WEBVTT 00:00:00.083 --> 00:00:03.253 You are seeing a patient four, five weeks earlier. 00:00:03.253 --> 00:00:04.963 Now, if you are stage four cancer, 00:00:04.963 --> 00:00:07.007 four, five weeks may mean, you know, 00:00:07.007 --> 00:00:08.883 another metastasis or not. 00:00:08.883 --> 00:00:11.136 Cancer centers right now are under a lot of pressure, 00:00:11.136 --> 00:00:13.763 staffing, volume, costs. 00:00:13.763 --> 00:00:17.142 Which area of that pressure hits the hardest, 00:00:17.142 --> 00:00:19.686 and where can AI make the biggest difference? 00:00:19.686 --> 00:00:21.563 From my perspective, personnel. 00:00:21.563 --> 00:00:25.108 That's the most expensive part. 00:00:25.108 --> 00:00:27.652 We need, not just Moffitt, 00:00:27.652 --> 00:00:31.656 we need, as providers, we need to leverage TAP, 00:00:31.656 --> 00:00:34.659 learn how to implement AI 00:00:34.659 --> 00:00:37.621 to continue to grow in term of volume, 00:00:37.621 --> 00:00:40.290 because we're going to see more and more patients, 00:00:40.290 --> 00:00:44.169 without having to grow in this same linear path 00:00:44.169 --> 00:00:46.546 of "I need to add personnel each time I grow. 00:00:46.546 --> 00:00:48.465 I get 100 more patients, 00:00:48.465 --> 00:00:51.093 I need 10 more people to manage that." 00:00:51.093 --> 00:00:55.680 So this is a promise of AI from an operational point of view 00:00:55.680 --> 00:00:57.599 and from a cost point of view. 00:00:57.599 --> 00:00:58.892 There's so many health centers right now 00:00:58.892 --> 00:01:01.561 who are running these pilots and they're so promising, 00:01:01.561 --> 00:01:04.064 but something is separating those 00:01:04.064 --> 00:01:06.024 who go from the pilot stage 00:01:06.024 --> 00:01:08.610 and never quite make it to scale. 00:01:08.610 --> 00:01:09.528 What is that? 00:01:09.528 --> 00:01:12.948 Yeah, the trap of pilots. 00:01:12.948 --> 00:01:15.450 Pilots after pilots, that's a word 00:01:15.450 --> 00:01:17.703 that I ask my team to not use. 00:01:17.703 --> 00:01:19.079 I don't want to use pilots, you know? 00:01:19.079 --> 00:01:21.164 I ask them to use phase one. 00:01:21.164 --> 00:01:22.624 Why phase one? 00:01:22.624 --> 00:01:24.585 Well, because what you're going to learn 00:01:24.585 --> 00:01:26.962 from your phase one means that you're going to either go 00:01:26.962 --> 00:01:28.881 to phase two or do something else. 00:01:28.881 --> 00:01:31.884 But if you do a pilot, it feels a little bit like isolated. 00:01:31.884 --> 00:01:34.052 We're trying it, no strings attached 00:01:34.052 --> 00:01:35.387 whether it works or not, you know? 00:01:35.387 --> 00:01:37.723 And so you have plenty of successful pilots 00:01:37.723 --> 00:01:39.641 that have never scaled up. 00:01:39.641 --> 00:01:44.438 Another issue that I do see also in the provider world 00:01:44.438 --> 00:01:46.899 is maybe the lack of a strategic roadmap 00:01:46.899 --> 00:01:49.818 to implement AI and digital. 00:01:49.818 --> 00:01:50.944 And so you do pilots, 00:01:50.944 --> 00:01:53.489 but you don't necessarily always connect them 00:01:53.489 --> 00:01:56.450 to this long term strategic roadmap. 00:01:56.450 --> 00:01:58.619 And so you don't really know if this is a time, 00:01:58.619 --> 00:02:00.996 the right time to implement it or not. 00:02:00.996 --> 00:02:04.500 Where do successful leaders invest in first? 00:02:04.500 --> 00:02:08.671 Great question, and at Moffitt, the path we've taken 00:02:08.671 --> 00:02:11.048 is to think about it through two main lenses. 00:02:11.048 --> 00:02:13.676 So we're looking at non-clinical application 00:02:13.676 --> 00:02:15.094 and clinical application. 00:02:15.094 --> 00:02:16.804 Sometimes these non-clinical applications 00:02:16.804 --> 00:02:19.139 have also clinical impact. 00:02:19.139 --> 00:02:24.103 And so that's our way to embrace AI 00:02:24.103 --> 00:02:27.398 and digital applications with digitizing, for example, 00:02:27.398 --> 00:02:29.525 our pathology right now. 00:02:29.525 --> 00:02:30.568 And we're trying to digitize 00:02:30.568 --> 00:02:33.445 as many line of services as we can. 00:02:33.445 --> 00:02:35.489 And then from there, you can build AI. 00:02:35.489 --> 00:02:37.616 So I'll give you a very concrete example. 00:02:37.616 --> 00:02:40.786 Very difficult to retrieve patient charts from patients 00:02:40.786 --> 00:02:43.956 who are coming to Moffitt from the community centers. 00:02:43.956 --> 00:02:45.499 And usually when they come to Moffitt, 00:02:45.499 --> 00:02:47.418 they're fairly advanced in their disease. 00:02:47.418 --> 00:02:49.754 So time is of the essence. 00:02:49.754 --> 00:02:53.132 We had an army of people taking phones, calls 00:02:53.132 --> 00:02:54.925 in and out, day in, day out 00:02:54.925 --> 00:02:57.136 to try to retrieve these charts. 00:02:57.136 --> 00:02:59.055 It was on average five weeks. 00:02:59.055 --> 00:03:01.682 When you get them, it's an average three, 400 pages. 00:03:01.682 --> 00:03:03.934 All format takes an average an hour 00:03:03.934 --> 00:03:06.270 and a half for a physician to flip through that 00:03:06.270 --> 00:03:10.107 before this physician can have a conversation with a patient 00:03:10.107 --> 00:03:12.193 and understand what the situation is. 00:03:12.193 --> 00:03:14.487 We implemented an AI platform. 00:03:14.487 --> 00:03:19.492 We can manage 70% of these patient charts. 00:03:19.825 --> 00:03:21.619 Through this platform, and on average, 00:03:21.619 --> 00:03:23.746 it's seven to 10 minutes per case. 00:03:23.746 --> 00:03:26.958 So that's a true, real impact on... 00:03:26.958 --> 00:03:31.295 Of AI allowing us to bring these patients 00:03:31.295 --> 00:03:34.840 to Moffitt way faster, and to engage with them 00:03:34.840 --> 00:03:36.592 in a much more effective way. 00:03:36.592 --> 00:03:38.469 And so you can imagine this is not just 00:03:38.469 --> 00:03:39.804 an operational benefit. 00:03:39.804 --> 00:03:42.181 Now, we can schedule patients faster. 00:03:42.181 --> 00:03:44.725 We don’t need an army of people on the phone. 00:03:44.725 --> 00:03:47.979 We can save physician's time, which is very important, 00:03:47.979 --> 00:03:49.981 but it's better preparing the physician 00:03:49.981 --> 00:03:52.483 so it can have a clinical impact as well. 00:03:52.483 --> 00:03:55.695 You're seeing a patient four, five weeks earlier. 00:03:55.695 --> 00:03:57.447 Now, if you are a stage four cancer, 00:03:57.447 --> 00:03:59.490 four, five weeks may mean, you know, 00:03:59.490 --> 00:04:01.325 another metastasis or not. 00:04:01.325 --> 00:04:02.160 You see what I mean? 00:04:02.160 --> 00:04:05.371 So this is real transformation at scale. 00:04:05.371 --> 00:04:06.873 Thank you so much.