1 00:00:00,080 --> 00:00:02,399 This is Rosie Talago with the Becker's Healthcare 2 00:00:02,399 --> 00:00:02,899 podcast. 3 00:00:03,359 --> 00:00:05,440 I'm thrilled today to be joined by doctor 4 00:00:05,440 --> 00:00:08,900 Eric Chang, chief medical informatics officer of UCLA 5 00:00:08,960 --> 00:00:09,699 Health Sciences 6 00:00:10,160 --> 00:00:12,480 and professor in the department of neurology at 7 00:00:12,480 --> 00:00:13,139 the David 8 00:00:13,439 --> 00:00:15,619 Geffen School of Medicine at UCLA. 9 00:00:16,125 --> 00:00:17,725 Doctor Chang, it's a pleasure to have you 10 00:00:17,725 --> 00:00:18,945 on the podcast today. 11 00:00:19,324 --> 00:00:20,704 Yeah. Thanks for the invitation. 12 00:00:21,244 --> 00:00:23,804 We have a great conversation in store, diving 13 00:00:23,804 --> 00:00:25,564 into some of the most exciting trends in 14 00:00:25,564 --> 00:00:28,364 health care, the innovations happening at UCLA Health, 15 00:00:28,364 --> 00:00:30,980 and how leaders like today's guests are thinking 16 00:00:30,980 --> 00:00:33,059 about growth in the year ahead. Before we 17 00:00:33,059 --> 00:00:35,059 get into it, doctor Chang, can you please 18 00:00:35,059 --> 00:00:36,820 introduce yourself and tell us a bit about 19 00:00:36,820 --> 00:00:37,479 your background? 20 00:00:38,340 --> 00:00:41,619 Oh, yeah. Sure. So what first, again, thanks 21 00:00:41,619 --> 00:00:44,259 for the invitation. I've really enjoyed listening to 22 00:00:44,259 --> 00:00:46,395 the podcast of some of my friends and 23 00:00:46,395 --> 00:00:48,954 former coworkers. I'll just call out Mike Pfeffer 24 00:00:48,954 --> 00:00:51,355 at Stanford and Clara Lynn at Seattle Children's 25 00:00:51,355 --> 00:00:52,655 who have done your podcast. 26 00:00:53,195 --> 00:00:56,335 I've interviewed some applicants for our clinical informatics 27 00:00:56,395 --> 00:00:58,575 fellowship program, and many describe 28 00:00:59,560 --> 00:01:02,520 a, like, childhood hobby in computing or an 29 00:01:02,520 --> 00:01:03,579 engineering background, 30 00:01:04,040 --> 00:01:05,959 but I don't have that. So I'll describe 31 00:01:05,959 --> 00:01:08,120 my background in a bit more detail just 32 00:01:08,120 --> 00:01:10,620 to show there's another path to my role. 33 00:01:11,079 --> 00:01:11,579 So 34 00:01:12,055 --> 00:01:15,015 I developed an interest in outcomes research during 35 00:01:15,015 --> 00:01:16,314 my neurology residency. 36 00:01:17,094 --> 00:01:19,754 UCLA is a national leader in health services 37 00:01:19,814 --> 00:01:21,894 research, so I did a fellowship in that 38 00:01:21,894 --> 00:01:24,215 area. And I spent over a decade as 39 00:01:24,215 --> 00:01:25,674 a grant supported 40 00:01:26,140 --> 00:01:26,880 health services researcher 41 00:01:27,500 --> 00:01:29,120 measuring quality of care 42 00:01:29,420 --> 00:01:31,359 for persons of neurologic conditions. 43 00:01:31,900 --> 00:01:34,799 And I also volunteer for my specialty society 44 00:01:35,020 --> 00:01:35,680 in developing 45 00:01:36,060 --> 00:01:38,240 quality measures for regulatory programs. 46 00:01:39,100 --> 00:01:39,840 I had 47 00:01:40,924 --> 00:01:42,465 ventured into informatics 48 00:01:43,244 --> 00:01:44,864 even though I really didn't know, 49 00:01:45,564 --> 00:01:47,104 what that word meant. 50 00:01:47,405 --> 00:01:48,704 So I'll give a brief, 51 00:01:49,084 --> 00:01:52,284 example. So we were analyzing the VA national 52 00:01:52,284 --> 00:01:54,704 database and whether blood pressures were 53 00:01:55,129 --> 00:01:57,549 well controlled the year after a stroke, 54 00:01:57,930 --> 00:01:59,950 and we showed that a stroke patient's 55 00:02:00,250 --> 00:02:00,750 antihypertensive 56 00:02:01,209 --> 00:02:01,709 regimen 57 00:02:02,250 --> 00:02:04,269 frequently wasn't intensified 58 00:02:04,810 --> 00:02:06,569 even though the recorded blood pressure in the 59 00:02:06,569 --> 00:02:07,629 clinic was high. 60 00:02:08,009 --> 00:02:08,989 But then we 61 00:02:09,444 --> 00:02:11,685 reviewed some charts, and we read that many 62 00:02:11,685 --> 00:02:13,064 physicians would remeasure 63 00:02:13,685 --> 00:02:16,004 a high blood pressure. They would document that 64 00:02:16,004 --> 00:02:18,324 value, which is typically a lower value, as 65 00:02:18,324 --> 00:02:19,705 free text in their notes. 66 00:02:20,165 --> 00:02:21,844 So at that time, I call that an 67 00:02:21,844 --> 00:02:22,344 administrative 68 00:02:22,724 --> 00:02:25,569 database error, but that's actually informatics. 69 00:02:25,870 --> 00:02:28,610 To fully understand data on the back end, 70 00:02:28,830 --> 00:02:30,509 you need to understand how it's entered on 71 00:02:30,509 --> 00:02:33,569 the front end. So using informatics vocabulary, 72 00:02:34,349 --> 00:02:37,229 there wasn't a workflow for physicians who enter 73 00:02:37,229 --> 00:02:39,969 blood pressure values in the same data field 74 00:02:40,245 --> 00:02:43,044 where the automated blood pressure machines were transmitting 75 00:02:43,044 --> 00:02:45,465 it. So when UCLA purchased an EHR, 76 00:02:45,925 --> 00:02:48,405 they sought some physicians who could help configure 77 00:02:48,405 --> 00:02:50,485 it. I did my clinical work at the 78 00:02:50,485 --> 00:02:52,564 VA, and we had implemented an EHR for 79 00:02:52,564 --> 00:02:53,064 years. 80 00:02:53,780 --> 00:02:56,420 UCLA also sought someone who could help interpret 81 00:02:56,420 --> 00:02:57,080 the federal 82 00:02:57,620 --> 00:03:00,340 regulations about meaningful use, and I was pretty 83 00:03:00,340 --> 00:03:02,740 familiar with government regulatory programs based on my 84 00:03:02,740 --> 00:03:04,120 volunteer work. So 85 00:03:04,500 --> 00:03:05,700 I guess they thought it was a good 86 00:03:05,700 --> 00:03:08,504 fit. It's been said, like, many times that 87 00:03:08,504 --> 00:03:09,405 life doesn't 88 00:03:09,705 --> 00:03:12,504 always make sense looking forward, and these aren't 89 00:03:12,504 --> 00:03:14,745 the career choices I would recommend to someone 90 00:03:14,745 --> 00:03:16,125 who would like to be a CMIO. 91 00:03:16,905 --> 00:03:18,125 But looking backwards, 92 00:03:18,479 --> 00:03:20,259 yeah, I can see how these experiences 93 00:03:20,560 --> 00:03:22,239 led to my position and my way of 94 00:03:22,239 --> 00:03:22,739 thinking. 95 00:03:23,840 --> 00:03:26,560 Absolutely. That's so interesting, and it's great to 96 00:03:26,560 --> 00:03:29,120 see another path, like you mentioned, opposed to 97 00:03:29,120 --> 00:03:31,699 that more exclusively or assumed 98 00:03:32,000 --> 00:03:34,180 engineering technical side of things. 99 00:03:34,575 --> 00:03:36,495 And I found it very interesting that you 100 00:03:36,495 --> 00:03:40,014 said you veered into informatics without truly knowing 101 00:03:40,014 --> 00:03:41,455 exactly what that word meant. 102 00:03:42,335 --> 00:03:44,415 But it's sounds like you bring a very 103 00:03:44,415 --> 00:03:46,835 valuable combination to the table, especially 104 00:03:47,375 --> 00:03:48,835 in today's health care environment. 105 00:03:50,110 --> 00:03:51,889 Yeah. It's it's a so there's, 106 00:03:52,430 --> 00:03:54,689 again, not a single path. And 107 00:03:55,469 --> 00:03:57,650 maybe one of the reasons why I was 108 00:03:57,870 --> 00:04:00,209 elected to be one of the first physician 109 00:04:00,349 --> 00:04:00,849 informaticists 110 00:04:01,229 --> 00:04:03,485 was that I was a little different from 111 00:04:03,485 --> 00:04:05,724 the other physician informaticists who did have a 112 00:04:05,724 --> 00:04:06,625 different background, 113 00:04:07,004 --> 00:04:09,644 maybe more business or may maybe more computer 114 00:04:09,644 --> 00:04:10,544 science related. 115 00:04:11,405 --> 00:04:13,185 Yeah. So with your 116 00:04:13,485 --> 00:04:14,669 new or fresh perspective 117 00:04:15,789 --> 00:04:17,189 perspective in mind, I'd love to hear how 118 00:04:17,189 --> 00:04:18,129 you're viewing the current 119 00:04:18,750 --> 00:04:20,589 landscape. So what are the top three trends 120 00:04:20,589 --> 00:04:22,850 that you're following in health care today? 121 00:04:23,949 --> 00:04:25,329 Okay. Well well, 122 00:04:25,870 --> 00:04:28,529 answer number one has got to be artificial 123 00:04:28,589 --> 00:04:31,865 intelligence, of course. Right? There's so much that 124 00:04:31,865 --> 00:04:34,444 people have said, and I follow, 125 00:04:35,225 --> 00:04:37,704 you know, the developments of both excitement and 126 00:04:37,704 --> 00:04:38,204 skepticism. 127 00:04:38,504 --> 00:04:40,204 I'll I'll just make this brief comment. 128 00:04:40,665 --> 00:04:43,464 Based on my health services research days, I 129 00:04:43,464 --> 00:04:44,204 know that 130 00:04:44,665 --> 00:04:45,564 there are 131 00:04:46,259 --> 00:04:46,759 methodologic 132 00:04:47,060 --> 00:04:50,500 flaws, especially like self selection bias, that can 133 00:04:50,500 --> 00:04:54,599 make it difficult to interpret observational studies. 134 00:04:55,139 --> 00:04:57,459 So when you really need confidence in an 135 00:04:57,459 --> 00:04:57,959 answer, 136 00:04:58,419 --> 00:05:00,439 there isn't a substitute for randomization. 137 00:05:01,379 --> 00:05:04,475 So I'll describe a project here for Ambient 138 00:05:04,475 --> 00:05:08,475 Scribes, very hot topic. We randomized 250 139 00:05:08,475 --> 00:05:09,694 physicians to 140 00:05:10,154 --> 00:05:12,634 two vendors and a control arm and then 141 00:05:12,634 --> 00:05:14,574 crossed them over at set intervals. 142 00:05:15,115 --> 00:05:16,175 So then we 143 00:05:16,789 --> 00:05:18,870 got some insights on how they were the 144 00:05:18,870 --> 00:05:21,610 vendors were similar and different from each other 145 00:05:21,829 --> 00:05:23,990 and when they did or did not differ 146 00:05:23,990 --> 00:05:25,449 from the control group. 147 00:05:26,069 --> 00:05:28,569 The overhead to conduct a randomized trial 148 00:05:29,414 --> 00:05:31,974 is too high to run for every AI 149 00:05:31,974 --> 00:05:32,474 implementation. 150 00:05:33,334 --> 00:05:35,735 But when a project has a, you know, 151 00:05:35,735 --> 00:05:38,954 potential seven digit price tag, it is worth 152 00:05:39,334 --> 00:05:40,634 slowing down the implementation 153 00:05:41,189 --> 00:05:42,949 just a bit so they can study it 154 00:05:42,949 --> 00:05:43,449 properly. 155 00:05:43,830 --> 00:05:46,389 So I definitely encourage others to do that 156 00:05:46,389 --> 00:05:47,050 as well. 157 00:05:47,589 --> 00:05:48,410 The second, 158 00:05:49,189 --> 00:05:51,050 I think, thing I'm following 159 00:05:51,350 --> 00:05:51,850 is 160 00:05:52,470 --> 00:05:53,689 related to AI. 161 00:05:54,230 --> 00:05:56,069 So I'm cheating a little bit, but I'll 162 00:05:56,069 --> 00:05:56,504 call it, 163 00:05:57,464 --> 00:05:57,964 systematic 164 00:05:58,264 --> 00:05:59,084 data collection. 165 00:05:59,545 --> 00:06:00,845 So there's a well known 166 00:06:02,105 --> 00:06:02,605 epidemiological, 167 00:06:03,384 --> 00:06:06,524 like, study that shows that about twenty percent 168 00:06:06,584 --> 00:06:07,645 of early mortality 169 00:06:08,264 --> 00:06:10,125 is attributable to health care, 170 00:06:10,425 --> 00:06:11,725 and the rest is attributed 171 00:06:12,079 --> 00:06:12,980 to factors 172 00:06:13,360 --> 00:06:14,420 we don't systematically 173 00:06:14,800 --> 00:06:18,000 collect, such as patient behaviors, social drivers of 174 00:06:18,000 --> 00:06:20,819 health, the built environments, and genetics. 175 00:06:21,439 --> 00:06:23,680 So I'll tell another story from my research 176 00:06:23,680 --> 00:06:26,944 days. We were wrapping up a trial. We 177 00:06:26,944 --> 00:06:27,925 had some difficulty 178 00:06:28,225 --> 00:06:31,524 getting the final survey results from some participants. 179 00:06:32,225 --> 00:06:33,444 So I asked a statistician, 180 00:06:34,144 --> 00:06:35,845 can't we run some fancy 181 00:06:36,305 --> 00:06:36,805 imputation 182 00:06:37,185 --> 00:06:39,605 imputation methods to overcome this? 183 00:06:40,250 --> 00:06:41,629 And he looked at me and said, 184 00:06:42,330 --> 00:06:43,230 you know, imputation 185 00:06:44,009 --> 00:06:46,110 is certainly better than no imputation, 186 00:06:46,730 --> 00:06:49,290 but there is no substitute for primary data 187 00:06:49,290 --> 00:06:49,790 collection. 188 00:06:50,170 --> 00:06:52,509 So likewise right now, if we don't collect 189 00:06:52,569 --> 00:06:53,230 the right 190 00:06:53,745 --> 00:06:54,245 predictors, 191 00:06:55,024 --> 00:06:58,004 then the outcomes will be inaccurate due to 192 00:06:58,225 --> 00:07:02,004 omitted variable bias, and AI can't overcome that. 193 00:07:02,145 --> 00:07:04,705 What AI does is generate better predictions for 194 00:07:04,705 --> 00:07:07,105 the data that we already have. And to 195 00:07:07,105 --> 00:07:09,205 fully leverage it, we just need 196 00:07:09,769 --> 00:07:12,810 accurate and more types of data. So I 197 00:07:12,810 --> 00:07:14,509 view that as complementary 198 00:07:14,810 --> 00:07:17,370 or kind of almost more fundamental to get 199 00:07:17,370 --> 00:07:19,550 AI to to work at its best. 200 00:07:20,089 --> 00:07:22,490 And then the third topic, I'll mention a 201 00:07:22,490 --> 00:07:23,949 topic that maybe 202 00:07:24,495 --> 00:07:26,675 no one else has mentioned in your podcast. 203 00:07:27,294 --> 00:07:28,834 But in 2025, 204 00:07:29,055 --> 00:07:32,095 it's gotta be government regulations. Right? It's it's, 205 00:07:32,334 --> 00:07:35,134 it's the world we live in. CMIOs have 206 00:07:35,134 --> 00:07:37,535 always needed to monitor this space, you know, 207 00:07:37,535 --> 00:07:40,410 meaningful use, the twenty first Century Cures Act, 208 00:07:40,410 --> 00:07:41,069 there's TEFCA, 209 00:07:41,689 --> 00:07:43,149 so all the quality measures. 210 00:07:44,410 --> 00:07:46,250 But it's not just at the federal level, 211 00:07:46,250 --> 00:07:48,189 but it's at the state level as well. 212 00:07:48,330 --> 00:07:51,230 So I'll give an example. So the federal 213 00:07:51,290 --> 00:07:53,995 twenty first Century Cures Act states that results 214 00:07:53,995 --> 00:07:56,654 need to be made available to patients 215 00:07:57,274 --> 00:07:58,175 without delay, 216 00:07:58,714 --> 00:08:01,435 but our state government passed a law stating 217 00:08:01,435 --> 00:08:03,855 that if you have radiology results that show 218 00:08:03,915 --> 00:08:05,454 new or recurrent malignancy, 219 00:08:06,449 --> 00:08:08,850 that should be verbally disclosed instead of through 220 00:08:08,850 --> 00:08:09,509 a portal. 221 00:08:09,970 --> 00:08:12,069 So we developed a process for the radiologist 222 00:08:12,209 --> 00:08:14,769 to manually tag such scans so they are 223 00:08:14,769 --> 00:08:17,269 held from the portal for a short time 224 00:08:17,569 --> 00:08:19,110 so that the ordering provider 225 00:08:19,595 --> 00:08:21,435 has the opportunity to reach out to the 226 00:08:21,435 --> 00:08:21,935 patients. 227 00:08:22,954 --> 00:08:24,894 But because that process is inconsistent, 228 00:08:25,514 --> 00:08:28,394 we just supplemented it with a homegrown machine 229 00:08:28,394 --> 00:08:29,294 learning algorithm 230 00:08:29,834 --> 00:08:32,654 that analyzes the text of the radiologist report 231 00:08:32,714 --> 00:08:33,294 to determine 232 00:08:33,940 --> 00:08:35,240 whether it meets a criteria 233 00:08:35,620 --> 00:08:37,540 mentioned in state law and then holds it 234 00:08:37,540 --> 00:08:38,920 back from the patient portal. 235 00:08:39,379 --> 00:08:41,379 Now did I need a state law to 236 00:08:41,379 --> 00:08:43,700 tell me that informing patients of a new 237 00:08:43,700 --> 00:08:45,879 diagnosis of cancer through the portal 238 00:08:46,420 --> 00:08:46,920 is 239 00:08:47,304 --> 00:08:48,524 now potentially problematic? 240 00:08:48,904 --> 00:08:50,125 No. But 241 00:08:50,585 --> 00:08:52,365 I couldn't champion and prioritize 242 00:08:52,745 --> 00:08:54,605 the developments of a machine learning 243 00:08:55,065 --> 00:08:55,565 algorithm 244 00:08:56,024 --> 00:08:57,545 if I didn't have the weight of the 245 00:08:57,545 --> 00:09:00,125 state law behind me. So it is 246 00:09:01,200 --> 00:09:04,320 important to follow, like, what's going on in 247 00:09:04,320 --> 00:09:05,379 government and 248 00:09:05,759 --> 00:09:08,320 possible, you know, really leverage it to drive 249 00:09:08,320 --> 00:09:08,820 change. 250 00:09:09,840 --> 00:09:10,340 Absolutely. 251 00:09:10,960 --> 00:09:14,754 I particularly liked your example of the connection 252 00:09:14,754 --> 00:09:16,934 between that systematic data collection 253 00:09:17,794 --> 00:09:20,455 and artificial intelligence and machine learning 254 00:09:21,075 --> 00:09:22,375 tools such as those. 255 00:09:23,075 --> 00:09:24,434 I feel like a lot of times it's 256 00:09:24,434 --> 00:09:27,410 easy to think of this artificial intelligence as 257 00:09:27,410 --> 00:09:28,230 this higher 258 00:09:28,690 --> 00:09:31,410 higher thing that's all knowing and super smart, 259 00:09:31,410 --> 00:09:32,389 which it is. 260 00:09:33,009 --> 00:09:35,330 But we gotta pay attention to the data 261 00:09:35,330 --> 00:09:37,330 that we are feeding it because at the 262 00:09:37,330 --> 00:09:38,710 end of the day, like you mentioned, 263 00:09:39,170 --> 00:09:40,070 if we're omitting 264 00:09:40,450 --> 00:09:40,950 variables, 265 00:09:41,705 --> 00:09:44,585 that's bias, and it's very important for the 266 00:09:44,585 --> 00:09:45,644 results that it produces. 267 00:09:46,825 --> 00:09:47,325 Yeah. 268 00:09:47,945 --> 00:09:50,044 People, yeah, others have commented 269 00:09:50,585 --> 00:09:52,664 it as well in a maybe a slightly 270 00:09:52,664 --> 00:09:54,745 different way. They point to the quality of 271 00:09:54,745 --> 00:09:56,284 the data. If we 272 00:09:57,139 --> 00:10:00,360 if the notes are somehow inaccurate, then certainly 273 00:10:00,420 --> 00:10:03,379 the predictions are inaccurate. But in addition to 274 00:10:03,379 --> 00:10:05,139 the quality of data, I think the types 275 00:10:05,139 --> 00:10:07,559 of data makes, should be expanded. 276 00:10:08,420 --> 00:10:08,920 Absolutely. 277 00:10:09,894 --> 00:10:12,394 So zooming in a bit now, UCLA 278 00:10:12,774 --> 00:10:14,774 is known for being on the leading edge 279 00:10:14,774 --> 00:10:15,434 of innovation. 280 00:10:15,815 --> 00:10:17,815 So I'm curious to know what are you 281 00:10:17,815 --> 00:10:20,154 most excited about right now at UCLA? 282 00:10:21,894 --> 00:10:25,059 So I've listened to this podcast, and a 283 00:10:25,059 --> 00:10:26,519 lot of people mention, 284 00:10:27,860 --> 00:10:30,120 burnout reduction, and I agree. 285 00:10:30,659 --> 00:10:34,120 And people point to several different factors, typically 286 00:10:34,340 --> 00:10:37,799 patient message volume and pajama time and EHR 287 00:10:37,860 --> 00:10:38,360 usability. 288 00:10:39,595 --> 00:10:41,054 And I agree with that too. 289 00:10:41,595 --> 00:10:43,995 But I wanna highlight maybe a personal interest 290 00:10:43,995 --> 00:10:46,095 of mine. It doesn't have a financial, 291 00:10:47,034 --> 00:10:49,274 ROI, but I still think it's just as 292 00:10:49,274 --> 00:10:49,774 important. 293 00:10:50,475 --> 00:10:52,154 I think one of the reasons why it's 294 00:10:52,154 --> 00:10:52,654 so 295 00:10:53,250 --> 00:10:54,549 cognitively taxing 296 00:10:55,090 --> 00:10:56,309 to use an EHR 297 00:10:56,769 --> 00:10:58,610 is because the notes are just too hard 298 00:10:58,610 --> 00:10:59,269 to read. 299 00:10:59,809 --> 00:11:01,509 They're long. They're repetitive. 300 00:11:02,850 --> 00:11:05,350 It mixes both past and current data. 301 00:11:05,815 --> 00:11:08,375 The information that's useful to the author is 302 00:11:08,375 --> 00:11:11,254 frequently not relevant to the reader. But but 303 00:11:11,254 --> 00:11:13,175 I get it. You know, the clinic has 304 00:11:13,175 --> 00:11:14,154 such time pressure. 305 00:11:14,535 --> 00:11:16,695 It's faster to start from a template or 306 00:11:16,695 --> 00:11:18,795 your prior notes than a blank page. 307 00:11:19,415 --> 00:11:22,750 Yet despite its long length, it doesn't always 308 00:11:22,750 --> 00:11:24,690 capture the essence of the patient. 309 00:11:25,470 --> 00:11:27,809 Like many medical students who chose neurology, 310 00:11:28,350 --> 00:11:30,850 you know, I was drawn by Oliver Sacks, 311 00:11:31,710 --> 00:11:32,929 narratives of patients. 312 00:11:33,629 --> 00:11:35,789 So when I'm reading a note now, I'm 313 00:11:35,789 --> 00:11:39,184 not expecting literature, you know, but it is 314 00:11:39,184 --> 00:11:41,345 hard to hear the patient's voice in many 315 00:11:41,345 --> 00:11:43,845 notes. So if the Ambient scribe 316 00:11:44,225 --> 00:11:46,245 can capture the essence of the conversation 317 00:11:46,784 --> 00:11:49,584 while the upcoming, like, summarization tools can bring 318 00:11:49,584 --> 00:11:51,824 in the relevant parts of the note, that 319 00:11:51,824 --> 00:11:54,600 frees the physician to focus on the activities 320 00:11:54,600 --> 00:11:56,519 we were trained to do, and people cite 321 00:11:56,519 --> 00:11:58,299 that as a reason for better satisfaction. 322 00:11:59,080 --> 00:12:00,379 But on top of that, 323 00:12:00,840 --> 00:12:02,919 it could lead to a better experience for 324 00:12:02,919 --> 00:12:05,559 the reader as well. We could achieve both 325 00:12:05,559 --> 00:12:06,059 a 326 00:12:06,434 --> 00:12:08,934 detailed rich note yet relatively 327 00:12:09,315 --> 00:12:09,815 concise. 328 00:12:10,595 --> 00:12:13,154 It's so nice to read a well written 329 00:12:13,154 --> 00:12:14,914 note, and I like to find a way 330 00:12:14,914 --> 00:12:15,975 to make that easier. 331 00:12:16,914 --> 00:12:18,754 That's a great point. I do hear a 332 00:12:18,754 --> 00:12:21,174 lot about how it can streamline the process 333 00:12:21,235 --> 00:12:21,735 for 334 00:12:22,379 --> 00:12:24,620 the providers and the nurses and the doctors, 335 00:12:24,620 --> 00:12:27,100 but that's a very interesting perspective to bring 336 00:12:27,100 --> 00:12:29,179 up for the reader and the patient bringing 337 00:12:29,179 --> 00:12:32,220 the patient's voice into the notes too. Very 338 00:12:32,220 --> 00:12:32,720 interesting. 339 00:12:33,660 --> 00:12:36,394 Yeah. There there's a tradition. Neurology at these 340 00:12:36,394 --> 00:12:37,215 long form, 341 00:12:38,955 --> 00:12:41,855 back to Luria and Sachs of, like, describing 342 00:12:42,235 --> 00:12:44,634 kind of how how a patient is, you 343 00:12:44,634 --> 00:12:45,455 know, thinking 344 00:12:46,154 --> 00:12:47,295 and doing. And, 345 00:12:47,750 --> 00:12:50,870 yeah, it's maybe impractical in everyday use, but 346 00:12:50,870 --> 00:12:53,350 I do I do enjoy reading things like 347 00:12:53,350 --> 00:12:53,850 that. 348 00:12:54,709 --> 00:12:55,209 Absolutely. 349 00:12:55,909 --> 00:12:59,350 And finally, looking ahead, growth is always on 350 00:12:59,350 --> 00:13:02,629 everyone's mind, especially in this industry that's adapting 351 00:13:02,629 --> 00:13:04,754 to new challenges and opportunities 352 00:13:05,134 --> 00:13:07,855 so quickly. So I'm wondering how you are 353 00:13:07,855 --> 00:13:10,514 thinking about growth over the next twelve months. 354 00:13:12,095 --> 00:13:12,595 So 355 00:13:13,215 --> 00:13:15,134 I'll think of it from the perspective of 356 00:13:15,134 --> 00:13:15,875 the organization. 357 00:13:17,054 --> 00:13:18,514 UCLA has always 358 00:13:19,110 --> 00:13:21,769 seems to be expanding their outpatient footprints, 359 00:13:22,389 --> 00:13:23,370 but they've recently 360 00:13:23,990 --> 00:13:26,169 expanded in other ways as well. 361 00:13:26,789 --> 00:13:29,129 I've been at UCLA now for thirty years, 362 00:13:29,350 --> 00:13:29,850 and 363 00:13:30,230 --> 00:13:32,970 we purchased new hospitals for the first time 364 00:13:33,215 --> 00:13:35,054 in the in the past year since I've, 365 00:13:35,455 --> 00:13:36,674 since I've been here. 366 00:13:36,975 --> 00:13:39,934 We just implemented our EHR there. We're gonna 367 00:13:39,934 --> 00:13:42,754 open a new psychiatric hospital next year. 368 00:13:43,375 --> 00:13:46,679 We're implementing our EHR in our university student 369 00:13:46,679 --> 00:13:48,299 health center this summer. 370 00:13:49,000 --> 00:13:50,459 We just rolled out our 371 00:13:50,839 --> 00:13:52,139 managed care medical 372 00:13:52,440 --> 00:13:54,139 Medicare Advantage health plan. 373 00:13:54,600 --> 00:13:55,980 So we're expanding 374 00:13:56,440 --> 00:13:56,940 into 375 00:13:57,720 --> 00:14:00,539 lots of different areas other than opening up 376 00:14:00,945 --> 00:14:01,845 outpatient clinics. 377 00:14:02,625 --> 00:14:04,004 So I think that's probably 378 00:14:04,625 --> 00:14:07,424 a main kind of focus of activity. But 379 00:14:07,424 --> 00:14:08,965 even if we weren't expanding, 380 00:14:09,504 --> 00:14:10,945 there's a lot of work to keep up 381 00:14:10,945 --> 00:14:13,924 with, you know, developments from our EHR vendor, 382 00:14:14,065 --> 00:14:16,799 from other third party vendors, and from our 383 00:14:16,799 --> 00:14:18,799 own community as well. And I I guess 384 00:14:18,799 --> 00:14:21,299 I wanna mention that. An academic medical center 385 00:14:21,919 --> 00:14:23,919 is in the position to build some of 386 00:14:23,919 --> 00:14:25,919 their own tools, especially when it comes to 387 00:14:25,919 --> 00:14:26,419 AI. 388 00:14:26,879 --> 00:14:28,820 And sometimes that's led by 389 00:14:29,165 --> 00:14:31,485 the IT organization itself, but sometimes it's led 390 00:14:31,485 --> 00:14:32,625 by our own faculty. 391 00:14:34,524 --> 00:14:36,365 I'll say it's a lot. I just wanna 392 00:14:36,365 --> 00:14:38,045 add that no one can do this alone. 393 00:14:38,045 --> 00:14:39,024 There are great people 394 00:14:39,565 --> 00:14:42,524 under me, next to me, above me, whom 395 00:14:42,524 --> 00:14:43,504 I interact in 396 00:14:44,100 --> 00:14:46,419 everyday basis, and that is one of the 397 00:14:46,419 --> 00:14:48,899 most enjoyable aspects of my job. There's kind 398 00:14:48,899 --> 00:14:50,500 of a lot of variety of the things 399 00:14:50,500 --> 00:14:51,879 that I'm exposed to. 400 00:14:52,659 --> 00:14:54,899 That's wonderful. And that's very cool to see 401 00:14:54,899 --> 00:14:57,134 that you've been there for so many years, 402 00:14:57,134 --> 00:14:59,375 but you just now are expanding in new 403 00:14:59,375 --> 00:15:00,674 ways you haven't seen, 404 00:15:01,134 --> 00:15:03,375 aside from the outpatient that you've been doing. 405 00:15:03,375 --> 00:15:05,855 But seeing this expansion and growth in new 406 00:15:05,855 --> 00:15:08,115 areas is very exciting and 407 00:15:08,440 --> 00:15:11,259 also very exciting to think of that possibility 408 00:15:11,320 --> 00:15:14,540 or the capability of an academic medical center 409 00:15:15,240 --> 00:15:17,180 to have faculty led 410 00:15:17,800 --> 00:15:18,620 new tools, 411 00:15:19,399 --> 00:15:20,920 be in that position to build some of 412 00:15:20,920 --> 00:15:23,019 their own tools is very, very exciting. 413 00:15:24,174 --> 00:15:26,914 Yeah. I think because of that, we are 414 00:15:27,054 --> 00:15:30,034 in the process of hiring a chief health 415 00:15:30,735 --> 00:15:32,355 AI AI officer to help 416 00:15:32,815 --> 00:15:34,894 kind of what's the best way to put 417 00:15:34,894 --> 00:15:37,075 it? Maybe organize some of those activities. 418 00:15:37,610 --> 00:15:39,290 You know, faculty, one of their strengths is 419 00:15:39,290 --> 00:15:40,190 that they are, 420 00:15:40,490 --> 00:15:43,529 you know, fiercely independent, but it it would 421 00:15:43,529 --> 00:15:46,170 help to have some way to connect them 422 00:15:46,170 --> 00:15:48,170 all together. And I think that's that'll be 423 00:15:48,170 --> 00:15:48,990 one of the 424 00:15:49,529 --> 00:15:51,470 functions of this, new, 425 00:15:52,009 --> 00:15:52,509 person. 426 00:15:53,714 --> 00:15:56,274 Absolutely. Like you said, you have great people 427 00:15:56,274 --> 00:15:57,634 all around you. No one can do it 428 00:15:57,634 --> 00:16:01,394 alone. So bringing all those great powerful minds 429 00:16:01,394 --> 00:16:03,174 together to do it together. 430 00:16:03,634 --> 00:16:04,134 Yeah. 431 00:16:04,514 --> 00:16:06,274 Well, thank you very much. That is all 432 00:16:06,274 --> 00:16:07,714 the time we have for today, but I 433 00:16:07,714 --> 00:16:10,459 wanna thank you, doctor Chang, for sharing such 434 00:16:10,459 --> 00:16:12,159 wonderful insights and experiences 435 00:16:12,539 --> 00:16:14,480 gained from your experience at UCLA. 436 00:16:15,019 --> 00:16:17,019 It's been really fascinating to hear about all 437 00:16:17,019 --> 00:16:19,659 the wonderful things that are underway, and we 438 00:16:19,659 --> 00:16:22,159 look forward to collaborating with you again soon. 439 00:16:23,151 --> 00:16:25,411 Thank you. So look forward to that.