WEBVTT

1
00:00:01.740 --> 00:00:03.480
<v ->Well, good afternoon everyone.</v>

2
00:00:03.480 --> 00:00:05.850
I want to talk today a little bit about my interest,

3
00:00:05.850 --> 00:00:10.650
and it looks like I need to pick up a different clicker,

4
00:00:10.650 --> 00:00:12.210
which is tuberculosis.

5
00:00:12.210 --> 00:00:13.800
So tuberculosis is the leading

6
00:00:13.800 --> 00:00:16.020
infectious cause of death worldwide.

7
00:00:16.020 --> 00:00:17.100
And we believe that last year

8
00:00:17.100 --> 00:00:20.010
it affected around 10.6 million people.

9
00:00:20.010 --> 00:00:23.220
The World WHO TB report comes out every year

10
00:00:23.220 --> 00:00:24.240
around this time,

11
00:00:24.240 --> 00:00:27.300
so the latest data that that we have is from 2021.

12
00:00:27.300 --> 00:00:28.320
And following the pandemic,

13
00:00:28.320 --> 00:00:30.420
there was a big decline in the number of people

14
00:00:30.420 --> 00:00:34.530
having access to quality assured diagnosis

15
00:00:34.530 --> 00:00:36.630
and initiation of treatment.

16
00:00:36.630 --> 00:00:37.740
But it's long been the case

17
00:00:37.740 --> 00:00:40.770
that a large proportion of people living with TB

18
00:00:40.770 --> 00:00:41.640
go undiagnosed.

19
00:00:41.640 --> 00:00:43.080
We think more than 4 million people.

20
00:00:43.080 --> 00:00:44.550
So one of the things that we're interested in

21
00:00:44.550 --> 00:00:46.680
is how can we identify those people?

22
00:00:46.680 --> 00:00:48.990
There's many reasons why they aren't diagnosed.

23
00:00:48.990 --> 00:00:52.470
One challenge of course, is that TB is often asymptomatic.

24
00:00:52.470 --> 00:00:54.900
So we need to identify better ways of screening

25
00:00:54.900 --> 00:00:56.250
and identifying people in the community

26
00:00:56.250 --> 00:00:57.720
who don't know they have symptoms.

27
00:00:57.720 --> 00:00:58.950
And then we need to improve access

28
00:00:58.950 --> 00:01:00.300
for people who do have his symptoms

29
00:01:00.300 --> 00:01:02.760
and encourage them to seek care

30
00:01:02.760 --> 00:01:05.640
in clinics where they can get diagnoses.

31
00:01:05.640 --> 00:01:08.040
The tool that we use in our laboratory for doing this work

32
00:01:08.040 --> 00:01:09.510
is called implementation science.

33
00:01:09.510 --> 00:01:11.130
I know there are probably a number of people in this room

34
00:01:11.130 --> 00:01:13.710
that have heard of this field,

35
00:01:13.710 --> 00:01:15.570
and there's a course that I teach in the fall,

36
00:01:15.570 --> 00:01:17.070
for those who are thinking about

37
00:01:17.070 --> 00:01:19.260
their courses for next fall,

38
00:01:19.260 --> 00:01:22.110
which goes over what implementation science is.

39
00:01:22.110 --> 00:01:25.530
But in short it is a systematic approach

40
00:01:25.530 --> 00:01:28.230
for the design, delivery and evaluation of strategies

41
00:01:28.230 --> 00:01:30.840
to improve the uptake of proven interventions

42
00:01:30.840 --> 00:01:31.920
into routine practice.

43
00:01:31.920 --> 00:01:34.620
And we think of it as a translational research discipline.

44
00:01:34.620 --> 00:01:36.030
Translational research

45
00:01:36.030 --> 00:01:38.130
involving multidisciplinary collaborations

46
00:01:38.130 --> 00:01:41.550
that move ideas from their foundational basis,

47
00:01:41.550 --> 00:01:42.840
whether that's in the laboratory,

48
00:01:42.840 --> 00:01:43.860
in fundamental biology

49
00:01:43.860 --> 00:01:47.100
as we were hearing about a little bit earlier from Brian.

50
00:01:47.100 --> 00:01:49.020
Or more on the other end of the spectrum

51
00:01:49.020 --> 00:01:50.100
here where our group works,

52
00:01:50.100 --> 00:01:51.270
which is thinking about how do we move

53
00:01:51.270 --> 00:01:52.920
from patients to practice

54
00:01:52.920 --> 00:01:54.600
and ultimately to public health.

55
00:01:54.600 --> 00:01:57.720
And it turns out that that discipline doesn't happen

56
00:01:57.720 --> 00:02:00.390
just when good ideas get published in journals

57
00:02:00.390 --> 00:02:01.410
showing that they're effective.

58
00:02:01.410 --> 00:02:03.300
It's really necessary to think about

59
00:02:03.300 --> 00:02:06.660
how would those ideas be adapted to a real world setting.

60
00:02:06.660 --> 00:02:08.880
So in the contexts of TB, I wanna talk about

61
00:02:08.880 --> 00:02:11.400
three different types of studies that we've done.

62
00:02:11.400 --> 00:02:13.080
Formative work to help us understand

63
00:02:13.080 --> 00:02:14.400
what the nature of the barriers are

64
00:02:14.400 --> 00:02:17.670
to delivering these evidence-based services for TB.

65
00:02:17.670 --> 00:02:19.440
Actual implementation studies

66
00:02:19.440 --> 00:02:21.750
where we introduce these ideas.

67
00:02:21.750 --> 00:02:22.920
And then evaluation studies

68
00:02:22.920 --> 00:02:24.780
where we try to figure out what works

69
00:02:24.780 --> 00:02:27.120
and why or why it doesn't work.

70
00:02:27.120 --> 00:02:28.740
And a lot of the work that I've done

71
00:02:28.740 --> 00:02:30.600
is in the area of contact investigation.

72
00:02:30.600 --> 00:02:31.770
So contact investigation,

73
00:02:31.770 --> 00:02:34.080
I think as everybody in the room is probably familiar,

74
00:02:34.080 --> 00:02:36.870
is a strategy where we look for high risk populations

75
00:02:36.870 --> 00:02:39.090
of having a disease like tuberculosis,

76
00:02:39.090 --> 00:02:40.800
among those that are close contacts

77
00:02:40.800 --> 00:02:42.420
to people with tuberculosis,

78
00:02:42.420 --> 00:02:44.490
knowing that to tuberculosis is transmitted

79
00:02:44.490 --> 00:02:48.720
from person to person through respiratory aerosols.

80
00:02:48.720 --> 00:02:50.280
And that work that we've done in the area

81
00:02:50.280 --> 00:02:51.690
of contact investigation

82
00:02:51.690 --> 00:02:53.280
has spanned different parts of the globe,

83
00:02:53.280 --> 00:02:55.260
because we know in different contexts

84
00:02:55.260 --> 00:02:56.640
there may be different reasons.

85
00:02:56.640 --> 00:02:57.840
We're interested in understanding

86
00:02:57.840 --> 00:02:59.280
both the generalizable reasons,

87
00:02:59.280 --> 00:03:01.872
why it's hard to do contact investigation,

88
00:03:01.872 --> 00:03:04.530
and also some of the setting specific regions.

89
00:03:04.530 --> 00:03:06.270
So for example here on the far left,

90
00:03:06.270 --> 00:03:08.970
in Columbia one of our scholars, Gustavo Diaz,

91
00:03:08.970 --> 00:03:11.730
who was a postdoctoral fellow at the GHS program,

92
00:03:11.730 --> 00:03:15.120
used routine public health data to map the barriers

93
00:03:15.120 --> 00:03:17.040
and characterize those.

94
00:03:17.040 --> 00:03:19.890
Here below, this is a qualitative research study

95
00:03:19.890 --> 00:03:21.900
that was looking at another aspect of TB care

96
00:03:21.900 --> 00:03:23.310
which is education and counseling,

97
00:03:23.310 --> 00:03:24.630
which is very important

98
00:03:24.630 --> 00:03:26.850
to get people newly diagnosed with TB

99
00:03:26.850 --> 00:03:29.550
to be willing to share the contact information

100
00:03:29.550 --> 00:03:30.480
of their close contacts,

101
00:03:30.480 --> 00:03:32.430
and also for them to be able to achieve

102
00:03:32.430 --> 00:03:34.020
good treatment outcomes on their own

103
00:03:34.020 --> 00:03:36.480
and get initiated in other therapies

104
00:03:36.480 --> 00:03:38.700
that may be important like HIV treatment.

105
00:03:38.700 --> 00:03:40.020
And so this is work that was carried out

106
00:03:40.020 --> 00:03:41.250
with Alisse Hannaford,

107
00:03:41.250 --> 00:03:43.050
who was a medical student at the time.

108
00:03:43.050 --> 00:03:45.870
Then the third example was looking at these similar methods,

109
00:03:45.870 --> 00:03:48.390
not just for TB case finding

110
00:03:48.390 --> 00:03:49.770
in the community that it's searching,

111
00:03:49.770 --> 00:03:53.130
but also for treatment of TB and for prevention

112
00:03:53.130 --> 00:03:56.040
as work that was done by Christina Parisi,

113
00:03:56.040 --> 00:03:57.600
who was an EMD student a few years back

114
00:03:57.600 --> 00:04:00.363
and is now a PhD student at the University of Florida.

115
00:04:01.380 --> 00:04:03.450
We're also interested in taking this information

116
00:04:03.450 --> 00:04:06.090
about barriers to delivery of care and try to combine it.

117
00:04:06.090 --> 00:04:07.500
So taking qualitative information,

118
00:04:07.500 --> 00:04:09.240
taking quantitative information,

119
00:04:09.240 --> 00:04:11.970
and developing strategies to try to change that.

120
00:04:11.970 --> 00:04:14.640
One of the strategies that we're very excited about

121
00:04:14.640 --> 00:04:18.030
is the use of collaborative design approaches,

122
00:04:18.030 --> 00:04:20.580
specifically human-centered design,

123
00:04:20.580 --> 00:04:22.080
where we try to come up with solutions

124
00:04:22.080 --> 00:04:24.030
that work for the communities that we're working with.

125
00:04:24.030 --> 00:04:26.940
And so here's an example in the middle of (indistinct)

126
00:04:26.940 --> 00:04:30.120
who's a laboratory manager in one of our projects in Uganda.

127
00:04:30.120 --> 00:04:32.430
We're using the human-centered design approach

128
00:04:32.430 --> 00:04:35.700
with the post-it notes to try to come up with good ideas.

129
00:04:35.700 --> 00:04:37.350
And than we've developed those

130
00:04:37.350 --> 00:04:39.090
into a couple of different types of strategies.

131
00:04:39.090 --> 00:04:41.940
So these are described in publications by Amanda Gupta,

132
00:04:41.940 --> 00:04:45.120
who's a research coordinator in our laboratory

133
00:04:45.120 --> 00:04:46.496
and a PhD student.

134
00:04:46.496 --> 00:04:48.930
And also by Joseph Tita,

135
00:04:48.930 --> 00:04:51.840
who's one of our qualitative scientists in Uganda.

136
00:04:51.840 --> 00:04:53.100
We're also a lot of times interested,

137
00:04:53.100 --> 00:04:54.240
even at the design phase,

138
00:04:54.240 --> 00:04:56.130
in thinking about what the costs are.

139
00:04:56.130 --> 00:04:58.110
A lot of times we wait until the end of a study

140
00:04:58.110 --> 00:04:59.250
to evaluate that.

141
00:04:59.250 --> 00:05:00.480
But in this particular project,

142
00:05:00.480 --> 00:05:02.400
we knew we learned from our prior work

143
00:05:02.400 --> 00:05:06.030
that it's very expensive to kind of develop strategy.

144
00:05:06.030 --> 00:05:07.560
So we looked at the cost strategy,

145
00:05:07.560 --> 00:05:11.340
and this was work led by Patricia Turimumahoro,

146
00:05:11.340 --> 00:05:14.400
who is currently a GHES scholar.

147
00:05:14.400 --> 00:05:16.380
So I'll just close by kind of describing

148
00:05:16.380 --> 00:05:18.060
the most exciting aspect of this work

149
00:05:18.060 --> 00:05:20.250
was actually evaluating where these strategies work.

150
00:05:20.250 --> 00:05:23.160
And we just recently completed a pragmatic trial

151
00:05:23.160 --> 00:05:25.230
using a stepped-wedge cluster-randomized design

152
00:05:25.230 --> 00:05:27.540
where we took a human-centered design approach

153
00:05:27.540 --> 00:05:29.400
that include multiple components.

154
00:05:29.400 --> 00:05:30.990
And these components include things

155
00:05:30.990 --> 00:05:34.590
that ease the multi-step process of contact investigation

156
00:05:34.590 --> 00:05:38.100
to kind of help participants, namely household contacts,

157
00:05:38.100 --> 00:05:40.140
better be able to engage in these services.

158
00:05:40.140 --> 00:05:43.740
Things like a motorcycle rider, to take them to the clinic.

159
00:05:43.740 --> 00:05:47.460
Educational strategies and also instructional information

160
00:05:47.460 --> 00:05:49.350
to help them produce sputum.

161
00:05:49.350 --> 00:05:51.780
And combining that with strategies of quality improvement

162
00:05:51.780 --> 00:05:54.300
to help community health workers better deliver that.

163
00:05:54.300 --> 00:05:57.030
And we are are just now looking at that preliminary data.

164
00:05:57.030 --> 00:05:58.620
It looks very promising

165
00:05:58.620 --> 00:06:00.720
that these strategies can be in fact impactful.

166
00:06:00.720 --> 00:06:02.340
And we're now testing those

167
00:06:02.340 --> 00:06:04.860
to other aspects like TB prevention.

168
00:06:04.860 --> 00:06:07.770
So with that, I'll close and see if there's any questions,

169
00:06:07.770 --> 00:06:09.030
or I can take them at the end.

170
00:06:09.030 --> 00:06:10.140
Thank you.

171
00:06:10.140 --> 00:06:11.259
(audience applauding)