1 00:00:02,240 --> 00:00:02,899 At athenahealth, 2 00:00:03,279 --> 00:00:06,480 we know your ambulatory practice wants healthier, a 3 00:00:06,480 --> 00:00:09,779 healthier business, healthier care teams, and healthier patients. 4 00:00:10,160 --> 00:00:12,480 But the complexities of modern health care tech 5 00:00:12,480 --> 00:00:14,240 make it hard for you and your care 6 00:00:14,240 --> 00:00:15,855 teams to focus on what matters 7 00:00:16,414 --> 00:00:19,295 most? That's where athenahealth can help. Our AI 8 00:00:19,295 --> 00:00:22,594 native all in one solutions reduce administrative burdens, 9 00:00:22,894 --> 00:00:25,875 streamline billing and payments, and deliver critical insights 10 00:00:25,934 --> 00:00:28,734 when clinicians need it most. That means fewer 11 00:00:28,734 --> 00:00:31,614 clicks, more time for patients, and stronger bottom 12 00:00:31,614 --> 00:00:31,829 lines. 13 00:00:32,710 --> 00:00:35,989 Practicing medicine is complex, but running a practice 14 00:00:35,989 --> 00:00:37,850 can be that much simpler with athenahealth. 15 00:00:38,549 --> 00:00:42,170 See how simpler is healthier@athenahealth.com. 16 00:00:43,925 --> 00:00:45,765 Hello, everyone. This is Scott King with the 17 00:00:45,765 --> 00:00:47,065 Becker's Healthcare podcast. 18 00:00:47,445 --> 00:00:49,865 Thrilled today to be joined by Brenna Lofek, 19 00:00:50,405 --> 00:00:53,765 director AI regulatory and quality center for digital 20 00:00:53,765 --> 00:00:55,685 health at Mayo Clinic. Brenna, how are you 21 00:00:55,685 --> 00:00:58,159 doing today? I'm doing great. Happy to be 22 00:00:58,159 --> 00:01:00,079 here. Thanks so much for joining us. I 23 00:01:00,079 --> 00:01:01,280 I we have a lot of big, you 24 00:01:01,280 --> 00:01:03,759 know, tech questions and and dealing with health 25 00:01:03,759 --> 00:01:05,200 care to get health care to get to. 26 00:01:05,200 --> 00:01:06,799 But, if you can please just share a 27 00:01:06,799 --> 00:01:08,640 little bit about your background and your work, 28 00:01:08,799 --> 00:01:09,780 that would be great. 29 00:01:10,305 --> 00:01:10,805 Certainly. 30 00:01:11,424 --> 00:01:14,245 I lead AI regulatory and quality 31 00:01:14,545 --> 00:01:17,265 strategy at Mayo Clinic in the Center for 32 00:01:17,265 --> 00:01:18,084 Digital Health, 33 00:01:18,465 --> 00:01:22,165 where our mission is to safely translate innovation 34 00:01:22,545 --> 00:01:23,844 into clinical practice. 35 00:01:24,569 --> 00:01:27,609 My work centers on building the infrastructure that 36 00:01:27,609 --> 00:01:30,909 allows AI technologies to to be deployed responsibly. 37 00:01:31,769 --> 00:01:35,769 We develop governance frameworks, ensuring regulatory compliance, and 38 00:01:35,769 --> 00:01:38,909 establishing evaluation process for digital health tools 39 00:01:39,474 --> 00:01:41,494 is across the entire enterprise. 40 00:01:42,194 --> 00:01:45,334 My background spans both the medical device industry 41 00:01:45,474 --> 00:01:46,774 and health care delivery, 42 00:01:47,155 --> 00:01:50,215 which gives me this really cool dual lens 43 00:01:50,274 --> 00:01:51,734 for balancing innovation 44 00:01:52,390 --> 00:01:55,049 with patient safety and ethical responsibility. 45 00:01:55,909 --> 00:01:58,790 From an academic perspective, I received my master's 46 00:01:58,790 --> 00:02:01,990 from Johns Hopkins in patient safety and health 47 00:02:01,990 --> 00:02:02,730 care quality, 48 00:02:03,109 --> 00:02:05,685 and I'm currently a doctoral student at that 49 00:02:05,685 --> 00:02:06,424 same institution 50 00:02:07,125 --> 00:02:08,585 focusing on the intersections 51 00:02:08,885 --> 00:02:09,865 of health policy 52 00:02:10,245 --> 00:02:10,985 and AI. 53 00:02:11,764 --> 00:02:13,705 My research is related to 54 00:02:14,004 --> 00:02:14,745 the translational 55 00:02:15,044 --> 00:02:17,064 practices across health care organizations 56 00:02:17,889 --> 00:02:21,729 in applying laws, regulations, and standards to AI 57 00:02:21,729 --> 00:02:23,509 enabled digital health technologies. 58 00:02:24,930 --> 00:02:26,689 Well, that's awesome. Thanks so much for sharing 59 00:02:26,689 --> 00:02:28,849 your your great background info. I would we're 60 00:02:28,849 --> 00:02:30,229 gonna get a lot for your 61 00:02:30,534 --> 00:02:32,534 role. I wanna ask you about specifically everything 62 00:02:32,534 --> 00:02:35,014 that's going on digitally and with tech and 63 00:02:35,014 --> 00:02:35,914 in health care. 64 00:02:36,615 --> 00:02:39,495 First, I wanna you know, our our research 65 00:02:39,495 --> 00:02:41,175 has told us that nearly half of medical 66 00:02:41,175 --> 00:02:42,235 practices reporting 67 00:02:43,020 --> 00:02:44,479 using AI in some capacity 68 00:02:44,939 --> 00:02:48,060 last year and, obviously, remains a key topic 69 00:02:48,060 --> 00:02:49,360 for health IT leaders. 70 00:02:49,819 --> 00:02:52,139 From your perspective, what are the use cases 71 00:02:52,139 --> 00:02:54,620 that are making a difference right now, and 72 00:02:54,620 --> 00:02:56,560 how are you leveraging them in your organization? 73 00:02:57,324 --> 00:02:57,824 Yeah. 74 00:02:58,444 --> 00:03:00,525 So in my opinion, use of AI can 75 00:03:00,525 --> 00:03:03,424 really be broken down into two key areas 76 00:03:03,484 --> 00:03:05,025 within health care organizations. 77 00:03:06,044 --> 00:03:07,025 The first one 78 00:03:07,324 --> 00:03:11,344 are technologies that are doing administrative support related 79 00:03:11,405 --> 00:03:13,259 tasks, And then the other area 80 00:03:13,959 --> 00:03:16,299 would be tools that are doing clinical decision 81 00:03:16,599 --> 00:03:17,099 support. 82 00:03:17,879 --> 00:03:20,840 Administrative support tools are gonna be technologies that 83 00:03:20,840 --> 00:03:23,560 are helping with workflow tasks within a health 84 00:03:23,560 --> 00:03:24,299 care organization, 85 00:03:24,759 --> 00:03:27,414 but they're not directly doing any sort of 86 00:03:27,414 --> 00:03:29,114 diagnosis, treatment, cure, 87 00:03:29,414 --> 00:03:30,074 or mitigation 88 00:03:30,455 --> 00:03:33,754 of any diseases or conditions within patients. 89 00:03:34,055 --> 00:03:36,455 Whereas the other category of tools, clinical decision 90 00:03:36,455 --> 00:03:37,354 support software, 91 00:03:37,750 --> 00:03:39,610 those are doing those activities. 92 00:03:40,150 --> 00:03:41,750 And there are many use cases that are 93 00:03:41,750 --> 00:03:42,650 making a difference 94 00:03:42,949 --> 00:03:43,449 for 95 00:03:43,750 --> 00:03:44,569 the practice, 96 00:03:45,349 --> 00:03:45,849 providers, 97 00:03:46,310 --> 00:03:47,129 and patients 98 00:03:47,669 --> 00:03:49,050 in both of these categories. 99 00:03:49,750 --> 00:03:52,664 In the admin support space, we see technologies 100 00:03:53,284 --> 00:03:56,185 that are aiding in scheduling and communication, 101 00:03:57,044 --> 00:03:58,104 which automate 102 00:03:58,485 --> 00:03:58,985 previously 103 00:03:59,284 --> 00:04:00,745 manual and tedious 104 00:04:01,044 --> 00:04:01,544 tasks. 105 00:04:02,004 --> 00:04:03,224 These are immensely 106 00:04:03,525 --> 00:04:04,745 impactful because 107 00:04:05,110 --> 00:04:07,849 they help not only with time savings, 108 00:04:08,310 --> 00:04:09,849 but also with burdensome 109 00:04:10,229 --> 00:04:11,370 tasks that 110 00:04:12,229 --> 00:04:13,689 are, health care professionals 111 00:04:14,469 --> 00:04:14,969 find, 112 00:04:15,590 --> 00:04:17,610 annoying or not really worthwhile 113 00:04:17,910 --> 00:04:18,649 or beneficial 114 00:04:18,949 --> 00:04:20,410 in the actual care delivery 115 00:04:21,084 --> 00:04:23,644 actions, which are the core of all health 116 00:04:23,644 --> 00:04:24,384 care institutions' 117 00:04:25,245 --> 00:04:27,105 mission, visions, and values. 118 00:04:27,805 --> 00:04:30,605 In the clinical decision support space, we see 119 00:04:30,605 --> 00:04:31,105 technologies 120 00:04:31,485 --> 00:04:31,985 that 121 00:04:32,285 --> 00:04:33,105 are either 122 00:04:33,790 --> 00:04:34,770 directly or indirectly 123 00:04:35,230 --> 00:04:36,610 assisting health care professionals 124 00:04:37,389 --> 00:04:38,210 with diagnosis 125 00:04:38,670 --> 00:04:41,550 and treatment of their patients. And this these 126 00:04:41,550 --> 00:04:43,330 really span everything from 127 00:04:44,110 --> 00:04:45,330 tools that are 128 00:04:45,704 --> 00:04:46,444 leveraging information 129 00:04:46,824 --> 00:04:48,365 found within the electronic 130 00:04:48,665 --> 00:04:49,485 health record 131 00:04:49,865 --> 00:04:50,605 to do 132 00:04:50,985 --> 00:04:51,485 detection 133 00:04:52,024 --> 00:04:54,104 of patients that are at risk of different 134 00:04:54,104 --> 00:04:55,564 diseases or conditions. 135 00:04:56,024 --> 00:04:59,165 Maybe they would benefit from some additional screening, 136 00:05:00,024 --> 00:05:00,524 or 137 00:05:01,080 --> 00:05:03,959 interacting with information in that electronic health record 138 00:05:03,959 --> 00:05:04,860 to give recommendations 139 00:05:05,560 --> 00:05:06,060 around 140 00:05:06,520 --> 00:05:07,980 treatments or diagnoses. 141 00:05:09,560 --> 00:05:10,779 I believe that 142 00:05:11,160 --> 00:05:14,120 around sixty percent of health care professionals are 143 00:05:14,120 --> 00:05:16,220 using a tool called open evidence, 144 00:05:16,564 --> 00:05:20,185 which is assisting with different differential diagnoses happening, 145 00:05:20,805 --> 00:05:22,805 and that is just highly beneficial for health 146 00:05:22,805 --> 00:05:23,544 care professionals. 147 00:05:24,404 --> 00:05:26,584 We also see more advanced technologies 148 00:05:27,284 --> 00:05:27,784 doing, 149 00:05:28,724 --> 00:05:30,024 really brilliant things 150 00:05:30,500 --> 00:05:31,879 where they're analyzing 151 00:05:32,819 --> 00:05:33,800 medical images, 152 00:05:34,180 --> 00:05:35,720 signals, or patterns 153 00:05:36,180 --> 00:05:36,839 to do, 154 00:05:37,699 --> 00:05:39,720 early detection and diagnosis 155 00:05:40,500 --> 00:05:43,274 or to automate different workflow steps, 156 00:05:43,754 --> 00:05:45,595 that take a lot of time by our 157 00:05:45,595 --> 00:05:48,394 health care professionals or that are not always 158 00:05:48,394 --> 00:05:48,894 done 159 00:05:49,354 --> 00:05:49,854 consistently 160 00:05:50,475 --> 00:05:53,675 and these are tools that are interacting with 161 00:05:53,675 --> 00:05:55,055 images signals or patterns 162 00:05:55,914 --> 00:05:57,854 to do things like 163 00:05:58,339 --> 00:06:00,279 segment medical images 164 00:06:01,060 --> 00:06:02,439 and do calculations 165 00:06:03,139 --> 00:06:05,720 of distance or volumetric measurements. 166 00:06:06,099 --> 00:06:08,759 There are also technologies that are interacting with 167 00:06:09,060 --> 00:06:11,399 patterns or signals like in cardiology 168 00:06:12,084 --> 00:06:15,064 where at Mayo Clinic, we have tools that 169 00:06:15,284 --> 00:06:15,784 analyze 170 00:06:16,164 --> 00:06:16,664 ECGs 171 00:06:17,204 --> 00:06:18,504 to flag patients 172 00:06:18,964 --> 00:06:21,305 that may be at risk of different diseases 173 00:06:21,524 --> 00:06:22,345 or conditions. 174 00:06:22,724 --> 00:06:25,764 And so these technologies, the clinical decision support 175 00:06:25,764 --> 00:06:26,264 tools, 176 00:06:26,689 --> 00:06:29,649 really help move the needle on patient care 177 00:06:29,649 --> 00:06:32,209 delivery, and we're really excited about these tools 178 00:06:32,209 --> 00:06:33,189 at Mayo Clinic. 179 00:06:33,889 --> 00:06:35,329 Yeah. A lot of those uses you just 180 00:06:35,329 --> 00:06:37,810 mentioned were are incredibly interested. I I feel 181 00:06:37,810 --> 00:06:39,269 like we're hearing about 182 00:06:39,735 --> 00:06:40,475 new tools 183 00:06:40,854 --> 00:06:44,134 and new uses every day now. Just with 184 00:06:44,134 --> 00:06:46,454 your with your role and, I guess, what 185 00:06:46,454 --> 00:06:48,774 you've seen and maybe even what you've heard. 186 00:06:48,774 --> 00:06:51,334 Do you do you remember ever health care 187 00:06:51,334 --> 00:06:51,834 having 188 00:06:52,300 --> 00:06:54,939 so much cutting edge technology involved and and 189 00:06:54,939 --> 00:06:55,680 how important 190 00:06:56,139 --> 00:06:57,819 tech is in in this moment in time 191 00:06:57,819 --> 00:06:59,340 to health care? Like, you ever heard that 192 00:06:59,340 --> 00:07:00,480 before in the industry? 193 00:07:01,340 --> 00:07:03,920 No. I think that the paradigm has really 194 00:07:03,980 --> 00:07:06,480 changed in the last five years. 195 00:07:07,004 --> 00:07:08,305 Whereas we saw 196 00:07:08,925 --> 00:07:09,745 a lot of 197 00:07:10,525 --> 00:07:12,225 the med tech innovation 198 00:07:12,764 --> 00:07:15,264 coming out of startup companies 199 00:07:15,965 --> 00:07:17,504 or established organizations 200 00:07:17,884 --> 00:07:20,949 like Medtronic or Johnson and Johnson where they 201 00:07:20,949 --> 00:07:22,329 were the ones really developing 202 00:07:22,709 --> 00:07:23,449 the technology 203 00:07:23,990 --> 00:07:25,529 often working in collaboration 204 00:07:26,310 --> 00:07:28,569 with health care institutions and providers, 205 00:07:28,949 --> 00:07:31,750 but they were the legal manufacturers of these 206 00:07:31,750 --> 00:07:34,569 tools. They were the ones identifying the opportunity, 207 00:07:35,264 --> 00:07:37,204 and being the ones legally responsible 208 00:07:37,584 --> 00:07:38,485 for developing, 209 00:07:38,865 --> 00:07:40,004 implementing, commercializing 210 00:07:40,785 --> 00:07:43,104 these different products. But as I said, the 211 00:07:43,104 --> 00:07:46,644 paradigm has really shifted now where AI enabled 212 00:07:46,704 --> 00:07:47,204 technologies 213 00:07:48,225 --> 00:07:49,365 are in the hands 214 00:07:49,904 --> 00:07:50,644 of healthcare 215 00:07:51,029 --> 00:07:52,089 professionals themselves. 216 00:07:52,629 --> 00:07:54,569 They are able to identify 217 00:07:55,189 --> 00:07:56,170 local problems 218 00:07:56,870 --> 00:07:59,850 that could be solved or local opportunities 219 00:08:00,310 --> 00:08:01,529 that could be addressed 220 00:08:01,910 --> 00:08:04,250 with AI enabled digital health tools, 221 00:08:04,584 --> 00:08:07,384 And they, in collaboration often with with data 222 00:08:07,384 --> 00:08:07,884 scientists, 223 00:08:08,264 --> 00:08:09,884 can develop these tools 224 00:08:10,264 --> 00:08:12,444 and implement them into their practice. 225 00:08:12,904 --> 00:08:15,964 So we have this space of booming innovation, 226 00:08:16,345 --> 00:08:18,444 but also highly localized 227 00:08:19,410 --> 00:08:20,629 development of technologies 228 00:08:21,009 --> 00:08:23,089 that are solving problems that the health care 229 00:08:23,089 --> 00:08:23,589 professionals 230 00:08:23,970 --> 00:08:26,370 themselves are seeing. And this is just going 231 00:08:26,370 --> 00:08:27,029 to continue 232 00:08:27,569 --> 00:08:29,589 to grow with generative 233 00:08:29,970 --> 00:08:31,669 AI and the accessibility 234 00:08:32,684 --> 00:08:35,904 of using established large language models 235 00:08:36,285 --> 00:08:38,384 to apply it to different activities 236 00:08:38,845 --> 00:08:40,065 within your organization. 237 00:08:40,764 --> 00:08:43,325 So in many ways, this is a massive 238 00:08:43,325 --> 00:08:45,024 boom, incredible expansion, 239 00:08:45,679 --> 00:08:48,579 but it's also a real shift in the 240 00:08:48,879 --> 00:08:50,019 types of individuals 241 00:08:50,399 --> 00:08:51,139 that are 242 00:08:51,519 --> 00:08:52,579 identifying opportunities, 243 00:08:53,360 --> 00:08:54,259 building tools, 244 00:08:54,559 --> 00:08:58,159 testing them, and deploying them where historically we 245 00:08:58,159 --> 00:08:59,325 saw a real shift, 246 00:08:59,965 --> 00:09:01,105 a real separation 247 00:09:01,565 --> 00:09:03,404 of those that were doing those activities and 248 00:09:03,404 --> 00:09:04,785 then those that were actually 249 00:09:05,245 --> 00:09:06,384 using the technology. 250 00:09:06,845 --> 00:09:08,605 So this is a really exciting time to 251 00:09:08,605 --> 00:09:11,585 be in health care and, helping innovators, 252 00:09:12,329 --> 00:09:13,309 healthcare professionals 253 00:09:13,769 --> 00:09:16,970 take tools from discovery or ideation to actual 254 00:09:16,970 --> 00:09:20,089 actual implementation into clinical practice. But it also 255 00:09:20,089 --> 00:09:22,809 opens up the door for a number of 256 00:09:22,809 --> 00:09:23,789 ethical considerations 257 00:09:24,169 --> 00:09:26,445 that we haven't thought of before, and it 258 00:09:26,445 --> 00:09:27,904 also puts health care institutions 259 00:09:28,605 --> 00:09:31,565 in the position of having to apply different 260 00:09:31,565 --> 00:09:34,445 governance activities that they historically have not been 261 00:09:34,445 --> 00:09:35,585 responsible for. 262 00:09:36,125 --> 00:09:37,644 Yeah. Yeah. I'm sure you just touched on 263 00:09:37,644 --> 00:09:39,565 it a little bit. But, yeah, obviously, new 264 00:09:39,565 --> 00:09:42,230 challenges come with these tools and new tech 265 00:09:42,230 --> 00:09:43,509 being introduced. If you if you had to, 266 00:09:43,509 --> 00:09:45,830 like, focus on your top piece of advice 267 00:09:45,830 --> 00:09:48,549 for leaders as they introduce this this new 268 00:09:48,549 --> 00:09:50,470 technology or or, you you know, kind of 269 00:09:50,470 --> 00:09:51,929 work it out, what would that be? 270 00:09:52,790 --> 00:09:53,610 Yeah. So 271 00:09:54,524 --> 00:09:56,704 your governance within your health care organization 272 00:09:57,404 --> 00:09:59,664 should be seen as a mechanism 273 00:10:00,044 --> 00:10:00,944 that accelerates 274 00:10:01,404 --> 00:10:02,304 and prioritizes 275 00:10:02,764 --> 00:10:03,264 innovation. 276 00:10:03,804 --> 00:10:06,299 It should not be perceived as a roadblock. 277 00:10:07,179 --> 00:10:09,919 At Mayo Clinic, we use a risk based 278 00:10:10,059 --> 00:10:13,839 assessment framework that allows us to evaluate these, 279 00:10:14,459 --> 00:10:17,039 internally developed solutions, but also vended solutions, 280 00:10:17,659 --> 00:10:19,120 quickly while maintaining 281 00:10:19,500 --> 00:10:20,000 rigor 282 00:10:20,355 --> 00:10:23,475 for that development and testing process that is 283 00:10:23,475 --> 00:10:26,274 proportionate to the risk of the tool. I 284 00:10:26,274 --> 00:10:29,075 previously mentioned that in my opinion, there's really 285 00:10:29,075 --> 00:10:30,455 kind of two main categories 286 00:10:30,914 --> 00:10:32,455 of AI enabled technologies. 287 00:10:33,120 --> 00:10:35,759 There's those that are in the admin support 288 00:10:35,759 --> 00:10:38,019 space, and then there are those that are 289 00:10:38,080 --> 00:10:39,220 directly or indirectly 290 00:10:39,600 --> 00:10:42,500 doing clinical decision making or aiding. 291 00:10:42,879 --> 00:10:45,360 We don't wanna apply the same level of 292 00:10:45,360 --> 00:10:46,740 rigor around requirements, 293 00:10:47,384 --> 00:10:49,325 within the software development life cycle, 294 00:10:49,865 --> 00:10:53,404 specifically with how we demonstrate quantitative or qualitative 295 00:10:53,545 --> 00:10:54,845 measure of safety, 296 00:10:55,225 --> 00:10:58,425 efficacy, and ethical use to tools that are 297 00:10:58,425 --> 00:11:00,825 being used for admin support that are, you 298 00:11:00,825 --> 00:11:02,285 know, doing basic scheduling 299 00:11:02,709 --> 00:11:04,730 compared with tools that are analyzing 300 00:11:05,190 --> 00:11:08,889 medical images and doing early detection of patients 301 00:11:09,110 --> 00:11:12,070 with pancreatic cancer. Those are two really polar 302 00:11:12,070 --> 00:11:14,789 opposite examples, but you need to make sure 303 00:11:14,789 --> 00:11:17,209 that the governance infrastructure that you develop 304 00:11:17,565 --> 00:11:20,705 accounts for that range of diversity of technologies 305 00:11:20,924 --> 00:11:23,725 that you certainly will be implementing into clinical 306 00:11:23,725 --> 00:11:24,225 practice, 307 00:11:24,764 --> 00:11:26,365 but you shouldn't apply the same level of 308 00:11:26,365 --> 00:11:27,745 rigor for those two. 309 00:11:28,684 --> 00:11:30,800 Under our chief AI implementation 310 00:11:31,180 --> 00:11:34,379 officer, Micky Tripathi here at Mayo Clinic, we 311 00:11:34,379 --> 00:11:37,120 are currently developing top down policies 312 00:11:37,660 --> 00:11:40,080 that standardize those evidence requirements. 313 00:11:40,700 --> 00:11:42,960 So our teams know in advance 314 00:11:43,375 --> 00:11:46,035 what is needed for safety and compliance. 315 00:11:46,654 --> 00:11:47,634 That transparency 316 00:11:48,175 --> 00:11:50,675 really shortens the path from concept 317 00:11:51,134 --> 00:11:54,035 to clinical use and implementation in the practice. 318 00:11:55,134 --> 00:11:58,019 Innovation must move at the speed of trust, 319 00:11:58,179 --> 00:12:01,720 and this is the trust with the practice, 320 00:12:02,259 --> 00:12:02,759 providers, 321 00:12:03,620 --> 00:12:05,160 patients, and the public. 322 00:12:05,860 --> 00:12:08,600 At Mayo Clinic, we encourage experimentation 323 00:12:09,460 --> 00:12:10,679 through suggested 324 00:12:10,980 --> 00:12:11,960 tiered deployment 325 00:12:12,545 --> 00:12:14,165 that allows for evaluating 326 00:12:14,705 --> 00:12:15,205 safety, 327 00:12:15,904 --> 00:12:19,125 equity, and value before scaling occurs. 328 00:12:19,904 --> 00:12:22,384 This approach allows us to be bold in 329 00:12:22,384 --> 00:12:23,365 trying new technologies 330 00:12:23,985 --> 00:12:27,379 while maintaining the discipline that patients expect from 331 00:12:27,379 --> 00:12:29,240 a world class health care institution. 332 00:12:30,419 --> 00:12:32,600 And sticking with it, I think that's that's 333 00:12:32,659 --> 00:12:34,120 fantastic. And sticking with 334 00:12:34,580 --> 00:12:37,059 advice, you know, the the a lot of 335 00:12:37,059 --> 00:12:38,820 the the use cases you work out, you 336 00:12:38,820 --> 00:12:40,795 you know what's coming in, you work with 337 00:12:40,795 --> 00:12:42,555 it. But the the unknown, you know, seeing 338 00:12:42,555 --> 00:12:44,955 all the advancements we have had recently, like 339 00:12:44,955 --> 00:12:46,555 you said, the last five years has been 340 00:12:46,555 --> 00:12:48,795 monumental. But, you know, just thinking of what 341 00:12:48,795 --> 00:12:50,154 could be out there down the road, what's 342 00:12:50,154 --> 00:12:52,555 your top piece of advice for leaders as 343 00:12:52,555 --> 00:12:53,855 they prepare for further 344 00:12:54,209 --> 00:12:55,350 advancements in technology 345 00:12:55,889 --> 00:12:57,649 and kinda weighing those with the rising demands 346 00:12:57,649 --> 00:12:58,389 for care? 347 00:12:58,769 --> 00:12:59,269 Mhmm. 348 00:13:00,449 --> 00:13:03,669 I think the future belongs to those organizations 349 00:13:04,289 --> 00:13:05,110 that effectively 350 00:13:05,569 --> 00:13:06,069 operationalize 351 00:13:06,610 --> 00:13:07,589 governance. Establish multidisciplinary review structures early and 352 00:13:09,294 --> 00:13:09,794 multidisciplinary 353 00:13:10,414 --> 00:13:14,334 review structures early, and ensure decisions about AI 354 00:13:14,334 --> 00:13:16,034 deployment are data driven, 355 00:13:16,414 --> 00:13:16,914 transparent, 356 00:13:17,615 --> 00:13:18,834 and clinically grounded. 357 00:13:19,454 --> 00:13:21,794 It is essential that you embed your organization's 358 00:13:22,574 --> 00:13:24,034 relative risk tolerance 359 00:13:24,570 --> 00:13:25,710 into this governance 360 00:13:26,090 --> 00:13:26,590 structure. 361 00:13:27,049 --> 00:13:29,129 What we've done here at Mayo Clinic may 362 00:13:29,129 --> 00:13:32,110 not be directly applicable to your healthcare organization 363 00:13:32,250 --> 00:13:34,670 because maybe you have a different risk tolerance. 364 00:13:35,610 --> 00:13:39,855 Technology can't transform care unless people trust and 365 00:13:39,855 --> 00:13:40,674 understand it. 366 00:13:41,134 --> 00:13:42,754 So leaders should prioritize 367 00:13:43,215 --> 00:13:44,115 digital literacy, 368 00:13:44,815 --> 00:13:46,034 cross functional collaboration, 369 00:13:46,575 --> 00:13:47,955 and clinician engagement 370 00:13:48,414 --> 00:13:50,995 just as much as the technical infrastructure. 371 00:13:51,990 --> 00:13:54,629 I really think that culture is the true 372 00:13:54,629 --> 00:13:56,970 enabler of responsible AI 373 00:13:57,269 --> 00:13:59,450 or really in any innovative 374 00:13:59,909 --> 00:14:00,409 technology. 375 00:14:01,669 --> 00:14:05,524 Every AI or digital tool must serve a 376 00:14:05,524 --> 00:14:08,105 clear clinical or operational purpose, 377 00:14:08,565 --> 00:14:11,764 and we should measure for outcomes that support 378 00:14:11,764 --> 00:14:14,644 that this tool is achieving that through post 379 00:14:14,644 --> 00:14:16,345 deployment monitoring activities. 380 00:14:17,490 --> 00:14:19,590 Lastly, my advice is to align 381 00:14:20,129 --> 00:14:21,990 innovation with your institution's 382 00:14:22,450 --> 00:14:22,950 mission, 383 00:14:23,410 --> 00:14:27,250 whether that's advancing healing, equity, or access for 384 00:14:27,250 --> 00:14:27,990 your patients, 385 00:14:28,290 --> 00:14:30,950 and then hold technology to the same ethical 386 00:14:31,009 --> 00:14:31,509 standards 387 00:14:32,129 --> 00:14:34,225 as the practice of medicine itself. 388 00:14:35,164 --> 00:14:36,764 Thank you so much. And the the last 389 00:14:36,764 --> 00:14:38,705 thing I wanna ask you is, 390 00:14:39,085 --> 00:14:39,585 obviously, 391 00:14:39,965 --> 00:14:42,524 all this tech that we're working with and 392 00:14:42,524 --> 00:14:44,144 and applying in health care, 393 00:14:44,845 --> 00:14:46,740 you and your team are, it takes it 394 00:14:46,740 --> 00:14:49,220 takes great leadership. Right? So I just wanna 395 00:14:49,220 --> 00:14:51,220 ask you on a human level, how have 396 00:14:51,220 --> 00:14:52,759 you evolved as a leader? 397 00:14:54,500 --> 00:14:56,279 So I started at Mayo Clinic 398 00:14:56,820 --> 00:14:59,779 as a manager of the same space in 399 00:14:59,779 --> 00:15:00,279 regulatory 400 00:15:00,659 --> 00:15:01,320 and quality 401 00:15:01,704 --> 00:15:03,945 where I was coming from the med tech 402 00:15:03,945 --> 00:15:04,445 industry 403 00:15:05,065 --> 00:15:08,044 really with the primary goal of getting technology 404 00:15:08,745 --> 00:15:09,245 into, 405 00:15:10,264 --> 00:15:13,865 the commercialized space. So my experience was really 406 00:15:13,865 --> 00:15:14,365 in 407 00:15:14,669 --> 00:15:18,110 submitting products to FDA and understanding what that 408 00:15:18,110 --> 00:15:19,089 pathway to 409 00:15:19,389 --> 00:15:19,889 commercialization 410 00:15:20,429 --> 00:15:20,929 would 411 00:15:21,389 --> 00:15:24,669 be. My formal and formative regulatory years were 412 00:15:24,669 --> 00:15:27,070 at Medtronic in their cardiac rhythm and heart 413 00:15:27,070 --> 00:15:27,970 failure division. 414 00:15:28,669 --> 00:15:30,049 And here at Mayo Clinic, 415 00:15:30,764 --> 00:15:33,964 my job is not so much to pursue 416 00:15:33,964 --> 00:15:34,704 the commercialization 417 00:15:35,485 --> 00:15:36,784 of these different technologies, 418 00:15:37,565 --> 00:15:39,644 but to see how we can build those 419 00:15:39,644 --> 00:15:40,144 roadmaps 420 00:15:40,684 --> 00:15:41,504 from research 421 00:15:41,804 --> 00:15:42,625 to implementation 422 00:15:43,245 --> 00:15:46,289 into clinical practice in a local way. So 423 00:15:46,289 --> 00:15:49,190 as a leader, I have really grown and 424 00:15:49,649 --> 00:15:50,149 expanded 425 00:15:50,769 --> 00:15:51,590 my understanding 426 00:15:52,049 --> 00:15:53,190 of what regulatory 427 00:15:53,570 --> 00:15:54,789 and quality means 428 00:15:55,169 --> 00:15:58,389 when building roadmaps or pathways to implementation, 429 00:15:59,009 --> 00:16:01,110 where I've really shifted my mindset 430 00:16:01,595 --> 00:16:02,495 from the angle 431 00:16:02,795 --> 00:16:05,915 of getting FDA clearance or approval for these 432 00:16:05,915 --> 00:16:08,174 technologies that are then going to be marketed 433 00:16:08,554 --> 00:16:10,654 to building pathways more generically 434 00:16:11,274 --> 00:16:14,095 for broader AI enabled digital health technologies, 435 00:16:14,940 --> 00:16:17,419 to be implemented in the clinical practice, but 436 00:16:17,419 --> 00:16:18,160 not necessarily 437 00:16:18,860 --> 00:16:19,360 commercialized. 438 00:16:19,820 --> 00:16:21,519 So I've had to root myself 439 00:16:22,059 --> 00:16:24,240 in the clinician's perspective 440 00:16:24,779 --> 00:16:25,519 of understanding 441 00:16:26,299 --> 00:16:29,279 what it really takes to develop safe, effective, 442 00:16:29,735 --> 00:16:30,795 and ethical products 443 00:16:31,175 --> 00:16:34,715 while leveraging information from laws, regulations, and standards, 444 00:16:34,935 --> 00:16:37,915 but really relying on their expertise to understand 445 00:16:38,615 --> 00:16:41,195 how and when can we develop AI technologies 446 00:16:41,735 --> 00:16:43,435 to solve these local problems. 447 00:16:43,815 --> 00:16:46,129 So I have I have grown from what 448 00:16:46,129 --> 00:16:48,149 I would say is a pure, 449 00:16:48,850 --> 00:16:50,070 formal regulatory 450 00:16:50,450 --> 00:16:53,329 expert to being someone who is more of 451 00:16:53,329 --> 00:16:54,070 a generalist 452 00:16:54,450 --> 00:16:55,269 in translation 453 00:16:55,649 --> 00:16:57,649 science. And I think that this is where 454 00:16:57,649 --> 00:16:59,190 we are going to see, 455 00:16:59,825 --> 00:17:00,325 opportunity 456 00:17:00,705 --> 00:17:02,325 and growth in leaders 457 00:17:02,785 --> 00:17:04,164 with effective implementation 458 00:17:04,704 --> 00:17:05,445 of AI, 459 00:17:05,904 --> 00:17:06,724 where we 460 00:17:07,025 --> 00:17:09,924 build leaders and, expert contributors 461 00:17:10,305 --> 00:17:11,445 who understand 462 00:17:11,984 --> 00:17:13,605 the regulatory ecosystem 463 00:17:14,549 --> 00:17:17,529 and what is written in laws and standards, 464 00:17:18,070 --> 00:17:20,710 but they take a generalist approach that really 465 00:17:20,710 --> 00:17:22,730 leans on the clinician's expertise 466 00:17:23,190 --> 00:17:25,529 to understand what it takes to pragmatically 467 00:17:25,910 --> 00:17:28,650 translate these tools into clinical practice. 468 00:17:29,714 --> 00:17:31,234 Prada, thank you so much for joining us 469 00:17:31,234 --> 00:17:32,355 on the podcast. I think it was a 470 00:17:32,355 --> 00:17:34,355 great conversation. I learned a lot personally. I 471 00:17:34,355 --> 00:17:36,275 know everyone listening did, and I hope to 472 00:17:36,275 --> 00:17:37,494 work with you again soon. 473 00:17:37,954 --> 00:17:39,095 Thank you so much. 474 00:17:42,130 --> 00:17:42,950 At athenahealth, 475 00:17:43,250 --> 00:17:45,910 we know your ambulatory practice wants healthier, 476 00:17:46,289 --> 00:17:49,170 a healthier business, healthier care teams, and healthier 477 00:17:49,170 --> 00:17:49,670 patients. 478 00:17:50,130 --> 00:17:52,369 But the complexities of modern health care tech 479 00:17:52,369 --> 00:17:54,289 make it hard for you and your care 480 00:17:54,289 --> 00:17:55,884 teams to focus on what matters 481 00:17:56,345 --> 00:17:58,284 most. That's where athenahealth can help. 482 00:17:58,664 --> 00:18:01,384 Our AI native all in one solutions reduce 483 00:18:01,384 --> 00:18:02,524 administrative burdens, 484 00:18:02,904 --> 00:18:05,865 streamline billing and payments, and deliver critical insights 485 00:18:05,865 --> 00:18:07,440 when clinicians need it most. 486 00:18:07,839 --> 00:18:10,480 That means fewer clicks, more time for patients, 487 00:18:10,480 --> 00:18:12,099 and stronger bottom lines. 488 00:18:12,720 --> 00:18:15,919 Practicing medicine is complex, but running a practice 489 00:18:15,919 --> 00:18:17,779 can be that much simpler with athenahealth. 490 00:18:18,480 --> 00:18:22,179 See how simpler is healthier at athenahealth.com.