WEBVTT

1
00:00:00.540 --> 00:00:09.380
N.I.H. Fails to Institute Health A root cause failure analysis to show why the world's largest medical research institution

2
00:00:09.380 --> 00:00:17.620
only allocates three point three percent of its budget to testing which of the thousands of known-safe treatments actually

3
00:00:17.620 --> 00:00:26.880
work in humans. Oxford tested COVID treatments for five hundred dollars per patient and saved a million lives in one hundred

4
00:00:26.880 --> 00:00:35.240
days. The N.I.H. spent one point six billion dollars and completed zero trials in four years.

5
00:00:35.960 --> 00:00:43.380
A Tale of Two Trials. The difference between efficient and inefficient trial design isn't theoretical.

6
00:00:43.980 --> 00:00:48.740
Your species accidentally ran a controlled experiment on itself during COVID.

7
00:00:49.440 --> 00:00:54.140
Two countries. Same disease. One spent like it was trying to cure death.

8
00:00:54.660 --> 00:01:03.120
The other actually did. RECOVERY Trial (U.K. Approach). The British spent twenty million dollars over six months.

9
00:01:03.700 --> 00:01:11.120
Enrolled forty-eight thousand patients. Found multiple effective treatments, including dexamethasone, a steroid that had

10
00:01:11.120 --> 00:01:20.140
been available since nineteen fifty-seven. It reduced COVID deaths by one-third and saved over one million lives.

11
00:01:21.000 --> 00:01:26.540
Cost per patient: five hundred dollars. RECOVER Initiative (N.I.H.

12
00:01:26.700 --> 00:01:32.080
Approach). Nearly six hundred potential treatments for Long COVID have been identified.

13
00:01:32.680 --> 00:01:39.880
At RECOVERY trial efficiency (five hundred dollars per patient, five thousand five hundred patients per treatment), one

14
00:01:39.880 --> 00:01:45.040
point six six five billion dollars buys six hundred completed trials.

15
00:01:45.880 --> 00:01:49.820
The N.I.H. received one point six six five billion dollars.

16
00:01:50.300 --> 00:01:59.740
Over four years, they completed zero. Only fifteen percent of the budget went to clinical trials; the rest funded observational

17
00:01:59.740 --> 00:02:08.300
studies, data management, and administration. Of those enrolled, roughly eleven percent were in interventional trials that

18
00:02:08.360 --> 00:02:15.620
actually tested treatments. The other eighty-nine percent were in observational studies, which is the medical equivalent

19
00:02:15.620 --> 00:02:24.100
of taking attendance at a funeral. Cost per interventional patient: five hundred and four thousand, five hundred dollars,

20
00:02:24.480 --> 00:02:28.860
the price of a house. Nobody got a house and nobody got cured.

21
00:02:29.820 --> 00:02:34.200
They got observed. Expensively. The observation was very thorough.

22
00:02:34.760 --> 00:02:38.320
It has been peer-reviewed. The peers also observed the patients.

23
00:02:38.920 --> 00:02:45.840
The patients are still sick, but they are the most well-documented sick people in the history of sickness.

24
00:02:46.540 --> 00:02:54.160
It gets worse. The treatments RECOVER eventually selected were called "truly absurd" by patient advocates.

25
00:02:54.320 --> 00:03:02.200
Low-dose naltrexone, the number one patient-requested treatment (generic, cheap, already showing results in independent

26
00:03:02.200 --> 00:03:08.760
studies), was excluded for over two years before finally being added in round two.

27
00:03:09.580 --> 00:03:15.940
David Putrino, director of rehabilitation innovation at Mount Sinai, called the spending "largely wasted".

28
00:03:16.760 --> 00:03:21.460
Researchers outside RECOVER had lists of drugs to test. The system ignored them.

29
00:03:22.720 --> 00:03:27.200
Six hundred trials were possible with that money. Zero were completed.

30
00:03:27.940 --> 00:03:33.430
That is the difference between a system that tests treatments and a system that tests patience.

31
00:03:34.460 --> 00:03:41.540
The Translation Crisis: All Theory, No Medicine. Basic research is not the problem.

32
00:03:42.240 --> 00:03:48.260
Five decades of N.I.H.-funded science built one of the largest drug candidate pipelines in history.

33
00:03:48.540 --> 00:03:53.680
Researchers identified targets, characterized molecules, established safety profiles.

34
00:03:54.080 --> 00:04:02.340
That work was essential, and it succeeded. The problem is what happens after the discoveries are made: almost nothing.

35
00:04:03.120 --> 00:04:11.200
The pipeline is full. The exit is closed. COVID made this visible for the first time, but it is the institution's default

36
00:04:11.200 --> 00:04:15.630
setting; the way you don't notice a leaking roof until it rains inside.

37
00:04:16.240 --> 00:04:23.560
It has been raining inside for fifty years. The N.I.H. has published several papers on indoor rain.

38
00:04:24.080 --> 00:04:28.530
There are over twenty thousand F.D.A.-approved drugs that are known to be safe in humans.

39
00:04:29.100 --> 00:04:34.740
Thousands more clinically-tested candidates are gathering dust, including over three thousand marketed drugs that could

40
00:04:34.740 --> 00:04:41.860
be repurposed for diseases nobody's tried them on. In Alzheimer's alone, researchers have identified hundreds of safe drugs

41
00:04:41.860 --> 00:04:45.740
worth testing, yet less than ten percent are in active trials.

42
00:04:46.280 --> 00:04:49.420
And the number of clinical trials available for your grandma in St.

43
00:04:49.560 --> 00:04:56.660
Louis? Zero. You have a warehouse of safe molecules that have never been systematically tested for new uses.

44
00:04:57.040 --> 00:05:01.260
The warehouse is full. The testing facility is closed for a team-building retreat.

45
00:05:02.180 --> 00:05:09.060
This is not hypothetical. In twenty twenty-three, researchers published case reports of severe long COVID patients experiencing

46
00:05:09.060 --> 00:05:13.960
complete remission within days of receiving monoclonal antibody infusions.

47
00:05:14.740 --> 00:05:20.760
Previously bedridden patients returned to normal lives, with remissions sustained over two years.

48
00:05:21.400 --> 00:05:28.080
The U.S. government had spent over five billion dollars purchasing millions of doses of these same antibodies for acute

49
00:05:28.080 --> 00:05:35.660
COVID. When Omicron rendered them ineffective for acute treatment, millions of unused doses sat in refrigerated storage.

50
00:05:36.340 --> 00:05:40.180
Long COVID patients, having read the case reports, begged for access.

51
00:05:40.780 --> 00:05:45.140
They organized petitions. They cited the remissions. The government said no.

52
00:05:45.520 --> 00:05:52.180
The doses were authorized for acute COVID only, and the authorization had been revoked because they didn't work for acute

53
00:05:52.180 --> 00:05:56.500
COVID, which meant they could not be given to anyone, for anything, ever.

54
00:05:57.400 --> 00:06:03.300
The antibodies matched the original variants; the long COVID patients had been infected by the original variants.

55
00:06:03.900 --> 00:06:09.100
Nobody tested this. Nobody ran trials. Nobody repurposed the stockpile.

56
00:06:09.780 --> 00:06:17.820
The first formal Phase two trial of monoclonal antibodies for long COVID is scheduled for mid-twenty twenty-six, four years

57
00:06:17.820 --> 00:06:21.940
after the doses were available and several years after they expired.

58
00:06:22.600 --> 00:06:27.880
Five billion dollars of medicine. Millions of doses. Patients who wanted them.

59
00:06:28.340 --> 00:06:35.200
Case reports showing they worked. The doses expired. The patients didn't; they just kept being sick.

60
00:06:35.920 --> 00:06:44.160
Your government threw away the cure and kept the disease. On Wishonia, we have a word for this, but it is a word I am not

61
00:06:44.160 --> 00:06:51.620
permitted to use in your language because it is considered a slur against the competence of your entire species, so I will

62
00:06:51.620 --> 00:06:56.820
simply note that the word exists and move on. Then there's metformin.

63
00:06:56.880 --> 00:07:05.240
A phase three randomized controlled trial, or R.C.T. (COVID-OUT) showed it reduces long COVID incidence by forty-one percent.

64
00:07:05.720 --> 00:07:10.170
It costs four dollars a month. It has been generic for decades.

65
00:07:10.340 --> 00:07:17.040
It is globally available. The N.I.H. guideline response: "insufficient evidence to recommend for or against".

66
00:07:17.960 --> 00:07:22.060
Long COVID costs one hundred and seventy billion dollars a year.

67
00:07:22.660 --> 00:07:30.660
The four dollar pill that cuts it by forty-one percent has no corporate sponsor, because nobody profits from a generic.

68
00:07:31.160 --> 00:07:38.460
This is the entire chapter in one sentence. To appreciate the scale: long COVID costs the U.S.

69
00:07:38.480 --> 00:07:46.580
an estimated one hundred and seventy billion dollars annually in lost wages alone, with roughly seventeen million Americans

70
00:07:46.580 --> 00:07:54.440
currently affected. A RECOVERY-style pragmatic trial would have cost approximately twenty-four million dollars and could

71
00:07:54.440 --> 00:08:01.900
have produced results in months. Instead, your government spent five billion dollars buying the medicine and is now paying

72
00:08:01.900 --> 00:08:08.260
one hundred and seventy billion dollars a year in economic damage from the disease it didn't bother to treat.

73
00:08:09.020 --> 00:08:14.800
The ratio of what you're losing to what the trial would have cost is seven thousand and eighty-three to one.

74
00:08:15.640 --> 00:08:21.240
The N.I.H. is, by its own metrics, a publishing house that occasionally dabbles in medicine.

75
00:08:21.680 --> 00:08:25.720
It publishes two point five million funded papers annually.

76
00:08:26.120 --> 00:08:29.860
It produces approximately fifty new drug approvals annually.

77
00:08:30.220 --> 00:08:35.419
That's a conversion rate from paper to patient of zero point zero zero two percent.

78
00:08:36.380 --> 00:08:44.780
For every fifty thousand papers, one drug reaches a human. The rest become citations in other papers, which become citations

79
00:08:44.780 --> 00:08:49.380
in other papers, in an infinite loop of academics reading each other's homework.

80
00:08:50.080 --> 00:08:55.620
On Wishonia, we call this a "knowledge ouroboros." It looks productive from the inside.

81
00:08:55.920 --> 00:09:00.760
From the outside, it's a snake eating its own tail and writing a paper about the nutritional content.

82
00:09:01.660 --> 00:09:08.280
Plenty of Knowledge. No Translation. About five percent of patients have participated in clinical trials.

83
00:09:08.740 --> 00:09:12.700
About forty-four point eight percent would participate if invited.

84
00:09:13.240 --> 00:09:20.480
About nine to eleven percent have ever been invited. Nearly half of all patients would join a trial if asked.

85
00:09:21.100 --> 00:09:28.540
Almost nobody asks. Your species has a room full of volunteers holding their hands up, and a system that looked at the room

86
00:09:28.540 --> 00:09:34.020
and decided to study the clipboard instead. The clipboard is very well-studied.

87
00:09:34.400 --> 00:09:38.530
It has been studied for decades. Multiple papers have been written about the clipboard.

88
00:09:38.980 --> 00:09:43.000
One of them won an award. The volunteers are still holding their hands up.

89
00:09:43.480 --> 00:09:46.560
Some of them have died. This has been noted on the clipboard.

90
00:09:47.400 --> 00:09:52.820
Understanding disease and testing treatments in humans are completely different activities.

91
00:09:53.280 --> 00:09:59.070
One requires microscopes. The other requires giving people medicine and seeing if they stop dying.

92
00:09:59.460 --> 00:10:05.230
In nineteen seventy, spending heavily on microscopes made sense; you didn't know enough to test intelligently.

93
00:10:06.020 --> 00:10:13.120
In twenty twenty-five, after two point five million papers a year for five decades, the ratio should have shifted.

94
00:10:13.780 --> 00:10:21.780
It didn't. The N.I.H. chose microscopes in nineteen seventy and has not reconsidered since, which is the institutional equivalent

95
00:10:21.780 --> 00:10:25.360
of still crawling because you learned to crawl before you learned to walk.

96
00:10:26.180 --> 00:10:34.380
The Allocation Problem. Your National Institutes of Health has the word "Health" in its name and an annual budget of forty-seven

97
00:10:34.380 --> 00:10:39.740
billion dollars. Fifty-five million of you die annually from disease.

98
00:10:40.220 --> 00:10:43.700
You would think those two facts would be related. They are not.

99
00:10:44.120 --> 00:10:51.100
Of two hundred and forty-seven billion dollars the N.I.H. spent over a decade, exactly three point three percent went to

100
00:10:51.140 --> 00:10:55.360
testing which of the thousands of known-safe treatments actually work in humans.

101
00:10:56.120 --> 00:11:02.720
The other ninety-six point seven percent funded understanding disease mechanisms, translating discoveries toward (but never

102
00:11:02.720 --> 00:11:10.180
into) clinical use, training scientists, observational programs that spend billions watching disease without treating it,

103
00:11:10.600 --> 00:11:17.080
and a side quest engineering bat viruses, which is called "research" the way watching someone drown from a boat is called

104
00:11:17.080 --> 00:11:24.320
"lifeguarding.". Basic research is not the problem. Understanding disease is how you find candidates to test.

105
00:11:25.020 --> 00:11:29.860
The problem is that after fifty years and a trillion dollars, the ratio hasn't moved.

106
00:11:30.700 --> 00:11:37.360
The marginal value of the eighty-fifth cent on basic research is near zero when you're spending three cents on the only

107
00:11:37.360 --> 00:11:41.850
activity that turns research into medicine. You don't have a knowledge deficit.

108
00:11:42.480 --> 00:11:49.660
You have an allocation deficit. Three point three percent on trials might be defensible if the testing queue were empty.

109
00:11:50.360 --> 00:11:57.560
It isn't. There are roughly nine thousand five hundred compounds with established safety profiles, and ninety-nine point

110
00:11:57.560 --> 00:12:02.220
seven percent of their possible disease applications have never been tested.

111
00:12:02.480 --> 00:12:09.380
Most are off-patent, which means no pharmaceutical company will fund trials for them, which means the N.I.H.

112
00:12:09.380 --> 00:12:16.700
is the only entity that could, which means it is specifically choosing not to do the one thing only it can do.

113
00:12:17.300 --> 00:12:21.460
The pipeline is overflowing with candidates. The bottleneck is trials.

114
00:12:21.980 --> 00:12:28.170
The drugs are sitting there. Safe. Ready. Unpatentable. They've been waiting for decades.

115
00:12:28.800 --> 00:12:32.400
Some of them have been waiting longer than most of your scientists have been alive.

116
00:12:33.080 --> 00:12:40.520
It's like spending ninety-six point seven percent of your grocery budget on cookbooks and three point three percent on food,

117
00:12:40.620 --> 00:12:47.020
then wondering why you're starving. Except the cookbooks are very prestigious and the starving people aren't on the committee

118
00:12:47.020 --> 00:12:53.660
that decides whether to buy more cookbooks. A J.A.M.A. study broke down two hundred and forty-seven point three billion

119
00:12:53.660 --> 00:12:57.380
dollars in N.I.H. spending (twenty ten to twenty nineteen).

120
00:12:58.100 --> 00:13:05.560
I have organized it from "largest amount spent not on clinical trials" to "smallest amount spent not clinical trials," though

121
00:13:05.560 --> 00:13:08.940
I should note that none of it is labeled "not clinical trials."

122
00:13:09.280 --> 00:13:15.140
It is labeled "investment in the future." The future has been invested in for fifty years.

123
00:13:15.700 --> 00:13:22.280
It has not yet arrived. Eighty-four point nine percent, or two hundred and nine point nine billion dollars, is dedicated

124
00:13:22.280 --> 00:13:29.940
to basic research for understanding disease mechanisms, molecular biology, and genomics; while essential for identifying

125
00:13:29.980 --> 00:13:36.000
new targets, eighty-five cents of every dollar is a lot of understanding and not much testing.

126
00:13:37.000 --> 00:13:44.320
Eleven point eight percent, or twenty-nine point three billion dollars, goes to non-trial applied research that translates

127
00:13:44.320 --> 00:13:50.080
discoveries toward practical use, but without the final step of testing them in patients.

128
00:13:51.020 --> 00:13:58.920
The Clinical and Translational Science Awards, or C.T.S.A., infrastructure program spends about one billion dollars per

129
00:13:58.920 --> 00:14:07.440
year in trial infrastructure, coordination, and training. Observational studies for population health data collection (such

130
00:14:07.440 --> 00:14:15.680
as large-scale observational cohorts costing two point one six billion dollars) are valuable for identifying patterns, but

131
00:14:15.680 --> 00:14:23.840
observation without intervention doesn't cure anyone. Other applied research includes drug characterization, health services

132
00:14:23.840 --> 00:14:33.060
research, and epidemiology. Only three point three percent, or eight point one billion dollars, is spent on phased clinical

133
00:14:33.060 --> 00:14:40.200
trials, which is the only category that directly tests whether treatments work in humans through phases one to three.

134
00:14:41.220 --> 00:14:47.280
Phase one accounts for one point five billion dollars, or eighteen point five percent of the clinical trial budget.

135
00:14:47.820 --> 00:14:52.440
Phase two accounts for three point five billion dollars, or forty-three point two percent.

136
00:14:53.180 --> 00:14:58.260
Phase three accounts for two point six billion dollars, or thirty-two point one percent.

137
00:14:59.100 --> 00:15:07.140
The N.I.H. covers only three point seven to four point three percent of Phase three costs, leaving completion to industry,

138
00:15:07.760 --> 00:15:10.800
which only funds trials with patent-protected profit potential.

139
00:15:12.000 --> 00:15:19.880
Other unclassified spending totals zero point five billion dollars, or six point two percent, for cross-phase overhead,

140
00:15:20.240 --> 00:15:23.480
pilot studies, and trials not classified to a single phase.

141
00:15:24.340 --> 00:15:31.800
Pragmatic trials (eighty-two times cheaper per patient) are severely underfunded despite proven efficiency.

142
00:15:32.920 --> 00:15:37.090
Why This Allocation Exists. This is not the fault of individual researchers.

143
00:15:37.980 --> 00:15:40.840
Most became scientists because they wanted to cure diseases.

144
00:15:41.440 --> 00:15:49.050
The system they work in rewards everything except curing diseases, and those rewards were designed by someone else entirely.

145
00:15:50.040 --> 00:15:54.220
Huge drug companies with political influence love the N.I.H.

146
00:15:54.360 --> 00:15:58.560
Taxpayers absorb the risk of basic research; industry patents whatever works.

147
00:15:59.080 --> 00:16:04.520
It's like if your neighbor paid for your kitchen renovation, and then you charged them rent to eat there.

148
00:16:05.360 --> 00:16:10.180
But independent trials comparing drugs head-to-head? Those are terrifying.

149
00:16:10.720 --> 00:16:16.060
They might reveal that the eighty thousand dollar drug works about as well as the four dollar generic.

150
00:16:16.780 --> 00:16:22.180
PhRMA, the pharmaceutical lobby, has consistently opposed comparative effectiveness research.

151
00:16:22.860 --> 00:16:31.000
When Congress created P.C.O.R.I. (the Patient-Centered Outcomes Research Institute) in twenty-ten, industry successfully

152
00:16:31.060 --> 00:16:36.440
lobbied to prohibit it from making coverage recommendations based on cost-effectiveness.

153
00:16:37.840 --> 00:16:43.520
Congress literally passed a law saying "don't fund research that might lower drug prices."

154
00:16:44.200 --> 00:16:49.760
They wrote it down. In legislation. On purpose. With their names on it.

155
00:16:50.340 --> 00:16:57.120
And then they went home and told their constituents they were fighting for affordable healthcare, because your species'

156
00:16:57.380 --> 00:17:04.660
capacity for doublethink is genuinely Olympian. This pattern held during the deadliest pandemic in a century.

157
00:17:05.339 --> 00:17:12.339
Operation Warp Speed initially selected fourteen vaccine candidates and planned head-to-head comparative studies in primates.

158
00:17:12.859 --> 00:17:20.099
Then dropped the comparative studies without explanation. Backed five companies with billions, gave others nothing.

159
00:17:20.800 --> 00:17:27.480
The selection process was opaque enough to draw Senate scrutiny; when Senator Tammy Baldwin asked the Biomedical Advanced

160
00:17:27.480 --> 00:17:35.260
Research and Development Authority, or BARDA's acting director to name the companies receiving public funds, he refused,

161
00:17:35.340 --> 00:17:41.480
citing "procurement sensitivity". No published cost-benefit analysis for excluding any candidate.

162
00:17:42.340 --> 00:17:49.340
The W.H.O. ran a Solidarity Trial for Vaccines doing exactly this kind of head-to-head comparison.

163
00:17:49.800 --> 00:17:59.060
The U.S. didn't join. Notice the pattern: the Patient-Centered Outcomes Research Institute, or PCORI is banned from comparing

164
00:17:59.060 --> 00:18:07.620
drug cost-effectiveness. Researching COVID to Enhance Recovery, or RECOVER ignores the treatments patients actually want.

165
00:18:08.820 --> 00:18:16.840
O.W.S. drops the comparative monkey studies. Every manufacturer has financial interests; Pfizer had them, Moderna had them,

166
00:18:17.120 --> 00:18:24.700
and that wasn't disqualifying. "Financial interest" only becomes a reason not to test when the treatment is cheap, generic,

167
00:18:24.940 --> 00:18:31.120
or threatens incumbent revenue. The system tests what's profitable and ignores what isn't.

168
00:18:31.360 --> 00:18:36.160
The revolving door helps. N.I.H. advisory committees include industry representatives.

169
00:18:36.800 --> 00:18:42.780
Former N.I.H. officials join pharmaceutical boards. Academic researchers depend on industry grants.

170
00:18:43.440 --> 00:18:50.380
Everyone's incentives align against cheap, fast trials testing whether expensive drugs outperform generics.

171
00:18:51.120 --> 00:18:55.770
It's not a conspiracy. Conspiracies require secrecy. This is all public.

172
00:18:56.280 --> 00:19:04.060
It's just that everyone involved happens to benefit from the same terrible outcome, purely by coincidence, repeatedly, for

173
00:19:04.060 --> 00:19:09.780
decades. On Wishonia, we have a word for this. The word is "conspiracy."

174
00:19:10.080 --> 00:19:15.380
The difference between a conspiracy and a system is whether the participants need to meet in a dark room.

175
00:19:15.860 --> 00:19:20.260
These people meet in conference rooms with catering. The outcome is identical.

176
00:19:20.520 --> 00:19:24.780
The three point three percent allocation isn't a mystery. It's a business model.

177
00:19:25.540 --> 00:19:29.220
Basic research produces publications. Publications produce tenure.

178
00:19:29.580 --> 00:19:34.040
Tenure produces more grants. Individual researchers want cures.

179
00:19:34.080 --> 00:19:37.760
The system rewards them for everything except producing one.

180
00:19:38.260 --> 00:19:45.520
Actual treatments are a byproduct that occasionally happens when someone forgets to optimize for papers and accidentally

181
00:19:45.520 --> 00:19:53.100
optimizes for results instead. It's like a restaurant where the chefs are passionate about food but get promoted based on

182
00:19:53.100 --> 00:19:57.710
how many menus they print. The menus are exquisite. They have won awards.

183
00:19:58.420 --> 00:20:02.280
The diners are eating the menus. Nobody has mentioned this.

184
00:20:02.660 --> 00:20:08.940
You Pay Twice. The N.I.H. claims to produce "public goods" like open science and training.

185
00:20:09.360 --> 00:20:16.020
What it actually produces is a free R. and D. department for pharmaceutical companies, funded by the same people who then

186
00:20:16.020 --> 00:20:24.600
pay full price for the drugs. Taxpayers funded research behind ninety-nine percent of new drugs (three hundred and fifty-six

187
00:20:24.600 --> 00:20:29.800
drugs, twenty-ten to twenty-nineteen). Industry patents the winners.

188
00:20:30.340 --> 00:20:37.060
The public pays again through high drug prices. When research leads to a dead end, the public eats the loss.

189
00:20:37.640 --> 00:20:43.700
You pay for the lottery ticket. They cash the winnings. You'd think someone would have mentioned this arrangement before

190
00:20:43.700 --> 00:20:52.000
now, and someone has, many times, but the people who cash the winnings also fund the campaigns of the people who could change

191
00:20:52.000 --> 00:20:57.560
the arrangement, so the arrangement continues. The only things the N.I.H.

192
00:20:57.560 --> 00:21:06.040
*should* fund are things private companies *cannot* profitably fund: open pragmatic trial platforms, repurposing of generic

193
00:21:06.040 --> 00:21:14.460
drugs where no patent means no profit means no interest, and comparative effectiveness data from head-to-head trials revealing

194
00:21:14.460 --> 00:21:20.180
which drugs actually work best (pharma's nightmare, your potential salvation).

195
00:21:20.760 --> 00:21:26.020
Instead, the N.I.H. explicitly avoids them to "not crowd out industry."

196
00:21:26.380 --> 00:21:33.600
Translation: "We won't run the trials that would lower drug prices, because the people who set drug prices asked us nicely."

197
00:21:34.160 --> 00:21:40.160
They asked nicely and also spent one hundred and twenty-seven million dollars on lobbying, but the niceness was noted.

198
00:21:40.680 --> 00:21:44.960
The Patient Disconnect: Zero Correlation with Health Outcomes.

199
00:21:45.440 --> 00:21:52.760
Correlation between N.I.H. funding priorities and actual disease burden: zero point zero seven.

200
00:21:53.580 --> 00:21:59.860
That is not a typo. On a scale of zero (random) to one (perfect), the N.I.H.

201
00:21:59.860 --> 00:22:05.620
scores "basically a coin flip." A dartboard would allocate research funding more rationally.

202
00:22:06.100 --> 00:22:13.080
A drunk person throwing darts at a list of diseases would produce a funding allocation that correlates more closely with

203
00:22:13.080 --> 00:22:19.940
actual disease burden than the National Institutes of Health, which employs twenty thousand people and has been doing this

204
00:22:19.940 --> 00:22:24.900
for fifty years. The dartboard doesn't even need to be in the building.

205
00:22:25.540 --> 00:22:33.600
Patients want medicine that works, access to experimental therapies, trials they can join, and a cure before they die.

206
00:22:34.220 --> 00:22:40.480
The N.I.H. funds molecular mechanisms, research empires, and institutional overhead.

207
00:22:41.340 --> 00:22:45.170
These two lists do not overlap. They don't even live in the same neighborhood.

208
00:22:45.980 --> 00:22:53.520
Patients don't control the money. Committees do. And committees optimize for the continued existence of committees.

209
00:22:53.960 --> 00:22:58.020
This is true of all committees, everywhere, since the invention of committees.

210
00:22:58.540 --> 00:23:02.600
Nobody has ever convened a committee that recommended fewer committees.

211
00:23:02.700 --> 00:23:08.300
On Wishonia, we have a word for an organism that exists solely to produce copies of itself.

212
00:23:08.940 --> 00:23:15.720
The word is "virus." You call them "advisory panels.". The Human Cost of Misallocation.

213
00:23:16.280 --> 00:23:24.340
When you allocate the vast majority of a forty-seven billion dollar annual budget into understanding disease instead of

214
00:23:24.340 --> 00:23:32.840
testing the safe, unpatentable treatments already gathering dust on the shelf, you get a very well-understood population

215
00:23:32.840 --> 00:23:40.440
of dead people. Cost Per Quality-Adjusted Life Year. A Quality-Adjusted Life Year is one year of healthy life.

216
00:23:41.280 --> 00:23:45.320
Lose a leg and live ten years, that's fewer than ten Q.A.L.Y.s.

217
00:23:46.060 --> 00:23:49.980
Die at forty instead of eighty, that's forty Q.A.L.Y.s lost.

218
00:23:50.580 --> 00:23:56.680
It's how you measure whether a medical system is keeping people alive and functional, or just keeping itself funded.

219
00:23:57.560 --> 00:24:04.700
The ADAPTABLE trial proved pragmatic trials cost forty-four point one times less per patient than traditional randomized

220
00:24:05.100 --> 00:24:11.020
controlled trials, or R.C.T.s (nine hundred and twenty-nine dollars versus forty-one thousand dollars).

221
00:24:11.780 --> 00:24:17.460
Same science. Fewer catered meetings. Forty-four point one times more patients enrolled.

222
00:24:18.020 --> 00:24:26.640
Fewer conference rooms rented. The Preventable Deaths. Conservative estimate: the N.I.H.'s misallocation costs approximately

223
00:24:26.640 --> 00:24:32.080
ten million Q.A.L.Y.s and seven hundred thousand preventable deaths every year.

224
00:24:32.700 --> 00:24:40.780
The upper bound is far worse (one hundred million Q.A.L.Y.s, seven million deaths), but even the floor is staggering.

225
00:24:41.400 --> 00:24:49.080
On Wishonia, if an institution's budget allocation was killing seven hundred thousand people a year, we would adjust the

226
00:24:49.080 --> 00:24:53.540
budget allocation. You adjusted the font on the annual report.

227
00:24:54.640 --> 00:24:59.640
Traditional clinical trials run at two point two seven percent the efficiency of pragmatic trials.

228
00:25:00.680 --> 00:25:04.820
And that's the ceiling, because it only accounts for the sliver of the N.I.H.

229
00:25:04.880 --> 00:25:12.580
budget that goes to trials at all. If your car ran at two point two seven percent efficiency, you would push it faster than

230
00:25:12.580 --> 00:25:19.060
it drives. If your heart pumped at two point two seven percent efficiency, you'd be dead before finishing this sentence.

231
00:25:19.700 --> 00:25:25.340
Your clinical trial system runs at two point two seven percent efficiency, and you gave it a raise.

232
00:25:25.920 --> 00:25:30.010
You gave it a raise and a new building. The building has a cafeteria.

233
00:25:31.000 --> 00:25:34.740
COVID is the most vivid example of what misallocation costs.

234
00:25:35.500 --> 00:25:41.600
So let me tell you what happened, and I will try to keep my composure, though I should warn you that I have been told I

235
00:25:41.600 --> 00:25:46.140
do not have composure. Your government knew this research was dangerous.

236
00:25:46.700 --> 00:25:53.940
In twenty fourteen, the Obama administration imposed a moratorium on gain-of-function research, specifically because engineering

237
00:25:53.940 --> 00:26:00.480
novel viruses posed pandemic risk. The government's own assessment: too dangerous to continue.

238
00:26:01.060 --> 00:26:06.500
In December twenty seventeen, the moratorium was lifted. There is evidence that N.I.H.

239
00:26:06.500 --> 00:26:10.820
officials may have effectively circumvented the pause the entire time regardless.

240
00:26:11.560 --> 00:26:19.820
At no point, before the moratorium, during it, or after lifting it, did anyone publish a cost-benefit analysis showing the

241
00:26:20.220 --> 00:26:26.020
expected benefits of engineering novel coronaviruses exceeded the risk of starting a pandemic.

242
00:26:26.680 --> 00:26:30.720
The question was never formally asked. It was not even formally avoided.

243
00:26:31.360 --> 00:26:38.180
It simply did not occur to anyone that the question should exist, which is the kind of oversight that is forgivable when

244
00:26:38.180 --> 00:26:44.140
you're deciding where to put a mailbox and less forgivable when you're deciding whether to make new viruses.

245
00:26:45.300 --> 00:26:54.800
Through a grant to EcoHealth Alliance, the N.I.H. sent six hundred thousand dollars to the Wuhan Institute of Virology for

246
00:26:54.800 --> 00:27:04.580
research engineering novel bat coronaviruses. In twenty eighteen and twenty nineteen, one of those engineered viruses turned

247
00:27:04.580 --> 00:27:14.460
out to be unexpectedly more infectious in mice. EcoHealth was required to report this immediately; they did not report it

248
00:27:14.460 --> 00:27:23.540
for nearly two years. The N.I.H. was required to oversee the grant; a federal watchdog found they did not.

249
00:27:24.240 --> 00:27:33.180
The N.I.H.'s own deputy director later admitted in congressional testimony that this was gain-of-function research.

250
00:27:33.680 --> 00:27:41.220
I am losing my composure. The F.B.I., the Department of Energy, and the C.I.A.

251
00:27:41.440 --> 00:27:50.660
(as of January twenty twenty-five) have all assessed that a lab leak is the most likely origin of COVID nineteen.

252
00:27:51.160 --> 00:28:01.140
A two-year congressional investigation concluded the same. This is not proven beyond doubt; four intelligence agencies still

253
00:28:01.140 --> 00:28:06.980
lean toward natural origin. But the question is not "who is to blame?"

254
00:28:07.440 --> 00:28:16.820
It is: what does it cost when an institution allocates six hundred thousand dollars to a side quest creating bat viruses

255
00:28:16.820 --> 00:28:26.860
instead of testing treatments? If these assessments are correct: fourteen to sixteen trillion dollars in economic damage

256
00:28:27.260 --> 00:28:37.760
to the United States alone. Seven million confirmed deaths, nineteen to thirty-six million by excess mortality estimates.

257
00:28:37.760 --> 00:28:44.880
An ongoing one trillion dollars per year in global long COVID costs.

258
00:28:45.000 --> 00:28:55.080
The N.I.H.'s entire budget over its fifty-year history, inflation-adjusted, is approximately one point one trillion dollars.

259
00:28:55.620 --> 00:29:05.380
A single unsupervised grant may have cost humanity roughly thirteen times more than everything the N.I.H.

260
00:29:05.440 --> 00:29:10.540
has ever spent, on everything, combined, since nineteen seventy.

261
00:29:11.260 --> 00:29:16.920
On Wishonia, we have a unit of measurement for this level of institutional failure.

262
00:29:17.580 --> 00:29:23.320
we have never had occasion to use it. Nobody intended this outcome.

263
00:29:23.380 --> 00:29:31.520
The researchers were studying bat coronaviruses to prevent pandemics, which is a goal I would applaud if it had not produced

264
00:29:31.520 --> 00:29:37.460
the opposite result. The grant administrators were following procedures.

265
00:29:37.480 --> 00:29:47.180
The system did not fail because of villains. It failed because nobody's job was to ask: "Is engineering novel coronaviruses

266
00:29:47.180 --> 00:29:56.540
in a city with inadequate biosafety protocols a better use of six hundred thousand dollars than testing whether any of the

267
00:29:56.540 --> 00:30:04.660
thousands of safe, untested compounds on the shelf might cure a disease that is killing people right now?".

268
00:30:05.300 --> 00:30:10.580
That question was never asked because the system is not designed to ask it.

269
00:30:11.240 --> 00:30:19.720
The system is designed to fund research. Whether that research helps anyone is a separate question, handled by a separate

270
00:30:19.720 --> 00:30:26.040
department, which does not exist. On Wishonia, we had a department like this once.

271
00:30:26.420 --> 00:30:30.720
It was called "the point." Every other department reported to it.

272
00:30:31.400 --> 00:30:37.960
Your species appears to have organized the entire building and forgotten to include the point.

273
00:30:38.580 --> 00:30:44.120
The building is very well organized. Then the pandemic happened.

274
00:30:44.280 --> 00:30:54.800
Then seven million people died. Then the N.I.H. received one point seven billion dollars to address the long-term damage.

275
00:30:55.200 --> 00:31:02.600
Then it spent eighty-five percent of that money watching people be sick instead of testing treatments.

276
00:31:03.020 --> 00:31:08.340
I have now fully lost my composure. In that timeline, the N.I.H.

277
00:31:08.340 --> 00:31:12.540
is too busy testing cures to fund gain-of-function research in Wuhan.

278
00:31:13.460 --> 00:31:19.180
There is no COVID nineteen pandemic. There is no sixteen trillion dollars in economic damage.

279
00:31:19.660 --> 00:31:25.500
There are no seven million dead. There is no long COVID costing one trillion dollars a year.

280
00:31:26.620 --> 00:31:35.060
The Track Record. You have already seen the numbers: one point one trillion dollars over fifty years, zero diseases eradicated

281
00:31:35.060 --> 00:31:40.380
(one possibly created). I asked the N.I.H. how many diseases they've eradicated.

282
00:31:40.780 --> 00:31:44.400
They sent me a forty-page document about their strategic vision.

283
00:31:44.940 --> 00:31:50.200
I read all forty pages. The number wasn't in it. The word "eradicate" wasn't in it either.

284
00:31:50.880 --> 00:31:55.120
The word "strategic" was in it thirty-one times. The W.H.O.

285
00:31:55.400 --> 00:32:03.280
eradicated smallpox for three hundred million dollars. Jonas Salk developed the polio vaccine in a university lab and gave

286
00:32:03.280 --> 00:32:09.820
it away for free, which is why nobody at the N.I.H. talks about him at parties because he makes everyone look bad by comparison

287
00:32:09.820 --> 00:32:13.840
and also he's dead, which is the thing they were supposed to be fixing.

288
00:32:14.420 --> 00:32:21.300
He gave away the patent. I asked an N.I.H. official what would happen if they gave away patents today and they made a sound

289
00:32:21.300 --> 00:32:30.240
I can only describe as a laugh, but sadder. Veterinarians have eradicated multiple animal diseases on budgets the N.I.H.

290
00:32:30.240 --> 00:32:35.560
would consider a rounding error. Your veterinarians eradicated hog cholera.

291
00:32:36.340 --> 00:32:44.080
Pigs. I should note the irony of humanity's failure to eradicate human diseases while more aggressively finding cures for

292
00:32:44.080 --> 00:32:51.240
animals you lock in metal crates too small to turn around in, standing in their own waste, whose tails you amputate without

293
00:32:51.240 --> 00:32:55.920
anesthesia because the confinement drives them insane enough to eat each other.

294
00:32:56.680 --> 00:33:03.540
But fixed that cholera for ya! What Would Actually Work. The RECOVERY trial already proved it.

295
00:33:04.420 --> 00:33:08.200
Five hundred dollars per patient instead of forty-one thousand dollars.

296
00:33:08.780 --> 00:33:14.920
First life-saving result in under one hundred days. Over one million lives saved.

297
00:33:15.520 --> 00:33:20.920
The blueprint exists. It is not theoretical. It happened. In England.

298
00:33:21.580 --> 00:33:26.020
During a pandemic. While Americans were arguing about horse dewormer.

299
00:33:27.020 --> 00:33:31.730
The British did not use a special technique. They did not have access to secret technology.

300
00:33:32.440 --> 00:33:36.840
They used the same drugs, the same disease, the same species of patient.

301
00:33:37.240 --> 00:33:42.780
They just tested the drugs on the patients instead of writing about the drugs near the patients.

302
00:33:43.600 --> 00:33:49.180
This is the breakthrough. It took zero breakthroughs. The evidence is right there.

303
00:33:49.620 --> 00:33:54.400
It's not hiding. It's waving at you. It has been waving at you for years.

304
00:33:55.120 --> 00:34:00.490
Now imagine the other timeline. Not as an indictment, but as a measure of what was lost.

305
00:34:01.200 --> 00:34:08.739
If the N.I.H. had allocated even twenty percent of its one point one trillion dollars to pragmatic trials over fifty years

306
00:34:09.239 --> 00:34:17.520
(the remaining eighty percent still on basic research), that produces four hundred and forty million patient-trials at RECOVERY

307
00:34:17.520 --> 00:34:26.580
efficiency (five hundred dollars per patient). That is enough to test every safe compound against every major disease, multiple

308
00:34:26.580 --> 00:34:34.620
times, with replication. The Oxford RECOVERY trial tested forty-eight thousand patients and found a treatment that saved

309
00:34:34.620 --> 00:34:39.560
over a million lives, in one hundred days, for twenty-four million dollars.

310
00:34:40.159 --> 00:34:47.600
Multiply that across four hundred and forty million patients and fifty years and the number of lives saved becomes difficult

311
00:34:47.600 --> 00:34:52.179
to calculate because the diseases start disappearing and the math changes.

312
00:34:52.920 --> 00:35:01.500
In that timeline, the N.I.H.'s one point one trillion dollars produces what a trillion dollars should produce: the systematic

313
00:35:01.500 --> 00:35:09.440
eradication of disease after disease after disease, because you finally pointed the money at the thing that actually works.

314
00:35:10.620 --> 00:35:14.500
The difference between that timeline and this one is not a matter of hindsight.

315
00:35:15.560 --> 00:35:21.020
The Oxford RECOVERY trial happened. Pragmatic trials existed the entire time.

316
00:35:21.460 --> 00:35:26.900
The efficiency data was available for decades. The counterfactual is not imaginary.

317
00:35:27.460 --> 00:35:33.840
It is the road you could see, that was paved, that had signs on it, that other countries were already driving on.

318
00:35:34.900 --> 00:35:42.620
You chose the other road. It led here. I am told that on your planet, describing this situation accurately is considered

319
00:35:42.620 --> 00:35:47.460
"rude." On Wishonia, not describing it would be considered "complicity."

320
00:35:48.300 --> 00:35:55.040
I will risk the rudeness. The recipe is simple enough that I can state it in four sentences: One.

321
00:35:55.500 --> 00:36:00.480
Pay for results. Treatment equals payment, no treatment equals no payment.

322
00:36:01.340 --> 00:36:09.260
Two. Publish everything, including failures. Especially failures, because the people who will die if the system fails are

323
00:36:09.260 --> 00:36:13.020
more motivated than the people who get tenure if the system continues.

324
00:36:13.940 --> 00:36:18.200
Three. Eliminate gatekeepers and let patients vote with participation.

325
00:36:18.740 --> 00:36:26.380
The volunteers are still holding their hands up. Four. Automate administration with transparent systems instead of committees.

326
00:36:27.020 --> 00:36:33.540
Committees cannot be trusted (see above, and also see the entire history of committees, and also see the committee that

327
00:36:33.540 --> 00:36:38.920
was formed to investigate whether committees can be trusted, which concluded that committees are essential).

328
00:36:40.080 --> 00:36:47.460
The cost of not doing this is measurable. Over fifty percent of trial results are never published, which means researchers

329
00:36:47.460 --> 00:36:56.200
unknowingly repeat experiments that already failed. This is how your species wastes an estimated eighty-five percent of

330
00:36:56.200 --> 00:37:04.020
its two hundred billion dollars annual research investment, and then passes the cost to patients as higher drug prices.

331
00:37:04.640 --> 00:37:11.700
The blueprint for this system is your decentralized F.D.A. The funding mechanism is the one percent treaty.

332
00:37:11.940 --> 00:37:13.480
The next chapters explain both.