1 00:00:00,160 --> 00:00:02,399 Hello, everyone. This is Jacob Emerson with the 2 00:00:02,399 --> 00:00:05,519 Becker's Healthcare Podcast. Thrilled today to be joined 3 00:00:05,519 --> 00:00:08,720 by 2 special guests. Andrew Lockhart is the 4 00:00:08,720 --> 00:00:12,000 CEO of Fathom, and doctor Yogan Patel is 5 00:00:12,000 --> 00:00:15,514 the CEO of Apollo MD. Doctor Patel, Andrew, 6 00:00:15,514 --> 00:00:17,114 thank you so much for taking the time 7 00:00:17,114 --> 00:00:18,574 to be with us on the podcast today. 8 00:00:19,035 --> 00:00:20,494 Thank you for having us. 9 00:00:20,954 --> 00:00:23,195 So, gentlemen, before we dive into everything we 10 00:00:23,195 --> 00:00:25,274 wanna talk with you about, can we hear 11 00:00:25,274 --> 00:00:26,730 a little bit more about yourselves? What do 12 00:00:26,730 --> 00:00:29,309 you do right now with your respective organizations? 13 00:00:29,609 --> 00:00:31,289 And what led you to this point in 14 00:00:31,289 --> 00:00:33,950 terms of health care career backgrounds? Doctor Patel? 15 00:00:35,049 --> 00:00:37,770 Thanks, Jacob. I serve as the chief executive 16 00:00:37,770 --> 00:00:40,835 officer for Apollo MD. I'm a practicing emergency 17 00:00:40,835 --> 00:00:41,074 physician. 18 00:00:41,715 --> 00:00:43,015 I've been in the space now, 19 00:00:44,034 --> 00:00:47,015 17 years with Apollo MD. So 20 00:00:47,554 --> 00:00:49,954 thrilled to be part of the organization and 21 00:00:49,954 --> 00:00:50,695 to serve. 22 00:00:51,074 --> 00:00:53,475 We have the distinct honor of serving over 23 00:00:53,475 --> 00:00:55,015 4,000,000 patients a year, 24 00:00:55,450 --> 00:00:57,530 mostly in emergency medicine. We also do some 25 00:00:57,530 --> 00:00:58,429 hospital medicine, 26 00:00:58,890 --> 00:01:00,190 anesthesia, and radiology. 27 00:01:00,649 --> 00:01:03,149 And we are, you know, a private practice 28 00:01:03,210 --> 00:01:05,530 model, private independent group. I've been in the 29 00:01:05,530 --> 00:01:08,090 space for over 40 years. How about you, 30 00:01:08,090 --> 00:01:08,590 Andrew? 31 00:01:09,594 --> 00:01:11,754 My name's Andrew Lockhart. I'm the cofounder and 32 00:01:11,754 --> 00:01:12,814 CEO of Fathom. 33 00:01:13,194 --> 00:01:15,995 My background in health care so I really 34 00:01:15,995 --> 00:01:16,894 got the bug, 35 00:01:17,194 --> 00:01:18,334 from Toronto originally 36 00:01:18,635 --> 00:01:20,734 and was working for a design firm there 37 00:01:20,954 --> 00:01:22,814 where are a lot of our 38 00:01:23,299 --> 00:01:25,640 our clients were US based health care organizations, 39 00:01:26,020 --> 00:01:27,239 Mayo Clinic, UHG, 40 00:01:28,099 --> 00:01:30,120 and was spending a lot of time, 41 00:01:30,579 --> 00:01:32,739 you know, exploring the US health care system. 42 00:01:32,739 --> 00:01:33,239 And 43 00:01:33,620 --> 00:01:34,840 the infinite complexity 44 00:01:35,459 --> 00:01:38,144 of the system was something that just was 45 00:01:38,144 --> 00:01:39,504 like a mind trap for me that got 46 00:01:39,504 --> 00:01:41,604 me very excited, interested in health care. 47 00:01:42,064 --> 00:01:43,765 Went on to do my MBA at Stanford, 48 00:01:44,545 --> 00:01:46,704 and then coming out of that, really had 49 00:01:46,704 --> 00:01:48,864 a distinct desire to do something in health 50 00:01:48,864 --> 00:01:49,364 care. 51 00:01:49,905 --> 00:01:51,990 My first swing there was in an on 52 00:01:51,990 --> 00:01:52,969 demand pharmacy, 53 00:01:53,670 --> 00:01:55,750 which, you know, I can give you lots 54 00:01:55,750 --> 00:01:59,829 of color on what's problem problematic about the 55 00:01:59,829 --> 00:02:02,650 way the pharmaceutical industry works. But, ultimately, 56 00:02:02,950 --> 00:02:05,430 you know, that journey eventually led us towards 57 00:02:05,430 --> 00:02:07,924 finding Fathom, a solution that we felt was 58 00:02:08,224 --> 00:02:08,965 much needed, 59 00:02:09,905 --> 00:02:10,405 timely, 60 00:02:10,705 --> 00:02:13,344 and, you know, ultimately, something that, you know, 61 00:02:13,344 --> 00:02:16,064 is very scalable. Sure. Absolutely. And, you know, 62 00:02:16,064 --> 00:02:17,745 for our listeners who might not be familiar, 63 00:02:17,745 --> 00:02:19,504 let's take a step back and talk about 64 00:02:19,504 --> 00:02:21,604 how the partnership between your two companies 65 00:02:21,909 --> 00:02:23,830 came about. How long you've been working together, 66 00:02:24,229 --> 00:02:25,669 and what were you trying to achieve when 67 00:02:25,669 --> 00:02:27,610 you when you first began working together? 68 00:02:28,550 --> 00:02:30,789 Yeah. I'll I'll start. Thanks, Jacob. I was 69 00:02:30,789 --> 00:02:32,550 actually just looking back at our emails. So 70 00:02:32,550 --> 00:02:33,210 we actually 71 00:02:33,750 --> 00:02:34,250 first, 72 00:02:35,205 --> 00:02:37,685 started talking to Fathom back in December of 73 00:02:37,685 --> 00:02:38,185 2019. 74 00:02:39,205 --> 00:02:40,504 So, we were, 75 00:02:41,125 --> 00:02:42,985 fortunately now an early adopter. 76 00:02:43,605 --> 00:02:45,685 And for us, you know, as a private 77 00:02:45,685 --> 00:02:47,525 group that's been in this space for a 78 00:02:47,525 --> 00:02:48,264 long time, 79 00:02:48,680 --> 00:02:49,580 we had always 80 00:02:49,879 --> 00:02:51,500 relied on kind of a human 81 00:02:51,879 --> 00:02:53,099 centered RCM 82 00:02:53,639 --> 00:02:55,500 model. So rep all of our revenue cycle 83 00:02:55,879 --> 00:02:57,340 required human coders. 84 00:02:58,040 --> 00:03:00,040 We actually do the entire revenue cycle in 85 00:03:00,040 --> 00:03:02,200 house. But the one component that we had 86 00:03:02,200 --> 00:03:04,379 historically outsourced was our coding. 87 00:03:04,895 --> 00:03:06,914 And what we found was over the years, 88 00:03:07,455 --> 00:03:08,594 that human workforce 89 00:03:08,974 --> 00:03:10,675 just didn't have the same resiliency. 90 00:03:11,375 --> 00:03:13,055 And the quality started to become a little 91 00:03:13,055 --> 00:03:16,254 bit more variable, especially as coding guidelines were 92 00:03:16,254 --> 00:03:16,754 changing. 93 00:03:17,215 --> 00:03:18,894 And so we were looking for a higher 94 00:03:18,894 --> 00:03:20,034 fidelity solution. 95 00:03:20,790 --> 00:03:22,469 And and ultimately, as we started to explore 96 00:03:22,469 --> 00:03:23,530 the idea of automation, 97 00:03:23,870 --> 00:03:24,370 we 98 00:03:24,710 --> 00:03:27,349 stumbled upon Fathom. And, you know, at the 99 00:03:27,349 --> 00:03:29,370 time, we probably looked at several 100 00:03:29,909 --> 00:03:31,210 different coding solutions, 101 00:03:32,069 --> 00:03:33,849 you know, computer assisted coding, 102 00:03:34,335 --> 00:03:36,335 you know, other other groups that were early 103 00:03:36,335 --> 00:03:37,155 in this space. 104 00:03:37,534 --> 00:03:39,375 But really we're looking for somebody that could 105 00:03:39,375 --> 00:03:39,875 deliver 106 00:03:40,495 --> 00:03:42,115 top quality and accuracy, 107 00:03:42,814 --> 00:03:43,314 sustainability, 108 00:03:44,014 --> 00:03:45,314 and and great value. 109 00:03:45,694 --> 00:03:48,094 And I think, even from the very first 110 00:03:48,094 --> 00:03:49,314 conversations, Fathom 111 00:03:49,810 --> 00:03:52,150 clearly kind of met those benchmarks. 112 00:03:52,849 --> 00:03:54,689 And then really as a organization that was 113 00:03:54,689 --> 00:03:55,189 naive 114 00:03:55,569 --> 00:03:57,750 to artificial intelligence and automation, 115 00:03:58,289 --> 00:03:59,889 they were great about kind of holding our 116 00:03:59,889 --> 00:04:02,289 hand as we kind of journey together. And 117 00:04:02,289 --> 00:04:02,870 it was, 118 00:04:03,215 --> 00:04:04,094 you know, it's been a, 119 00:04:04,655 --> 00:04:07,215 it's been a good collaboration. Sure. Absolutely. And 120 00:04:07,215 --> 00:04:08,414 I wanna get into what some of those 121 00:04:08,414 --> 00:04:10,334 outcomes are in just a second. But, Andrew, 122 00:04:10,334 --> 00:04:12,254 do you remember the initial goals you you 123 00:04:12,254 --> 00:04:14,114 set out to achieve with your partnership? 124 00:04:15,055 --> 00:04:17,714 Yeah. I mean, like, when we started Fathom, 125 00:04:18,550 --> 00:04:19,290 I think, 126 00:04:19,750 --> 00:04:21,290 we would have thought that 127 00:04:21,590 --> 00:04:25,189 50% automation for emergency department coding, which, you 128 00:04:25,189 --> 00:04:27,910 know, I'd put in the, like, moderate high 129 00:04:27,910 --> 00:04:30,550 complexity on the scale Mhmm. Was something that 130 00:04:30,550 --> 00:04:32,009 was maybe within reach. 131 00:04:32,545 --> 00:04:33,985 And then by the time we were able 132 00:04:33,985 --> 00:04:36,064 to connect with Apollo MD, we felt pretty 133 00:04:36,064 --> 00:04:37,605 confident we could get to 70. 134 00:04:38,384 --> 00:04:38,884 And 135 00:04:39,345 --> 00:04:41,585 the fact that we're able to jump to 136 00:04:41,585 --> 00:04:44,004 where we are today is really, 137 00:04:44,479 --> 00:04:46,800 you know, just a testament to the to 138 00:04:46,800 --> 00:04:48,479 the to the engineering prowess of the team, 139 00:04:48,479 --> 00:04:49,459 but also partners 140 00:04:49,919 --> 00:04:51,519 who were able to work with us and, 141 00:04:51,519 --> 00:04:53,680 you know, iron out edge cases. Sure. And 142 00:04:53,680 --> 00:04:55,459 you're at 90% today. Correct? 143 00:04:55,919 --> 00:04:58,479 Yeah. I think 91. Wow. So even even 144 00:04:58,479 --> 00:05:00,225 higher than that. I mean, can you discuss 145 00:05:00,225 --> 00:05:02,705 that journey of of improving automation rates to 146 00:05:02,705 --> 00:05:04,004 90 over 90%? 147 00:05:05,425 --> 00:05:08,305 And, also, you know, doctor Patel, why is 148 00:05:08,305 --> 00:05:10,405 such high automation? Why is that an important, 149 00:05:11,105 --> 00:05:12,564 achievement for the organization? 150 00:05:13,540 --> 00:05:15,379 Yeah. I mean, as I mentioned, you know, 151 00:05:15,379 --> 00:05:16,899 this has been a 4 year journey for 152 00:05:16,899 --> 00:05:18,500 us. And at the start of the journey, 153 00:05:18,500 --> 00:05:20,420 I think we would have been, you know, 154 00:05:20,420 --> 00:05:22,600 we would have been thrilled at 70%. So 155 00:05:22,740 --> 00:05:25,620 it really is a testament to how well 156 00:05:25,620 --> 00:05:26,279 the platform 157 00:05:26,580 --> 00:05:29,144 has learned and developed over the years. 158 00:05:29,704 --> 00:05:31,324 It also allowed us 159 00:05:31,704 --> 00:05:33,564 for the more complex cases 160 00:05:34,185 --> 00:05:36,685 to really focus the human coders 161 00:05:37,384 --> 00:05:38,365 on those cases. 162 00:05:38,904 --> 00:05:41,404 And so we're really deploying the right resource 163 00:05:42,264 --> 00:05:44,529 to the to the right work workflow. 164 00:05:45,069 --> 00:05:47,470 And so for us, that has been kind 165 00:05:47,470 --> 00:05:48,990 of a game changer. I mean, we have 166 00:05:48,990 --> 00:05:50,750 a what I'd like to think of as 167 00:05:50,750 --> 00:05:52,830 a kind of 6 Sigma process almost where 168 00:05:52,830 --> 00:05:55,904 the auditing process is best of breed now. 169 00:05:55,985 --> 00:05:58,225 And really the partnership with Fathom allowed for 170 00:05:58,225 --> 00:06:00,144 that. Mhmm. You know, in addition to kind 171 00:06:00,144 --> 00:06:02,865 of automation accuracy, I would say is even 172 00:06:02,865 --> 00:06:03,524 more important. 173 00:06:04,225 --> 00:06:05,204 And for us, 174 00:06:05,504 --> 00:06:07,024 it's been a pleasure to see that that 175 00:06:07,024 --> 00:06:08,964 has been a priority for Fathom. 176 00:06:09,319 --> 00:06:11,479 And every time we've had kind of a 177 00:06:11,479 --> 00:06:11,979 question, 178 00:06:12,279 --> 00:06:12,939 a query, 179 00:06:13,479 --> 00:06:14,919 they have been right there to kind of 180 00:06:14,919 --> 00:06:15,419 investigate, 181 00:06:15,959 --> 00:06:18,120 walk that path with us, and improve the 182 00:06:18,120 --> 00:06:20,680 technology. Yeah. Any other outcomes you'd you'd point 183 00:06:20,680 --> 00:06:22,519 out, Andrew? Well, maybe to answer your question 184 00:06:22,519 --> 00:06:23,979 about the journey to 90%. 185 00:06:24,279 --> 00:06:24,524 Sure. 186 00:06:25,485 --> 00:06:25,985 One 187 00:06:26,604 --> 00:06:29,024 extraordinarily challenging to get there with any organization. 188 00:06:29,165 --> 00:06:31,084 Yeah. But Appol MD is a physician group 189 00:06:31,084 --> 00:06:33,485 servicing, what, how many emergency departments? Like, over 190 00:06:33,485 --> 00:06:36,865 a 100? Yeah. You've got disparate coding guidelines, 191 00:06:37,389 --> 00:06:40,050 disparate EHR formats, disparate documentation 192 00:06:40,430 --> 00:06:40,930 practices 193 00:06:41,470 --> 00:06:43,230 that you need to cover go all the 194 00:06:43,230 --> 00:06:43,889 way across. 195 00:06:44,430 --> 00:06:47,330 And that's where, you know, really the benefit 196 00:06:47,389 --> 00:06:50,689 of a technology platform that's continuously snowballing data. 197 00:06:51,389 --> 00:06:53,524 So we're able to get going with with 198 00:06:53,524 --> 00:06:55,204 clients like ApolloMD. As we've got more and 199 00:06:55,204 --> 00:06:56,884 more clients, the AI became more and more 200 00:06:56,884 --> 00:06:58,245 aware. And I'd say at this point, you 201 00:06:58,245 --> 00:07:00,884 know, our system has seen north of 3 202 00:07:00,884 --> 00:07:02,964 quarters of a 1000000000 encounters. Wow. 203 00:07:03,444 --> 00:07:05,685 And that's enabled it to get this very 204 00:07:05,685 --> 00:07:07,685 nuanced understanding of the ins and outs of 205 00:07:07,685 --> 00:07:09,064 all these different data formats 206 00:07:09,970 --> 00:07:13,189 and let it work across that so effectively. 207 00:07:14,050 --> 00:07:15,649 And then, also, like, I I wanna give 208 00:07:15,649 --> 00:07:18,050 credit to the team. Like, we've been very 209 00:07:18,050 --> 00:07:20,290 fortunate that we've been able to pull in. 210 00:07:20,290 --> 00:07:21,970 We've got engineers on our team who have 211 00:07:21,970 --> 00:07:22,949 joined us from 212 00:07:23,365 --> 00:07:25,845 AI teams at meta, AI teams at, like, 213 00:07:25,845 --> 00:07:28,985 teams at Neuralink, literally building brain computer interfaces, 214 00:07:29,604 --> 00:07:30,104 Amazon, 215 00:07:31,125 --> 00:07:32,344 Microsoft. And 216 00:07:32,644 --> 00:07:35,284 these folks, you know, have been really attracted 217 00:07:35,284 --> 00:07:37,830 to the idea. Interestingly, like, basically, probably half 218 00:07:37,830 --> 00:07:39,529 our engineering team is pre med. 219 00:07:41,110 --> 00:07:41,610 And 220 00:07:42,069 --> 00:07:44,389 so they're interesting in this idea that, one, 221 00:07:44,389 --> 00:07:46,709 enormous amount of data, 2, something that has 222 00:07:46,709 --> 00:07:48,089 a really scalable impact, 223 00:07:48,870 --> 00:07:51,029 and then 3, like, a really hard problem. 224 00:07:51,029 --> 00:07:51,495 Like, 225 00:07:51,894 --> 00:07:53,814 you know, our CTO has a lot of, 226 00:07:53,814 --> 00:07:56,214 like, vision and really pushes the team to 227 00:07:56,214 --> 00:07:58,794 continually work with the latest, greatest. Sure. Sure. 228 00:07:58,854 --> 00:07:59,354 Absolutely. 229 00:07:59,735 --> 00:08:01,735 So, doctor Patel, you're in over a 100 230 00:08:01,735 --> 00:08:02,235 emergency 231 00:08:02,694 --> 00:08:03,194 departments. 232 00:08:03,894 --> 00:08:05,254 What would you say have been the direct 233 00:08:05,254 --> 00:08:07,949 and indirect impacts of autonomous coding on your 234 00:08:07,949 --> 00:08:08,449 physician's 235 00:08:09,069 --> 00:08:10,689 day to day experiences 236 00:08:11,790 --> 00:08:13,009 and work life satisfaction? 237 00:08:13,949 --> 00:08:16,350 Yeah. We are unique in the industry, I 238 00:08:16,350 --> 00:08:17,730 think, in that we actually 239 00:08:18,589 --> 00:08:19,089 compensate 240 00:08:19,470 --> 00:08:20,290 our physicians, 241 00:08:21,014 --> 00:08:23,834 on a almost a purely performance driven model. 242 00:08:24,134 --> 00:08:26,535 And what that requires is our ability to 243 00:08:26,535 --> 00:08:29,415 understand their productivity in a given month, and 244 00:08:29,415 --> 00:08:32,715 within 2 weeks, translate that into payroll. 245 00:08:33,254 --> 00:08:35,754 And so what we're talking about is their, 246 00:08:35,960 --> 00:08:36,860 you know, RVU 247 00:08:37,399 --> 00:08:37,899 productivity, 248 00:08:38,759 --> 00:08:39,580 their collaboration 249 00:08:39,960 --> 00:08:40,460 with, 250 00:08:41,000 --> 00:08:43,740 advanced practice clinicians, so PAs and nurse practitioners. 251 00:08:44,279 --> 00:08:47,000 And at some practices, their collaboration with resident 252 00:08:47,000 --> 00:08:47,500 physicians. 253 00:08:48,174 --> 00:08:51,054 And so it's a fairly complex box. Right? 254 00:08:51,054 --> 00:08:53,235 And it does require this kind of precision 255 00:08:54,014 --> 00:08:54,835 and execution 256 00:08:55,295 --> 00:08:56,595 on a very tight timeline. 257 00:08:57,215 --> 00:08:59,375 And with human coders that had always historically 258 00:08:59,375 --> 00:09:00,195 been a challenge, 259 00:09:00,600 --> 00:09:03,339 we have some unique workflows that that complicated 260 00:09:03,399 --> 00:09:04,220 that further. 261 00:09:04,679 --> 00:09:06,360 And what we found is as we partnered 262 00:09:06,360 --> 00:09:08,919 with Fathom, you know, it took a lag 263 00:09:08,919 --> 00:09:10,379 of several days, 264 00:09:11,159 --> 00:09:13,559 that human coders required down till, like, less 265 00:09:13,559 --> 00:09:16,455 than 2 hours. And so really, it just 266 00:09:16,455 --> 00:09:19,414 closed the feedback loop to our team, you 267 00:09:19,414 --> 00:09:20,235 know, dramatically. 268 00:09:20,774 --> 00:09:22,695 And that allowed us to, you know, improve 269 00:09:22,695 --> 00:09:25,174 our chart delinquency rates. That allowed us to 270 00:09:25,174 --> 00:09:26,134 improve the accuracy and capture of the procedures 271 00:09:26,134 --> 00:09:26,634 and 272 00:09:29,590 --> 00:09:30,250 and documentation, 273 00:09:30,870 --> 00:09:33,669 and really ultimately translated, I think, into a 274 00:09:33,669 --> 00:09:34,169 more 275 00:09:34,549 --> 00:09:37,769 accurate compensation model for our clinicians. So 276 00:09:38,230 --> 00:09:39,990 it has been, you know, I think we're 277 00:09:39,990 --> 00:09:41,830 we're thrilled with where we are today. The 278 00:09:41,830 --> 00:09:43,290 system continues to improve. 279 00:09:43,725 --> 00:09:45,725 Andrew, other things that kinda stood out to 280 00:09:45,725 --> 00:09:46,625 you in that journey? 281 00:09:47,485 --> 00:09:49,245 No. I mean, I like, I think the 282 00:09:49,245 --> 00:09:50,865 one thing that gets maybe understated 283 00:09:52,044 --> 00:09:54,365 with AI solutions is everyone's like, oh, this 284 00:09:54,365 --> 00:09:56,845 is, like, a thing that's getting rid of 285 00:09:56,845 --> 00:09:57,345 people. 286 00:09:58,204 --> 00:09:58,704 But, 287 00:09:59,070 --> 00:10:01,410 I I think, really, it just empowers people. 288 00:10:01,470 --> 00:10:03,090 Sure. You know, the coders 289 00:10:03,629 --> 00:10:05,970 on, you know, both our teams and 290 00:10:06,429 --> 00:10:08,509 Paul and Dee's team, they're able to rather 291 00:10:08,509 --> 00:10:10,350 than focus on things that are pretty rote 292 00:10:10,350 --> 00:10:11,009 and monotonous 293 00:10:11,945 --> 00:10:12,445 and 294 00:10:12,904 --> 00:10:15,705 really dive into the details and, like, build 295 00:10:15,705 --> 00:10:18,424 better systems rather than just, like, basically plugging 296 00:10:18,424 --> 00:10:20,184 holes in the dam. Yeah. And I think 297 00:10:20,184 --> 00:10:22,424 that that is something that is, you know, 298 00:10:22,424 --> 00:10:24,745 really exciting. And, you know, ultimately, the promise 299 00:10:24,745 --> 00:10:26,105 of this is, like, you 300 00:10:26,970 --> 00:10:28,669 with these tools, you can basically 301 00:10:29,529 --> 00:10:32,750 maximize your executive function rather than just, like, 302 00:10:32,889 --> 00:10:34,889 getting tasks done. Yeah. And to hit home 303 00:10:34,889 --> 00:10:37,049 on that where you use empowerment of these 304 00:10:37,049 --> 00:10:39,370 physicians, are you finding that that really fits 305 00:10:39,370 --> 00:10:42,589 into how Apollo MD prefers to operate? 306 00:10:43,424 --> 00:10:43,924 Absolutely. 307 00:10:44,384 --> 00:10:45,904 You know, we are we're kind of a 308 00:10:45,904 --> 00:10:47,924 clinician cooperative. So we have, 309 00:10:48,544 --> 00:10:50,865 an ownership model that allows our physicians, our 310 00:10:50,865 --> 00:10:53,264 PAs, our nurse practitioners, our company employees to 311 00:10:53,264 --> 00:10:54,485 all own in the organization. 312 00:10:55,100 --> 00:10:57,279 So that need for empowerment 313 00:10:57,899 --> 00:10:59,840 and autonomy of work, 314 00:11:00,379 --> 00:11:02,540 is kind of ingrained in us. Sure. And 315 00:11:02,540 --> 00:11:05,279 frankly, the ability to have better transparency 316 00:11:05,740 --> 00:11:07,519 and reliability on the coding 317 00:11:08,875 --> 00:11:11,195 really kind of facilitates that. It also, as 318 00:11:11,195 --> 00:11:14,075 Andrew pointed out, rather than displace people, we 319 00:11:14,075 --> 00:11:15,855 were a lot we were able to focus 320 00:11:16,394 --> 00:11:19,595 our coding folks into an audit program that 321 00:11:19,595 --> 00:11:20,575 really has taken, 322 00:11:20,960 --> 00:11:22,799 I think, the fidelity of the entire system 323 00:11:22,799 --> 00:11:24,960 to another level. So, you know, and when 324 00:11:24,960 --> 00:11:26,419 we onboard new clinicians, 325 00:11:26,960 --> 00:11:29,120 we're able to focus those auditors on their 326 00:11:29,120 --> 00:11:31,679 work and give them early feedback, which, you 327 00:11:31,679 --> 00:11:33,759 know, is kind of a it's it's a 328 00:11:33,759 --> 00:11:36,019 lifelong journey for them. Right? So, 329 00:11:37,414 --> 00:11:39,995 yeah, very excited about the future with Fathom. 330 00:11:40,054 --> 00:11:41,514 Sure. So, Andrew, 331 00:11:41,975 --> 00:11:43,274 let's talk about another 332 00:11:44,534 --> 00:11:46,634 administrative favorite for for physicians, 333 00:11:47,014 --> 00:11:49,574 and I use favorite lately. The issue with 334 00:11:49,574 --> 00:11:50,970 payers administrative processes, 335 00:11:51,370 --> 00:11:52,829 processes, prior authorization, 336 00:11:53,689 --> 00:11:54,189 denials, 337 00:11:55,129 --> 00:11:56,189 submitting claims, 338 00:11:57,289 --> 00:11:59,610 all considering all of that, how does Fathom 339 00:11:59,610 --> 00:12:02,490 help Apollo MD navigate what we know are 340 00:12:02,490 --> 00:12:05,370 increasing challenges with payer denials and audits right 341 00:12:05,370 --> 00:12:08,035 now? So this touches nicely on your earlier 342 00:12:08,095 --> 00:12:11,375 question around why automation rates matter. Mhmm. The 343 00:12:11,375 --> 00:12:14,014 biggest benefit to AI is your ability to 344 00:12:14,014 --> 00:12:14,514 programmatically 345 00:12:15,055 --> 00:12:16,894 perform certain tasks. And a lot of the 346 00:12:16,894 --> 00:12:18,675 denials, they come about because, 347 00:12:19,055 --> 00:12:21,149 you know, what a lot of payers work 348 00:12:21,149 --> 00:12:22,910 on is they have these sort of disparate 349 00:12:22,910 --> 00:12:25,470 guidelines that, you know, this will never reimburse 350 00:12:25,470 --> 00:12:28,509 this procedure with this diagnosis code, etcetera. And 351 00:12:28,509 --> 00:12:30,750 these can be insanely hard to track and 352 00:12:30,750 --> 00:12:33,250 enforce against a large, you know, manual 353 00:12:33,855 --> 00:12:36,495 pool of workers, but, like, basically, like, the 354 00:12:36,495 --> 00:12:38,414 Paul and d MD team comes to us 355 00:12:38,414 --> 00:12:39,934 and says, like, hey. For this payer, we 356 00:12:39,934 --> 00:12:42,274 can't submit something that looks like this anymore. 357 00:12:42,575 --> 00:12:44,274 You know, that's a that's a code change. 358 00:12:44,414 --> 00:12:46,034 And then it's enforced across 359 00:12:46,620 --> 00:12:49,259 not 50% of the volume, 90% of the 360 00:12:49,259 --> 00:12:51,200 volume. And so you're inherently 361 00:12:51,660 --> 00:12:54,000 your your coding related denials go down. 362 00:12:54,620 --> 00:12:57,259 And, also, you know, the other benefit is 363 00:12:57,259 --> 00:13:00,220 with that high automation rate, your compliance processes 364 00:13:00,220 --> 00:13:02,504 go up because you've you've created these very 365 00:13:02,504 --> 00:13:05,625 robust coding guidelines and documentation on how you're 366 00:13:05,625 --> 00:13:08,424 gonna approach things, and it's enforced consistently. So 367 00:13:08,424 --> 00:13:09,705 that if you are in an audit and 368 00:13:09,705 --> 00:13:11,304 they're like, hey. Why did something happen this 369 00:13:11,304 --> 00:13:13,144 way? There's not a ton of inconsistencies in 370 00:13:13,144 --> 00:13:15,399 the system. There basically is like, hey. This 371 00:13:15,399 --> 00:13:16,299 is very consistent. 372 00:13:16,759 --> 00:13:18,200 The same way we wrote this down, the 373 00:13:18,200 --> 00:13:19,559 same thing that we agreed to, and this 374 00:13:19,559 --> 00:13:20,459 is our logic. 375 00:13:21,000 --> 00:13:23,159 Yeah. No. That makes sense. And, doctor Patel, 376 00:13:23,159 --> 00:13:24,919 what's your experience? And what are you hearing 377 00:13:24,919 --> 00:13:25,820 from your employees? 378 00:13:26,644 --> 00:13:28,404 Yeah. I mean, I for us, you know, 379 00:13:28,404 --> 00:13:30,105 it has taken a lot of the variability 380 00:13:30,404 --> 00:13:31,065 of documentation 381 00:13:31,605 --> 00:13:33,285 out of the work. And, you know, an 382 00:13:33,285 --> 00:13:34,965 earlier point that Andrew made is I'd say 383 00:13:34,965 --> 00:13:37,144 half of our practices are epic sites. 384 00:13:37,764 --> 00:13:40,165 The other half are are spread across a 385 00:13:40,165 --> 00:13:43,309 smattering of EHR. So, you know, Cerner, Meditech, 386 00:13:43,449 --> 00:13:44,269 Med Host, 387 00:13:44,730 --> 00:13:47,309 and and each has its own documentation 388 00:13:47,769 --> 00:13:48,669 kind of idiosyncrasies. 389 00:13:49,289 --> 00:13:51,289 So the ability for us to message to 390 00:13:51,289 --> 00:13:52,669 our teams to use 391 00:13:53,049 --> 00:13:54,110 standard language 392 00:13:54,490 --> 00:13:56,085 in the way they talk about 393 00:13:57,125 --> 00:13:57,945 certain procedures. 394 00:13:58,804 --> 00:14:00,884 FATHOM, you know, picks that up with such 395 00:14:00,884 --> 00:14:02,184 a high degree of accuracy 396 00:14:02,565 --> 00:14:05,384 that it has minimized, you know, lost procedures, 397 00:14:06,085 --> 00:14:06,585 missing, 398 00:14:07,125 --> 00:14:08,184 you know, electrocardiograms 399 00:14:08,644 --> 00:14:09,304 or CPTs. 400 00:14:09,925 --> 00:14:12,149 So for our clinicians, it's, you know, I 401 00:14:12,149 --> 00:14:14,329 I think net net has been a huge 402 00:14:14,789 --> 00:14:16,490 validation of their work 403 00:14:16,870 --> 00:14:18,889 because they're doing the work at the bedside. 404 00:14:19,350 --> 00:14:20,649 But, you know, documentation 405 00:14:21,350 --> 00:14:22,569 documentation challenges, 406 00:14:23,269 --> 00:14:25,644 sometimes leave them without getting full credit for 407 00:14:25,644 --> 00:14:27,485 the work they do. Sure. So fair to 408 00:14:27,485 --> 00:14:31,164 say you've seen payer denials decrease then? Yes. 409 00:14:31,164 --> 00:14:32,764 And and, you know, some of that is 410 00:14:32,764 --> 00:14:34,764 we've we've layered on other technologies as well, 411 00:14:34,764 --> 00:14:36,284 but Fathom has been a big part of 412 00:14:36,284 --> 00:14:37,424 that for us. Absolutely. 413 00:14:37,830 --> 00:14:38,330 It's 414 00:14:38,710 --> 00:14:41,769 attribution can be difficult because pair denials, generically, 415 00:14:41,910 --> 00:14:42,570 are increasing. 416 00:14:42,950 --> 00:14:45,129 And so to, you know, know 417 00:14:45,750 --> 00:14:48,790 exactly what's what's what and what's because Fathom's 418 00:14:48,790 --> 00:14:50,790 doing a better job versus, you know, what 419 00:14:50,790 --> 00:14:53,924 is some unnamed plan tightening screws? Hard to 420 00:14:53,924 --> 00:14:55,924 know. Yeah. That is a great point. So, 421 00:14:55,924 --> 00:14:59,284 Andrew, for for providers considering autonomous coding that 422 00:14:59,284 --> 00:15:01,365 really haven't dipped their toes into this space 423 00:15:01,365 --> 00:15:01,865 yet, 424 00:15:02,485 --> 00:15:05,019 advice that you'd give for all the leaders 425 00:15:05,019 --> 00:15:07,019 listening in in terms of selecting the right 426 00:15:07,019 --> 00:15:07,519 partner, 427 00:15:07,899 --> 00:15:08,960 managing the implementation, 428 00:15:09,740 --> 00:15:11,040 what what goes into this? 429 00:15:11,660 --> 00:15:13,820 I mean, I'd say the biggest thing in 430 00:15:13,820 --> 00:15:14,960 selecting a partner 431 00:15:15,980 --> 00:15:16,480 is 432 00:15:17,179 --> 00:15:19,759 make the contract reflect the promises made. 433 00:15:20,565 --> 00:15:22,965 You know, the industry has evolved a lot. 434 00:15:22,965 --> 00:15:25,605 It's much more mature. The technology is much 435 00:15:25,605 --> 00:15:28,085 more mature. Sure. You know, serve vendors like 436 00:15:28,085 --> 00:15:29,785 Fathom know what we can deliver. 437 00:15:30,725 --> 00:15:32,889 And, yeah, if we say x, like, put 438 00:15:32,889 --> 00:15:34,570 in the contract. We get punitive if we 439 00:15:34,570 --> 00:15:36,250 didn't. Like, when we got started with Apollo 440 00:15:36,250 --> 00:15:37,709 MD, it was like, hey. Like, 441 00:15:38,089 --> 00:15:39,929 this is new. Like, we think this, but 442 00:15:39,929 --> 00:15:41,309 the range was pretty wide. 443 00:15:41,690 --> 00:15:43,610 But, you know, we're now in a place 444 00:15:43,610 --> 00:15:45,949 where, you know, you should hold your vendors 445 00:15:46,009 --> 00:15:48,644 to SLAs on accuracy, SLAs on 446 00:15:49,205 --> 00:15:51,925 automation. Because automation, it, you know, drives all 447 00:15:51,925 --> 00:15:54,165 these compliance benefits. It drives your ROI. And 448 00:15:54,165 --> 00:15:56,105 if you're not getting that, then, you know, 449 00:15:56,485 --> 00:15:58,404 there should be a penalty. The other thing 450 00:15:58,404 --> 00:16:01,240 is make them demonstrate their wares. Like, see 451 00:16:01,240 --> 00:16:03,420 the AI work across a large volume, 452 00:16:04,360 --> 00:16:05,820 before you commit to anything 453 00:16:06,200 --> 00:16:09,480 or make it contingent on seeing that because, 454 00:16:09,480 --> 00:16:11,000 you know, the nice thing is is, like, 455 00:16:11,000 --> 00:16:12,860 most of these systems can run at relatively 456 00:16:12,920 --> 00:16:15,100 low marginal cost. And so, 457 00:16:15,434 --> 00:16:18,075 you know, push on that. Yeah. Absolutely. And 458 00:16:18,075 --> 00:16:19,835 doctor Patel, let me reframe the question for 459 00:16:19,835 --> 00:16:21,695 you. What do you wish you had known 460 00:16:21,754 --> 00:16:22,254 before 461 00:16:22,634 --> 00:16:23,455 you had entered, 462 00:16:23,995 --> 00:16:26,654 looking for a partner within autonomous coding? 463 00:16:27,355 --> 00:16:29,774 Yeah. I mean, 5 years ago for us, 464 00:16:30,154 --> 00:16:30,815 we were 465 00:16:31,589 --> 00:16:33,929 relatively novice with the whole idea of automation 466 00:16:34,230 --> 00:16:35,529 and artificial intelligence. 467 00:16:35,990 --> 00:16:38,250 So I think there's been a huge leap 468 00:16:38,629 --> 00:16:39,929 in just familiarity 469 00:16:40,470 --> 00:16:43,769 and comfort with with this type of engagement. 470 00:16:44,389 --> 00:16:46,309 What we did do well, I think, was 471 00:16:46,309 --> 00:16:48,235 that quality was our North Star. 472 00:16:48,615 --> 00:16:50,855 And we kind of had this relentless pursuit 473 00:16:50,855 --> 00:16:51,514 of accuracy 474 00:16:52,215 --> 00:16:54,154 and sustainability in this process. 475 00:16:54,535 --> 00:16:57,514 And I think we continually return to that 476 00:16:57,815 --> 00:16:59,674 to say as we're measuring milestones, 477 00:17:00,134 --> 00:17:02,710 are we are we more accurate? Are we 478 00:17:03,330 --> 00:17:05,730 are we higher quality than we were? And 479 00:17:05,730 --> 00:17:06,950 I think that has informed 480 00:17:07,809 --> 00:17:09,349 much of the journey for us. 481 00:17:09,970 --> 00:17:11,570 You know, I we spent we spent our 482 00:17:11,570 --> 00:17:13,809 time on this. Right? And and again, as 483 00:17:13,809 --> 00:17:15,970 Andrew pointed out in the earlier days of 484 00:17:15,970 --> 00:17:19,164 this, we were very methodical about a phase 485 00:17:19,164 --> 00:17:20,625 rollout across our sites, 486 00:17:21,005 --> 00:17:21,505 about 487 00:17:22,045 --> 00:17:22,545 validation, 488 00:17:22,924 --> 00:17:25,484 validation of the model, parallel coding, you know, 489 00:17:25,484 --> 00:17:26,785 third party audits, 490 00:17:27,565 --> 00:17:30,224 and then really developing an auditing system 491 00:17:31,220 --> 00:17:33,080 that could be continued going forward. 492 00:17:33,619 --> 00:17:36,099 I think they have that playbook now. And 493 00:17:36,099 --> 00:17:37,720 and I think for their newer clients, 494 00:17:38,500 --> 00:17:40,500 can expedite a lot of those things, including 495 00:17:40,500 --> 00:17:42,820 the onboarding of new clients. So we're proud 496 00:17:42,820 --> 00:17:44,340 to have been part of that journey. That's 497 00:17:44,340 --> 00:17:46,944 great advice. What else are we missing? Anything 498 00:17:46,944 --> 00:17:48,625 we didn't get to today that our listeners 499 00:17:48,625 --> 00:17:49,285 should know? 500 00:17:49,744 --> 00:17:51,904 I would say if you're a a a 501 00:17:51,904 --> 00:17:54,304 health system out there that still is relying 502 00:17:54,304 --> 00:17:54,804 on 503 00:17:55,424 --> 00:17:55,924 humans 504 00:17:56,224 --> 00:17:58,704 or in particular, physicians to do their own 505 00:17:58,704 --> 00:17:59,204 coding, 506 00:17:59,859 --> 00:18:02,179 you you need to run, not walk to 507 00:18:02,179 --> 00:18:02,679 a 508 00:18:03,299 --> 00:18:04,119 more sustainable, 509 00:18:05,220 --> 00:18:06,679 more compliant solution. 510 00:18:06,980 --> 00:18:10,019 And I think the promise of automating revenue 511 00:18:10,019 --> 00:18:11,799 cycle is already here. 512 00:18:12,434 --> 00:18:14,454 Yeah. I dovetail on that. I think 513 00:18:14,835 --> 00:18:16,274 one of the things that has me most 514 00:18:16,274 --> 00:18:18,674 excited about Fathom is the recent movement in 515 00:18:18,674 --> 00:18:19,654 clients that 516 00:18:20,274 --> 00:18:21,174 are in 517 00:18:21,634 --> 00:18:24,514 practices health systems where clinician coding is being 518 00:18:24,514 --> 00:18:26,660 performed. And what we're seeing in terms of 519 00:18:26,660 --> 00:18:29,160 impact of clinician experience and bandwidth 520 00:18:29,940 --> 00:18:33,059 and ability to better deliver care because an 521 00:18:33,059 --> 00:18:35,059 AI is sitting there performing the coding on 522 00:18:35,059 --> 00:18:37,779 their behalf. Mhmm. And further, what we are 523 00:18:37,779 --> 00:18:39,720 discovering is, generally speaking, 524 00:18:40,625 --> 00:18:42,884 most health care organizations are undercoding. 525 00:18:43,664 --> 00:18:45,505 And so not only do you get a 526 00:18:45,505 --> 00:18:46,945 step up in compliance, but you used to 527 00:18:46,945 --> 00:18:48,244 get a step up in acuity 528 00:18:48,625 --> 00:18:50,945 and ensure that you're being fully reimbursed or 529 00:18:50,945 --> 00:18:53,345 accurately reimbursed for all the work that your 530 00:18:53,345 --> 00:18:54,325 team is performing. 531 00:18:55,210 --> 00:18:56,890 Absolutely. Well, I think that's a great place 532 00:18:56,890 --> 00:18:59,130 to leave it. Gentlemen, thank you so much 533 00:18:59,130 --> 00:19:00,170 for taking the time to be with us 534 00:19:00,170 --> 00:19:00,910 on the podcast, 535 00:19:01,210 --> 00:19:03,690 for sharing your insights with our listeners. We 536 00:19:03,690 --> 00:19:06,330 really appreciate it. Thank you. Thank you. If 537 00:19:06,330 --> 00:19:07,125 you'd like to listen 538 00:19:07,765 --> 00:19:10,085 more podcasts from Becker's HealthCare, you can visit 539 00:19:10,085 --> 00:19:11,865 our podcast page at beckershospitalreview.com. 540 00:19:13,765 --> 00:19:15,924 I'd also like to thank Fathom as well 541 00:19:15,924 --> 00:19:17,704 for sponsoring this episode.