,669
I would find this quite interesting invited for screening, risk assessment and lifestyle counselling up to four times over a five year period.

149
00:14:20,670 --> 00:14:29,220
So it's not just one, it's four times over five years. It's you then have your tailored lifestyle and counselling all visits.

150
00:14:30,430 --> 00:14:36,969
And those at high risk of IHT have fixed further suggestions of group based lifestyle counselling.

151
00:14:36,970 --> 00:14:39,420
So now we really trying to make the intervention work online.

152
00:14:40,050 --> 00:14:44,740
We are definitely making sure compared to doing nothing you are going to have some effect.

153
00:14:45,640 --> 00:14:52,060
Yeah, and if there were any changes, final counselling session, you were referred to GP for medical treatment if relevant.

154
00:14:52,480 --> 00:14:56,860
The control group was not invited for screening. Okay, that's pretty intensive, isn't it?

155
00:14:56,860 --> 00:14:57,610
And look what happen.

156
00:14:58,630 --> 00:15:06,490
So even if you go to that intensity level, know the different thing between intervention and control groups in terms of ischaemic heart disease,

157
00:15:06,640 --> 00:15:12,770
even if you work really hard with people. And the combined endpoint, total mortality, no difference.

158
00:15:12,770 --> 00:15:20,150
And you can see the complete conclusion is a community based, individually tailored intervention program with screening have no effect.

159
00:15:22,050 --> 00:15:29,070
We though have rolled out a health screening programme in the UK and I can come back to that at the end.

160
00:15:29,400 --> 00:15:34,290
There are interesting issues why this might never work and what the connotations and what the issues.

161
00:15:34,890 --> 00:15:39,540
But at the moment I am ignoring my health check screen.

162
00:15:40,020 --> 00:15:43,530
Okay. I'm quite happy to not know where I'm at.

163
00:15:43,530 --> 00:15:47,399
What's going on. I do not know my cholesterol level at the moment and I'm quite happy.

164
00:15:47,400 --> 00:15:54,020
Not tonight. Okay. So there's a second one here, the third one.

165
00:15:54,230 --> 00:15:57,270
Okay. Now talking Jeff at the back about this.

166
00:15:57,350 --> 00:15:59,239
This is I was feeling a bit on well, you know,

167
00:15:59,240 --> 00:16:03,590
when you get the cold and flu and you feel in a bit miserable that everybody is not in, you can feel like that.

168
00:16:03,590 --> 00:16:05,450
You feel a bit under the weather.

169
00:16:05,690 --> 00:16:14,270
I felt while I'm doing that, I'm going to do my Q nine score, which is use of the Q nine to make a tentative depression diagnosis.

170
00:16:15,350 --> 00:16:18,950
So these are the Q nine school. So I thought, well, I'll do my score.

171
00:16:19,670 --> 00:16:24,380
Little interest of pleasure in doing things. Well, actually, I was about one for that several days.

172
00:16:25,880 --> 00:16:29,960
I would also have trouble falling asleep, staying asleep, or sleeping too much.

173
00:16:30,110 --> 00:16:33,620
In fact, I gave myself a free for all dependent with feeling pretty grotty.

174
00:16:33,980 --> 00:16:37,610
I couldn't really sleep. And then sometimes in the day I felt like sleeping all the time.

175
00:16:37,970 --> 00:16:41,630
When it came to my appetite, I was feeling tired or having little energy.

176
00:16:42,080 --> 00:16:47,900
I had a poor appetite and I wasn't all overeating and I was overeating at some points as well.

177
00:16:48,050 --> 00:16:52,730
You know that feeling when you did feel a bit thick, but then when you feel or you overeat, that happens to me on occasion.

178
00:16:53,300 --> 00:16:58,070
And I also had trouble concentrating on things such as reading the newspaper or watching TV.

179
00:16:59,690 --> 00:17:06,230
And so I ended up with a score of nine. And my nine gave me minimal symptoms.

180
00:17:06,470 --> 00:17:13,250
The poor educate to call it worth returning one month treatment recommendation if I scored one more point.

181
00:17:14,130 --> 00:17:18,480
I could have had a diagnosis of minor depression at that point in time,

182
00:17:20,130 --> 00:17:26,370
and we started to use all of these schools a significant amount of time because come out of research and somehow

183
00:17:26,370 --> 00:17:32,910
we've translated them into clinical practice without even thinking of what's the connotation of what we're doing.

184
00:17:34,020 --> 00:17:37,020
And this is a really interesting issue because when you look at it,

185
00:17:37,830 --> 00:17:43,660
you can start to look at depression and look at the validity of the diagnosis and what you don't see.

186
00:17:44,520 --> 00:17:50,640
Identification of depression is strongly felt created with increased familiarity with the patient,

187
00:17:51,870 --> 00:17:59,450
to the point that people who don't have depression are labelled as depressed, false, positive people.

188
00:17:59,480 --> 00:18:03,420
Both these people who do have depression but are not labelled false negative.

189
00:18:04,650 --> 00:18:10,020
The only thing that starts to distinguish them is they're identical in terms of

190
00:18:10,020 --> 00:18:14,100
the clinician and the clinical characteristics is how many times you turn up.

191
00:18:15,630 --> 00:18:22,530
So what happens if you think about it in depression when we've seen huge increases is if you go and see a GP quite a few times,

192
00:18:22,530 --> 00:18:27,180
at some point you're going to go, I think you're depressed, have a treatment, have the medication.

193
00:18:28,590 --> 00:18:34,320
Because it's really easy to do that. And so that has connotations for us as GP's to really think about.

194
00:18:34,530 --> 00:18:38,010
Why is this person here? Has their symptoms changed in any way?

195
00:18:38,250 --> 00:18:44,580
Are they really depressed? Now this is really interesting when you start to look at it, you look at the evidence.

196
00:18:45,810 --> 00:18:50,370
Now I've looked at the evidence and I've kept looking for five or six years for the evidence in primary care to evolve.

197
00:18:50,820 --> 00:18:54,540
And what I've noticed is actually this is published in 2009,

198
00:18:54,540 --> 00:19:05,610
there were only 14 studies and only 2000 people globally involved in randomised trials for anti-depressant in the community in primary care.

199
00:19:05,820 --> 00:19:11,430
That just seems outrageous to me. Not least we're of short duration, typically 6 to 8 weeks.

200
00:19:12,630 --> 00:19:16,110
There is very limited evidence beyond eight weeks of what to do.

201
00:19:17,210 --> 00:19:21,500
This is in an area when you marked in the product, you want to show some effect and get it out there quick.

202
00:19:22,250 --> 00:19:29,330
Now, this was updated again in 2014. What they said that actually the number of studies with observation period of longer than 12 weeks,

203
00:19:30,770 --> 00:19:35,720
comparative analysis of long term effects was not possible. Okay.

204
00:19:35,780 --> 00:19:38,960
So that makes me concerned about the use of antidepressants in primary care.

205
00:19:39,270 --> 00:19:43,790
There's a small effect and that small effect is over 6 to 8 weeks and beyond.

206
00:19:43,790 --> 00:19:51,500
About 12 weeks is very little to inform what to do. Yet most people I see a month guideline say you should give it at least six month treatment.

207
00:19:52,820 --> 00:19:58,820
And these treatments are really interesting. So here's some more interesting that I've always noticed about the way we prescribe, the way we medicate.

208
00:19:59,240 --> 00:20:04,190
This is FDA data. This is the specific odds ratio for suicidal ideation and behaviour.

209
00:20:04,190 --> 00:20:14,030
You can see below 18 on an antidepressant, the odds ratio is about to 18 to 24, it's about 1.5 and then it starts to decrease.

210
00:20:14,210 --> 00:20:15,590
Happy with that? Yep.

211
00:20:16,370 --> 00:20:24,920
So you would look at that and think, well, okay, anybody who's young under the age of 25, I will be more vigilant about using these treatments.

212
00:20:26,060 --> 00:20:29,360
Yeah. You particularly know some of you about paroxetine.

213
00:20:29,360 --> 00:20:38,240
That was a particular issue in adolescents. Yeah, I get these sorts of pieces, which I just cannot fathom what's going on.

214
00:20:38,870 --> 00:20:41,870
This is the is published in 2016.

215
00:20:41,870 --> 00:20:49,970
So only just last month there's been a rise in prescribing of anti-depressants in children of 50% in the last seven or eight years.

216
00:20:50,360 --> 00:20:58,550
20% of taking tricyclic despite guidance, the 2009 guidance against the use of these agents in children and young people.

217
00:21:00,320 --> 00:21:04,130
So we're using treatments more and more that we're being told not to use,

218
00:21:04,760 --> 00:21:09,440
and we use them in the age group where we're having warnings about not using them.

219
00:21:11,200 --> 00:21:18,760
And this is what we start to see. Globally in a decade, huge rises of antidepressants.

220
00:21:19,660 --> 00:21:25,130
This one here is citalopram. Quadrupled over time, doubled over time.

221
00:21:26,700 --> 00:21:29,820
And then you look at the prescribing around the UK of antidepressants.

222
00:21:30,840 --> 00:21:34,890
Now at the back there, just look, this is the grape. It is.

223
00:21:35,640 --> 00:21:44,520
This is 50,000 people in the grape bit. So Oxford's just in there is about 70 to 141 items per thousand people.

224
00:21:45,500 --> 00:21:49,150
Yeah. The blue bits are not the Tory party.

225
00:21:49,960 --> 00:21:53,560
The blue bits are 141 to 185.

226
00:21:54,160 --> 00:21:59,680
Yellow bits are next. But when you get to the red, you're talking 225 to 331.

227
00:22:00,130 --> 00:22:03,370
You're talking one in five people are on an antidepressant.

228
00:22:04,860 --> 00:22:08,390
Yeah. Yeah. But they are in Norfolk. Mhm. Uh huh.

229
00:22:10,020 --> 00:22:13,070
Even if in particular about Norfolk they're.

230
00:22:14,090 --> 00:22:18,980
Well it's interesting I put the. You get the lowest. These are the same IMD for deprivation.

231
00:22:19,040 --> 00:22:24,350
You can look at areas like that and show a little here's Brent that's actually it's not not deprivation

232
00:22:24,950 --> 00:22:31,310
this is the data today from open describing which we do in the IBM data lab 39.5 per thousand.

233
00:22:31,660 --> 00:22:35,360
This is over here. This is 176 per thousand.

234
00:22:36,170 --> 00:22:37,730
It's about fourfold variation.

235
00:22:37,970 --> 00:22:46,130
Oxford right there in the middle, you can explain some of this variation probably by age subgroups, but boy, can you explain all of that?

236
00:22:47,960 --> 00:22:50,000
Can you explain the difference between that?

237
00:22:50,000 --> 00:22:56,600
And it's a really interesting but what's happening is one in seven people is on an antidepressant in this country right now.

238
00:22:57,870 --> 00:23:07,620
That's staggering in the 21st century. That suggests we've got a crisis in mental health, that despite all of our growth, everything we're doing,

239
00:23:07,890 --> 00:23:16,530
we are becoming sadder and worse off or the summit structurally going wrong in how we use medications, how we diagnose people.

240
00:23:17,580 --> 00:23:21,300
In fact, the prescribing is slightly higher than the rate for diagnosis.

241
00:23:21,450 --> 00:23:27,240
Whether that's because there are some factors, but actually we're even giving treatments to people that we don't think are that depressed.

242
00:23:29,020 --> 00:23:32,620
And I can see I get a lot into the jeep.

243
00:23:32,650 --> 00:23:38,830
I get a lot of students who come out and they go into the workforce and they'll be 21, 22, and they'll turn up at my practice in the go.

244
00:23:39,370 --> 00:23:46,300
I'm really tired. Can't get up in the morning and I'm a bit fed up with the world and I'm like, Yeah, that's what the real world like, isn't it?

245
00:23:47,680 --> 00:23:52,600
But you can be careful if you come and do that. A couple of times I've told you going to end up on anti-depressant at some point, so.

246
00:23:53,230 --> 00:23:59,709
And the issue is when you get on them, if you can't get off of them. And this is Peter Grotius position, who writes a lot about this,

247
00:23:59,710 --> 00:24:06,820
is that the with discontinuation symptoms are so bad that it's very hard to get off them, so you end up staying on them.

248
00:24:07,630 --> 00:24:12,010
So we have this epidemic where it continues in time, where you'd expect,

249
00:24:12,010 --> 00:24:15,850
if they were working, that at some point people would actually be coming off these drug.

250
00:24:16,810 --> 00:24:22,270
So although they may be coming on it, you get better. Then you come off. When you feel a bit better, you come down.

251
00:24:22,480 --> 00:24:27,580
That's what we do a lot in medications, unless it's an irreversible that actually just on a continuum of depression.

252
00:24:27,580 --> 00:24:31,330
And we can't do anything about it unless you're unless you're on the drugs.

253
00:24:33,010 --> 00:24:38,950
Okay. So that's on to depression. We've done we've done glaucoma, we've done health screening.

254
00:24:38,950 --> 00:24:44,040
And I've talked to you about antidepressants. Now we're going to just look a little bit at cancer.

255
00:24:45,090 --> 00:24:50,700
And what's interesting about the cancer is the advert is how emotive they are.

256
00:24:52,200 --> 00:24:58,559
This little kit saved lives from bowel cancer. The three vehicle screening test.

257
00:24:58,560 --> 00:25:03,320
Put it on your left. Have a look at the picture.

258
00:25:03,860 --> 00:25:07,549
My mom missed a smear test. Now I miss my mom. Really?

259
00:25:07,550 --> 00:25:11,900
Get into your heartstrings, isn't it? Why? You should have a test? That's what we have to rely on.

260
00:25:11,900 --> 00:25:15,830
We're marketing you to have your test. And I see this.

261
00:25:15,830 --> 00:25:19,990
If you start to become aware, you'll notice these adverts are everywhere in your life.

262
00:25:20,000 --> 00:25:27,830
They're all over the place and you get these ones. Text Test 802 Somebody could have a go if they want.

263
00:25:27,870 --> 00:25:32,659
It'll cost you 8 to 12 p subsequent tax charge and it's done network, right?

264
00:25:32,660 --> 00:25:37,130
And somebody is probably making some money somewhere. But then you get things like this.

265
00:25:37,160 --> 00:25:40,160
I got this. This was on the London Underground.

266
00:25:41,390 --> 00:25:45,320
First of all, it was some TV that I could buy at £23 for a plasma screen.

267
00:25:45,560 --> 00:25:52,219
And so when I went and looked up at this, I nearly died because I nearly fell on the track at this point because I couldn't read it.

268
00:25:52,220 --> 00:25:53,750
And it said the myriad to test it.

269
00:25:53,780 --> 00:26:00,620
The first over-the-counter blood test that allows you to screen yourself and quickly indiscreetly for a range of 14 different health conditions.

270
00:26:01,430 --> 00:26:07,970
So you can do that now. You can go and screen yourself, you can pick all these tests, you can have a CT scan,

271
00:26:08,420 --> 00:26:14,090
you can go and see your doctor and you'll definitely get anti-depressant. By the time you're 40, you can have a health check.

272
00:26:15,780 --> 00:26:19,890
But this is what seemingly is happening and this is data from Cancer Research UK.

273
00:26:20,190 --> 00:26:27,630
While we are increasing the incidence of thyroid cancer per phone fans and people, this is what's happening to mortality.

274
00:26:28,510 --> 00:26:31,590
Okay, it's flatlining. Very little effect.

275
00:26:32,340 --> 00:26:38,100
So we diagnose in more conditions and this is what the term overdiagnosis is referred to.

276
00:26:39,720 --> 00:26:44,940
Now, this is how they see it, the diagnosis of conditions that will never cause symptoms or harm during the patient's long time.

277
00:26:45,510 --> 00:26:50,760
The consequences are unnecessary anxiety, harms from excessive treatment, and increased health care costs.

278
00:26:51,120 --> 00:26:55,320
Risk that those who don't need treatment are prioritised over those who do.

279
00:26:58,310 --> 00:27:00,650
And this is the thought diagram that you get seen a lot.

280
00:27:01,400 --> 00:27:07,520
And Jason here will know this diagram very well in the cans and the stats people, you get diagnosed with an abnormal cell.

281
00:27:07,760 --> 00:27:12,410
We don't know whether that's going to be a fast growing cancer, whether it's going to be slow on that,

282
00:27:12,530 --> 00:27:15,910
get you to this at which line this is where it causes symptoms.

283
00:27:15,920 --> 00:27:19,850
This is where you die or is it very slow cells? You'll die something else in your lifetime?

284
00:27:20,120 --> 00:27:26,000
Or actually, is it just going to be non-core, progressive or actually going to be regressive over time and you'll return to normal?

285
00:27:26,420 --> 00:27:34,190
That can happen in breast cancer and cervical cancer and all these areas where you make a diagnosis, you have abnormal cells.

286
00:27:34,430 --> 00:27:40,760
We know very little about this right now. And so this is what it looks like for breast cancer.

287
00:27:41,600 --> 00:27:45,770
The incidence has been going up and the mortality has been coming down.

288
00:27:46,040 --> 00:27:54,319
People would like to say that screenings introduced. This is the effects that we see for screening, that this is what they like to say.

289
00:27:54,320 --> 00:28:02,450
For every 10,000 women screened, 43 deaths from breast cancer would be prevented and 129 cases of breast cancer would be overdiagnosed.

290
00:28:02,780 --> 00:28:10,110
So for about every one where we find a case, three of three cases of overdiagnosed are detected, somebody that something wouldn't happen to.

291
00:28:10,730 --> 00:28:13,600
But that's assuming that all of this mortality through screening,

292
00:28:13,610 --> 00:28:18,460
it's not it's not actually based on treatment effects, both treatments getting better.

293
00:28:19,130 --> 00:28:24,350
So we don't really know about what's really happening with breast cancer screening the effect both

294
00:28:24,350 --> 00:28:31,879
of the trial for 40 years old and I actually did this when I went to the W.H.O. believe it or not,

295
00:28:31,880 --> 00:28:38,060
the W.H.O. figured in all the world problems that actually one of the priorities was to produce a guideline on breast cancer screening,

296
00:28:39,140 --> 00:28:41,050
because there are only two camps there.

297
00:28:41,060 --> 00:28:47,300
Either people who believe it's the best thing to do things like spread, or those people who think you should never do breath screening.

298
00:28:48,170 --> 00:28:53,060
And in the middle, when I looked at the evidence, all I could see with huge amounts of uncertainty.

299
00:28:54,400 --> 00:28:59,920
Evidence that with 40 years old, where the mortality reductions were based on whether you had a mastectomy or not,

300
00:28:59,920 --> 00:29:04,660
which is what we don't even do anymore for the treatment changed and the data is very old.

301
00:29:04,900 --> 00:29:09,850
So I would say that there is saying, wow, this is a real myth. How do we make sense of this?

302
00:29:10,180 --> 00:29:13,850
Where do we go next? Okay.

303
00:29:14,930 --> 00:29:21,710
And this is the sorts of things you see a lot of, don't you? Our survival rate lags years behind those in Europe.

304
00:29:23,660 --> 00:29:29,120
In fact, the easiest way to improve survival rates is to just diagnose something earlier.

305
00:29:29,750 --> 00:29:36,560
But it is not the same as improving overall mortality, because if you count it in five year survival,

306
00:29:36,800 --> 00:29:42,080
you want to diagnose things much earlier and you can improve five year survival really quickly.

307
00:29:42,350 --> 00:29:48,740
And that's when you hear about socialised medicine in America. What they're doing is, for instance, they diagnosed prostate cancer earlier.

308
00:29:49,340 --> 00:29:53,210
A five year survival is better, but the overall mortality is no different.

309
00:29:53,220 --> 00:29:58,430
People die at the same age from prostate cancer in the UK and in the U.S.

310
00:29:59,150 --> 00:30:03,650
And that's important to know. And this is what's happening now.

311
00:30:04,580 --> 00:30:08,870
We are referring more patients with low risk cancer systems under new knife guidance.

312
00:30:09,560 --> 00:30:16,820
So, for instance, if you have a cough, how many people here have had a cough that's gone on for three weeks?

313
00:30:18,530 --> 00:30:26,900
Okay. On the ninth guidance, you should have all been referred for a chest x ray because that's what knife is starting to tell it.

314
00:30:27,200 --> 00:30:33,500
And in that position is saying what's happened in the last five years is the number of x rays in Oxford have doubled.

315
00:30:34,130 --> 00:30:38,120
And Jack's just walked in the room and he knows that because that's what is felt about.

316
00:30:38,330 --> 00:30:45,880
And somehow we're good. Who's accounted for all this extra work? Why is it people are not paying attention to when we do things like this?

317
00:30:46,270 --> 00:30:49,240
You massively increase the amount of work in the NHS.

318
00:30:50,920 --> 00:30:57,490
And everybody is now telling us, despite all the extra investment, the nature of even more crisis than it's ever been.

319
00:31:00,130 --> 00:31:05,920
And that's interesting. And before I came here, I was having the question with Geoff at the back about this talk.

320
00:31:06,580 --> 00:31:17,230
And in 1997, 20 years ago, there were 1437 billion into the health service and I worked with Jeff directly in the NHS.

321
00:31:18,220 --> 00:31:21,460
20 years later, we now spend 120 billion.

322
00:31:22,880 --> 00:31:26,000
And we have the same structural problems that we had 20 years.

323
00:31:26,030 --> 00:31:34,070
In some ways, they're even worse. Now, why is that occurring if we put all that extra investment in as a huge amount of money, isn't it?

324
00:31:34,670 --> 00:31:40,730
We've doubled the number of doctors in that town. We train twice as many doctors now than we did then.

325
00:31:42,720 --> 00:31:45,870
And what has happened. You can do this.

326
00:31:46,140 --> 00:31:51,330
And there's Dylan at the back. You can do this all the time. You can go and get yourself tested all the time.

327
00:31:51,990 --> 00:31:56,460
You can be advertised to. You can be marketed to. You can be at the airport.

328
00:31:56,700 --> 00:32:01,440
And you can do your blood pressure from the fly healthy seat on the machine.

329
00:32:01,500 --> 00:32:07,020
And what does all that mean? Here's where I'm at. When you have things like this, you have to think about.

330
00:32:07,170 --> 00:32:10,230
Do you want to know? Would you be happy to live with the disease or not?

331
00:32:10,260 --> 00:32:18,600
Would you happy to be live with the risk or not? Does it ultimately help you to know whether you've got a high risk or not?

332
00:32:18,630 --> 00:32:22,860
Are you the sort of person that can tolerate that or will it make you take more time off work?

333
00:32:22,890 --> 00:32:26,580
Do you want to know? Do we want to know? We had that debate.

334
00:32:27,000 --> 00:32:33,060
And so when you do things like that, often go back in time and think, well, somebody must have said something sensible about this in the past.

335
00:32:34,110 --> 00:32:40,320
And in going back, this is the first thing I think most information relates to tested patients.

336
00:32:41,130 --> 00:32:45,050
It seems you've already had the test to get you to screening.

337
00:32:45,060 --> 00:32:53,430
You have to be most information is emotive relying on scare tactics and you most individuals are not equipped to decide whether to take tests or not.

338
00:32:54,840 --> 00:33:02,909
That fair comment you might be once you've had the test, but beforehand I'm showing you all the illustrations where I'm in a position where I'm going.

339
00:33:02,910 --> 00:33:10,500
I'm not quite sure about this. And so I look back at the World Health Organisation in 1971, the screening criteria,

340
00:33:10,560 --> 00:33:20,160
and there's one thing that sticks out to me screening most lead to an improvement in end results defined in terms of mortality,

341
00:33:20,790 --> 00:33:29,790
physical, social and emotional function, pain and satisfaction among those in whom early diagnosis is achieved or in other members of the community.

342
00:33:31,290 --> 00:33:36,090
If I applied that definition to health screening, would I have achieved that?

343
00:33:37,630 --> 00:33:43,540
Ask yourself, then, if we haven't achieved that, why have we rolled it out on a national basis for everybody to make it available?

344
00:33:44,680 --> 00:33:49,990
Have I achieved that for depression? Am I sure about that when it comes to the glaucoma?

345
00:33:50,800 --> 00:33:53,920
And there are many instances where I don't think we are in that position.

346
00:33:54,610 --> 00:34:01,870
So this is where I am with the overdiagnosis that too much demand, too much screening,

347
00:34:02,140 --> 00:34:08,050
too much testing, too much treatment, and therefore overdiagnosis is really just too much health care.

348
00:34:08,110 --> 00:34:11,200
And that's where we are today. Thank you very much.