1 00:00:00,719 --> 00:00:04,000 Exciting things are happening at Becker's Healthcare. Stay 2 00:00:04,000 --> 00:00:06,240 ahead of industry trends with the new Becker's 3 00:00:06,240 --> 00:00:08,419 CFO plus Revenue Cycle podcast, 4 00:00:08,800 --> 00:00:11,119 your go to source for insights from top 5 00:00:11,119 --> 00:00:14,154 healthcare finance leaders. Tune in wherever you get 6 00:00:14,154 --> 00:00:15,054 your podcasts. 7 00:00:15,914 --> 00:00:18,475 And don't miss the tenth annual health IT 8 00:00:18,475 --> 00:00:22,154 plus digital health plus RCM conference, happening September 9 00:00:22,154 --> 00:00:24,474 30 to 10/03/2025 10 00:00:24,474 --> 00:00:25,214 in Chicago. 11 00:00:25,755 --> 00:00:26,879 Join thousands of 12 00:00:27,279 --> 00:00:30,079 executives, engage with industry leaders, and explore the 13 00:00:30,079 --> 00:00:31,619 future of health care innovation. 14 00:00:32,320 --> 00:00:36,719 Learn more about our upcoming events at beckershospitalreview.com. 15 00:00:36,719 --> 00:00:37,619 See you there. 16 00:00:38,479 --> 00:00:40,559 This is Laura Dierdo with the Becker's Healthcare 17 00:00:40,559 --> 00:00:42,515 podcast. I'm thrilled today to be joined by 18 00:00:42,515 --> 00:00:43,655 doctor Yasser Terabishi, 19 00:00:44,195 --> 00:00:47,075 chief medical informatics officer at Ovation and chief 20 00:00:47,075 --> 00:00:49,554 health AI officer at MetroHealth. Doctor Terabishi, it's 21 00:00:49,554 --> 00:00:50,914 a pleasure to have you on the podcast 22 00:00:50,914 --> 00:00:52,755 today. Thank you. It's a pleasure to be 23 00:00:52,755 --> 00:00:53,250 here. 24 00:00:53,649 --> 00:00:55,890 Absolutely. Now I'm excited to have you here 25 00:00:55,890 --> 00:00:57,489 because I know there's so many innovative and 26 00:00:57,489 --> 00:00:59,329 cool things that you're doing, especially on the 27 00:00:59,329 --> 00:01:02,770 digital health side at, MetroHealth system. But before 28 00:01:02,770 --> 00:01:03,969 we dive into that, can you tell us 29 00:01:03,969 --> 00:01:06,370 a little bit more about yourself and the 30 00:01:06,370 --> 00:01:07,989 organization? What makes it unique? 31 00:01:09,134 --> 00:01:11,134 Sure. So, I'll start with my role at 32 00:01:11,134 --> 00:01:13,774 MetroHealth. So I'm chief health AI officer at 33 00:01:13,774 --> 00:01:14,274 MetroHealth. 34 00:01:14,814 --> 00:01:17,155 And in in that role, it's really to 35 00:01:17,454 --> 00:01:18,594 guide the responsible 36 00:01:19,054 --> 00:01:19,554 implementation 37 00:01:20,015 --> 00:01:23,375 and strategic planning around AI initiatives within within 38 00:01:23,375 --> 00:01:24,034 the organization. 39 00:01:24,719 --> 00:01:27,520 And for reference, MetroHealth is Cleveland's safety net 40 00:01:27,520 --> 00:01:29,760 health care system that serves a diverse patient 41 00:01:29,760 --> 00:01:30,260 population 42 00:01:30,640 --> 00:01:32,340 in Cleveland and and beyond. 43 00:01:33,040 --> 00:01:34,480 The other hat that I wear is I 44 00:01:34,480 --> 00:01:35,219 am CMIO 45 00:01:35,520 --> 00:01:38,340 of Ovation. And Ovation is actually a virtual 46 00:01:38,400 --> 00:01:40,954 care start up that's spun out of MetroHealth 47 00:01:41,174 --> 00:01:42,075 and our partners 48 00:01:42,694 --> 00:01:45,655 in South Carolina, namely the Medical University of 49 00:01:45,655 --> 00:01:46,475 South Carolina. 50 00:01:47,015 --> 00:01:49,594 The goal behind this platform is to create 51 00:01:49,975 --> 00:01:53,674 a single solution that integrates into several different 52 00:01:53,814 --> 00:01:56,509 partner settings where patients can get integrated 53 00:01:57,049 --> 00:01:59,609 primary and urgent care. And so what that 54 00:01:59,609 --> 00:02:02,329 means specifically is we can take care of 55 00:02:02,329 --> 00:02:03,310 connecting patients 56 00:02:03,689 --> 00:02:05,629 to on the ground brick and mortar solutions 57 00:02:06,009 --> 00:02:08,509 at the partner organization and their home organization 58 00:02:08,944 --> 00:02:10,705 in a way that other point solutions that 59 00:02:10,705 --> 00:02:12,325 exist today do not. 60 00:02:13,425 --> 00:02:15,585 That's really fantastic to hear. You know, cool 61 00:02:15,585 --> 00:02:17,444 to hear about that kind of collaboration 62 00:02:17,825 --> 00:02:18,325 between, 63 00:02:19,504 --> 00:02:20,564 MetroHealth and, 64 00:02:21,025 --> 00:02:23,344 and USC. It certainly seems like a great 65 00:02:23,344 --> 00:02:23,844 opportunity 66 00:02:24,145 --> 00:02:27,340 to, work on a a digital technology that 67 00:02:27,340 --> 00:02:30,060 is really meaningful and beneficial for patients and 68 00:02:30,060 --> 00:02:31,819 can make a big difference for the health 69 00:02:31,819 --> 00:02:33,579 system as well in getting for the right 70 00:02:33,579 --> 00:02:35,579 people at the right places more quickly. How 71 00:02:35,579 --> 00:02:37,280 did that partnership come about? 72 00:02:37,735 --> 00:02:39,895 Actually, it's been, several years in the making, 73 00:02:39,895 --> 00:02:41,194 mostly a collegial, 74 00:02:41,894 --> 00:02:42,394 friendship 75 00:02:43,254 --> 00:02:43,834 and collaboration 76 00:02:44,215 --> 00:02:46,694 that started off at the highest levels of 77 00:02:46,694 --> 00:02:47,354 each organization, 78 00:02:47,974 --> 00:02:50,659 namely the executives at the time. And it 79 00:02:50,659 --> 00:02:51,860 was a it was a bit of a 80 00:02:51,860 --> 00:02:53,560 natural, I think, collaboration. 81 00:02:53,860 --> 00:02:55,000 So MetroHealth 82 00:02:55,379 --> 00:02:58,360 is known for being very innovative in the 83 00:02:58,580 --> 00:03:00,979 digital and informatics space. And I would say 84 00:03:00,979 --> 00:03:04,114 we're scrappy innovators. We we generally punch above 85 00:03:04,114 --> 00:03:06,194 our weights for for the organization that that 86 00:03:06,194 --> 00:03:06,694 we're 87 00:03:06,995 --> 00:03:09,074 at, the, you know, the size of our 88 00:03:09,074 --> 00:03:09,574 organization 89 00:03:10,194 --> 00:03:11,474 and and the fact that we're a safety 90 00:03:11,474 --> 00:03:14,514 net system that's typically more resource resource strapped 91 00:03:14,514 --> 00:03:15,314 than the average, 92 00:03:15,715 --> 00:03:17,094 you know, health care system. 93 00:03:18,060 --> 00:03:19,520 And MUSC really 94 00:03:19,980 --> 00:03:22,400 similar patient population that they serve. 95 00:03:23,260 --> 00:03:24,400 They also are 96 00:03:24,700 --> 00:03:27,900 relatively well known in the virtual care space. 97 00:03:27,900 --> 00:03:29,360 They have a very strong, 98 00:03:30,300 --> 00:03:30,800 asynchronous 99 00:03:31,775 --> 00:03:34,414 care program, actually, which is nationally recognized. And 100 00:03:34,414 --> 00:03:36,914 so the leadership on that side felt that, 101 00:03:37,455 --> 00:03:39,394 the leadership on this side 102 00:03:39,854 --> 00:03:41,555 could work together to achieve 103 00:03:41,854 --> 00:03:44,275 a share a common goal, really, a shared 104 00:03:44,520 --> 00:03:46,539 platform here between both organizations. 105 00:03:47,240 --> 00:03:49,080 And I think the the really cool thing 106 00:03:49,080 --> 00:03:50,060 about that opportunity 107 00:03:50,680 --> 00:03:52,780 is, you know, if you bring two organizations 108 00:03:53,000 --> 00:03:56,120 together that are trying to solve virtual care, 109 00:03:56,120 --> 00:03:58,300 integrated virtual care within their systems, 110 00:03:58,895 --> 00:04:01,294 When we start comparing the problems we're trying 111 00:04:01,294 --> 00:04:01,955 to solve, 112 00:04:02,335 --> 00:04:05,615 we start to align on shared problems and 113 00:04:05,615 --> 00:04:07,395 start to actually co solution. 114 00:04:08,094 --> 00:04:10,354 And what that does is create a platform 115 00:04:10,415 --> 00:04:13,090 or an environment where we start solving problems 116 00:04:13,090 --> 00:04:14,310 and providing solutions, 117 00:04:15,090 --> 00:04:16,709 that are much more scalable. 118 00:04:17,410 --> 00:04:19,889 Right? So we're pouring in resources together, co 119 00:04:19,889 --> 00:04:20,870 solutioning, collaborating, 120 00:04:21,410 --> 00:04:23,670 and we're creating a platform and a system 121 00:04:23,889 --> 00:04:25,490 that we can take to other partners. And 122 00:04:25,490 --> 00:04:27,845 so that's that's kind of our opportunity there 123 00:04:27,845 --> 00:04:29,365 to scale this out to a to a 124 00:04:29,365 --> 00:04:31,685 higher, you know, to a higher level across 125 00:04:31,685 --> 00:04:32,904 different states and different 126 00:04:33,285 --> 00:04:33,785 organizations. 127 00:04:34,404 --> 00:04:35,865 Our solution is built 128 00:04:36,324 --> 00:04:39,120 on the Epic platform and takes advantage of 129 00:04:39,120 --> 00:04:40,800 all of the things that the Epic platform 130 00:04:40,800 --> 00:04:42,879 provides. And we've built on top of that 131 00:04:42,879 --> 00:04:44,259 with Epic and in addition, 132 00:04:44,879 --> 00:04:46,420 internally with our developers 133 00:04:47,120 --> 00:04:49,520 to kind of complete that experience so that 134 00:04:49,520 --> 00:04:51,839 it really is as integrated and seamless as 135 00:04:51,839 --> 00:04:54,160 can be even though it it's truly a 136 00:04:54,160 --> 00:04:54,660 separate 137 00:04:55,014 --> 00:04:55,514 entity. 138 00:04:56,055 --> 00:04:57,975 That's fascinating to hear. Thank you so much 139 00:04:57,975 --> 00:05:00,714 for explaining that to us. Now I'm curious. 140 00:05:00,775 --> 00:05:02,615 What is the accomplishment that you're most proud 141 00:05:02,615 --> 00:05:03,835 of from the last year? 142 00:05:04,535 --> 00:05:06,855 Right. So I'll say, you know, I'll I'll 143 00:05:06,855 --> 00:05:08,855 use my two hats here to to go 144 00:05:08,855 --> 00:05:09,595 through these. 145 00:05:09,980 --> 00:05:11,500 When I think about my c m I 146 00:05:11,660 --> 00:05:15,740 CMIO hat innovation, so we have successfully launched 147 00:05:15,740 --> 00:05:18,540 a virtual care service for for patients that 148 00:05:18,540 --> 00:05:21,100 are in Ohio that are attributed to the 149 00:05:21,100 --> 00:05:24,080 MetroHealth system, both established and new. 150 00:05:24,675 --> 00:05:27,154 We've we've had, at this point, well over 151 00:05:27,154 --> 00:05:30,194 10 providers that we've hired on. We've seen 152 00:05:30,194 --> 00:05:33,415 thousands of patients, and we've had just fantastic 153 00:05:33,475 --> 00:05:34,935 results, raving reviews, 154 00:05:35,634 --> 00:05:37,980 a lot of emphasis on the ease of 155 00:05:37,980 --> 00:05:38,480 access, 156 00:05:39,020 --> 00:05:41,759 lot of a lot of good positive feedback 157 00:05:41,819 --> 00:05:42,720 on the integration. 158 00:05:43,020 --> 00:05:44,540 So, you know, I'll I'll give you an 159 00:05:44,540 --> 00:05:45,040 example. 160 00:05:45,900 --> 00:05:46,960 You have a fever. 161 00:05:47,339 --> 00:05:48,800 You call into our services. 162 00:05:49,504 --> 00:05:50,865 We figure out that you really ought to 163 00:05:50,865 --> 00:05:52,164 get swabbed for COVID, 164 00:05:52,625 --> 00:05:55,345 and flu, and we can go from a 165 00:05:55,345 --> 00:05:56,324 virtual visit 166 00:05:56,625 --> 00:05:58,704 to a swab in a brick and mortar 167 00:05:58,704 --> 00:06:00,805 facility that's a MetroHealth facility, 168 00:06:01,745 --> 00:06:02,724 to the prescription 169 00:06:03,264 --> 00:06:05,125 and even delivery, home delivery, 170 00:06:05,879 --> 00:06:06,939 for instance, Paxlovid, 171 00:06:07,479 --> 00:06:08,699 if if you had COVID, 172 00:06:09,399 --> 00:06:10,699 within several hours, 173 00:06:11,240 --> 00:06:13,399 which is just amazing compared to the usual 174 00:06:13,399 --> 00:06:15,240 experience of going into an express care, an 175 00:06:15,240 --> 00:06:17,500 emergency room and waiting for much longer. 176 00:06:18,095 --> 00:06:20,175 And so it's been really rewarding to see 177 00:06:20,175 --> 00:06:20,835 that integration 178 00:06:21,295 --> 00:06:22,035 play out. 179 00:06:22,895 --> 00:06:24,335 The other, you know, so I'd say that's 180 00:06:24,335 --> 00:06:26,014 a really big win we've had over the 181 00:06:26,014 --> 00:06:27,475 last year that's 182 00:06:28,095 --> 00:06:30,095 developed and evolved. And actually, we're hoping to 183 00:06:30,095 --> 00:06:31,634 go live in South Carolina 184 00:06:32,254 --> 00:06:32,754 shortly 185 00:06:33,060 --> 00:06:35,139 actually to to match the level of service 186 00:06:35,139 --> 00:06:36,759 that we've achieved in Ohio. 187 00:06:37,379 --> 00:06:38,759 And so that's been a phenomenal 188 00:06:39,139 --> 00:06:39,639 opportunity, 189 00:06:40,100 --> 00:06:41,240 lots of growth potential, 190 00:06:42,180 --> 00:06:45,060 for Ovation and also MetroHealth and MUSC in 191 00:06:45,060 --> 00:06:47,560 that collaboration. So that's really coming to fruition. 192 00:06:48,254 --> 00:06:49,935 We also have several other partners that are 193 00:06:49,935 --> 00:06:52,175 lined up thinking about, you know, launching in 194 00:06:52,175 --> 00:06:53,855 other states and other systems. So I think 195 00:06:53,855 --> 00:06:56,035 that the value of what we're developing 196 00:06:56,335 --> 00:06:57,634 has been well received. 197 00:06:58,095 --> 00:06:59,955 As far as the AI work, 198 00:07:00,415 --> 00:07:02,194 that I do at MetroHealth, so, 199 00:07:02,539 --> 00:07:04,860 you know, I'm a physician informaticist first. I'm 200 00:07:04,860 --> 00:07:06,800 a pulmonary critical care doc at MetroHealth. 201 00:07:07,660 --> 00:07:08,160 And 202 00:07:08,860 --> 00:07:11,259 my growth into the AI space really came 203 00:07:11,259 --> 00:07:12,639 from several implementations 204 00:07:13,099 --> 00:07:15,660 of AI solutions that maybe didn't always have 205 00:07:15,660 --> 00:07:17,120 something to do with my specialty, 206 00:07:18,045 --> 00:07:19,964 but got a lot of attention for being 207 00:07:19,964 --> 00:07:22,625 really pragmatic, responsible, ways to 208 00:07:23,004 --> 00:07:25,665 validate, vet, and implement a solution. 209 00:07:26,045 --> 00:07:28,144 And the takeaway from that has always been, 210 00:07:28,285 --> 00:07:29,745 you know, there's a large responsibility 211 00:07:31,330 --> 00:07:34,129 in implementing these solutions, especially in in very 212 00:07:34,129 --> 00:07:36,770 diverse, different environments where these models may not 213 00:07:36,770 --> 00:07:38,389 have been validated or developed. 214 00:07:39,250 --> 00:07:40,930 We always want to make sure we are 215 00:07:40,930 --> 00:07:41,430 improving, 216 00:07:42,129 --> 00:07:43,650 care for all of our patients. We wanna 217 00:07:43,650 --> 00:07:45,189 make sure that we are not 218 00:07:45,964 --> 00:07:48,524 worsening disparities that we know exist in all 219 00:07:48,524 --> 00:07:49,664 aspects of care. 220 00:07:50,285 --> 00:07:51,664 And, you know, beyond that, 221 00:07:52,125 --> 00:07:54,764 thinking strategically about how do we bring these 222 00:07:54,764 --> 00:07:56,444 solutions in, you know, how do we make 223 00:07:56,444 --> 00:07:58,620 sure that they're actually having an impact? And 224 00:07:58,699 --> 00:08:00,459 I'll say that a lot of our learning 225 00:08:00,459 --> 00:08:02,939 is consistent with with others and that the 226 00:08:02,939 --> 00:08:04,639 vast majority of the value 227 00:08:05,180 --> 00:08:07,660 of any AI or tech implementation for that 228 00:08:07,660 --> 00:08:08,160 matter 229 00:08:08,779 --> 00:08:10,139 is more to do with the people in 230 00:08:10,139 --> 00:08:12,464 the process and the why and less to 231 00:08:12,464 --> 00:08:14,064 do with the actual solution. So we've been 232 00:08:14,064 --> 00:08:17,745 building predictive models and analytics for really decades 233 00:08:17,745 --> 00:08:18,805 at this point in medicine. 234 00:08:19,185 --> 00:08:21,824 And our opportunity has really been harnessing them 235 00:08:21,824 --> 00:08:23,204 in a way that makes sense 236 00:08:23,584 --> 00:08:24,720 at the point of care 237 00:08:25,040 --> 00:08:27,519 and measuring and understanding that impact, both in 238 00:08:27,519 --> 00:08:30,180 terms of positive impact and also unanticipated 239 00:08:30,480 --> 00:08:30,980 consequences. 240 00:08:32,080 --> 00:08:34,320 And there are several examples that actually that 241 00:08:34,320 --> 00:08:35,220 we have published. 242 00:08:35,519 --> 00:08:37,279 You can find in the literature today if 243 00:08:37,279 --> 00:08:37,860 you looked. 244 00:08:39,225 --> 00:08:41,785 So, you know, we've taken that framework and 245 00:08:41,785 --> 00:08:43,945 we've built governance on top of that. So 246 00:08:43,945 --> 00:08:45,245 we have healthy, 247 00:08:45,625 --> 00:08:46,125 multidisciplinary 248 00:08:47,625 --> 00:08:49,225 AI governance in our in the form of 249 00:08:49,225 --> 00:08:50,764 our AI advisory committee. 250 00:08:51,690 --> 00:08:53,690 And I'll say one thing that I thought 251 00:08:53,690 --> 00:08:57,049 we've done relatively well is not jump onto 252 00:08:57,049 --> 00:08:59,610 the wagon too quickly. And so we did 253 00:08:59,610 --> 00:09:01,370 not want to be at the extreme cutting 254 00:09:01,370 --> 00:09:02,590 edge of these solutions. 255 00:09:03,290 --> 00:09:05,129 You know, we don't have that much flex 256 00:09:05,129 --> 00:09:06,590 money to spend on 257 00:09:07,345 --> 00:09:08,485 experimental concepts. 258 00:09:08,945 --> 00:09:11,024 And so we waited just long enough to 259 00:09:11,024 --> 00:09:13,184 know that, for instance, ambient AI needs to 260 00:09:13,184 --> 00:09:15,105 happen, and we already know who the top 261 00:09:15,105 --> 00:09:16,785 players in the space are. And so we've 262 00:09:16,785 --> 00:09:18,785 made a selection now for for our first 263 00:09:18,785 --> 00:09:21,105 partner in the ambient AI space without having 264 00:09:21,105 --> 00:09:23,870 to do six or seven pilots like other 265 00:09:23,870 --> 00:09:25,090 sites were were doing. 266 00:09:25,950 --> 00:09:28,429 And, you know, from the the the rest 267 00:09:28,429 --> 00:09:29,889 of the generative AI perspective, 268 00:09:30,750 --> 00:09:32,370 we've just partnered with, 269 00:09:33,430 --> 00:09:36,129 a relatively mature but still evolving 270 00:09:36,750 --> 00:09:37,250 vendor 271 00:09:37,575 --> 00:09:38,475 who will be 272 00:09:38,934 --> 00:09:40,634 conducting general advice summarization 273 00:09:41,335 --> 00:09:44,215 within the electronic health records. So it's not 274 00:09:44,215 --> 00:09:45,274 a native solution, 275 00:09:45,735 --> 00:09:47,894 but we found an opportunity where we could 276 00:09:47,894 --> 00:09:49,995 partner and collaborate and build 277 00:09:50,330 --> 00:09:52,250 for our own, you know, our own purposes 278 00:09:52,250 --> 00:09:53,309 in a relatively, 279 00:09:53,690 --> 00:09:56,429 I would say, affordable way that both feeds 280 00:09:57,049 --> 00:09:58,570 what we hope to do with the core 281 00:09:58,570 --> 00:10:00,649 product as it exists and also gives us 282 00:10:00,649 --> 00:10:01,149 opportunity 283 00:10:01,529 --> 00:10:04,009 to develop what we think we need as, 284 00:10:04,009 --> 00:10:06,384 you know, as MetroHealth individually 285 00:10:06,684 --> 00:10:07,504 or independently. 286 00:10:08,684 --> 00:10:10,524 That's great to hear. And really exciting on 287 00:10:10,524 --> 00:10:13,004 both ends, especially as you mentioned, just building 288 00:10:13,004 --> 00:10:14,924 out that AI infrastructure and how you're able 289 00:10:14,924 --> 00:10:15,424 to, 290 00:10:15,884 --> 00:10:18,929 understand the advisory council, looking at the ways 291 00:10:18,929 --> 00:10:21,090 that, you know, it will be most meaningful 292 00:10:21,090 --> 00:10:23,190 to bring AI and generative AI into, 293 00:10:23,889 --> 00:10:26,769 the broader organization, I think, is something that 294 00:10:26,769 --> 00:10:28,610 so many hospitals and health systems are still 295 00:10:28,610 --> 00:10:30,629 trying to work through today and, you know, 296 00:10:30,690 --> 00:10:32,825 much a position like yourself and looking at, 297 00:10:32,825 --> 00:10:34,585 you know, what partnerships makes the most sense, 298 00:10:34,585 --> 00:10:36,665 what products then you know, what applications are 299 00:10:36,665 --> 00:10:39,304 gonna be most meaningful for them. Given this 300 00:10:39,304 --> 00:10:41,625 background you you've, shared with us in terms 301 00:10:41,625 --> 00:10:42,764 of the digital transformation, 302 00:10:43,144 --> 00:10:44,264 where do you see some of the big 303 00:10:44,264 --> 00:10:46,424 growth opportunities over the next twelve months or 304 00:10:46,424 --> 00:10:46,924 so? 305 00:10:47,225 --> 00:10:49,680 I I think it's leading into exactly those 306 00:10:49,680 --> 00:10:52,100 spaces. So if we focus on the generative 307 00:10:52,160 --> 00:10:53,460 AI applications, 308 00:10:54,960 --> 00:10:56,879 I'll say that there's, you know, there's a 309 00:10:56,879 --> 00:10:58,639 lot that we know that this technology can 310 00:10:58,639 --> 00:11:01,620 do. It's probably strongest in terms of information 311 00:11:01,759 --> 00:11:02,259 retrieval. 312 00:11:04,384 --> 00:11:06,705 The opportunities beyond that, I think, you know, 313 00:11:06,705 --> 00:11:08,945 there's a lot of questions that come my 314 00:11:08,945 --> 00:11:11,264 way as chief health officer, like why can't 315 00:11:11,264 --> 00:11:13,345 we just put the entire electronic health record 316 00:11:13,345 --> 00:11:15,105 into, you know, a large language model? Like 317 00:11:15,105 --> 00:11:16,544 why can't you just copy and paste once 318 00:11:16,544 --> 00:11:17,605 all trying to chat 319 00:11:18,129 --> 00:11:20,449 in a protected setting, obviously? And those of 320 00:11:20,449 --> 00:11:21,730 us that live in the space know it's 321 00:11:21,730 --> 00:11:23,970 a lot more complicated. So what information we 322 00:11:23,970 --> 00:11:26,789 submit to a large language model, the context, 323 00:11:26,850 --> 00:11:27,509 the prompting. 324 00:11:27,809 --> 00:11:29,569 Right. That's that's the hard part. That's the 325 00:11:29,569 --> 00:11:31,970 part that really requires a lot of diligence 326 00:11:31,970 --> 00:11:32,709 and oversight. 327 00:11:33,415 --> 00:11:33,995 And so, 328 00:11:34,455 --> 00:11:36,455 you know, we've gotten to a point now 329 00:11:36,455 --> 00:11:38,315 where we, I would say, as a community, 330 00:11:38,375 --> 00:11:41,014 have started to understand where this technology does 331 00:11:41,014 --> 00:11:43,254 well, where it does not. I think the 332 00:11:43,254 --> 00:11:44,875 obvious parts are, like I said, 333 00:11:45,254 --> 00:11:47,595 information retrieval. So what happened to this patient? 334 00:11:47,720 --> 00:11:50,040 You know, in this last hospital encounter, for 335 00:11:50,040 --> 00:11:51,720 instance, or write me a discharge summary. I 336 00:11:51,720 --> 00:11:54,299 think those are all fairly achievable things. 337 00:11:55,000 --> 00:11:57,000 What we'd like to see more of and 338 00:11:57,000 --> 00:11:59,100 I think the literature starting sport, 339 00:11:59,559 --> 00:12:02,345 it's more of that kind of physician's copilot 340 00:12:02,644 --> 00:12:04,345 where if I can identify, 341 00:12:04,965 --> 00:12:07,205 I'll use the pulmonary and critical care example 342 00:12:07,205 --> 00:12:09,365 here, but I can identify when somebody looks 343 00:12:09,365 --> 00:12:11,365 like they could have an infection so that 344 00:12:11,365 --> 00:12:12,884 I can determine that they may or may 345 00:12:12,884 --> 00:12:13,784 not be septic. 346 00:12:14,259 --> 00:12:16,500 That's information that can come from rich notes, 347 00:12:16,500 --> 00:12:18,980 that can come from imaging, for instance, that 348 00:12:18,980 --> 00:12:21,620 isn't immediately available in the electronic health record 349 00:12:21,620 --> 00:12:24,100 today. So I think that there's opportunities to 350 00:12:24,100 --> 00:12:26,440 develop phenotypes from the data 351 00:12:26,784 --> 00:12:29,664 that can serve both our operational and clinical 352 00:12:29,664 --> 00:12:33,205 need and then beyond that, actually, our quality 353 00:12:33,504 --> 00:12:36,384 focus. And so being able to capture or, 354 00:12:36,384 --> 00:12:37,684 you know, identify 355 00:12:38,304 --> 00:12:38,804 potentially 356 00:12:39,710 --> 00:12:41,090 reversible or preventable 357 00:12:41,470 --> 00:12:43,090 conditions early in the state. 358 00:12:43,549 --> 00:12:45,149 I think that's gonna be really crucial. And 359 00:12:45,149 --> 00:12:48,110 another example, for instance, is someone might have 360 00:12:48,110 --> 00:12:50,350 a Foley catheter or a line that does 361 00:12:50,350 --> 00:12:51,870 not need to be there. It's a lot 362 00:12:51,870 --> 00:12:53,889 more difficult to identify that 363 00:12:54,615 --> 00:12:57,335 in real time when you're not trying to 364 00:12:57,335 --> 00:12:57,835 analyze 365 00:12:58,615 --> 00:13:01,754 context through documentation or a note. 366 00:13:02,134 --> 00:13:04,375 So I'm excited about that prospect, and that's 367 00:13:04,375 --> 00:13:06,054 the kind of development that we're going to 368 00:13:06,054 --> 00:13:08,634 be working on. In addition to those things, 369 00:13:08,899 --> 00:13:10,600 we always have to focus on 370 00:13:11,059 --> 00:13:13,620 what can potentially improve our bottom line. So 371 00:13:13,620 --> 00:13:16,019 if we're paying for this expensive technology, then 372 00:13:16,019 --> 00:13:18,580 naturally after that, start asking, can we can 373 00:13:18,580 --> 00:13:20,899 we use the technology for appropriate billing? Right. 374 00:13:20,899 --> 00:13:22,039 So we don't miss diagnosis 375 00:13:22,934 --> 00:13:24,695 to make sure that we're submitting the right 376 00:13:24,695 --> 00:13:26,955 information for for CDI, for instance. 377 00:13:27,415 --> 00:13:30,875 We're also using the technology for, prior authorizations, 378 00:13:31,735 --> 00:13:34,075 for example. So that does help us 379 00:13:34,695 --> 00:13:37,195 solve a unique health care problem faster. 380 00:13:37,600 --> 00:13:38,899 Although I'm usually 381 00:13:39,279 --> 00:13:41,299 really nervous about using technology 382 00:13:41,840 --> 00:13:43,220 to solve broken processes. 383 00:13:44,240 --> 00:13:46,000 Obviously, as a single system, we're not going 384 00:13:46,000 --> 00:13:47,519 to be able to solve a lot of 385 00:13:47,519 --> 00:13:49,279 these problems. But if they can help us 386 00:13:49,279 --> 00:13:50,340 reduce the effort 387 00:13:50,644 --> 00:13:52,404 and get the patients the outcomes that we 388 00:13:52,404 --> 00:13:54,485 need, I think that's that's fair game at 389 00:13:54,485 --> 00:13:55,144 this point. 390 00:13:56,325 --> 00:13:57,945 Absolutely. That sounds great. 391 00:13:58,325 --> 00:14:00,644 And really truly will be cool when you 392 00:14:00,644 --> 00:14:02,725 have that system in place to use the 393 00:14:02,725 --> 00:14:04,264 technology, use the AI, 394 00:14:05,125 --> 00:14:06,509 and get the most out of, 395 00:14:06,910 --> 00:14:08,830 the team members that you have in in 396 00:14:08,830 --> 00:14:10,210 in really truly elevating, 397 00:14:10,590 --> 00:14:12,350 throughout the space. So that's really cool to 398 00:14:12,350 --> 00:14:14,190 hear about, and thank you for sharing with 399 00:14:14,190 --> 00:14:16,029 us. I I know there's a lot of 400 00:14:16,029 --> 00:14:18,590 opportunities, especially with technology and AI, but I 401 00:14:18,590 --> 00:14:20,745 can imagine some big challenges as well. What 402 00:14:20,904 --> 00:14:22,504 are you anticipating for the next year or 403 00:14:22,504 --> 00:14:24,985 so? Anything that could be potential roadblocks for 404 00:14:24,985 --> 00:14:25,485 you? 405 00:14:25,865 --> 00:14:27,464 Yeah. I'll I'll do the the two hat 406 00:14:27,464 --> 00:14:28,524 thing now. So, 407 00:14:28,985 --> 00:14:30,664 when I when I think about my role 408 00:14:30,664 --> 00:14:32,845 in in virtual care and CMI of Invesha, 409 00:14:33,784 --> 00:14:34,365 I think 410 00:14:35,290 --> 00:14:37,450 I'll just say that the the telehealth extension, 411 00:14:37,450 --> 00:14:38,590 while reassuring 412 00:14:38,970 --> 00:14:40,269 and, you know, 413 00:14:41,210 --> 00:14:43,070 definitely a positive move, 414 00:14:43,850 --> 00:14:46,190 only gives us so much runway to build 415 00:14:46,570 --> 00:14:48,889 our our plat to continue to build our 416 00:14:48,889 --> 00:14:50,029 platform and solution 417 00:14:50,424 --> 00:14:51,725 in a comfortable way. 418 00:14:52,105 --> 00:14:54,524 So I think most of us have now 419 00:14:54,825 --> 00:14:57,304 reached this conclusion that virtual care is here 420 00:14:57,304 --> 00:15:00,504 to stay. It's a necessary component. It's it 421 00:15:00,504 --> 00:15:01,004 obviously 422 00:15:01,384 --> 00:15:03,545 works best at the right patient time and 423 00:15:03,545 --> 00:15:05,340 place. It's not for all care settings. 424 00:15:05,740 --> 00:15:07,600 It's really can you right size it? 425 00:15:07,980 --> 00:15:08,960 Patients expect 426 00:15:09,340 --> 00:15:11,820 it. Clinicians enjoy it. At least, you know, 427 00:15:11,820 --> 00:15:13,600 based on the surveys we've done internally, 428 00:15:14,139 --> 00:15:17,019 it does provide more access. It likely is 429 00:15:17,019 --> 00:15:19,500 similar or better quality in terms of keeping 430 00:15:19,500 --> 00:15:21,394 patients engaged to their health care 431 00:15:21,695 --> 00:15:23,554 journey. And so I would love to see 432 00:15:24,254 --> 00:15:26,115 more of a realization of that, 433 00:15:26,654 --> 00:15:28,574 again, at the kind of the policy level. 434 00:15:28,574 --> 00:15:31,214 Right. To continue to fund that mechanism. Because, 435 00:15:31,214 --> 00:15:33,615 again, it's I think it's it's here to 436 00:15:33,615 --> 00:15:35,534 stay. It's not going to go back to 437 00:15:35,534 --> 00:15:36,355 COVID levels 438 00:15:36,710 --> 00:15:38,470 anytime soon. But there is a time and 439 00:15:38,470 --> 00:15:39,670 a place for it. And I think we 440 00:15:39,670 --> 00:15:40,889 need to acknowledge that 441 00:15:41,350 --> 00:15:43,429 when I put on my C H I 442 00:15:43,429 --> 00:15:44,090 O hat. 443 00:15:44,870 --> 00:15:46,730 The biggest concern that I have 444 00:15:47,429 --> 00:15:49,910 continues to be spending money in the wrong 445 00:15:49,910 --> 00:15:51,850 places and on the wrong solutions. 446 00:15:52,445 --> 00:15:54,365 I still think we're figuring these things out, 447 00:15:54,365 --> 00:15:56,284 so we are still treading lightly. We are 448 00:15:56,284 --> 00:15:58,705 trying to, at least when we commit to 449 00:15:59,004 --> 00:16:00,144 partners in the space, 450 00:16:00,524 --> 00:16:02,125 come in in a way where we have 451 00:16:02,125 --> 00:16:05,504 leeway. You know, our EHR vendors developing solutions 452 00:16:05,725 --> 00:16:06,544 and opportunities, 453 00:16:07,789 --> 00:16:09,970 when it comes to generative AI specifically 454 00:16:10,750 --> 00:16:12,209 that start to conflict. 455 00:16:12,669 --> 00:16:14,209 I'm really concerned about 456 00:16:14,509 --> 00:16:16,669 AI scope creep. So when we pick a 457 00:16:16,669 --> 00:16:19,389 solution for a particular pain point or if 458 00:16:19,389 --> 00:16:21,089 it's even part of a larger platform, 459 00:16:21,914 --> 00:16:23,914 it starts to grow in terms of scope. 460 00:16:23,914 --> 00:16:26,154 So, you know, first, it's summarizing notes. Next, 461 00:16:26,154 --> 00:16:28,315 it's listening to your conversation. Well, hold on. 462 00:16:28,315 --> 00:16:30,095 I already have an AMP and AI solution. 463 00:16:30,794 --> 00:16:32,315 These things are starting to bump into each 464 00:16:32,315 --> 00:16:34,875 other. So I am worried about the overall 465 00:16:34,875 --> 00:16:35,375 spend. 466 00:16:35,690 --> 00:16:37,309 From the clinician perspective, 467 00:16:38,250 --> 00:16:39,789 I think there's a lot of assumption 468 00:16:40,570 --> 00:16:42,089 about the fact that all of this is 469 00:16:42,089 --> 00:16:44,809 going to make clinicians' lives easier. But it's 470 00:16:44,809 --> 00:16:47,049 still out. You know, the question is still 471 00:16:47,049 --> 00:16:49,789 out there as to whether or not this 472 00:16:50,355 --> 00:16:51,014 really decreases 473 00:16:51,394 --> 00:16:53,955 a provider's cognitive burden or if it just 474 00:16:53,955 --> 00:16:56,115 kind of changes the level of work that 475 00:16:56,115 --> 00:16:57,495 they have to do. Right? So 476 00:16:57,955 --> 00:16:59,095 writing a note versus 477 00:16:59,554 --> 00:17:01,554 reviewing a note that's been written by a 478 00:17:01,554 --> 00:17:02,934 generative AI solution, 479 00:17:03,539 --> 00:17:04,599 looking for hallucinations, 480 00:17:04,980 --> 00:17:07,539 being always worried about something that crept in 481 00:17:07,539 --> 00:17:09,059 there that's not supposed to be in there 482 00:17:09,059 --> 00:17:11,460 or even worse, potentially something that was that 483 00:17:11,460 --> 00:17:13,940 was missed. Right? So none of these solutions 484 00:17:13,940 --> 00:17:15,220 are perfect. And so we have to be 485 00:17:15,220 --> 00:17:16,359 really careful about 486 00:17:16,825 --> 00:17:18,744 how much more work we're throwing on our 487 00:17:18,744 --> 00:17:20,525 care team and making assumptions 488 00:17:20,984 --> 00:17:23,465 about whether or not we're truly, you know, 489 00:17:23,465 --> 00:17:24,365 truly enhancing 490 00:17:25,225 --> 00:17:25,965 their workload. 491 00:17:26,984 --> 00:17:28,585 So, yeah, I think that those are those 492 00:17:28,585 --> 00:17:30,059 are the things that I don't know that 493 00:17:30,059 --> 00:17:31,740 they keep me up at night, but they're 494 00:17:31,740 --> 00:17:33,839 definitely concerns that I have as we step 495 00:17:34,140 --> 00:17:35,759 more, you know, into the space. 496 00:17:36,940 --> 00:17:39,339 Absolutely. I think, definitely, it makes a lot 497 00:17:39,339 --> 00:17:41,179 of sense and, you know, especially in bringing 498 00:17:41,179 --> 00:17:41,980 in AI, 499 00:17:42,619 --> 00:17:43,200 you know, 500 00:17:43,525 --> 00:17:45,365 not really a blueprint of how that should 501 00:17:45,365 --> 00:17:46,744 all work. And so I think, 502 00:17:47,125 --> 00:17:48,644 it seems like you've got a a great 503 00:17:48,644 --> 00:17:50,325 approach and strategy to just keep an eye 504 00:17:50,325 --> 00:17:52,724 on that and understanding the the risks as 505 00:17:52,724 --> 00:17:54,484 well. Before we wrap up here with our 506 00:17:54,484 --> 00:17:56,164 last minute, what is the number one thing 507 00:17:56,164 --> 00:17:58,184 you're doing right now to set your organization 508 00:17:58,244 --> 00:17:59,039 up for long term 509 00:18:00,160 --> 00:18:02,100 success? I think we've we've echoed, 510 00:18:02,480 --> 00:18:04,000 a lot of these points here, but it's 511 00:18:04,000 --> 00:18:05,220 it's it's worth rehashing. 512 00:18:05,920 --> 00:18:07,859 So making sure we have a solid 513 00:18:08,640 --> 00:18:09,460 AI governance 514 00:18:09,840 --> 00:18:11,759 strategy and plan, making sure we have the 515 00:18:11,759 --> 00:18:13,940 right people, a diverse set of experiences, 516 00:18:14,714 --> 00:18:15,454 at the table, 517 00:18:15,835 --> 00:18:18,335 and making sure that we upscale our clinicians 518 00:18:18,794 --> 00:18:21,515 and what these technologies can do and what 519 00:18:21,515 --> 00:18:23,755 they can't do. And also empowering them. Right. 520 00:18:23,755 --> 00:18:25,515 To speak up when it's not working the 521 00:18:25,515 --> 00:18:27,694 way that it's supposed to be working. 522 00:18:28,075 --> 00:18:29,855 So continuing down that road, 523 00:18:30,470 --> 00:18:32,250 being really careful about 524 00:18:32,789 --> 00:18:34,950 the the life cycle of all of these 525 00:18:34,950 --> 00:18:36,009 solutions, meaning 526 00:18:36,710 --> 00:18:38,650 being ready from a financial 527 00:18:39,190 --> 00:18:40,490 responsibility perspective 528 00:18:41,029 --> 00:18:43,450 and from even just an overall governance perspective 529 00:18:43,509 --> 00:18:45,769 to say this tool or this solution 530 00:18:46,365 --> 00:18:48,625 is not working. We need to decommission it. 531 00:18:49,085 --> 00:18:50,924 I think there's this natural assumption of more 532 00:18:50,924 --> 00:18:53,005 technology is always good. So we have to 533 00:18:53,005 --> 00:18:54,924 keep balancing that out. Right. To make sure 534 00:18:54,924 --> 00:18:57,565 that we're actually literally getting what we pay 535 00:18:57,565 --> 00:18:59,644 for and making sure that we're not making 536 00:18:59,644 --> 00:19:01,825 anything harder than it has to be. 537 00:19:02,250 --> 00:19:03,769 So those are the things that I focused 538 00:19:03,769 --> 00:19:06,190 on, at least from the AI perspective. 539 00:19:07,369 --> 00:19:08,029 From the 540 00:19:08,490 --> 00:19:09,630 virtual care 541 00:19:10,170 --> 00:19:10,670 perspective, 542 00:19:11,289 --> 00:19:13,369 I think we're just keeping our heads down 543 00:19:13,369 --> 00:19:13,869 and 544 00:19:14,265 --> 00:19:16,924 trying to build a solid solution 545 00:19:17,305 --> 00:19:19,565 or I should say building a solid solution 546 00:19:19,945 --> 00:19:20,765 and a product 547 00:19:21,545 --> 00:19:23,545 that really scales well. And the most important 548 00:19:23,545 --> 00:19:25,464 thing I think that that I've learned in 549 00:19:25,464 --> 00:19:27,779 that space is you can build the most 550 00:19:27,779 --> 00:19:30,659 exquisite piece of technology and platform that is, 551 00:19:30,659 --> 00:19:31,720 to a high degree, 552 00:19:32,259 --> 00:19:35,240 seamlessly integrated into others, EHRs or solutions. 553 00:19:35,859 --> 00:19:37,240 But really, what's driving 554 00:19:37,619 --> 00:19:38,119 our 555 00:19:38,579 --> 00:19:39,079 energy 556 00:19:39,380 --> 00:19:42,034 and driving our positive results, it's the people 557 00:19:42,034 --> 00:19:43,015 that we've hired. 558 00:19:43,394 --> 00:19:44,294 And so we've 559 00:19:44,835 --> 00:19:48,674 continued to, you know, make sure that people 560 00:19:48,674 --> 00:19:50,755 understand people that we bring out, understand our 561 00:19:50,755 --> 00:19:52,694 culture, that they're good fits for our culture. 562 00:19:52,994 --> 00:19:54,534 And I think a lot of the positive 563 00:19:54,595 --> 00:19:57,700 responses we've gotten from from patients of our 564 00:19:57,700 --> 00:19:59,940 services really come from the stellar care that 565 00:19:59,940 --> 00:20:02,659 they've that they've had independent of how easy 566 00:20:02,659 --> 00:20:04,900 and, you know, seamless the integration of the 567 00:20:04,900 --> 00:20:06,919 virtual care experience truly was. 568 00:20:08,019 --> 00:20:10,259 That's amazing to hear. Doctor Terabishi, thank you 569 00:20:10,259 --> 00:20:11,744 so much for joining us on the podcast 570 00:20:11,744 --> 00:20:13,184 today. This has been such a fun and 571 00:20:13,184 --> 00:20:15,345 informative conversation, and I look forward to connecting 572 00:20:15,345 --> 00:20:17,345 with you again soon. Yeah. Thank you for 573 00:20:17,345 --> 00:20:18,005 having me.