WEBVTT 1 00:00:00.200 --> 00:00:03.960 Welcome to Farmer Talk Radio. I'm Andrew Goldstein. In this podcast, 2 00:00:04.040 --> 00:00:07.320 I'll be joined by raw Lima, SVP of Strategic Clinical 3 00:00:07.360 --> 00:00:12.320 Operations of Inception Group, about clinical trial rescue. In this conversation, 4 00:00:12.480 --> 00:00:15.839 we're going to address a scaffolding approach where chief medical 5 00:00:15.839 --> 00:00:19.039 officers of biotech companies can prioritize the areas that need 6 00:00:19.079 --> 00:00:22.679 immediate support without sacrificing the entirety of the trial and 7 00:00:22.719 --> 00:00:26.399 thereby safe time, money and energy rall. Why don't you 8 00:00:26.399 --> 00:00:27.160 introduce yourself. 9 00:00:27.800 --> 00:00:30.559 Hi, my name is raw Lima. I'm the executive vice 10 00:00:30.559 --> 00:00:33.880 president of Clinical Operations for a company called Inception Group. 11 00:00:34.679 --> 00:00:37.280 I've got about twenty five years of experience in the industry, 12 00:00:37.359 --> 00:00:40.960 most of that as a VP of clinups for very 13 00:00:41.000 --> 00:00:42.479 small biotech sponsors. 14 00:00:42.719 --> 00:00:45.399 To get us started, can you define what you mean 15 00:00:45.520 --> 00:00:48.359 when you talk about a rescue and what you mean 16 00:00:48.399 --> 00:00:49.719 when you talk about scaffolding. 17 00:00:50.439 --> 00:00:52.799 Yeah, I know this one really touched a nerve at 18 00:00:52.840 --> 00:00:55.520 the CMO conference when we first started talking about it. 19 00:00:55.640 --> 00:00:58.520 What I'm really getting at with a rescue or when 20 00:00:58.600 --> 00:01:02.200 people historically and my experience talk about rescues, they're really 21 00:01:02.200 --> 00:01:06.239 talking about completely changing their vendor, right they're talking about 22 00:01:06.680 --> 00:01:09.560 ripping their study up, study out from the roots and 23 00:01:10.239 --> 00:01:14.560 almost starting fresh right. That is very much a nuclear option, 24 00:01:15.079 --> 00:01:19.680 and that is very expensive, very difficult operationally to pull 25 00:01:19.719 --> 00:01:22.799 off without tanking the whole study. And so what I'm 26 00:01:22.840 --> 00:01:26.640 what I'm actually talking about is coming in and rescuing 27 00:01:26.680 --> 00:01:29.439 a study right doing what is in the word without 28 00:01:29.480 --> 00:01:34.040 going nuclear and the the or without I'm going to 29 00:01:34.159 --> 00:01:35.920 use a building analogy and we're going to beat this 30 00:01:36.000 --> 00:01:40.519 analogy up throughout this whole podcast. You don't have to 31 00:01:40.560 --> 00:01:42.719 demo the building to build a new building, right the 32 00:01:42.760 --> 00:01:46.120 way that we refab buildings in New York City and all, 33 00:01:46.519 --> 00:01:49.840 we're not knocking everything down. We're putting up scaffolding and 34 00:01:49.879 --> 00:01:52.519 we're repairing it from the insides, and we're targeting what 35 00:01:52.680 --> 00:01:55.040 needs to be fixed. And so if you use that 36 00:01:55.079 --> 00:01:58.480 analogy that it really is the what I am talking about. 37 00:01:58.519 --> 00:02:02.760 It's it's putting some controls around what you do have 38 00:02:03.319 --> 00:02:06.480 and targeted fixing the things that you might need to fix. 39 00:02:07.079 --> 00:02:10.439 But what that really comes down to is identifying what 40 00:02:10.599 --> 00:02:13.919 needs to be fixed rather than focusing on the symptoms, 41 00:02:14.520 --> 00:02:18.080 which are often bad enrollment. The costs are getting out 42 00:02:18.120 --> 00:02:20.759 of control. Those are the sorts of things that get 43 00:02:20.800 --> 00:02:24.360 people to say we need a rescue, but really identifying 44 00:02:24.400 --> 00:02:28.400 what is the problem and getting to the core issue 45 00:02:28.719 --> 00:02:31.680 so that you can keep what's good and fix what's 46 00:02:31.719 --> 00:02:35.319 wrong and support the greater infrastructure that is your study. 47 00:02:35.960 --> 00:02:39.960 So how can biotechs better identify the root causes of 48 00:02:40.000 --> 00:02:43.560 these issues so that they're treating those rather than treating 49 00:02:43.599 --> 00:02:44.120 a symptom. 50 00:02:44.680 --> 00:02:48.680 This is very much an art form and it is 51 00:02:48.759 --> 00:02:51.360 the million dollar question, right, But if there are going 52 00:02:51.439 --> 00:02:55.199 to be problems, they really come down to first you'll 53 00:02:55.240 --> 00:02:58.120 hear me. Start, there is a fundamental truth to the 54 00:02:58.159 --> 00:03:01.639 operational reality of your study. So sit down and map 55 00:03:01.719 --> 00:03:05.199 your study. Who are the vendors, Where is that information 56 00:03:05.400 --> 00:03:08.240 flowing right, Where does the information get collected, How does 57 00:03:08.280 --> 00:03:10.319 it flow into a database, How does it flow into 58 00:03:10.319 --> 00:03:14.680 your clinical database? Who is engaging with the patients? Right? 59 00:03:14.719 --> 00:03:18.280 So really map your study out so that you understand 60 00:03:18.719 --> 00:03:22.039 how it works. And when I'm saying this, I'm really 61 00:03:22.039 --> 00:03:27.159 speaking to the executives, right, the chief medical offices, the CEOs. 62 00:03:27.840 --> 00:03:30.000 I'm assuming here that you have a clinops team who 63 00:03:30.080 --> 00:03:32.680 might have already done this, and if you don't, you 64 00:03:33.039 --> 00:03:35.680 should probably start there. And get one who can help 65 00:03:35.719 --> 00:03:38.479 you with this. But very often, and I've had this 66 00:03:38.599 --> 00:03:42.840 experience being a lifelong clinops person, that very often that 67 00:03:42.960 --> 00:03:47.280 work gets done at the ground level but really doesn't 68 00:03:47.360 --> 00:03:51.560 get fed up into management in sort of the detail 69 00:03:51.639 --> 00:03:54.039 that it probably needs to be. Particularly when things are 70 00:03:54.039 --> 00:03:57.560 going wrong. What you end up getting is opinions of 71 00:03:57.599 --> 00:04:00.759 what is going wrong rather than sort of really trying 72 00:04:00.800 --> 00:04:04.439 to identify yourself. But the targets are you know, ultimately 73 00:04:04.479 --> 00:04:08.120 they come down to a few targets, the sites, right, 74 00:04:08.159 --> 00:04:11.639 So are the sites able to execute the protocol? Are 75 00:04:12.360 --> 00:04:14.639 the patients able to execute the protocol? Right? 76 00:04:14.719 --> 00:04:14.800 Like? 77 00:04:14.840 --> 00:04:17.079 What is the burden you are putting on? And I 78 00:04:17.079 --> 00:04:20.160 don't mean theoretical burden in terms of, you know, this 79 00:04:20.199 --> 00:04:23.759 assessment takes fifteen minutes, this assessed assessment takes thirty minutes, 80 00:04:24.399 --> 00:04:27.480 But there's an there's a real burden that comes Sometimes 81 00:04:27.560 --> 00:04:30.480 those assessments aren't done in the same place in a hospital, 82 00:04:30.800 --> 00:04:34.079 and so if you stack them up, and sometimes if 83 00:04:34.120 --> 00:04:36.920 you stack them up, it becomes very burdensome. And if 84 00:04:36.920 --> 00:04:40.399 you stack those things up with timeframes and they are 85 00:04:40.439 --> 00:04:44.040 done on opposite sides of a hospital campus, that is 86 00:04:45.519 --> 00:04:48.560 virtually impossible for sites to do, and that will will 87 00:04:48.560 --> 00:04:52.600 come out in things like protocol deviations. At best. At best, 88 00:04:52.680 --> 00:04:55.639 people will try and you will get failures. But what 89 00:04:55.839 --> 00:04:59.399 at worst what you see here is people don't do 90 00:04:59.439 --> 00:05:02.199 your protocol. They just say it's impossible and they quit 91 00:05:02.279 --> 00:05:05.519 on you. And so having that discussion, and not just 92 00:05:05.600 --> 00:05:09.360 with the pis, but with the research nurses, the study coordinators, 93 00:05:09.600 --> 00:05:12.680 finding ways to touch other parts of the institutions so 94 00:05:12.680 --> 00:05:16.839 that you really understand what is going on. So that's 95 00:05:16.839 --> 00:05:19.680 the first is really understand how this is happening at sites. 96 00:05:20.120 --> 00:05:22.759 The second is, well, who's interacting with those sites and 97 00:05:22.800 --> 00:05:27.160 who's asking those questions? And very often, particularly small biotechs 98 00:05:27.399 --> 00:05:30.040 will be very hands off with this. They hire a 99 00:05:30.079 --> 00:05:33.800 big cro they hire some contractors, and then they get 100 00:05:33.800 --> 00:05:36.959 hands off because they say it is their job. At 101 00:05:36.959 --> 00:05:38.839 the end of the day. That might be right, it 102 00:05:38.920 --> 00:05:41.959 is their job, but it is ultimately your responsibility and 103 00:05:42.000 --> 00:05:45.199 it is clearly when it goes wrong, your problem. So 104 00:05:45.560 --> 00:05:48.399 spend some energy there and do it yourself, or get 105 00:05:48.839 --> 00:05:52.439 very experienced people who you trust, know, maybe have experience with, 106 00:05:52.639 --> 00:05:55.199 have worked with before, and bring them in. And this 107 00:05:55.240 --> 00:05:57.720 is what I'm talking about is scaffolding right? You may 108 00:05:57.720 --> 00:06:00.040 have hired somebody for it, but you don't have to 109 00:06:00.079 --> 00:06:02.759 fire them and hire somebody new. Maybe just get a 110 00:06:02.759 --> 00:06:05.759 targeted approach to go do things like talk to your sites. 111 00:06:06.279 --> 00:06:10.600 Maybe have a targeted approach of having an independent third 112 00:06:10.600 --> 00:06:13.879 party person come in and talk to both your clinical 113 00:06:13.920 --> 00:06:18.120 operations group and your vendors and figure out what's going wrong. 114 00:06:18.240 --> 00:06:21.839 The last piece or the second to last piece that 115 00:06:22.040 --> 00:06:25.240 could be wrong. Is the vendors themselves right? So is 116 00:06:25.279 --> 00:06:28.600 their staffing appropriate or do they have the right kinds 117 00:06:28.600 --> 00:06:30.879 of people doing the work? Have you had a lot 118 00:06:30.879 --> 00:06:34.439 of turnover? Right? Those are sort of red flag signals. 119 00:06:34.600 --> 00:06:37.399 And then the last really is in the contract. And 120 00:06:37.480 --> 00:06:41.360 so I've met again, I've been in experience of this 121 00:06:41.959 --> 00:06:45.399 where we as biotechs have done it to ourselves, where 122 00:06:45.439 --> 00:06:47.600 we put out an RFP, it came back, we didn't 123 00:06:47.639 --> 00:06:50.079 like the price, so we redlined it, and then it 124 00:06:50.120 --> 00:06:52.079 turns out we needed all of those things, and then 125 00:06:52.120 --> 00:06:54.800 that shows up later, right, And when that shows up later, 126 00:06:55.040 --> 00:06:57.800 we tend to be surprised by it, but it's very 127 00:06:57.839 --> 00:07:01.120 often shouldn't be surprised by it. Really looking at that 128 00:07:01.199 --> 00:07:05.000 contract upfront, like being honest with yourself about what you 129 00:07:05.079 --> 00:07:08.480 really do need, and then keeping a documentation of what 130 00:07:08.519 --> 00:07:11.519 we've readlined so we're not surprised by it in the future. 131 00:07:11.800 --> 00:07:14.399 Keeping them in things like a risk log are important. 132 00:07:14.439 --> 00:07:16.519 Way is to sort of mitigate this in the future. 133 00:07:16.560 --> 00:07:19.160 I'm not saying don't readline. I'm not saying don't try 134 00:07:19.160 --> 00:07:22.879 to be efficient, but just be honest with yourself. Be 135 00:07:22.959 --> 00:07:26.519 honest with your future self about some of the compromises 136 00:07:26.560 --> 00:07:29.120 you might have made, because it does make it easier 137 00:07:29.120 --> 00:07:32.319 to reckon with them and identify what those problems might 138 00:07:32.360 --> 00:07:34.279 be if you have a strong record of it. And 139 00:07:34.959 --> 00:07:38.079 today it's never too late to do that activity. And 140 00:07:38.160 --> 00:07:40.560 so if your contract was written poorly two years ago, 141 00:07:40.680 --> 00:07:43.639 or maybe you're a new team on a failing study 142 00:07:43.680 --> 00:07:45.759 and you're trying to make that decision, today is a 143 00:07:45.759 --> 00:07:48.600 great day to start looking in the mirror and being 144 00:07:48.639 --> 00:07:51.240 honest about the contracts you have in front of you 145 00:07:51.560 --> 00:07:54.639 and what they do and do not include, and trying 146 00:07:54.680 --> 00:07:57.240 to reckon with that all at once, because it is 147 00:07:57.319 --> 00:08:00.839 very painful to do it incrementally. That that is painful 148 00:08:00.839 --> 00:08:02.720 for the vendors, and it is painful for the sites. 149 00:08:02.759 --> 00:08:06.639 It is painful for the overall study. So getting at 150 00:08:06.680 --> 00:08:12.759 the sites, vendors, contracts are really the three and the 151 00:08:12.879 --> 00:08:15.480 last pillar, if we're going to beat the house metaphor, 152 00:08:15.639 --> 00:08:19.680 the last pillar is the protocol itself, right, So the protocol, 153 00:08:19.759 --> 00:08:22.680 the sites, the vendors themselves, and the contracts. 154 00:08:23.199 --> 00:08:27.319 So I understand the contracts, and I understand the protocol. 155 00:08:27.560 --> 00:08:29.680 But you know, when you're talking about site burden, or 156 00:08:29.680 --> 00:08:32.360 you're talking about vendors who aren't educating off actively, or 157 00:08:32.360 --> 00:08:34.919 don't have the right staffing or have turnover, you know, 158 00:08:34.960 --> 00:08:38.120 why is the answer scaffolding rather than you know, a 159 00:08:38.159 --> 00:08:40.360 rescue and going to a different group to work. 160 00:08:40.200 --> 00:08:43.080 With that the ultimate answer might be going to a 161 00:08:43.120 --> 00:08:45.519 different group. What you want to be able to do, though, 162 00:08:45.600 --> 00:08:48.600 is a soft landing, right, And so cutting across into 163 00:08:48.679 --> 00:08:51.840 a new organization that has all sorts of new rules 164 00:08:51.919 --> 00:08:54.720 and a new way of working might be a similar 165 00:08:54.759 --> 00:08:59.200 shock to the system as before. So what I'm actually 166 00:08:59.240 --> 00:09:02.559 recommending is putting up scaffolding to make sure you're solving 167 00:09:02.600 --> 00:09:05.240 the right problem. Because, for example, if the problem is 168 00:09:05.240 --> 00:09:08.440 the protocol, but you think that it is the staff, 169 00:09:09.159 --> 00:09:11.600 and you go and you fire your cro and then 170 00:09:11.639 --> 00:09:13.879 you hire a new one that is a very costly 171 00:09:13.919 --> 00:09:18.320 and painful activity that ultimately won't get you anything because 172 00:09:18.840 --> 00:09:22.159 your actual real problem was the burden on the sites 173 00:09:22.600 --> 00:09:26.360 from protocol assessments. And then the new CRO is going 174 00:09:26.399 --> 00:09:29.960 to come in and have very similar challenges and you 175 00:09:30.039 --> 00:09:32.600 might have very similar outcomes, and so if you're not 176 00:09:32.799 --> 00:09:36.960 identifying the right problem, then you can't actually solve for it. 177 00:09:37.399 --> 00:09:41.279 And sometimes what you have with for example, staffing issues, 178 00:09:41.519 --> 00:09:43.960 Let's say with cras, this is a very common one 179 00:09:44.399 --> 00:09:46.679 where you know a lot of the cras out in 180 00:09:46.480 --> 00:09:50.080 the field, particularly with larger organizations, don't have as much 181 00:09:50.120 --> 00:09:53.879 experience the company. Those companies are built for scale and 182 00:09:54.799 --> 00:09:59.440 they have some lesser experience staff, and if the issue 183 00:09:59.440 --> 00:10:03.480 is with those, then finding ways to support those just 184 00:10:03.559 --> 00:10:07.039 firing those staff and putting a new inexperienced staff, you 185 00:10:07.120 --> 00:10:08.960 might not get the kinds of people you want with 186 00:10:09.039 --> 00:10:11.840 the next one, and you're going to have to train 187 00:10:11.960 --> 00:10:13.720 that new person, and so you could get into this 188 00:10:13.840 --> 00:10:17.360 cycle of finding where what you really need is somebody 189 00:10:17.399 --> 00:10:21.039 to do some oversight and education of those people which 190 00:10:21.159 --> 00:10:25.440 may or may not exist inside of that CRO right 191 00:10:25.799 --> 00:10:28.080 because the CRO is never going to be an expert 192 00:10:28.120 --> 00:10:30.320 at your clinical trial. They're never going to be an 193 00:10:30.320 --> 00:10:34.000 expert at sort of the intricacies of your assessments and 194 00:10:34.039 --> 00:10:36.399 the way you want things done. They're just not right. 195 00:10:36.480 --> 00:10:39.639 They're not living your trials every day. Some of the 196 00:10:39.679 --> 00:10:43.600 boutique CROs do and engage differently, but the bigger ones don't. 197 00:10:43.879 --> 00:10:47.240 Like they're there to check boxes and to do source 198 00:10:47.279 --> 00:10:51.200 stark verification and collect documents, right, They're not really there 199 00:10:51.320 --> 00:10:54.879 to live your study. And so have somebody there to 200 00:10:54.919 --> 00:10:58.080 live your study and to help them through their day, 201 00:10:58.279 --> 00:11:01.639 help them be more efficient. And that's a real value 202 00:11:01.639 --> 00:11:04.279 add that you can put inside of your organization who 203 00:11:04.279 --> 00:11:06.600 can go across your programs or across all of your 204 00:11:06.639 --> 00:11:10.799 sites and across all of your vendors to create that consistency. 205 00:11:11.240 --> 00:11:14.159 And that really does look like your clinops group. But 206 00:11:14.240 --> 00:11:16.919 again people don't think of it this way. They think 207 00:11:16.919 --> 00:11:19.639 of that as extra staff that you don't need and 208 00:11:19.679 --> 00:11:22.399 that you can redline. But that is a function. If 209 00:11:22.440 --> 00:11:26.200 you separate everything into its little component pieces, then if 210 00:11:26.240 --> 00:11:29.120 there really is nobody to tie it all together, then 211 00:11:29.159 --> 00:11:32.440 what you often get is disassociated pieces. And it's not 212 00:11:32.519 --> 00:11:36.120 that those pieces are wrong, they're not coming together the 213 00:11:36.120 --> 00:11:40.159 way you want, and that often is what I'm talking 214 00:11:40.159 --> 00:11:42.480 about with scaffolding, Like, don't fire the people at the 215 00:11:42.480 --> 00:11:44.840 bottom who are making the pieces and get new people 216 00:11:44.879 --> 00:11:47.279 to make pieces, because they're just going to make those pieces. 217 00:11:47.480 --> 00:11:49.000 You need something different. 218 00:11:50.279 --> 00:11:54.080 So how can companies implement scaffolding effectively? 219 00:11:54.639 --> 00:11:56.120 You know, there's a lot of ways to go about this. 220 00:11:56.879 --> 00:12:01.600 Some have an instinct to go internal, right, so just 221 00:12:01.639 --> 00:12:05.759 to hire some extra people internally to oversee those activities. 222 00:12:05.919 --> 00:12:09.360 What I would say to most CEOs and CFOs cmos 223 00:12:09.480 --> 00:12:13.320 out there is that clinops is a very very difficult job. 224 00:12:13.559 --> 00:12:17.799 That is, you know, forty percent skill and like sixty 225 00:12:17.840 --> 00:12:21.320 percent solving problems right that come up emergently, and the 226 00:12:21.360 --> 00:12:24.879 more skilled people are prepared for those emergent problems. In 227 00:12:24.960 --> 00:12:27.519 order to do this well, you need somebody who's seen 228 00:12:27.559 --> 00:12:32.840 those problems before, who are prepared in have backup plans. 229 00:12:33.000 --> 00:12:37.039 It is a very hard job. Don't under resource it internally, right, 230 00:12:37.120 --> 00:12:39.879 Let's start there. Don't underresource it internally. There is a 231 00:12:39.919 --> 00:12:42.399 lot of information, there's a lot of data to look at. 232 00:12:42.679 --> 00:12:44.879 There's a lot that can go wrong, and when those 233 00:12:44.879 --> 00:12:48.720 things go wrong, you're already like two months too late 234 00:12:48.759 --> 00:12:51.399 to solve them. And so you really need people there 235 00:12:51.919 --> 00:12:55.360 living and understanding it in order to engage it well. 236 00:12:55.440 --> 00:12:59.679 So everybody wants to keep it into a tight internal staff. 237 00:12:59.799 --> 00:13:02.480 Just have some slack in your systems so that you 238 00:13:02.559 --> 00:13:07.759 can handle what I call the unexpected, the expected unexpected problems. Right, 239 00:13:07.759 --> 00:13:09.600 things are going to go wrong, make sure you have 240 00:13:09.639 --> 00:13:11.879 slack in the system to cover that. Also, make sure 241 00:13:11.919 --> 00:13:17.200 you have people who can talk at multiple levels to vendors. Right, 242 00:13:17.279 --> 00:13:23.159 So very often we biotechs are built with very very 243 00:13:23.240 --> 00:13:26.720 high level people in mind, and they have high science 244 00:13:27.120 --> 00:13:31.600 and high technology at their forefront. But if you if 245 00:13:31.639 --> 00:13:34.679 you skip down a few levels, the people who are 246 00:13:34.799 --> 00:13:39.159 cras don't necessarily have that same sort of skill set, right, 247 00:13:39.279 --> 00:13:42.399 They're just there's they're not going to have that skill 248 00:13:42.440 --> 00:13:45.919 set because their job is to be broad bridging the 249 00:13:45.960 --> 00:13:49.559 gap to understanding sort of what is coming up from 250 00:13:49.600 --> 00:13:54.799 the field, like the complaints and the difficulties, and then 251 00:13:54.879 --> 00:13:57.679 what the science is saying and how to implement that. 252 00:13:58.080 --> 00:14:02.120 You need people who can bridge that gap and so 253 00:14:02.960 --> 00:14:06.519 really focus on that. Those sorts of i'll call translation 254 00:14:06.679 --> 00:14:10.840 skill sets. It's not translations of language, it's translation of 255 00:14:11.000 --> 00:14:14.360 jargon that that you really you really need, and so 256 00:14:14.519 --> 00:14:19.519 you can find this potentially through consultants, through boutique CROs, 257 00:14:19.679 --> 00:14:23.799 through through internal hiring with the right sort of skill 258 00:14:23.879 --> 00:14:27.559 set approach, right, but be really mindful that you have 259 00:14:27.679 --> 00:14:31.639 different people with different levels of skill that all need 260 00:14:31.679 --> 00:14:35.279 to be communicated and messaged to, and that in and 261 00:14:35.320 --> 00:14:38.600 of itself is a skill set. I'd also say this 262 00:14:38.840 --> 00:14:42.519 on this point that as technology has become more and 263 00:14:42.600 --> 00:14:46.320 more infused into what we do, many of people, many 264 00:14:46.320 --> 00:14:50.120 of the people who deliver that technology are technology companies, 265 00:14:50.240 --> 00:14:53.679 not clinical companies, and so you really need somebody to 266 00:14:53.679 --> 00:14:57.159 do And I'm not talking data management in the classic 267 00:14:57.279 --> 00:15:00.720 sense i am I'm literally here talking about programming in 268 00:15:00.840 --> 00:15:06.679 IT and back end data in you know, architecture. You 269 00:15:06.759 --> 00:15:09.960 really need somebody who can speak that language as well, 270 00:15:10.039 --> 00:15:14.639 because you know, nowadays it is very common to have four, five, six, 271 00:15:16.360 --> 00:15:20.039 maybe even up to ten different data sets feeding your 272 00:15:20.159 --> 00:15:24.519 final analyzable data set. Getting all that information to be 273 00:15:24.679 --> 00:15:28.879 brought in and talk to each other is actually extraordinarily complex, 274 00:15:29.039 --> 00:15:33.360 and you really should have somebody dedicated to speaking that language, 275 00:15:33.360 --> 00:15:36.000 because if you don't, you will have problems tying the 276 00:15:36.039 --> 00:15:40.480 information together, and that is very hard to unwind. Once 277 00:15:40.559 --> 00:15:42.559 it is built right, it's a lot easier to build 278 00:15:42.559 --> 00:15:45.320 it right. Kind of add on to it, do transfers. 279 00:15:45.399 --> 00:15:48.240 If you build it wrong and the transfers don't work, 280 00:15:48.639 --> 00:15:51.120 that can be very painful and that often doesn't show 281 00:15:51.159 --> 00:15:54.080 itself until you are trying to get data out and 282 00:15:54.159 --> 00:15:57.360 it becomes a crisis. So really really focus on these 283 00:15:57.399 --> 00:16:01.360 points of translation, and don't underestimate both that they exist 284 00:16:01.480 --> 00:16:04.519 and that they are getting more and more complicated as 285 00:16:04.600 --> 00:16:07.720 the as sort of our industry fragments. 286 00:16:08.240 --> 00:16:10.919 Seems like you're looking for unicorn here who can speak 287 00:16:10.919 --> 00:16:13.679 the technology, you can speak broad and speak the jargon. 288 00:16:14.720 --> 00:16:15.759 How do you find these people? 289 00:16:16.519 --> 00:16:19.279 Well, I guess that's my sort of point is that 290 00:16:19.600 --> 00:16:22.559 it might not be It might not be a single person, 291 00:16:22.879 --> 00:16:26.720 and don't underestimate that it might need to be multiple people. 292 00:16:26.759 --> 00:16:30.080 And I get for small companies that that that may 293 00:16:30.120 --> 00:16:34.039 be difficult. But that's where I'm saying, borrow this from 294 00:16:34.279 --> 00:16:38.440 boutique CROs or from consulting companies. Get your one or 295 00:16:38.480 --> 00:16:42.039 two good clinops, people who know enough to kind of 296 00:16:42.039 --> 00:16:44.600 put their tendrils in and to really help to help 297 00:16:44.679 --> 00:16:48.080 you as an organization speak the language. But then borrow 298 00:16:48.159 --> 00:16:54.159 some independent experience from just bringing in your vendors, right 299 00:16:54.200 --> 00:16:58.080 because they might ask questions that you don't really understand, 300 00:16:58.639 --> 00:17:01.879 but you might think you understand, and that's where you 301 00:17:01.919 --> 00:17:04.559 go wrong. So just borrow that and make sure you 302 00:17:04.680 --> 00:17:08.720 have sort of solutions for checking your instinct and then 303 00:17:08.920 --> 00:17:12.160 if something does go wrong, borrow more of it. And 304 00:17:12.200 --> 00:17:15.880 that's the scaffolding approach, right like early on, try to 305 00:17:15.920 --> 00:17:18.279 infuse this into the way you build your program. But 306 00:17:18.319 --> 00:17:20.200 if it goes wrong and you're at a rescue and 307 00:17:20.240 --> 00:17:24.160 you need to scaffold, then go out and find independent 308 00:17:24.160 --> 00:17:27.240 people who are good at this particular problem that you 309 00:17:27.319 --> 00:17:29.640 need to solve, or the set of problems that you 310 00:17:29.720 --> 00:17:33.279 might need solve, and go and engage with that and 311 00:17:33.400 --> 00:17:37.000 start with fact finding and then move yourself up. As 312 00:17:37.039 --> 00:17:40.119 we talked about early on, Andrew, the answer might be fire. 313 00:17:40.319 --> 00:17:42.440 You know, the ultimate answer might be fire your cro 314 00:17:43.119 --> 00:17:46.480 and demo and rebuild. I'm not saying it doesn't happen, 315 00:17:47.039 --> 00:17:50.279 but I'm saying that it should be a process to 316 00:17:50.319 --> 00:17:52.200 get you there. That's where scaffolding can help. 317 00:17:52.960 --> 00:17:55.079 So to drill down a little bit, you know, if 318 00:17:55.079 --> 00:17:56.960 you're going to need to bring another vendor in any 319 00:17:57.000 --> 00:17:59.680 way you know are there time and cost savings is 320 00:17:59.680 --> 00:18:01.400 that you know you don't want to end up in 321 00:18:01.480 --> 00:18:03.799 a situation where you're firing all your cras and hiring 322 00:18:03.799 --> 00:18:05.759 a new batch of cras who and you're going to 323 00:18:05.839 --> 00:18:08.839 end up in the same situation, or you know where, 324 00:18:08.920 --> 00:18:12.799 where are the time and cost savings for switching the approach? 325 00:18:13.319 --> 00:18:16.759 The goal is to identify your problems and target them 326 00:18:16.799 --> 00:18:19.480 so that you are not just blowing everything up, and 327 00:18:19.559 --> 00:18:23.160 so you're walking your way there smartly, right, like we 328 00:18:23.200 --> 00:18:25.559 are not demoing the building because we wanted to change 329 00:18:25.559 --> 00:18:29.599 an electrical plug. The time and cost savings come from 330 00:18:29.920 --> 00:18:35.720 not actually demoing the whole thing and so incrementally walking 331 00:18:35.759 --> 00:18:38.680 your way to solving the problem, to the solving the 332 00:18:38.680 --> 00:18:40.680 problems in front of you and making sure that you're 333 00:18:40.720 --> 00:18:44.079 solving the right problem is the smart way to go 334 00:18:44.160 --> 00:18:47.039 about it. Right. It goes slow to be smart, but 335 00:18:47.400 --> 00:18:51.240 being smart goes fast. Right. It's that sort of mindset, 336 00:18:51.400 --> 00:18:54.680 right like you might feel that that need to go 337 00:18:54.880 --> 00:18:58.759 do something, break something, and keep moving, But often what 338 00:18:58.839 --> 00:19:01.720 that does is it is it compounds the complexity and 339 00:19:01.759 --> 00:19:06.839 compounds the problems and makes it very difficult to continue 340 00:19:06.880 --> 00:19:08.960 to operate long term, right, and you don't want to 341 00:19:09.000 --> 00:19:13.279 tank everything and so identify and I would say, if 342 00:19:13.319 --> 00:19:17.480 you are are held bent on actually changing over your CRO, 343 00:19:18.200 --> 00:19:22.000 it does actually make sense to overlap your new CRO 344 00:19:22.160 --> 00:19:24.400 and your old CRO, and you might want to do 345 00:19:24.519 --> 00:19:27.720 that in a staged approach, right, So bring in your 346 00:19:27.720 --> 00:19:34.079 new CRO's project manager and maybe CRA manager at first 347 00:19:34.160 --> 00:19:37.160 to really get a sense right and engage them as 348 00:19:37.519 --> 00:19:41.039 internal consultants and let them get a feel for the 349 00:19:41.079 --> 00:19:43.480 study and what they're going to need and make sure 350 00:19:43.480 --> 00:19:46.400 they really understand it before you pull the trigger and 351 00:19:46.519 --> 00:19:50.160 change everything that the sites see. So you might end up, 352 00:19:50.200 --> 00:19:52.279 like I said, you might end up in those same places. 353 00:19:53.000 --> 00:19:55.759 It might cost you more in the short run, But 354 00:19:55.880 --> 00:19:59.680 the goal is to smooth out the operations, make things 355 00:20:00.000 --> 00:20:04.200 mean for your customers, and make make things organized on 356 00:20:04.240 --> 00:20:08.279 the back end right, so you don't have sheer chaos everywhere. 357 00:20:08.680 --> 00:20:12.960 So at the cmosumm it this past year, Ken gets 358 00:20:13.079 --> 00:20:14.599 brought up and I think he's brought this up in 359 00:20:14.759 --> 00:20:18.279 years prior as well, that we're seeing more and more 360 00:20:18.640 --> 00:20:23.160 protocol amendments happening, which add more money to trials and 361 00:20:23.200 --> 00:20:26.319 add more time to trials and earlier in our conversation, 362 00:20:26.440 --> 00:20:30.559 you spoke about expected unexpected? Are these inevitable? Are these 363 00:20:30.599 --> 00:20:33.960 rescues in scaffiolding approaches inevitable based on the way that 364 00:20:33.960 --> 00:20:36.880 we currently run clinical trials? You know, he's there that 365 00:20:36.880 --> 00:20:39.079 we could be doing from the beginning so that we 366 00:20:39.119 --> 00:20:42.000 don't have to do an amendment, so that we don't 367 00:20:42.000 --> 00:20:46.039 have to, you know, hire another group to watch over everything, 368 00:20:46.359 --> 00:20:48.599 or hire another internal person to watch over everything. 369 00:20:49.160 --> 00:20:52.519 I don't have this data, but I'm gonna I do 370 00:20:52.599 --> 00:20:57.799 believe that some of these amendments are inevitable because what's 371 00:20:57.880 --> 00:21:01.039 driving particularly for biotechs, right, biotech aren't really in the 372 00:21:01.079 --> 00:21:04.160 phase three space. They're in the phase one phase two 373 00:21:04.200 --> 00:21:07.640 space largely, right. Some do move on, but by the 374 00:21:07.640 --> 00:21:10.400 time they move there, they themselves become more mature, their 375 00:21:10.440 --> 00:21:13.400 programs become more mature. So really what drives a lot 376 00:21:13.400 --> 00:21:16.079 of these amendments early on is phase one and phase 377 00:21:16.119 --> 00:21:19.440 two research. And I think it has gotten worse over 378 00:21:19.519 --> 00:21:23.400 time because the medicines we are doing don't have clean 379 00:21:23.599 --> 00:21:28.000 and clear regulatory pathways. Right, we actually don't know what 380 00:21:28.079 --> 00:21:30.319 the answers are, and we don't really know what we 381 00:21:30.359 --> 00:21:33.599 are doing, and we're using phase one in phase two 382 00:21:33.640 --> 00:21:36.839 to learn and I don't mean this in a bad way. 383 00:21:36.880 --> 00:21:39.720 That is good science. That is exactly the way science 384 00:21:39.759 --> 00:21:44.920 is supposed to work. It's that the propulsion of these 385 00:21:44.960 --> 00:21:48.640 personalized medicines out of pre clinical and into human it 386 00:21:48.680 --> 00:21:52.480 goes a lot faster than in the with small molecules. 387 00:21:52.519 --> 00:21:55.559 And so what you're seeing is we're sort of learning 388 00:21:55.599 --> 00:22:00.960 on the job, right and we're learning with diseases that 389 00:22:01.000 --> 00:22:05.359 aren't well studied in academia, right, Like they don't have 390 00:22:05.440 --> 00:22:07.880 big data sets, so we don't know the answers. And 391 00:22:07.920 --> 00:22:10.720 so I do think that we have to expect that 392 00:22:10.920 --> 00:22:15.880 some amount of amending the protocol, some amount of what 393 00:22:15.920 --> 00:22:19.799 we think is good science, hits the wall of practicality. 394 00:22:19.880 --> 00:22:23.240 That's the site the protocol or site issue where it 395 00:22:23.279 --> 00:22:25.359 seems like it's a good idea to do this thing 396 00:22:25.440 --> 00:22:28.000 and then send them to do a scale and then 397 00:22:28.039 --> 00:22:31.519 send them to an MRI. But those two things happen 398 00:22:31.599 --> 00:22:35.880 in vastly different places inside of the inside of the institution. 399 00:22:36.000 --> 00:22:37.640 So if you're really trying to do them at the 400 00:22:37.680 --> 00:22:42.119 same time, that takes inordinate amount of infrastructure for an 401 00:22:42.160 --> 00:22:46.640