1 00:00:00,235 --> 00:00:02,205 - Prepare to dive into the forefront 2 00:00:02,225 --> 00:00:05,525 of healthcare innovation at our 14th annual meeting coming 3 00:00:05,545 --> 00:00:06,645 up on April 8th 4 00:00:06,645 --> 00:00:10,765 through 11th at the Hyatt Regency in Chicago, with thousands 5 00:00:10,765 --> 00:00:13,965 of industry leaders converging over four dynamic days 6 00:00:13,985 --> 00:00:16,805 of discussions on crucial topics from health IT 7 00:00:17,185 --> 00:00:18,685 to executive leadership. 8 00:00:18,995 --> 00:00:21,325 It's where the future of healthcare takes shape. 9 00:00:21,865 --> 00:00:24,085 We can't wait to connect with you in person 10 00:00:24,305 --> 00:00:26,685 and engage in these important conversations. 11 00:00:27,195 --> 00:00:29,605 - Welcome back to the Becker's Healthcare Podcast, devote 12 00:00:29,605 --> 00:00:31,165 to the people who power us healthcare. 13 00:00:31,465 --> 00:00:32,845 I'm Molly Gamble with Becker's, 14 00:00:32,845 --> 00:00:34,885 and delighted to have you join us for another episode filled 15 00:00:34,885 --> 00:00:36,485 with insights from healthcare leaders. 16 00:00:37,135 --> 00:00:39,645 Today we're spending some time with Dr. Thomas Maddox. 17 00:00:40,185 --> 00:00:42,525 Dr. Maddox is the Vice President of Digital Products 18 00:00:42,525 --> 00:00:44,485 and Innovation with BJC Healthcare 19 00:00:44,665 --> 00:00:46,765 and Washington University School of Medicine, 20 00:00:47,145 --> 00:00:48,525 and also Professor of Medicine 21 00:00:48,525 --> 00:00:50,565 with Washington University School of Medicine. 22 00:00:51,145 --> 00:00:52,725 Dr. Maddox, welcome to the podcast. 23 00:00:52,785 --> 00:00:55,565 Thanks so much for being my guest today. How are you? 24 00:00:55,625 --> 00:00:57,285 And where does the podcast find you? 25 00:00:58,655 --> 00:01:00,665 - Well, thank you so much for the invitation. 26 00:01:01,325 --> 00:01:02,985 I'm doing well. It's a Monday, 27 00:01:04,035 --> 00:01:05,705 which is always a little treacherous, 28 00:01:05,805 --> 00:01:07,025 but, uh, so far so good. 29 00:01:07,645 --> 00:01:10,105 And I'm in St. Louis, Missouri, which is where, uh, 30 00:01:10,865 --> 00:01:12,865 BJC Healthcare and the Washington University 31 00:01:12,865 --> 00:01:13,905 School of Medicine are based. 32 00:01:14,795 --> 00:01:16,305 - Great. Yeah. 33 00:01:16,425 --> 00:01:18,745 I should mention I'm extra grateful to start the week off 34 00:01:18,895 --> 00:01:20,825 with some afternoon catching up with you. 35 00:01:21,325 --> 00:01:22,785 Um, Dr. Maddox, you know, 36 00:01:22,885 --> 00:01:25,105 as I highlighted in my introduction, 37 00:01:25,125 --> 00:01:26,625 you wear a couple different hats 38 00:01:26,725 --> 00:01:27,865 in your work and your roles. 39 00:01:28,645 --> 00:01:30,665 Can you start us off by sharing a bit more 40 00:01:30,855 --> 00:01:33,305 with our listeners about your responsibilities 41 00:01:33,445 --> 00:01:37,425 and the scope of your work, both with the BJC health system 42 00:01:37,565 --> 00:01:39,425 and then Wash U School of Medicine? 43 00:01:40,455 --> 00:01:42,385 - Sure, sure. So you're right. 44 00:01:42,505 --> 00:01:45,945 I wear several hats, um, all of which I enjoy. 45 00:01:46,515 --> 00:01:48,265 First and foremost, I'm a cardiologist. 46 00:01:48,565 --> 00:01:51,865 Um, I've, I've been practicing for a couple decades now 47 00:01:51,965 --> 00:01:53,185 and, and just love it. 48 00:01:53,485 --> 00:01:56,385 You know, medicine is a, is a wonderful profession, 49 00:01:56,965 --> 00:01:58,305 and it's really cool to be able 50 00:01:58,305 --> 00:02:00,385 to pull on the white coat periodically 51 00:02:00,405 --> 00:02:02,985 and, um, do what I've trained to do for years. 52 00:02:03,125 --> 00:02:06,505 And that is, is try to take really good care of patients, 53 00:02:06,505 --> 00:02:08,065 particularly those with cardiac issues. 54 00:02:08,925 --> 00:02:12,915 So that's part of my job. And then, um, I arrived in St. 55 00:02:12,915 --> 00:02:14,275 Louis about seven years ago 56 00:02:14,895 --> 00:02:18,595 to help found their healthcare innovation laboratory. 57 00:02:19,415 --> 00:02:22,595 And so at that time, 2017, uh, 58 00:02:22,595 --> 00:02:24,755 digital health was really starting to take off. 59 00:02:25,525 --> 00:02:26,825 Uh, and the presence 60 00:02:26,825 --> 00:02:30,105 of digital data in our healthcare system through EHRs 61 00:02:30,405 --> 00:02:33,745 and other methodologies really was starting to take hold. 62 00:02:34,605 --> 00:02:38,105 So the school felt like, um, it was a good time 63 00:02:38,165 --> 00:02:41,105 to have a group focused exclusively on 64 00:02:41,845 --> 00:02:43,665 how might these digital health tools 65 00:02:43,805 --> 00:02:45,105 and this digital information 66 00:02:45,925 --> 00:02:50,865 and the, uh, the communications technologies, uh, be 67 00:02:51,505 --> 00:02:53,905 engineered in a way that could be maximally useful 68 00:02:54,325 --> 00:02:56,265 for our patients, for their families, 69 00:02:56,645 --> 00:02:58,305 and for the care teams to take care of them. 70 00:02:59,275 --> 00:03:01,215 So they brought me on to kind of help think 71 00:03:01,215 --> 00:03:02,415 through how that might work. 72 00:03:03,035 --> 00:03:05,455 And for the last seven years, we've really been trying 73 00:03:05,475 --> 00:03:09,735 to tackle some of these, uh, new approaches to healthcare, 74 00:03:10,475 --> 00:03:12,895 be it, um, predictive analytics, uh, 75 00:03:12,895 --> 00:03:14,975 remote patient monitoring, digital tools 76 00:03:15,205 --> 00:03:18,255 that can be used at the point of care by our care teams, 77 00:03:18,395 --> 00:03:20,735 all the myriad ways that, um, 78 00:03:20,735 --> 00:03:22,695 digital health can provide some innovative 79 00:03:22,935 --> 00:03:24,055 approaches to better healthcare. 80 00:03:24,595 --> 00:03:26,815 Mm-hmm, . So that's, uh, a role that I've had 81 00:03:26,815 --> 00:03:29,175 for seven years, continue to have, enjoy that immensely. 82 00:03:29,205 --> 00:03:31,535 It's, uh, it's, it's fun to be charged 83 00:03:31,535 --> 00:03:34,895 with thinking down the road of what could be coming 84 00:03:35,475 --> 00:03:37,495 and to have that creative component to my job. 85 00:03:37,835 --> 00:03:40,775 Mm-Hmm. , about three years ago, um, 86 00:03:40,935 --> 00:03:45,415 I was asked by our CIO to then, uh, take on, uh, 87 00:03:45,615 --> 00:03:47,215 a digital products effort. 88 00:03:47,875 --> 00:03:49,655 So I think most of your listeners 89 00:03:49,655 --> 00:03:52,135 who are involved in it will recognize that, you know, 90 00:03:52,135 --> 00:03:53,655 digital products and a, 91 00:03:53,715 --> 00:03:58,375 and a product, uh, philosophy is really sort of, um, 92 00:03:58,715 --> 00:04:00,695 uh, the standard now in technology. 93 00:04:00,835 --> 00:04:02,975 And really it was pioneered by a lot 94 00:04:02,975 --> 00:04:05,695 of the big tech companies, the, the Googles and Metas 95 00:04:05,695 --> 00:04:08,495 and others, uh, in the, in the big tech sector. 96 00:04:08,915 --> 00:04:11,455 And it's basically the idea of 97 00:04:11,875 --> 00:04:15,335 how can you really get deep into your user's problems 98 00:04:15,995 --> 00:04:19,535 and then build a product, build a solution, um, 99 00:04:19,755 --> 00:04:23,015 in a rapid way that is continually checking in 100 00:04:23,015 --> 00:04:24,695 with those users to make sure 101 00:04:24,695 --> 00:04:26,135 that it truly is solving the problem 102 00:04:26,485 --> 00:04:28,015 that it's ostensibly designed to solve. 103 00:04:28,075 --> 00:04:30,535 And you use various techniques to develop 104 00:04:30,565 --> 00:04:33,055 that human-centered design, uh, 105 00:04:33,265 --> 00:04:35,495 agile delivery methods, rapid sprints. 106 00:04:36,035 --> 00:04:38,215 And it really is sort of bringing, in my opinion, 107 00:04:38,715 --> 00:04:42,295 the scientific method of inquiry into the development 108 00:04:42,715 --> 00:04:44,255 of technology products 109 00:04:44,255 --> 00:04:46,655 and solutions into, in how we, 110 00:04:46,715 --> 00:04:48,175 in our case and how we deliver care. 111 00:04:48,995 --> 00:04:53,375 So, uh, those teams have product managers, uh, 112 00:04:53,775 --> 00:04:56,255 scrum masters, product designers, 113 00:04:56,955 --> 00:04:58,535 and our first sort of set 114 00:04:58,535 --> 00:05:00,615 of products have focused on our digital front door. 115 00:05:01,075 --> 00:05:04,575 So how do we have digital tools like find a doctor 116 00:05:04,715 --> 00:05:05,975 and online scheduling 117 00:05:05,975 --> 00:05:08,975 and other things to make it easier for patients 118 00:05:08,995 --> 00:05:10,135 to get in to see us. 119 00:05:10,835 --> 00:05:14,135 And then we've also been working a lot on our nursing, um, 120 00:05:14,265 --> 00:05:17,575 front and building some digital tools to help our nurses 121 00:05:17,795 --> 00:05:21,775 who are, who are really at capacity, if you will, um, to, 122 00:05:21,915 --> 00:05:23,455 to enable their work 123 00:05:24,035 --> 00:05:25,655 in a little bit more of an effective way. 124 00:05:25,675 --> 00:05:28,015 As we built a virtual nursing product, 125 00:05:28,155 --> 00:05:31,335 we built our scheduling has to make it easier for them 126 00:05:31,395 --> 00:05:34,455 to pick up shifts and sort of, um, uh, be able 127 00:05:34,455 --> 00:05:36,215 to provide maximum care to our patients. 128 00:05:36,635 --> 00:05:38,495 And we've really had some good success with that. 129 00:05:38,495 --> 00:05:40,975 And, and now, um, I think we're on the cusp 130 00:05:40,975 --> 00:05:44,375 of introducing this product methodology more broadly across 131 00:05:44,435 --> 00:05:45,655 our IT portfolio. 132 00:05:47,055 --> 00:05:50,555 And then finally, most recently, um, with the rise of ai, 133 00:05:50,555 --> 00:05:52,435 both analytic and generative ai, 134 00:05:52,935 --> 00:05:55,355 and seeing some of the possibilities it can provide 135 00:05:55,355 --> 00:05:58,525 for healthcare, um, I've been asked to help lead some 136 00:05:58,525 --> 00:06:01,405 of those AI development efforts in conjunction 137 00:06:01,435 --> 00:06:03,925 with washus chief data scientist, Philip Hane. 138 00:06:04,465 --> 00:06:07,045 And together with our teams, we're really trying to think 139 00:06:07,045 --> 00:06:10,885 through how can we bring the best of what AI's developing 140 00:06:11,345 --> 00:06:13,205 to our patients into our care teams. 141 00:06:13,945 --> 00:06:16,285 And right now we're exploring a couple 142 00:06:16,285 --> 00:06:18,525 of different predictive models that use AI 143 00:06:18,525 --> 00:06:21,285 to help us understand how patients are progressing 144 00:06:21,285 --> 00:06:23,605 through their inpatient stay, trying to make that 145 00:06:23,605 --> 00:06:24,965 as most efficient as possible. 146 00:06:25,545 --> 00:06:27,165 And then we're also experimenting with some 147 00:06:27,165 --> 00:06:28,925 of the new generative AI technologies, 148 00:06:29,545 --> 00:06:32,045 and particularly how they might help our clinicians be more 149 00:06:32,285 --> 00:06:34,085 efficient in their documentation, 150 00:06:34,255 --> 00:06:36,485 which is often a real time suck 151 00:06:36,545 --> 00:06:38,565 and a, a real dissatisfier 152 00:06:38,565 --> 00:06:40,085 to our care teams drives a lot of burnout. 153 00:06:40,665 --> 00:06:44,125 So we're trying to see if generative AI can help, um, 154 00:06:44,675 --> 00:06:47,245 with them generating notes in clinic visits 155 00:06:47,345 --> 00:06:49,885 and really knock down some of the time needed to, uh, 156 00:06:49,905 --> 00:06:52,605 to record, uh, those clinical encounters. 157 00:06:53,215 --> 00:06:56,605 We're also using it to help the clinicians respond 158 00:06:57,305 --> 00:07:00,125 to the sometimes hundreds of secure message 159 00:07:00,145 --> 00:07:03,525 and end baskets, um, email responses they receive daily 160 00:07:04,075 --> 00:07:06,485 from, uh, from, uh, patients and care teams. 161 00:07:07,185 --> 00:07:08,445 So it's a lot of different hats. 162 00:07:08,445 --> 00:07:10,605 They're all connected in my mind. Mm-Hmm. 163 00:07:10,645 --> 00:07:13,005 , um, and, and sort of speak to one another. 164 00:07:13,065 --> 00:07:15,045 So it's not like they're totally disparate jobs, 165 00:07:15,705 --> 00:07:17,645 but it does make for full days and weeks. 166 00:07:17,825 --> 00:07:19,125 But I, I really enjoy it. 167 00:07:20,195 --> 00:07:22,115 - I imagine, you know, as, as you're walking us 168 00:07:22,115 --> 00:07:24,955 through this, I mean, not to go too deep on the, 169 00:07:24,975 --> 00:07:28,035 the hat metaphor, but some of these hats that you described, 170 00:07:28,075 --> 00:07:29,115 I mean, these are new hats. 171 00:07:29,245 --> 00:07:32,195 These are roles that you were asked to take on. 172 00:07:32,335 --> 00:07:36,475 It was a first of its kind opportunity or position. And Dr. 173 00:07:36,535 --> 00:07:38,515 Max, can I ask you a question for other leaders 174 00:07:38,695 --> 00:07:41,835 who might one day be invited to head up such opportunities, 175 00:07:41,965 --> 00:07:44,155 which I think at academic medical centers 176 00:07:44,295 --> 00:07:45,755 and large integrative health systems, 177 00:07:45,905 --> 00:07:47,075 it's really interesting. 178 00:07:47,075 --> 00:07:49,715 It's like, directionally, we know where we're headed, 179 00:07:50,265 --> 00:07:53,395 it's time to formalize a lot of the idea generation we have 180 00:07:53,395 --> 00:07:55,195 around this, a lot of the potential. 181 00:07:55,255 --> 00:07:57,435 We want someone to head it up and take it on 182 00:07:57,495 --> 00:07:59,955 and lead formally and properly. 183 00:08:00,015 --> 00:08:02,955 And that in several instances was you, what, 184 00:08:02,955 --> 00:08:05,515 what's really critical as a leader in those times? 185 00:08:06,105 --> 00:08:07,515 What leadership traits 186 00:08:07,515 --> 00:08:10,635 or qualities really emerged as you were working 187 00:08:10,635 --> 00:08:14,155 through these new enterprises with BJC 188 00:08:14,155 --> 00:08:17,275 and Washington University that you relied on most? 189 00:08:19,275 --> 00:08:20,935 - That's a great question. Um, 190 00:08:21,435 --> 00:08:25,525 and I'm not sure there's, you know, a, a 191 00:08:26,105 --> 00:08:28,285 pat answer at least I don't have one. 192 00:08:28,425 --> 00:08:31,905 But, um, things that have really helped me 193 00:08:33,455 --> 00:08:36,755 in starting to understand a new area on behalf 194 00:08:36,755 --> 00:08:39,975 of our organization is, you know, 195 00:08:40,315 --> 00:08:42,415 one, don't go it alone. 196 00:08:43,155 --> 00:08:45,615 Um, I think both internally 197 00:08:45,675 --> 00:08:50,045 and externally, I've relied a lot on folks 198 00:08:50,105 --> 00:08:52,965 who maybe have deeper knowledge in certain areas like 199 00:08:52,975 --> 00:08:56,685 technology than I've had, and just the, the type of training 200 00:08:56,745 --> 00:08:58,285 and experiences I've had to date. 201 00:08:58,825 --> 00:09:00,285 So, for example, I have a lot 202 00:09:00,285 --> 00:09:03,405 of peers in our technology group who've just taught me a ton 203 00:09:03,975 --> 00:09:08,445 about the, the technical mechanics of digital tools 204 00:09:08,545 --> 00:09:10,645 and infrastructure and information security 205 00:09:10,705 --> 00:09:12,525 and all the other things you need to have 206 00:09:12,525 --> 00:09:14,765 to have an effective technology solution. 207 00:09:15,845 --> 00:09:17,895 I've also learned a ton from our care 208 00:09:17,895 --> 00:09:18,975 teams and our patients. 209 00:09:18,975 --> 00:09:20,495 Like what are they experiencing? 210 00:09:21,325 --> 00:09:23,375 What are they running into issues with 211 00:09:24,005 --> 00:09:25,775 that innovations may help them with? 212 00:09:25,915 --> 00:09:27,775 And I think that's where some of the methodologies 213 00:09:27,835 --> 00:09:30,135 as human-centered design, which I've, you know, 214 00:09:30,455 --> 00:09:33,575 recently learned, have been really powerful, particularly, 215 00:09:33,915 --> 00:09:36,295 um, the, so-called ethnographic interviewing, 216 00:09:36,545 --> 00:09:38,895 where you really kind of walk along with somebody, 217 00:09:38,895 --> 00:09:41,375 literally walk along with them as they're trying 218 00:09:41,375 --> 00:09:42,455 to get care if they're a patient 219 00:09:42,515 --> 00:09:44,495 or trying to deliver care if they're a doctor or nurse, 220 00:09:45,115 --> 00:09:46,215 and really sort of see 221 00:09:46,315 --> 00:09:49,015 and ask like, why are you doing it that way? 222 00:09:49,125 --> 00:09:50,135 What are you trying to solve? 223 00:09:51,235 --> 00:09:54,935 And really sort of getting deep into the user journeys 224 00:09:54,935 --> 00:09:57,095 and experiences that ultimately fuel the 225 00:09:57,095 --> 00:09:58,175 solutions you need to develop. 226 00:09:58,515 --> 00:10:00,775 Mm-hmm, . And then externally, you know, 227 00:10:00,775 --> 00:10:02,495 there's a lot of people trying to tackle all 228 00:10:02,495 --> 00:10:04,695 of this other people at other am, uh, 229 00:10:04,935 --> 00:10:06,735 academic medical centers, um, 230 00:10:06,965 --> 00:10:10,955 other people even in industries outside of healthcare who, 231 00:10:11,295 --> 00:10:13,835 um, who are also trying to harness technology 232 00:10:13,975 --> 00:10:15,195 to do a better job for them. 233 00:10:15,775 --> 00:10:17,595 And I have found that just being open 234 00:10:17,615 --> 00:10:19,115 and connecting to those folks 235 00:10:19,855 --> 00:10:23,485 and making time to just sort of run ideas by them 236 00:10:23,545 --> 00:10:26,045 or talk through problems really has sort 237 00:10:26,045 --> 00:10:28,565 of expanded my ability to be more effective in my job. 238 00:10:29,225 --> 00:10:30,405 And then the final thing I'll say, 239 00:10:30,405 --> 00:10:33,805 and this has been true my entire life, read voraciously. 240 00:10:34,345 --> 00:10:36,845 Um, I have a couple of ways that I maintain reading lists 241 00:10:36,905 --> 00:10:40,365 and receive input about interesting things to stay on top 242 00:10:40,365 --> 00:10:43,285 of, either on the web or books or just other materials. 243 00:10:44,145 --> 00:10:47,965 And even though you can quickly run out of time to read, 244 00:10:47,985 --> 00:10:49,485 if you aren't conscious 245 00:10:49,485 --> 00:10:52,725 and like block out time, which I do into your calendar, 246 00:10:52,965 --> 00:10:55,925 I find that your ability to get smart about things 247 00:10:56,025 --> 00:10:58,205 and make connections about things 248 00:10:58,205 --> 00:11:01,645 that ultimately fuel innovations is greatly accelerated. 249 00:11:02,025 --> 00:11:03,285 Um, the more that you can read 250 00:11:03,285 --> 00:11:04,645 and the more broadly that you read. 251 00:11:05,265 --> 00:11:07,325 So those are just some of the things that I have found, 252 00:11:07,705 --> 00:11:10,805 you know, both internal and external community, um, 253 00:11:10,905 --> 00:11:12,285 taking advantage of some 254 00:11:12,285 --> 00:11:14,445 of the human-centered design insights, 255 00:11:14,985 --> 00:11:16,845 and then just being a voracious reader. 256 00:11:17,125 --> 00:11:20,965 I find all of those to be really helpful in some of these, 257 00:11:21,135 --> 00:11:23,125 these roles I've had to date. Can 258 00:11:23,435 --> 00:11:24,515 - You talk about reading broadly? 259 00:11:24,695 --> 00:11:26,115 Is there something you've read recently 260 00:11:26,115 --> 00:11:27,355 that might surprise me? 261 00:11:29,565 --> 00:11:34,035 - Hmm. That, oh, that's a got got you. 262 00:11:34,275 --> 00:11:36,435 Question, isn't it? , you told me you were gonna ask 263 00:11:36,435 --> 00:11:38,715 me one of those . Well, if you 264 00:11:38,715 --> 00:11:41,675 - Need a minute, I don't, I'm a reader myself always 265 00:11:41,675 --> 00:11:42,715 looking for suggestions. 266 00:11:44,625 --> 00:11:49,035 - Yeah. You know, I, um, I'm a little bit of a history buff, 267 00:11:50,555 --> 00:11:54,645 and I just find it fascinating to read 268 00:11:55,455 --> 00:11:57,925 about different parts, different times in history. 269 00:11:58,045 --> 00:12:01,365 I, I just recently finished the Truman biography, 270 00:12:01,365 --> 00:12:02,845 which I felt like I needed to do now 271 00:12:02,845 --> 00:12:04,885 that I'm here in Missouri and he's, uh, 272 00:12:04,885 --> 00:12:06,685 obviously a Missouri Missouri son. 273 00:12:07,575 --> 00:12:11,715 And as, as the listeners in, uh, no Doubt know, he was, 274 00:12:12,055 --> 00:12:14,915 he was president at, uh, a specific time in our history 275 00:12:14,915 --> 00:12:17,955 where a ton was being developed not only in the, 276 00:12:17,975 --> 00:12:19,635 in the aftermath of World War ii, 277 00:12:20,175 --> 00:12:23,355 but then, you know, America was starting to reemerge out of 278 00:12:23,355 --> 00:12:26,195 that war and develop in some ways the foundations 279 00:12:26,195 --> 00:12:31,025 of the modern economy and sort of the various things that he 280 00:12:31,025 --> 00:12:32,985 and his administration had to tackle 281 00:12:33,125 --> 00:12:37,105 and deal with, not only sort of the rise of the Soviet block 282 00:12:37,125 --> 00:12:39,905 and some of the, the foreign matters that were going on, 283 00:12:39,905 --> 00:12:42,305 then the rise of nuclear weapons, um, 284 00:12:42,325 --> 00:12:44,665 but also some of the other sort of societal things. 285 00:12:44,725 --> 00:12:46,905 How do you, how do you handle the influx 286 00:12:46,905 --> 00:12:48,225 of all the GIS coming home 287 00:12:48,805 --> 00:12:50,625 and stand up the economic supports 288 00:12:50,625 --> 00:12:52,265 to make sure they can get off to a good start? 289 00:12:52,765 --> 00:12:54,185 The baby boom was starting to happen. 290 00:12:54,285 --> 00:12:56,825 How do you think about supporting all the, the children 291 00:12:56,825 --> 00:12:59,505 that were coming, just starting to sort of think about 292 00:12:59,505 --> 00:13:00,905 how do you take these massive problems 293 00:13:00,905 --> 00:13:02,825 that in many cases people had never seen before. 294 00:13:03,385 --> 00:13:05,345 IE needed an innovative approach. 295 00:13:05,925 --> 00:13:08,065 And just see how smart people at 296 00:13:08,065 --> 00:13:09,705 that time would sort of tackle the issues. 297 00:13:10,325 --> 00:13:14,065 Um, I, I think everybody's always had this challenge. 298 00:13:14,335 --> 00:13:16,705 They've got, they've got new challenges in front of them, 299 00:13:17,205 --> 00:13:19,305 and just sort of seeing in history 300 00:13:19,885 --> 00:13:21,465 how different figures have tackled that 301 00:13:21,605 --> 00:13:22,705 to me, it's both interesting. 302 00:13:23,125 --> 00:13:26,665 But then, um, you certainly get, get lessons about 303 00:13:26,665 --> 00:13:29,345 how you might not apply that to your own set of problems 304 00:13:29,345 --> 00:13:31,145 and challenges you might have professionally. 305 00:13:31,445 --> 00:13:34,065 Mm-Hmm. . So history's, history's a big thing. 306 00:13:34,605 --> 00:13:36,825 Um, you know, all sorts of industry reads that I do. 307 00:13:36,905 --> 00:13:40,385 I stay on top of the medical literature for my clinical job, 308 00:13:40,405 --> 00:13:42,065 but also to see kind of what's coming. 309 00:13:42,805 --> 00:13:45,105 And then every now and then I'll sprinkle it in with, 310 00:13:45,125 --> 00:13:46,545 um, with fiction. 311 00:13:46,925 --> 00:13:50,025 And my current, my current plug is for, um, 312 00:13:50,305 --> 00:13:53,585 a more tolls book, A Gentleman in Moscow, 313 00:13:54,075 --> 00:13:56,345 which I just finished and absolutely loved. 314 00:13:56,605 --> 00:13:59,145 So that's a good, that's a good escape from it all. 315 00:13:59,365 --> 00:14:00,465 So maybe more than you wanted. 316 00:14:00,785 --> 00:14:01,945 I don't know if I surprised you, 317 00:14:02,445 --> 00:14:03,585 but that's kind of what I'm reading. 318 00:14:04,415 --> 00:14:06,305 - Very well rounded. You know, 319 00:14:06,725 --> 00:14:08,585 the last thing you touched on when walking us 320 00:14:08,585 --> 00:14:10,985 through your roles was the, the rise of ai. 321 00:14:11,445 --> 00:14:13,585 And I'm curious if there's one aspect 322 00:14:13,605 --> 00:14:17,905 or one dimension of AI that has been demanding most 323 00:14:17,905 --> 00:14:19,905 of your time and energy as of late, is, 324 00:14:19,905 --> 00:14:22,385 is there one singular thing you could highlight 325 00:14:22,445 --> 00:14:23,905 for myself and listeners? 326 00:14:25,905 --> 00:14:29,315 - Yeah, I think, um, obviously AI 327 00:14:30,125 --> 00:14:33,275 moved into a whole nother realm of public consciousness 328 00:14:34,145 --> 00:14:38,195 when open AI put chat GPT in the public domain. 329 00:14:38,855 --> 00:14:41,425 And I think it's, it's an interesting phenomenon 330 00:14:41,425 --> 00:14:43,465 to just think about almost sociologically, 331 00:14:43,465 --> 00:14:46,945 because why a has been around for decades, 332 00:14:47,525 --> 00:14:49,585 why is this an inflection point? 333 00:14:50,485 --> 00:14:52,845 And I, my theory, I don't know if it's right, 334 00:14:52,945 --> 00:14:56,845 but my theory is that for the first time, 335 00:14:57,395 --> 00:15:00,875 this technology is speaking 336 00:15:01,015 --> 00:15:04,115 to us quite literally in a language that we understand. 337 00:15:04,875 --> 00:15:07,275 Hmm. You know, previously we, we, the lab 338 00:15:07,275 --> 00:15:10,795 and our organization been working with predictive ai, 339 00:15:11,155 --> 00:15:12,715 analytic AI for a couple years now, 340 00:15:13,745 --> 00:15:17,885 but when you build a predictive model, often the output of 341 00:15:17,885 --> 00:15:20,005 that is, is like a numerical score 342 00:15:20,385 --> 00:15:23,285 or maybe a risk level like this, oh, this patient 343 00:15:23,865 --> 00:15:26,205 is at a high risk for readmission, for example. 344 00:15:26,785 --> 00:15:28,405 And even though that's an important thing to know 345 00:15:28,465 --> 00:15:31,845 and, you know, certain teams would act on that information, 346 00:15:32,425 --> 00:15:36,085 it doesn't really capture the imagination, particularly 347 00:15:36,085 --> 00:15:39,245 of the lay public as much as generative ai. 348 00:15:39,315 --> 00:15:42,405 When you ask it a question, it gives you a response 349 00:15:42,405 --> 00:15:45,165 that like, feels like a human is answering you. 350 00:15:45,585 --> 00:15:48,285 Mm-Hmm. . And I think it's that sort of factor, 351 00:15:48,395 --> 00:15:51,845 that sort of interaction with language, that use of language 352 00:15:52,355 --> 00:15:54,725 that this particular approach of AI uses 353 00:15:54,955 --> 00:15:58,485 that has really captured the imagination, both in terms of, 354 00:15:58,785 --> 00:16:02,085 oh, how exciting it can help us in all these myriad ways, 355 00:16:02,665 --> 00:16:05,005 and frankly, probably causes some 356 00:16:05,005 --> 00:16:08,085 of the anxiety we've heard too about, oh my God, here it is, 357 00:16:08,225 --> 00:16:10,565 how he's gonna lock us out of the bay doors 358 00:16:10,705 --> 00:16:11,885 and we're all gonna die. 359 00:16:12,705 --> 00:16:16,085 Um, so I, I think 360 00:16:16,265 --> 00:16:19,485 that's probably the biggest driver of why it's 361 00:16:19,705 --> 00:16:21,085 so front of mind for everybody now. 362 00:16:21,785 --> 00:16:26,675 And now my job is to then take that somewhat hypey situation 363 00:16:27,095 --> 00:16:31,295 and say, okay, that's cool, but in reality, how can it help? 364 00:16:31,395 --> 00:16:34,785 How can it really solve problems that matter to us? 365 00:16:35,765 --> 00:16:38,945 And so I think, um, I'm, I'm excited about 366 00:16:38,945 --> 00:16:42,825 what we're focusing on because clinical documentation, 367 00:16:43,865 --> 00:16:46,685 the amount of information that we generate as we visit 368 00:16:46,685 --> 00:16:48,965 with patients that need to record those encounters 369 00:16:49,785 --> 00:16:53,925 has never been more, um, voluminous than it is right now. 370 00:16:54,795 --> 00:16:58,815 And it's really just a huge challenge for our clinicians. 371 00:16:59,435 --> 00:17:03,135 You know, a huge number of them are spending hours a day 372 00:17:03,855 --> 00:17:06,455 documenting sometimes obviously, to just make sure 373 00:17:06,455 --> 00:17:08,295 that we're staying on top of the care 374 00:17:08,305 --> 00:17:09,535 we're providing for our patients. 375 00:17:09,955 --> 00:17:12,735 But sometimes it's just a lot of regulatory requirements. 376 00:17:12,835 --> 00:17:13,895 Mm-Hmm. , which, you know, 377 00:17:13,895 --> 00:17:15,175 each has their own rationale, 378 00:17:15,475 --> 00:17:17,735 but when you add it all up, it's an enormous burden, 379 00:17:18,135 --> 00:17:21,175 particularly on, you know, a doc who kind of has to pull 380 00:17:21,175 --> 00:17:22,775 that all on their own shoulders. 381 00:17:23,755 --> 00:17:27,815 So even though it's early innings, if this tool, 382 00:17:27,945 --> 00:17:30,175 which does seem to be well-suited 383 00:17:30,875 --> 00:17:32,375 for summarizing information 384 00:17:32,395 --> 00:17:35,255 and putting it in the formats that we typically use 385 00:17:35,255 --> 00:17:39,175 for clinical notes, if this is an opportunity to really kind 386 00:17:39,175 --> 00:17:42,735 of offload some of that clerical burden, which 387 00:17:42,915 --> 00:17:45,495 to be honest is like a necessary evil, 388 00:17:45,595 --> 00:17:48,655 but not really the reason that any of us get into medicine, 389 00:17:49,405 --> 00:17:52,005 I think that could be a huge satisfier 390 00:17:52,585 --> 00:17:55,205 and a huge help to some of the burnout 391 00:17:55,345 --> 00:17:57,605 and dissatisfaction that I 392 00:17:57,605 --> 00:17:59,805 and my colleagues are seeing in some 393 00:17:59,805 --> 00:18:01,965 of the day-to-day practice of modern medicine. 394 00:18:02,065 --> 00:18:05,265 Mm-Hmm. . So I'm hopeful that as we continue 395 00:18:05,265 --> 00:18:06,505 to build this technology out 396 00:18:06,505 --> 00:18:09,065 and it matures in our ability to work with I tours 397 00:18:09,735 --> 00:18:12,225 that we can really see some productive inroads made 398 00:18:12,575 --> 00:18:13,785 into that particular problem. 399 00:18:15,595 --> 00:18:17,115 - I think that's a really interesting theory 400 00:18:17,115 --> 00:18:20,315 that you put forward on why cha EPT this public 401 00:18:21,515 --> 00:18:23,675 fascination took hold, especially early 2023. 402 00:18:24,435 --> 00:18:26,395 I just remember so many of the headlines 403 00:18:26,415 --> 00:18:27,795 and you know, what could be, 404 00:18:27,935 --> 00:18:29,635 and a lot of hypotheticals around it. 405 00:18:29,855 --> 00:18:31,355 And it seems like, 406 00:18:31,375 --> 00:18:33,515 and I'd be curious what your reaction is to this, 407 00:18:33,615 --> 00:18:36,195 but it seems like in this time we've learned 408 00:18:36,705 --> 00:18:38,635 it's really skillful and good 409 00:18:38,775 --> 00:18:41,755 and helpful in doing a lot of aspects of my job 410 00:18:42,185 --> 00:18:43,675 that I can't even bear to talk about 411 00:18:43,675 --> 00:18:46,035 because it's like the most boring aspects 412 00:18:46,155 --> 00:18:48,755 of my job can now be done by chat EPT. 413 00:18:48,815 --> 00:18:50,955 It doesn't really make for interesting conversation 414 00:18:50,955 --> 00:18:52,075 though, if that makes sense. 415 00:18:52,695 --> 00:18:55,235 Um, compared to the headlines early on that were like, 416 00:18:55,295 --> 00:18:58,075 you know, a a lot of provocative questions being asked 417 00:18:58,075 --> 00:19:00,595 around chat EPT, who it will displace and so forth. 418 00:19:01,215 --> 00:19:03,595 And right now at least it seems like it's really good at 419 00:19:03,595 --> 00:19:05,955 some basic routine admin things. 420 00:19:06,535 --> 00:19:08,235 Um, I don't know if that's similar for, for 421 00:19:08,235 --> 00:19:09,315 what you're seeing in medicine. 422 00:19:10,615 --> 00:19:11,905 - Yeah, I think that's right. 423 00:19:12,125 --> 00:19:14,065 And you know, now the hard work begins, right? 424 00:19:14,065 --> 00:19:16,945 Like, we've gotta figure out how to take this sort of 425 00:19:17,795 --> 00:19:21,125 cool technology and apply it to real world problems. 426 00:19:22,145 --> 00:19:24,565 And I think, um, sure what, 427 00:19:24,565 --> 00:19:27,005 especially when you don't know much about something, 428 00:19:27,765 --> 00:19:30,105 you tend to extrapolate either for, you know, 429 00:19:30,275 --> 00:19:33,185 exorbitant ends, you know, really optimistic ends 430 00:19:33,185 --> 00:19:34,545 or really pessimistic ends. 431 00:19:35,085 --> 00:19:36,545 And there was a lot written about that. 432 00:19:36,685 --> 00:19:38,465 And, and you're right, it was very hypey 433 00:19:38,675 --> 00:19:40,025 right when it first hit the market. 434 00:19:41,745 --> 00:19:45,935 But I think now, um, you know, reality's starting 435 00:19:45,935 --> 00:19:47,255 to set in, it's not perfect. 436 00:19:47,995 --> 00:19:52,255 Um, it takes some doing this, this idea of the right prompts 437 00:19:52,275 --> 00:19:53,775 to direct the models 438 00:19:53,915 --> 00:19:56,615 to generate relevant output is becoming more 439 00:19:56,615 --> 00:19:57,855 and more important is just tricky. 440 00:19:58,555 --> 00:20:00,375 Uh, I think we're getting a little bit better at it, 441 00:20:01,535 --> 00:20:05,995 but I, I, I think this idea that it will 442 00:20:06,635 --> 00:20:09,715 displace jobs we've heard 443 00:20:09,715 --> 00:20:11,555 before, you know, this gets back to history. 444 00:20:12,335 --> 00:20:15,995 Um, there are technological innovations that have happened 445 00:20:16,415 --> 00:20:20,715 for centuries, and at each point society has almost the, 446 00:20:20,955 --> 00:20:23,155 a same question and the same word. 447 00:20:23,155 --> 00:20:26,315 They're like, oh my God, now all this is gonna go away. 448 00:20:26,495 --> 00:20:29,795 Now the Gutenberg designs the printing press, oh my God, 449 00:20:29,795 --> 00:20:30,795 what about all the scribes? 450 00:20:30,795 --> 00:20:31,915 They're gonna lose all their jobs 451 00:20:33,155 --> 00:20:34,895 and you know, so on and so forth. 452 00:20:34,995 --> 00:20:37,015 And I think what we have seen time 453 00:20:37,015 --> 00:20:40,465 and time again is sure, some jobs 454 00:20:41,685 --> 00:20:45,985 can now be replaced, if you will, by the technology, 455 00:20:46,445 --> 00:20:48,645 but the people doing those jobs 456 00:20:49,545 --> 00:20:51,845 almost inevitably have other things to do. 457 00:20:52,145 --> 00:20:54,965 Mm-Hmm. . And one of the examples 458 00:20:54,965 --> 00:20:59,075 that made the most sense to me recently was the idea that, 459 00:20:59,455 --> 00:21:01,035 uh, electronic calculators, 460 00:21:01,355 --> 00:21:03,875 whenever they came out, everybody thought 461 00:21:03,985 --> 00:21:06,995 that is now the end of being an accountant. 462 00:21:07,445 --> 00:21:09,955 There will no longer be a need to be an accountant, 463 00:21:10,535 --> 00:21:11,555 and that's just not true. 464 00:21:12,415 --> 00:21:14,675 The fact of the matter is, being an accountant changed 465 00:21:15,135 --> 00:21:17,475 and you no longer needed to go through the drudgery 466 00:21:17,775 --> 00:21:19,955 of summing up massive rows of numbers 467 00:21:20,135 --> 00:21:21,395 and doing everything 468 00:21:21,395 --> 00:21:23,835 that an electronic calculator can now do in seconds. 469 00:21:25,175 --> 00:21:29,075 But now the job in an accountant became much more strategic 470 00:21:29,655 --> 00:21:31,595 and taking advantage of that technology 471 00:21:31,735 --> 00:21:32,955 to further their thinking 472 00:21:32,955 --> 00:21:36,435 and productivity in the overall world of accounting. 473 00:21:36,575 --> 00:21:38,915 Mm-Hmm. . And I think that'll be true in the 474 00:21:38,915 --> 00:21:41,355 various ways that AI gets applied for healthcare too. 475 00:21:41,765 --> 00:21:43,595 It'll change the nature of my job. 476 00:21:44,335 --> 00:21:46,675 And, and somebody once said, you know, 477 00:21:46,975 --> 00:21:48,675 AI is not gonna replace to the doctors, 478 00:21:49,695 --> 00:21:53,645 but doctors using AI will replace doctors 479 00:21:53,865 --> 00:21:54,925 not using ai. 480 00:21:55,625 --> 00:21:57,245 And I think that's true. Mm-Hmm. . 481 00:21:57,465 --> 00:21:59,285 So we'll see how it plays out. 482 00:21:59,345 --> 00:22:00,445 But if history's a guide 483 00:22:00,465 --> 00:22:03,125 and it very often is, I suspect that's 484 00:22:03,125 --> 00:22:04,365 where we'll land. Yeah. 485 00:22:04,735 --> 00:22:06,125 - We're coming full circle now. 486 00:22:06,265 --> 00:22:08,445 Dr. Maddox, uh, this has been such an interesting 487 00:22:08,555 --> 00:22:09,845 dynamic conversation. 488 00:22:09,985 --> 00:22:12,165 Is, is there anything we didn't mention that you'd like 489 00:22:12,165 --> 00:22:14,085 to leave our listeners with as we wind down here? 490 00:22:16,115 --> 00:22:18,135 - No, I think, um, you know, most of what I, 491 00:22:18,135 --> 00:22:21,415 what I've been saying around the use of AI as an assistant, 492 00:22:21,675 --> 00:22:24,575 as a co-pilot, as an augmentation in medicine, I, 493 00:22:24,655 --> 00:22:27,215 I do feel strongly about, um, I haven't, well, 494 00:22:27,215 --> 00:22:28,655 I should say I haven't seen any signals 495 00:22:28,875 --> 00:22:30,695 to dissuade me from that point of view. 496 00:22:31,195 --> 00:22:33,415 And so I'm, I'm a bit of a techno optimist, 497 00:22:33,515 --> 00:22:36,015 but I also think I'm realistic and, uh, 498 00:22:36,015 --> 00:22:38,335 and I'm just excited to do the hard work. 499 00:22:38,395 --> 00:22:41,615 But I think ultimately the productive work to, um, 500 00:22:41,915 --> 00:22:45,215 to really mature the technology into something 501 00:22:45,215 --> 00:22:48,375 that can be useful for our organization and, 502 00:22:48,475 --> 00:22:50,215 and for the patients that we serve. 503 00:22:51,315 --> 00:22:54,115 I think the other thing that, um, is exciting 504 00:22:54,115 --> 00:22:58,235 to me about the technology is I've mentioned that so much 505 00:22:58,235 --> 00:22:59,355 of a clinician's day, 506 00:22:59,485 --> 00:23:01,275 day-to-day experience right now is filled 507 00:23:01,275 --> 00:23:03,995 with clerical drudgery and administrative tasks 508 00:23:04,175 --> 00:23:06,035 and note generation, et cetera. 509 00:23:06,735 --> 00:23:08,515 And I think it's really exciting to think about 510 00:23:08,515 --> 00:23:11,395 how technology can kind of remove that, at least 511 00:23:11,395 --> 00:23:15,235 to a degree from the day-to-day experience of healthcare. 512 00:23:15,535 --> 00:23:19,275 And, and in some ways re-enable the humanity of healthcare, 513 00:23:19,745 --> 00:23:22,555 that I can finally take my eyes off the computer screen 514 00:23:23,225 --> 00:23:25,315 instead, look at my patient, talk to them, 515 00:23:26,095 --> 00:23:28,835 and really, you know, continue to build the relationship 516 00:23:29,545 --> 00:23:33,195 that frankly underpins all effective, uh, 517 00:23:33,305 --> 00:23:35,275 clinician patient relationships. 518 00:23:36,175 --> 00:23:39,875 And so if this, if this technology can help to that end, 519 00:23:40,745 --> 00:23:42,235 then um, you know, the work 520 00:23:42,295 --> 00:23:45,035 and sometimes the aggravation of all the work is, 521 00:23:45,305 --> 00:23:46,605 is certainly gonna be well worth it. 522 00:23:48,405 --> 00:23:50,935 - Well, Dr. Max, I wanna wish you continued luck in, 523 00:23:50,955 --> 00:23:54,255 in success in your role as a techno optimist, 524 00:23:54,315 --> 00:23:56,055 but a realistic one at that . 525 00:23:56,135 --> 00:23:58,175 I think always a, a real pleasure to catch up 526 00:23:58,175 --> 00:24:00,175 with people like you and, and get your read on things, 527 00:24:00,755 --> 00:24:01,815 um, to set us straight. 528 00:24:01,995 --> 00:24:04,335 So listeners, this has been Dr. Thomas Maddox, 529 00:24:04,365 --> 00:24:05,695 vice President of Digital Products 530 00:24:05,695 --> 00:24:07,335 and Innovation with BJC, healthcare 531 00:24:07,475 --> 00:24:09,415 and Washington University School of Medicine. 532 00:24:09,525 --> 00:24:10,695 Also Professor of Medicine 533 00:24:10,755 --> 00:24:13,055 and Cardiology at Washington University School 534 00:24:13,055 --> 00:24:14,575 of Medicine, Dr. Maddox. 535 00:24:14,575 --> 00:24:16,135 Hope we can catch up again sometime soon. 536 00:24:17,475 --> 00:24:17,925 - Love it.