1 00:00:00,640 --> 00:00:03,199 Philips is a health tech leader focused on 2 00:00:03,199 --> 00:00:06,000 innovation that improves the health and well-being of 3 00:00:06,000 --> 00:00:09,460 people. Our health care technology and informatics solutions 4 00:00:09,599 --> 00:00:12,559 help care teams diagnose, treat, and manage more 5 00:00:12,559 --> 00:00:16,244 patients with greater precision, speed, and confidence across 6 00:00:16,244 --> 00:00:19,225 the care journey. With Philips, clinicians are empowered 7 00:00:19,285 --> 00:00:21,765 with streamlined insights in the moments that matter 8 00:00:21,765 --> 00:00:24,904 for every patient. Better care for more people. 9 00:00:25,204 --> 00:00:25,704 Philips. 10 00:00:29,410 --> 00:00:31,329 Hello, and welcome to the Becker's Health Care 11 00:00:31,329 --> 00:00:33,570 podcast recorded at the 9th annual health IT 12 00:00:33,570 --> 00:00:36,469 and digital health with RCM conference in Chicago. 13 00:00:36,929 --> 00:00:39,350 I'm joined today by Chris Carmody, chief technology 14 00:00:39,570 --> 00:00:40,710 officer at UPMC. 15 00:00:41,315 --> 00:00:42,914 Chris, to get us started today, will you 16 00:00:42,914 --> 00:00:44,594 please share a bit about yourself, your background, 17 00:00:44,594 --> 00:00:45,655 and your role at your organization? 18 00:00:46,515 --> 00:00:49,655 Sure. I've, been with UPMC for over 26 19 00:00:49,715 --> 00:00:52,354 years. I've done pretty much everything in IT 20 00:00:52,354 --> 00:00:53,174 you can imagine. 21 00:00:53,480 --> 00:00:55,320 But the easiest way to describe my role 22 00:00:55,320 --> 00:00:55,820 today 23 00:00:56,120 --> 00:00:57,020 is I 24 00:00:57,480 --> 00:00:59,880 am responsible for everything except for our payer 25 00:00:59,880 --> 00:01:00,380 applications 26 00:01:01,240 --> 00:01:02,940 and our oncology EHR. 27 00:01:03,560 --> 00:01:04,859 But everything from cybersecurity 28 00:01:05,400 --> 00:01:07,420 to our currently ten EHRs, 29 00:01:08,834 --> 00:01:09,734 to our epic, 30 00:01:10,114 --> 00:01:12,834 single EHR implementation that we're going through right 31 00:01:12,834 --> 00:01:13,334 now. 32 00:01:14,034 --> 00:01:15,974 Among everything else you can imagine, infrastructure, 33 00:01:16,834 --> 00:01:19,715 just everything kinda falls under, my responsibilities. But 34 00:01:19,715 --> 00:01:21,314 I have a great team that I work 35 00:01:21,314 --> 00:01:22,055 with closely, 36 00:01:22,650 --> 00:01:24,250 to make sure that things work well at 37 00:01:24,250 --> 00:01:24,750 UTMC. 38 00:01:25,450 --> 00:01:26,430 Awesome. Well 39 00:01:26,890 --> 00:01:28,810 well, in your role, obviously, you're aware, you 40 00:01:28,810 --> 00:01:31,310 know, AI adoption is exploding in health care. 41 00:01:31,609 --> 00:01:33,850 In your view, you know, what's the most 42 00:01:33,850 --> 00:01:36,810 significant or promising application of this technology right 43 00:01:36,810 --> 00:01:39,575 now, and how is this informing your organization's, 44 00:01:39,954 --> 00:01:41,094 innovation strategy? 45 00:01:41,715 --> 00:01:44,515 They're you're right. AI is exploding. I think 46 00:01:44,515 --> 00:01:47,314 it's a tremendous opportunity for health care. A 47 00:01:47,314 --> 00:01:49,234 couple areas that we're really focused on right 48 00:01:49,234 --> 00:01:50,614 now have to deal with 49 00:01:51,079 --> 00:01:52,939 efficiency opportunities. So 50 00:01:53,319 --> 00:01:56,040 great example is the Ambient AI that our 51 00:01:56,040 --> 00:01:56,520 physicians 52 00:01:57,000 --> 00:02:00,040 in fact, over 1700 currently use Ambient AI 53 00:02:00,040 --> 00:02:01,640 in their interactions with patients to help with 54 00:02:01,640 --> 00:02:02,780 the clinical documentation. 55 00:02:03,240 --> 00:02:05,079 So it's really about that efficiency to get 56 00:02:05,079 --> 00:02:07,844 them away from turning away from the computer 57 00:02:07,844 --> 00:02:10,664 and typing things in or even dictating after 58 00:02:10,884 --> 00:02:12,185 they have that patient encounter 59 00:02:12,564 --> 00:02:15,125 to engaging with a patient, having a a 60 00:02:15,125 --> 00:02:17,284 better experience, and, again, getting them out of 61 00:02:17,284 --> 00:02:19,125 the documentation business. They still have to review 62 00:02:19,125 --> 00:02:21,240 things and then and, you know, accept it 63 00:02:21,240 --> 00:02:23,259 into the electronic health record, but, 64 00:02:23,639 --> 00:02:26,439 we're seeing great positive feedback with that. We've 65 00:02:26,439 --> 00:02:27,259 used AI, 66 00:02:27,560 --> 00:02:28,780 from a data acquisition 67 00:02:29,080 --> 00:02:31,959 and aggregation perspective. So from a clinical analytics, 68 00:02:31,959 --> 00:02:33,500 so we look at predictive modeling 69 00:02:34,105 --> 00:02:35,405 and and trying to, 70 00:02:36,265 --> 00:02:39,224 achieve our goals around higher quality and more 71 00:02:39,224 --> 00:02:40,844 precision medicine for patients. 72 00:02:41,224 --> 00:02:43,385 We use these predictive models that are informed 73 00:02:43,385 --> 00:02:45,485 on all this data that we've generated over, 74 00:02:46,025 --> 00:02:47,645 honestly, over 2 decades, 75 00:02:48,060 --> 00:02:50,319 with electronic health records and other sources 76 00:02:51,900 --> 00:02:54,319 to predict out different models for, 77 00:02:55,099 --> 00:02:55,599 readmission 78 00:02:55,980 --> 00:02:57,360 and even presurgical, 79 00:02:58,379 --> 00:03:00,219 preventative actions that we can take with patients 80 00:03:00,219 --> 00:03:02,140 before they go through a surgical process to 81 00:03:02,140 --> 00:03:04,275 help improve the quality and outcomes. And then 82 00:03:04,275 --> 00:03:05,474 lastly, I think you see a lot of 83 00:03:05,474 --> 00:03:07,735 this from a diagnostic perspective around 84 00:03:08,114 --> 00:03:10,294 images. A lot of things like auto contouring, 85 00:03:10,514 --> 00:03:12,594 things that are saving time and very accurate 86 00:03:12,594 --> 00:03:14,534 at this point in time. But I think 87 00:03:14,675 --> 00:03:17,590 our challenge still ahead is, again, training the 88 00:03:17,590 --> 00:03:20,469 the generative AIMO models and how that fits 89 00:03:20,469 --> 00:03:20,969 into 90 00:03:21,590 --> 00:03:23,750 the delivery of care and and and, again, 91 00:03:23,750 --> 00:03:25,990 creating the efficiencies, but improving the quality and 92 00:03:25,990 --> 00:03:28,009 and the outcomes for our patients. Mhmm. 93 00:03:28,655 --> 00:03:29,155 So 94 00:03:29,534 --> 00:03:31,455 on a daily basis, health care leaders are 95 00:03:31,455 --> 00:03:34,014 managing greater volumes of data, you know, on 96 00:03:34,014 --> 00:03:35,935 more devices and in a growing number of 97 00:03:35,935 --> 00:03:37,794 care settings and growing number of populations. 98 00:03:38,574 --> 00:03:41,534 In this complex environment, what clinical data integration 99 00:03:41,534 --> 00:03:44,129 tools or practices are you seeing drive improvements 100 00:03:44,129 --> 00:03:46,209 in patient outcomes and operations? And can you 101 00:03:46,209 --> 00:03:47,810 share an example or 2? Yeah. I I 102 00:03:48,049 --> 00:03:49,989 getting back to the use of AI, 103 00:03:50,289 --> 00:03:52,310 we actually have used a tool 104 00:03:52,769 --> 00:03:54,469 for the last few years. 105 00:03:55,090 --> 00:03:56,175 I don't know if you want me to 106 00:03:56,415 --> 00:03:58,254 name vendors, but I'll stay away from it 107 00:03:58,254 --> 00:04:00,175 from now. But Mhmm. It's a it's an 108 00:04:00,175 --> 00:04:02,514 AI machine learning tool that actually 109 00:04:02,895 --> 00:04:06,014 takes the unstructured data. So, you know, it's 110 00:04:06,014 --> 00:04:08,254 very easy to take a structured field and 111 00:04:08,254 --> 00:04:10,269 make sense of it. Think about if you're 112 00:04:10,269 --> 00:04:12,349 unfamiliar with the one an electronic health record 113 00:04:12,349 --> 00:04:13,709 looks like. Think of it, like, as a 114 00:04:13,709 --> 00:04:15,569 cell, like, in in Excel or 115 00:04:15,949 --> 00:04:17,790 Access or something like that to where it's 116 00:04:17,790 --> 00:04:19,550 it's structured. You can, you know, mail, theme, 117 00:04:19,550 --> 00:04:20,129 or whatever, 118 00:04:20,910 --> 00:04:23,229 versus the the unstructured data where there's a 119 00:04:23,229 --> 00:04:26,444 lot of rich content that wasn't always readily 120 00:04:26,504 --> 00:04:28,044 accessible in the EHRs 121 00:04:28,504 --> 00:04:31,225 in helping inform those models. So we use 122 00:04:31,225 --> 00:04:34,345 AI to extract out things like symptoms and 123 00:04:34,345 --> 00:04:37,245 social determinants of health, which weren't in EHRs 124 00:04:37,384 --> 00:04:38,985 10 years ago or 5 years ago or 125 00:04:38,985 --> 00:04:41,279 maybe even 3 years ago to help get 126 00:04:41,279 --> 00:04:43,759 that data and and inform the models in 127 00:04:43,759 --> 00:04:45,300 terms of our predictive, 128 00:04:46,240 --> 00:04:47,620 analytics that we've created 129 00:04:48,000 --> 00:04:50,960 around our ecosystem of delivering health care. So 130 00:04:50,960 --> 00:04:52,420 it's been it's been wily, 131 00:04:53,275 --> 00:04:55,754 helpful for us to actually do something productive 132 00:04:55,754 --> 00:04:58,314 with that data other than just creating an 133 00:04:58,314 --> 00:04:59,995 electronic record of something that used to be 134 00:04:59,995 --> 00:05:01,535 on paper. Mhmm. Mhmm. 135 00:05:01,835 --> 00:05:03,595 So, you know, you talked about things that 136 00:05:03,595 --> 00:05:05,694 weren't possible 5, 10 years ago. 137 00:05:06,019 --> 00:05:07,300 You know, what do you think might be 138 00:05:07,300 --> 00:05:09,379 possible in in 5 to 10 years that's 139 00:05:09,379 --> 00:05:11,639 not possible at all today? Well, I think 140 00:05:12,100 --> 00:05:13,399 AI is gonna be 141 00:05:13,779 --> 00:05:16,259 a a big driver towards that. But, again, 142 00:05:16,259 --> 00:05:19,074 I'm a technologist at heart. I'm very excited 143 00:05:19,074 --> 00:05:21,634 about quantum computing and what that can do 144 00:05:21,634 --> 00:05:24,214 in terms of solving some very complex problems. 145 00:05:24,435 --> 00:05:26,594 Hopefully, down the road, it'll actually help, you 146 00:05:26,594 --> 00:05:27,735 know, cure diseases, 147 00:05:28,435 --> 00:05:30,534 and help drive that that discovery. 148 00:05:30,914 --> 00:05:32,754 So I'm I'm I'm very excited what that 149 00:05:32,754 --> 00:05:35,509 might mean 5 or maybe maybe it's closer 150 00:05:35,509 --> 00:05:37,910 to 10 years from now, that intersection of 151 00:05:37,910 --> 00:05:39,529 all this data we're we're generating, 152 00:05:40,389 --> 00:05:42,230 the focus on health care, not just within 153 00:05:42,230 --> 00:05:43,910 the four walls of a hospital or physician 154 00:05:43,910 --> 00:05:46,150 office, but how we're reaching in and and 155 00:05:46,150 --> 00:05:49,415 and being present for the patient wherever they 156 00:05:49,415 --> 00:05:50,074 may be. 157 00:05:50,454 --> 00:05:52,855 Obviously, all the wearable devices that that we 158 00:05:52,855 --> 00:05:54,154 have today, which Mhmm. 159 00:05:55,014 --> 00:05:57,894 Isn't really well adopted in terms of ingesting 160 00:05:57,894 --> 00:05:59,574 that into our records, but I I think 161 00:05:59,574 --> 00:06:01,649 there's an opportunity to improve that as well. 162 00:06:01,810 --> 00:06:03,750 When you see this sort of intersection 163 00:06:04,209 --> 00:06:06,209 up ahead on the horizon of all this 164 00:06:06,209 --> 00:06:07,029 coming together, 165 00:06:07,410 --> 00:06:09,089 I just think the sky's the limit as 166 00:06:09,089 --> 00:06:10,610 we look into the future or what's what's 167 00:06:10,610 --> 00:06:12,370 possible from a health care perspective. Right. And 168 00:06:12,370 --> 00:06:13,729 and I I often hear it described too 169 00:06:13,729 --> 00:06:15,009 as similar, like, when the iPhone hit the 170 00:06:15,009 --> 00:06:16,764 market, you know, you couldn't even imagine some 171 00:06:16,764 --> 00:06:18,305 of the things we're doing with it today. 172 00:06:18,685 --> 00:06:19,964 And that's gonna be kinda the same way 173 00:06:19,964 --> 00:06:21,805 we look at, like, like, say, for example, 174 00:06:21,805 --> 00:06:23,165 AI and health care and its uses in 175 00:06:23,165 --> 00:06:24,524 the future. A lot of them don't even 176 00:06:24,524 --> 00:06:26,444 have any idea yet. Absolutely. And I and 177 00:06:26,444 --> 00:06:27,745 I think, obviously, 178 00:06:28,600 --> 00:06:30,600 we'll we'll figure out the future down the 179 00:06:30,600 --> 00:06:32,680 road, but I at least within the next 180 00:06:32,680 --> 00:06:34,519 couple years, I think AI is gonna get 181 00:06:34,519 --> 00:06:36,519 better. It's gonna become more accurate. We'll have 182 00:06:36,519 --> 00:06:37,420 less hallucinations. 183 00:06:37,879 --> 00:06:40,120 We'll figure out better ways to govern the 184 00:06:40,120 --> 00:06:40,904 use of it, 185 00:06:41,305 --> 00:06:43,704 to where not just our patients or, in 186 00:06:43,704 --> 00:06:45,625 our case, at UPMC, our our health plan 187 00:06:45,625 --> 00:06:48,185 members feel comfortable interacting with a Mhmm. A 188 00:06:48,185 --> 00:06:51,404 generative AI type of technology, but our physicians, 189 00:06:51,544 --> 00:06:53,944 our nurses, our technicians that how that can 190 00:06:53,944 --> 00:06:56,605 help better enable them doing their job. Again, 191 00:06:56,830 --> 00:06:59,230 all geared towards the outcome of higher quality 192 00:06:59,230 --> 00:07:01,230 of care in a more efficient and effective 193 00:07:01,230 --> 00:07:01,730 way. 194 00:07:02,189 --> 00:07:02,689 Perfect. 195 00:07:03,150 --> 00:07:05,650 Well, how can health care leaders and organizations, 196 00:07:06,670 --> 00:07:08,750 better support IT and clinical teams as they 197 00:07:08,750 --> 00:07:10,590 carry out innovation efforts? And, you know, what 198 00:07:10,590 --> 00:07:12,029 are some of the common pitfalls you see 199 00:07:12,029 --> 00:07:13,384 here? I think 200 00:07:13,845 --> 00:07:15,685 probably the biggest thing has nothing to do 201 00:07:15,685 --> 00:07:18,004 with technology. It has everything to do with 202 00:07:18,004 --> 00:07:20,824 change. And how how do we, as leaders, 203 00:07:22,004 --> 00:07:24,404 put our organization in a position that our 204 00:07:24,404 --> 00:07:26,644 our our users, the people actually using the 205 00:07:26,644 --> 00:07:29,100 technology, again, doctors, nurses, technicians, 206 00:07:29,560 --> 00:07:31,339 operational folks, revenue cycle, 207 00:07:31,639 --> 00:07:34,199 how can how can they adopt and change 208 00:07:34,199 --> 00:07:36,199 how they did things in the past? So 209 00:07:36,199 --> 00:07:37,959 as we see that improvement, that's a that's 210 00:07:37,959 --> 00:07:39,720 tough to get human beings to change how 211 00:07:39,720 --> 00:07:41,835 they did things in the past. So we're 212 00:07:41,835 --> 00:07:44,395 successful with that. Again, sky's the limit to 213 00:07:44,395 --> 00:07:47,275 what what's possible in leveraging technology and actually 214 00:07:47,275 --> 00:07:49,035 getting the most out of the investment because 215 00:07:49,035 --> 00:07:51,514 technology is not cheap. AI is definitely not 216 00:07:51,514 --> 00:07:53,435 cheap. Mhmm. So we need to be effective 217 00:07:53,435 --> 00:07:55,435 in how we manage that change and how 218 00:07:55,435 --> 00:07:57,120 do we get it institutionalized 219 00:07:57,500 --> 00:07:58,639 into our organizations. 220 00:07:59,019 --> 00:07:59,919 Mhmm. Mhmm. 221 00:08:00,539 --> 00:08:02,060 And what would you say your top piece 222 00:08:02,060 --> 00:08:03,819 of advice for health care leaders as they 223 00:08:03,819 --> 00:08:06,300 prepare for further advancements in technology and greater 224 00:08:06,300 --> 00:08:08,639 demands for care? That's a great question. 225 00:08:09,579 --> 00:08:10,879 A lot of it is 226 00:08:11,915 --> 00:08:15,035 participating in events like Becker's, where you get 227 00:08:15,035 --> 00:08:18,074 to hear from different technology companies. You get 228 00:08:18,074 --> 00:08:20,254 to talk to your peers across the industry. 229 00:08:20,634 --> 00:08:22,074 You get to hear their problems and how 230 00:08:22,074 --> 00:08:24,790 they're potentially solving them. You might have very 231 00:08:24,790 --> 00:08:26,810 similar ones that, you know, you can basically 232 00:08:26,870 --> 00:08:27,370 beg, 233 00:08:27,830 --> 00:08:30,790 beg, borrow, and steal their ideas, which isn't 234 00:08:30,790 --> 00:08:32,330 a bad thing. But, 235 00:08:32,870 --> 00:08:34,809 you know, an aspect of that is listening, 236 00:08:34,870 --> 00:08:37,769 you know, talking, engaging with your your clinicians, 237 00:08:37,910 --> 00:08:39,825 with with the folks that are utilizing the 238 00:08:39,825 --> 00:08:42,304 technology and trying to help come together and 239 00:08:42,304 --> 00:08:44,465 figure out what that right solution is from 240 00:08:44,465 --> 00:08:47,184 a technology perspective versus, hey. We're IT, and 241 00:08:47,184 --> 00:08:48,865 you're gonna use this. We're gonna force this 242 00:08:48,865 --> 00:08:51,904 upon you. It's that's not a successful way 243 00:08:51,904 --> 00:08:53,830 to especially in health care to get that 244 00:08:53,830 --> 00:08:55,750 level of adoption and the outcomes that we're 245 00:08:55,750 --> 00:08:57,190 striving for. So we have to be in 246 00:08:57,190 --> 00:08:59,110 this together. Mhmm. So a lot of that, 247 00:08:59,110 --> 00:09:01,429 again, is building relationships that's listening to our 248 00:09:01,429 --> 00:09:01,929 customers 249 00:09:02,470 --> 00:09:04,629 and then working with them to Mhmm. Again, 250 00:09:04,629 --> 00:09:05,929 deal with that change management 251 00:09:06,235 --> 00:09:09,054 aspect and and accepting and adopting the technology. 252 00:09:09,115 --> 00:09:11,355 So we actually fulfill what we think will 253 00:09:11,355 --> 00:09:12,654 happen with a new technology. 254 00:09:13,274 --> 00:09:15,934 Well, Chris, I really appreciate this conversation today. 255 00:09:16,315 --> 00:09:17,355 Just wanna know, do you have any final 256 00:09:17,355 --> 00:09:18,715 thoughts you'd like to leave with our listeners? 257 00:09:18,715 --> 00:09:20,095 No. Just, again, 258 00:09:20,700 --> 00:09:22,059 it starts out to be a a great 259 00:09:22,059 --> 00:09:24,220 conference again at Becker's, and looking forward to 260 00:09:24,220 --> 00:09:25,660 the rest of the rest of the time 261 00:09:25,660 --> 00:09:27,660 here and looking forward to to learn a 262 00:09:27,660 --> 00:09:29,500 lot more about what others are doing and 263 00:09:29,500 --> 00:09:31,179 what some of these technology partners are up 264 00:09:31,179 --> 00:09:33,019 to. Well, thank you so much for joining 265 00:09:33,019 --> 00:09:34,059 us today, and I hope you have a 266 00:09:34,059 --> 00:09:35,679 great rest of your day. Thank you.