What is stopping IVF labs from becoming fully automated? Tune in to this week’s episode of Inside Reproductive Health, as Griffin Jones sits down with Eva Schenkman and Helena Russell of ARTLAB to breakdown the six main barriers to automating the IVF lab.
Listen to Hear About:
Why automation isn’t happening in certain areas of the IVF lab.
Risk and inefficiency of data entry.
Lack of trust that comes from business intelligence software.
Lack of adoption of the Vienna consensus.
Which metrics are meaningful for safety that don’t necessarily improve clinical outcomes, but are required to improve safety and productivity.
Delivery vs operations- what needs to be prioritized now vs. what should be prioritized for the future.
Website: www.artlabconsulting.com
Eva’s LinkedIn: https://www.linkedin.com/in/eva-schenkman-ms-phd-cc-eld-hcld-6121778/
Helena’s LinkedIn: https://www.linkedin.com/in/helena-russell-5aa60214/
Transcript
Eva Schenkman 00:00
They're missing the point that you know I think UCSF did some data where they showed that having an embryo scope in their lab saves them the equivalent of one embryologist time per day. And if you look at the cost of an embryo scope which is probably akin to about you know, one year embryologist salary that is becoming more efficient with these devices will in the long run, save you money, especially now when there is no embryologist to be found.
Griffin Jones 00:32
All of the change that is not happening in the IVF lab we talk all about the automation is coming to the field and seemingly every talk at every conference many episodes on, I want to know why hasn't it happened already? Why isn't it happening faster. And so I explore those obstacles and barriers with my two guests on today's program. That's Dr. Eva Schenkman. She was a lab manager for a number of years to different practices. She has been a consultant. She now runs a program called ART Lab. And I bring in her colleague Helena Russell, and we talk about the barriers to implementing automation categorically. In the IVF lab, we talked about the risk and inefficiency of data entry, we talked about the lack of trust in the data that comes from business intelligence software, if estimates that fewer than 10% of IVF labs have fully automated their data entry with business intelligence software, we talk about the Vienna consensus. Why has there been a lack of adoption in the Vienna consensus again, I asked Helena and Eva just a ballpark how many labs they think have adopted the Vienna consensus. And I'm asking them to do this off the top of their head, but they think it's about half that have adopted some meaningful level of the Vienna consensus. We talk about other metrics that are meaningful for efficiency and safety that don't necessarily improve clinical outcome, but are necessary for improving safety efficiency. And for activity. We talked about this person dynamic between delivery and operations where you are on the hook for doing a certain number of IVF cycles, you're on the hook for serving a certain number of patients, you have to do that to make payroll to keep the lights on to keep the patients happy. Meanwhile, there's the operational systems behind that which are another entity another chore to solve. And those two things are at odds of each other in terms of what is prioritized now in the moment, but what needs to be prioritized and improved for the future and for ongoing delivery. Finally, Helena and Eva say that some solutions are not ready for primetime and boy do they go to town on naming who those folks are? Now they don't try to get them to but of course they go hard and ideas and soft on people as is generally good advice. So it was a constellation for myself, I have to detail what they would like to see from RCTs what they think is missing from solutions that are coming to the via what they think needs to be proved in order for solutions to merit much wider adoption and what IVF centers could do in the meantime to help prove the concept. Enjoy today's episode with Helena Russell and Dr. Eva, Schenkman, Dr. Schenkman, Eva, Ms. Russell, Helena, welcome to Inside Reproductive Health.
Helena Russell 03:19
Thank you, it's great to be here.
Eva Schenkman 03:20
Thank you.
Griffin Jones 03:22
I've finally fulfilled the promise or I'm living up to a promise where I said it was going to create more IVF lab content than I have in the past. I think, this year, we've already done more episodes about the lab than we did in the first three years of the show, combined. So I'm starting to have a rudimentary level of knowledge to where I can maybe start to ask more interesting questions. And one of the things that I want to talk about today is the obstacles behind the automation for the lab. So at a high level, on the show before I've talked about the automation that's coming to the lab, and like to take advantage, speaking with each of you about why it isn't happening faster, and probably have you unpack and give specific examples as we go. But maybe we start at a high level, with just the automation that you're seeing in the lab happening right now that you weren't seeing five years ago, and maybe not even two years ago, what's happening with regard automation.
Eva Schenkman 04:25
Now, one of the ways in which, you know, I've been involved in some of my consulting activities in some of the automation is through data analysis. You know, we spend an awful lot of time in the lab, you know, crunching numbers. And in most labs, we still do it the same way we did 30 years ago, which is, you know, we've usually got two or three different Excel spreadsheets, we've got one for data, we've got one for cryo, you know, we may also be entering something 20 or more, and we used to sit there at the end of the month or the end of a quarter and spend, you know, 234 days to crunch all those numbers. So not only counting the amount of time that embryol Just spending putting in all that data, you know, risking all those data transcription errors, you know, now we've been using things, you know, business intelligence software, like Power BI, to pull that data automatically out of the IVF EMRs, to run that data in real time, so kind of call that real time analytics. So that I see is one of the key ways into which we can save, you know, an enormous amount of time making the labs, you know, a lot more efficient, is on a data analysis standpoint, you know, one of the big talks now with a lot of the meetings or on automation in the lab and efficiencies in the lab, and, and, you know, I think we can talk a little bit more more about that, what the roadblocks are, you know, to those. And, you know, to a long way, I think a lot of the roadblocks are One is cost, you know, a lot of these devices, things like, you know, an embryo scope, for example, are very expensive. And, you know, a lot of physicians or a lot of practices expect to see, oh, I'm gonna get this device, it's going to increase my pregnancy rates, oh, it doesn't increase my pregnancy rates, well, that I'm not investing that kind of, you know, money into it. But they're missing the point that, you know, I think UCSF did some data where they showed that having an embryo scope in their lab saves them the equivalent of one embryologist time per day. And if you look at the cost of an embryo scope, which is probably akin to about, you know, one year embryologist salary, that it becoming more efficient with these devices, will in the long run, save you money, especially now when there is no embryologist to be found. You know, and I think some of the other issues I see with the automation is things are rushed to market quickly, you know, at at a very high price, and they don't necessarily have you know, a lot of the data behind it yet, that you know, that it is going to be you know, just just to save for just the same as a senior embryologist. So I think kind of got, you know, a couple of issues there, you know, between the cost and, and the efficiency, and, you know, making sure that you know, that we can get get current staff to adopt, you know, this new technologies,
Griffin Jones 06:59
because you give me a couple of different avenues that I could further explore. Let's start with the spreadsheets. You mentioned, having two or three Excel spreadsheets previously, for which you need for your data analysis. What were they what what were their roles, those those spreadsheets and the information that they contain
Eva Schenkman 07:19
everything from, you know, you're doing your pregnancy rates, your competency assessments, also your CRO inventory, you know, we typically, for the most part, still keep paper worksheets in the lab, very few of us are using, you know, tablets or have gone paperless. So, you know, we've got that paper, you know, we're either scanning that paper into an EMR or, you know, retyping that data into an EMR. And then typically, a lot of the EMRs, don't do data analysis very well. A lot of them don't have reports that follow the Vienna consensus, you know, guidelines. So we're then keeping separate spreadsheets, so we're putting things into the EMR, putting things into, you know, Excel spreadsheet for data analysis, and then typically having a third sheet for, you know, cryo inventory. So we're entering everything, you know, typically three times, and then taking having somebody you know, typically higher up, then do all of that data analysis, like I said, usually typically the end of the month, sometimes at the end of the quarter,
Griffin Jones 08:17
how is QA done in this instance, when you have three different sources of information, but they're all in different places? How, how is QA done so that the duplicate of information is correct, because anytime you have information, different sources that isn't uniformly exported, you always risk you
Eva Schenkman 08:37
typically an Excel worksheet, you hope you catch it, there's not really a lot of a lot of formulas in there to kind of automate to to pick that up. You're always gonna get data, transcription errors, some of the things like Power BI can can pick that up for you. But I think, you know, honestly, a lot of times it gets caught when you're giving a patient data off of your cryo Inventory spreadsheet and a patient, you know, or nurse, correct shoe, you know, will will that's, that's wrong. That's not what we had, you know, so that that is a problem, you know, with data entry errors, is we really don't have a good mechanism to ensure that the data is accurate.
Griffin Jones 09:14
So when you have three sources of info like that, you got your spreadsheet for cryo inventory, you're scanning into the EMR, and then you've got a separate spreadsheet for the data analysis. There generally isn't like an overarching QA for the data entry to make sure they're all uniform. Now, okay, so even without regard to efficiency, there's still there's a risk there.
Eva Schenkman 09:36
Yeah, absolutely. You know, your data is only as good as the information you're putting in.
Griffin Jones 09:41
You mentioned that is an area where clinics are starting to automate more and those spreadsheets are being supplanted or that's something that you envisioned in
Eva Schenkman 09:51
the know there actually is is a few systems out there. Several of the EMRs have been using business intelligence software either through Tableau or through Power BI and linking those with their EMRs to that automatically pull that data out of the EMR. So as soon as you've done your first check, you know, as soon as you've done, you know, your, you know, your observation or the pregnancy data is entered in, it's pulling it into those Power BI sheets. And those not only that are automated, but they can even be set up to then watch you when there's a problem. So they can send you notifications that, you know, Hey, your XC three P and rate is starting to creep up. So you can, you know, definitely not only from an efficiency standpoint, but also from a troubleshooting standpoint. So I know, you know, recently one of the media companies had an issue with with some oil, for example, you know, and that, you know, typically tends to take a little bit of time until you're able to pinpoint what the problem is. And you know, the hope is that these automated systems would be able to pick up on something like that much quicker than you'd notice by eye or, you know, you got to wait till the end of the month, you know, obviously, something's killing all your embryos, you'll notice that pretty quickly, but let's just say you've got, you know, 25%, drop and blast conversion rates, that may not be something you pick up so easily, maybe you had some bad patients in there. But you can use a lot of that business intelligence software, it's been used by the, you know, financial industry and other industries for for years, you know, now we can kind of harvest the power of that, and and use for the IVF labs,
Griffin Jones 11:20
do you have even a ballpark guess, of what percentage of IVF labs are now automating their data entry with business intelligence software?
Helena Russell 11:30
Automating? I'd say, single digits?
Griffin Jones 11:33
That's a very, very low, yep. What's stopping it from being at 90 100%?
Eva Schenkman 11:39
I think one is trusting in the data. Two is, is, you know, we, for as much as we like to think we're ever changing, we don't actually like to change that much. You know, we don't want to let go of our paper worksheets, we, you know, this is, this is what we've done for 30 years, you know, we don't want to make mistakes, and what we do we know that, you know, an Excel spreadsheet, you know, as long as it's not, you know, sorted wrong or tampered with, you know, it will get you the, you know, the data that that you need, you know, a lot of the EMRs aren't necessarily don't necessarily have the best fertility modules. So, you know, even, you know, a lot of people in the lab, they're, they're still using the paper worksheets, and they're only scanning in their sheets. So one is, is, you know, if you're going to use something like Power BI or Tableau, you really have to have a dynamic EMR, to be able to use that with so. So that's something a lot of the clinics struggle with, you know, and I think just just trusting, trusting in the data is a bit of a learning curve, you know, to to get going with it. And, you know, I think slowly it's, it's starting to come come about, but, you know, slowly,
Griffin Jones 12:46
by the way, Helena, anytime that you want to jump in, I tend to just riff off questions, because I
Helena Russell 12:51
just want to say a couple of things to, to kind of, you know, kind of chime in with Eva, one thing, that's what's really challenging is learning curve, because it's not just trust, it's taking somebody who works with their hands, and putting them into a situation where they're going to have to be working with computers more. And that can be a little daunting. But again, having the right tool and the right support from that tool, helps us something else that even just said, is that they're not, not all of these EMRs are created the same. And that's true across healthcare industry, in general, you know, they're very unique, there are so many out there. And they do different things differently. And so there may be some that are a little bit better for gathering all the information that needs to be gathered, and also to be flexible enough. One thing that you may or may not realize about IVF is that not all IVF centers do things exactly the same way. So you have to be flexible. And the learning curve is one of the one of the things that I think is challenging for people and trust, like Eva said, another way of automating that kind of tails into EMRs. And specifically EMRs built for IVF is witnessing, which is an automated system these days with barcode reading or with radio frequency. And even might want to chime in on this one as well. She has a lot of familiarity with these. And those are also tying in with some of these IVF databases, or electronic medical record systems. And again, pulling a lot of really good valuable information from the lab into that system helps with once we get to that point where we can do the analysis via you know, Power BI, what we can then do is really target quality control, quality enhancement, and quality assurance.
Griffin Jones 14:56
Let's stay on that thread for a second before we get into workflow variance and And the barrier of change. You mentioned one of the issues apart from that is trusting the data itself. So what is the cause for mistrust and data? Or what is the risk of inaccurate or incorrect data in using business intelligence software for data entry,
Eva Schenkman 15:18
when you're pulling data from from an EMR, you know, one of the problems is, these EMRs are all structured differently, you know, they're usually large back end SQL databases, they may not be, so you can't take, you know, three different EMRs take the same Power BI software setup and plug it into these three different systems, they won't work, you know, so these things have to be customized, you know, unless it's something your EMR is already offering, they, they would then have to be customized to each setup. And a lot of it is just in that analysis, knowing you might have two or 3000 different fields on the back end, to pull from, you know, how are you? How is each lab recording that data? Where are they? Where is that data sitting in the SQL? databases for analysis? I think some of it might be generational, you know, I think, you know, the first first generation of embryologist, you know, even though we're we're, you know, we are pretty good at using computers, you know, we, for the most part for the last 30 years have done everything on paper, have done everything, you know, simply the second we have to trust, setting up those scripts and setting up something to to the IT department, you know, it's these things are very difficult to validate. So it's a lot of time, and one of the things we don't have right now is a lot of time in the lab. So I think part of that is, is having the time to validate these systems to trust them, it would be very hard for company to come in to develop, you know, a Power BI software, that's, that's applicable to all EMRs. Because the EMRs are all structured differently. So they need to be done, you know, on a customized or bespoke, you know, level between between each system. But I think it's just as I said, I think it'll be different with this new generation of embryologist coming through, I think they expect it, you know, they practically live with a phone, you know, in their hand, you know, I think they're going to be a bit more comfortable with with having this data. Automated?
Griffin Jones 17:11
Tell me a little bit more about what you mean, by the time it takes to validate systems? Does it mean to like pilot the program to check the…
Eva Schenkman 17:20
Yeah, you know, I'm actually involved with one, you know, right now looking at at some of these, these automated reports, and I have to go into the EMR and I put in test cycles, and I'm putting in, you know, different complicated ones with day one xe or with late for some with thaw biopsy, refreezes, combination cycles with fresh and frozen eggs. And all of these data sets are stored in different tables in the back end of the CMR. So that I have to sit with the IT people and structure each of these queries. And, you know, we tested on these cycles, and, you know, these, how do you tell an IT person, you know, when they're doing a competency for, you know, good day three cleavage rate? You know, for example, you know, what does the word good mean? You know, if you asked, you know, for embryologist, you're gonna get five different answers, you know, and that's part of why, you know, we rely on things like the Vienna consensus, you know, as a standard, you know, guideline to go through, but then, you know, each and every clinic, we roll these things out to, has to validate it on their own, because none of us are doing recording data the same way, you know, there's, you know, we all record it a little bit differently, we're all using different templates, we're all using, you know, different embryo grading criteria. So I think that's part of, you know, a bit of a problem with it, you know, I think but, you know, as clinic start to see the benefit of these systems, I think it'd be easier and easier, you know, we get these things validated, we get a couple of hopefully, key key labs, you know, incorporating them into their workflow. You know, I think we'll, you know, we'll kind of get the message out there, that the systems are, you know, are reliable or trustworthy. And, you know, that'll go a long way to really making the labs, you know, more efficient. Everybody's talking about, you know, lab on a chip and everything else. But, you know, I think, you know, when you're embryologist are spending a significant amount of their time being admins, you know, hand entering data is still using paper worksheets. Were a long way away from talking about, you know, lab on a chip.
Griffin Jones 19:18
How much chicken and egg is happening here, like, if part of the reason why labs are slow to adopt the technology, they're slow to validate the systems because there's so much variance in workflow, people report data differently, they grade embryos differently, how much of so that's the barrier, but it's also the result, isn't it? Like if you had the universal systems implemented, that you might have a more universal way of recording data, you might have a more universal Is that happening?
Eva Schenkman 19:51
We have the Vienna consensus, you know, the paper that was written for KPIs. I think that goes you know, along A great deal.
Griffin Jones 20:01
Okay, what is stopping people from categorically adopting this Vienna consensus across all labs?
Eva Schenkman 20:10
I think for the most part, it's been very well, you know, received, I think it's just it's that the woods that way, we've been doing it for 30 years. You know, it's, it's that belief, it's, it's worked for all this time, you know, this is, you know, in that belief that, that, you know, we're kind of all homegrown cooks in each of our labs, that, you know, we kind of, we kind of do it our way, these are the KPIs that, that that worked for us, there are still some labs that are doing d3 biopsy, you know, as opposed to, you know, blastocyst biopsy and slow freezing, it's just that ingrained, you know, because we don't want to make mistakes and in what we do, so in some ways, we're very reluctant to try new things. And, and part of that comes with doing it the same way it's worked, we don't want to change it, but and
Helena Russell 20:54
so much hinges on it, right? Yeah.
Eva Schenkman 20:59
And that first generation of embryologist is retiring. They're leaving the field. So, you know, I think it's, it's, it's important to, you know, this new generation, they're not going to sit there for the, you know, the amount of hours and hours and hours that we spent typing into three, you know, three databases, they want to enter things on a tablet, you know, they don't want to enter things on on paper and then transcribe so, you know, I think there is a lot of push from, from these newer embryologist to to automate things, you know, and, and hopefully, you know, we'll get some significant changes. They're
Helena Russell 21:31
more comfortable trusting the data, as Eva has said,
Griffin Jones 21:35
what percentage of labs is, if you can even ballpark it? Do you suppose have adopted the Vienna consensus to? If not to the letter, you know, 90%?
Eva Schenkman 21:46
I'd probably have to say, maybe, what do you think Elena, close to 50? Probably
Helena Russell 21:53
I still they're not accepting all of them. They're probably focusing in on a few Don't you think? Eva?
Eva Schenkman 21:58
I think so. I'm still surprised how many lab people I speak to who haven't heard of it. And, you know, as I said, each one typically has their own KPIs.
Griffin Jones 22:06
Thank you, Eva. Now, I don't feel as dumb for asking.
Helena Russell 22:08
Yep. It's unfortunate. And I think it's a lack of communication in our field. But I also think that what we're doing is very difficult. And so the challenge is making sure that we continue to be able to produce what it is our patients need. And to meet our patients needs. I mean, there, there's, there's no excuse for failure. And so when you have something working, it's difficult to hear what somebody else is saying, if it doesn't mean an improvement, which I think you've kind of hit on earlier, unless you can show a, you know, a positive outcome. And it may be that they'd rather spend that extra money to have somebody do something in a less efficient way, then trust in something that may not may or may not give them the outcomes that they are looking for. Yeah, is
Eva Schenkman 23:06
it’s difficult to trust in the scripts that are written by, you know, by someone with a computer background that, you know, you as an embryologist don't really understand. So as I said, that's why the validation of it is so important, get them seeing that this data is accurate, and is pulling correctly. And, you know, I think, you know, to be able to have an automated system like that, then alert you, not only when something is out of range, but as deviating towards being out of range, I think will be you know, will be invaluable. And, you know, this, you know, one issue that recently developed with oil is now resulting in a class potentially, you know, class action lawsuit. So, I think, you know, anytime we can develop something that would pick up on these things, not only tell us our what our pregnancy rate is and what our our individual embryologist competency rates are, but to be able to then alert us to any troubleshooting issues in the lab, that we don't have to wait six weeks, you know, now we see something in our data analysis. Now we have to try to figure out, you know, figure out what it is, you know, that's where we're using AI is also going to help at some point, you know, with analyzing this data.
Griffin Jones 24:11
So I'm understanding if there's not a clear clinical outcome that lab directors can see of in terms of success rates, that there often isn't the impetus to impose a change, and I see the agents working against change. We've done it this way forever. It's worked this way forever. We have a big variance in workflow from one place to another. So just because it worked for these guys over here doesn't mean that I know that it's going to work over here, but at this point, why isn't the shortage of embryol embryologist and the constraint on embryologist time enough to have made a bigger catalyst for change? seems like to me it seems like okay, if success rates are equal, but I can get back an embryologist day. Every time that we use this solution, or I can get back this many hours of embryologist time, why is that not enough of a catalyst to be seen way more automation than we're currently seeing?
Helena Russell 25:22
Part of it has to do with time, it takes time to train somebody to do something new. You know, if you're so overwhelmed in your lab or your IVF facility, and you don't have enough time to train a new person, you don't have time to learn something new, don't you think? Eva?
Eva Schenkman 25:44
I think so. And I think it's just that you know, exactly that you don't have time to train something new, it's that chicken and egg, you know, scenario, again, you know, I'm so overwhelmed, I not only have time to not train somebody, and then you say, Oh, well, you know, get this piece of equipment or whatever, for automation, there is going to be a period of time where that, you know, system is going to actually take you more time, until you you know, you wreck it, you know, you're able to be proficient at it and you're able to, to realize its efficiency. And, you know, not all people have the patience for that much time for adopting it and the cost, you know, all of these, these automated systems are very expensive. So getting physicians in groups and practices, it's easy to say, I need another embryologist and they'll pay, you know, six figures. Plus, for an embryologist who see a body sitting there, you know, to pay six figures plus for a piece of equipment sitting on the counter, you don't see the efficiency savings as easily as you see another body sitting there. So I think that's part of it. And without them seeing, you know, like, as I said it, you know, I go back to time lapse, you know, they there was just, you know, paper recently that, you know, basically is, you know, we shouldn't be, you know, looking at time lapse, because there's we didn't see an improvement in pregnancy rate, but you're missing, you know, the picture of it, you're missing, you know, the safety of it, you're not having to take the embryos out to look at them, you can monitor embryos remotely, you know, so if there is, you know, more COVID outbreaks or another pandemic, you know, you can check fertilization from from home. And, you know, just that
Griffin Jones 27:18
you could centralize embryologist could knew or at least part of that workflow,
Eva Schenkman 27:23
you could do you have offsite lab directors could monitor things remotely, they can log in and look at the embryos look at how they're growing, you know, pull the data, you can see these Power BI apps, you can see all of your data on your mobile device, you can even see the images of your embryos on your mobile device. So I think it's, it's, it's, it's that cost barrier, but it is that learning barrier, that it's just not something new that we've done. And, you know, I think you'll I think next years, there'll be some workshops, at some of the meetings that are going to be focusing on future of technology and innovation, and where where things are going to be, but not just theoretical, but actual practical, what's here, what's now you know, what can we kick the tires on now, and part of that is, is training and having these new innovative systems launched at the at training centers, and having a rail just come in and use them because nobody wants to practice on a real patient. You know, you need to be able to have a place that's comfortable, that you can go in and you know, learn this in an environment that's not stressful, you know, not while you're you're trying to, you know, to do real patient samples, that you have a place to get comfortable with these devices and, and to you know, learn how they work.
Helena Russell 28:36
And we're all monitoring is integrated. And I mean, yeah, looking at your incubator, your temperature, your co2 level, your oxygen level, looking to see if your liquid nitrogen tank is got enough liquid nitrogen tank, liquid nitrogen in it, making sure your refrigerators are performing up to par. And having those be part of your automated, automated integrated system so that you literally have every function that you would normally assigned to possibly, you know, an intern or a novice embryologist, somebody who's a junior who's just coming in. Instead, you can have continuous monitoring, which I think is extraordinarily reassuring. Probably there's a role for someone or company out there to help clinics bundle and to become efficiency experts. I think one of the things that our training center does is helped expose new embryologist and even in workshops where we're opening up our center to experienced embryologist to come in to have one or two day workshops, they will be exposed to those kinds of integrated systems as well. And you know, a lot of it has to do with you know, I can I can hear about it all day long. I can read about it all day long. But if I can touch it, and I can move the dials and nobody's sample is going to get hurt by that. And I can actually download an app and do it on my own phone or my, you know, my iPad, while I'm in this Training Center. You know, the
Griffin Jones 30:13
exposure that you're talking about in the training center accounts for some of the issues, the distrust in the data, the lack of familiarity, the validation of the system counts, for some of them. Some of the things that it doesn't like, what you've been talking about is something that I've been obsessing over with regard to my own business and business in general. And I think we can apply it to the IVF lab, and that is delivery versus operations. And often when you hear business books, or you hear business talks, operations, and delivery are almost used interchangeably, like delivery, meaning the fulfillment of the good or service, which we've sold or promise and operations is really the system behind it. So we're roofers, our delivery is we're going to have a new tear off roof on your house by the end of April. That's the delivery. And we have an obligation once that roof is sold to fulfill that deliver, you could use delivery and fulfillment interchangeably. But operations is the system behind that delivery. So delivery is getting the roof on the darn house getting it done by the date, we said we were going to get it done by but operations is what types of materials we buy the workflow behind it, who we assigned to the job, how the job is assigned and accounted for and reported on the QA that comes after it the what what we automate or don't automate. And, and all of that is operations. And there's a tension between delivery and operations, because you have delivery obligations that you have patients cycling through, and you have a finite number of embryologist that can work on those embryos, while those patients are being served while you need to make this institutional change at the operational level. So how do you solve for that how, in this specific to the IVF lab, how do you begin to relieve some delivery obligations, while investing in the operations that will ultimately result in a virtuous cycle.
Eva Schenkman 32:35
Part of what we have here as opposed to just also having, you know, kind of a training facility is is you know, our training facilities a fully functioning mock IVF lab. So one to have all of these different systems communicating here. So that when people do come and try them, it's not just trying one piece of it, it's kind of seeing, you know, the entire system working as if this, this was a functioning lab, the other thing we have to convince them of is, is you know what to do when it goes down, because that's one of the most common things, you know, I hear that if we're going to be entering things on a tablet, or we're going to be entering things, you know, when our mobile device, you know, data patient data is potentially going up into the cloud, you know, nobody trusts that. So, you know, it's, it's the redundancy that's built in, you know, are we going to do you know, backups to, you know, to, to our local desktop, or we're going to print out, you know, a daily report, because what are you going to do when, you know, there's a hurricane that comes through retreating, like, what are you going to do, if a natural disaster comes through, I always have my paper, I always have my paper chart, you know, but there's that trust and what you can't see. And you know, we're all used to the internet going down the Wi Fi going down. But as an embryologist, you still have to do your job. And if everything is up in the cloud, and you come in, you got no Wi Fi, you know, how do you know what patients to do the first checks on or how do you know what patients to, you know, to do the freeze on or which embryos to thaw. So, you know, we do need to get better at that, you know, ensuring you know, what we're going to do from redundancy standpoint, to be sure that those concerns are addressed. And, you know, I think is, is, you know, manufacturers out there, we need to play a bit better in the sandbox with each other, and, you know, working on ways to get these systems communicating better with each other, because each one, you know, is kind of fine on its own, but there are these own little islands that aren't interacting very well with each other. They're very clunky, you know, not not not very quick. So, you know, we do need a lot of development still in those areas. But and I think, you know, the only way is to have kind of testing labs, you know, where where we can kind of kick the tires on these things and bring embryologist in to use them?
Helena Russell 34:40
Well, just to add to the you know, a lot of what we see in other industries, like the banking industry, a lot of what they do is done in the cloud. And you know, they have to have their very, very strict rules and regulations and other health care branches of health care industry. These people are doing a lot of commerce in the cloud, a lot of data storage in the cloud, and those redundancies have to be backed up by a robust IT support system. So they do exist for some of the systems that, you know, we've been talking about, you know, sort of loosely, but the really good ones are going to have that kind of support and structure so that you can, you know, assure those who are using it, hey, that information is going to be there. And they have to have an offline, you know, like a holding place at their own facility, a server that that information can be stored on,
Eva Schenkman 35:36
I still see a lot of doctors practices, their servers are in a closet down the hall. Yeah, and, you know, a lot of clouds. Yeah, that, you know, and, you know, we don't really hear it's not really openly discussed, but you know, we get a lot of clinics, there's a lot of clinics that are hit with ransomware. And, you know, a lot of that is kind of kept swept under the rug. And that's something that we need to, you know, why why do we not have a strict regulations as the financial industry, as far as how we're keeping this data, you know, where we're keeping this data redundancy,
Helena Russell 36:05
if you're thinking about automating, and you're thinking about going down this road with an EMR ask the really important question. And that is, how is this stored? What is your security structure? How is it done and who's handling that? Because, I mean, you have to, you have to have a very robust system, and it has to be redundant, can't just be stored in one place and must be stored in multiple places. And how that is done is actually critical, not only to the, you know, the security of your data, how you trust your data, the validation of the systems, but also whether or not you can move forward and practice one day, you know, if somebody holds you for ransom, you're stuck.
Griffin Jones 36:47
Well, that solves for the issue of redundancy, it solves for a lot of the issue of implementation. But a lot of what you described is still the challenge of delivery versus operations. A lot of the reason why people have their server in a closet down the hall is because they've been so busy fulfilling delivery commitments, meaning seeing patients doing retrievals doing transfers, and all of the lab work on the other side of that, that they have not had the time, money energy, to focus on the overall operation systems, you happen to have a program that takes care of a lot of the risk that allows people to visit allows people to do this without putting their own things at at risk or and taking their own, you know, having to test everything within their own system. But they still have to say, alright, well, I've got you know, maybe I've got four embryologist and I need seven. And so how am I going to send you one of my foreign biologists when I'm already half staffed? And, and so how do you how do you begin to solve for that
Eva Schenkman 37:56
one of the things we've been doing is offering you know, several, kind of intensive lengthy courses a year, you know, we, we, you know, and Elena primarily has been going out to to the universities we have someone who's also worked with us doing you know, on tick tock, you know, doing tick tock videos of getting those students out here to, you know, for training, so they typically come to us for for 10 weeks and we teach them everything from Andrology to biopsy, you know, we don't expect that these these these, these new embryologist could go back to their clinic and you know, be doing biopsy on day one. But you know, the typical in the old school apprenticeship style, it would take between two and four years to train one embryologist then we're losing embryologist at a much quicker rate than we can replace them. So if not only, you know, the training school that we have, but the other ones that exist in the country. You know, we are we believe that we're able to now get that training, once they're at the clinic down to under 12 months, so that we can speed up their training. So if you've got four you need seven. Well we can send you you know, you know, we're churning out embryologist, every embryologist that has been through here. I know everyone else had been through, you know, the, you know, one of the other firms California has had a job offer, you know, they're all you know, getting employed. And you know, we need to to, you know, bring through more embryologist and you know, and replace somebody even even a faster clip and that's the only way you know, we can't any longer do this, this apprenticeship, where it takes two to four years to get one new embryologist it's, it's not it's not sustainable. You know, we need a better way of of bringing them bringing them up, bringing them through quicker getting them trained. And you know, the style that we do it here which is very intensive, you know, they spend probably close to about 500 hours, you know, doing every literally every procedure and you know, over the course about two and a half months,
Helena Russell 39:52
hundreds of times they do each procedure hundreds of time. So what we're doing is set adding them up to make it easier for those who are doing the training on site in the IVF lab, making it easier for them to get the embryologist they need. I do think that part of the operational pushback is there needs to be kind of somebody who could bundle I really do believe that there's a there's another role out there for it, an IT biologist or something, you know, somebody who could go into a lab and do a consultation and say, you know, an EVA really has that kind of perspective, she may not be the IT expert, but she has, you know, a really good perspective on, you know, hey, you're doing this, this, this, and this, here are some products and, you know, we can put all these things together and deliver them to you. And you know, here's our IT redundancy expert, you know, can come in, look at your system right now, and say what needs to happen? And what tools can we bring in here that are going to meet your needs? What need do you have? Do you want to do all your quality control remotely? Do you want to do your embryo analysis remotely your embryo culture analysis remotely? Do you want to bring all your data together so that you can meet your KPI with a click of a button, review your your KPIs, and then bring all of those things together, and act as a liaison between all these different groups? Because it is a little mind boggling when you look at what is happening in the IVF field. And you have you know, this automated system and this automated system and this automated system and this automated system, how do you bring all of those things together? That's the challenge. And not everybody's going to want all those things. So how do you do that? That's that part of that operation could be someone who's an expert at all these different things, helping to give advice, consulting, and charging a fee to bring it all together for them and stitch it together.
Griffin Jones 42:01
Helena, you were talking about the challenges in having so many different automation solutions, one solution to that problem of having so many is having a consultant or an umbrella solution of some kind that can bring them together. How much of the problem is also those solutions not integrating with each other not integrating with the EMR? How common is that
Helena Russell 42:28
it's happens all the time. And Eva spoke to that earlier that people in these different realms need to play well in the sandbox, they need to be able to open up their their systems a little bit, so that they can speak to each other push and pull data, because a lot of times you'll see, well, one company will let you do one thing, but not the other. And you need both. And, you know, I think it's a little that's an operational hurdle. And again, an integrator, somebody who really is quite savvy and knows, you know, how to communicate with these folks could hopefully bring some of this together, I know of, you know, at least one company who's doing things like that. I'm sure there are plenty of others that are attempting that, you know, it's it's a daunting task, we know that we know it's very difficult to change. But one of the things that the light at the end of the tunnel, you're never going to stop changing. And IVF though that's just plain and simple, it, you're not going to reach a pinnacle and say, Oh, we're done. Now we've reached the pinnacle, because something new is going to happen down the road, something new, some new way of doing analysis. And so you're going to always have to change you're going to have to learn to live with that. And like Eva has said some of the newer generation, they're used to maybe looking at things a little differently, maybe not so much always changing. But at least the electronic aspect of it doesn't seem like it's so that was daunting, not as daunting not as as much of a trust issue. Now I can't trust my computer gets viruses, right, or I can get malware. So I think that, you know, if you if you have the right systems and the right checks and balances the right security systems and redundancies, as we've said, you will begin to you know, get over that hurdle. That's one of the biggest ones.
Griffin Jones 44:20
But if they don't integrate, aren't we back to the same challenge of the spreadsheets?
Helena Russell 44:25
A lot of them are integrating. Yes, we are if they don't integrate a lot of them are seeing the handwriting on the wall. I think Eva, wouldn't you agree?
Eva Schenkman 44:35
I think so. Now,
Griffin Jones 44:37
seeing the handwriting on the wall and that they're not being adopted, if they don't integrate
Helena Russell 44:42
They’ve got to make themselves a lot more malleable in order to be adopted. Like you just said, if if we're trying to show people how to use a KPI and the system that is is giving you your best data and is not you No handing it over that you have to actually export it and upload it a different way that may be not as user friendly, you might do it. But if somebody else down the street will integrate, guess who's gonna get pot?
Griffin Jones 45:14
So there might be a market response that forces people to integrate more you had in the beginning of the conversation, you alluded to some solutions, maybe not coming to market, but not having the scientific proof that they have a great benefit. What are some examples of that?
Helena Russell 45:36
Well, I think even would agree that there are some products out there that we need to more closely scrutinize and names. I'm not going to do that. But I will say that their artificial intelligence base, but the the issue with some of these is, you know, the gold standard in scientific medical research is the randomized control trial. And some of these products, they may have them in progress, but as far as I know, not really have published as much as they should, or at all. And so one of the things that I think we need to as a scientific community, which is what IVF is a part of, is that before we fully buy in, or spend an awful lot of money on something, that I mean, maybe we volunteer to be part of that study, you know, if you're an IVF center, and you're interested, you know, say, okay, all I'll be part of this study in order to help advance this field so that we'll know one way or the other, what they're promising may not be that we have better outcomes, necessarily, but that we might have more efficient outcomes, which might lead to better outcomes, because maybe your embryologist won't be so incredibly stressed out all the time, because they can't function because they can't get all their work done. Because there's not enough of them. And this automation could become part of the workflow that holds an answer for them, at least part of an answer.
Eva Schenkman 47:13
And I think that I agree with Helena, you know, the biggest issue is, is you know, especially, you know, right now, you know, the flavor of the month is kind of anything AI. And you know, each of them have some some papers coming out that they're showing that that, you know, this system is the best or that system is the best. But there's really a lack of well, plans. Well, well, rigorous setup. Yeah, what very rigorous those randomized controlled studies. And that's really, because what happens is people that adopt it, and they don't see the same benefit in their hands. So there's a big distrust of it, when you have for profit companies, who are then also sponsors of these papers, we're putting out data saying that this is the best thing ever. And then when somebody pays the money and adopts the system, they're not seeing, you know, the same, you know, Return, return to there. And so, you know, I think, you know, that's probably the one thing in this field that that I think is hurt us that we don't do, you know, as many well planned RCT studies, that, you know, we do a lot of retrospective, a lot of, you know, prospective, but not necessarily a gold standard, you know, stuff, which is hard to do.
Helena Russell 48:22
I mean, in IVF, it's very difficult to do that. Now, it's very difficult to do certain kinds of randomized control trials, because you do not have, you know, that many chances for fertility, in many cases who are coming to you for treatment. You know, if you're going to do a randomized control trial, it's got to be planned in such a way to limit the harm or potential harm for the patient. What's harm harm is, maybe they didn't get pregnant. And so, you know, in these cases, when you're looking at artificial intelligence, as long as you have a good check and balance, like you're having, you're having your own technicians review, and re and, you know, respect what's coming out, but review what's coming out of the AI. And make sure that well, whatever it is, it's telling you, you have the human aspect that you've learned to, you know, know, you know, and love, and you trust, then, you know, oversight is good, but what does randomized control trial mean? And what is blinded mean? Because a lot of times bias, unfortunately, you know, enters into these things and how do you create a study where there's limited bias, meaning that you're not overtly influencing the people who are conducting the study? The doctors, the even the patients, and certainly the embryologist, how are you ever going to blind the embryologist? Probably not never, you're probably never going to blind them because they're going to have to keep the numbers straight. Somebody has to protect the patient's embryos and make sure they really truly understand they know this is embryo 1234. And this is embryo 3456 and make sure everything is working properly. So blinding, the embryologist is almost impossible.
Griffin Jones 50:07
Which RCTs? Would you like to see happen with regard to AI companies entering the lab space? Like, can you detail what you would like to see an RCT or a couple of RCTs?
Helena Russell 50:18
I mean, even you talked about this the other day with the AI that you were thinking about that, that I think one of the things that we need to see is more numbers, also consistency and how the training database is working. So how you build that artificial intelligence is by having, you know, a large enough number of input and outcomes, you know, so you have something that you're observing, right, and you're applying an algorithm to it. And then what comes out the end is, hey, do it this way, or, or select this embryo. And so if you have a large enough database, you could potentially apply that one of the biggest problems that we have, is applying it across the entire world, probably not doable, because in each and every lab or each and every IVF. Center, there may be some variables that we really have no control over, that we have to kind of focus in on that particular lab and having enough data to have an artificial intelligence algorithm built may not be possible on a center by central basis. So some of these things, I think it takes time to develop the algorithm and then apply that to a randomized controlled trial, where you're looking at either isolating the artificial intelligence and doing it with sibling embryos, for example. So you have to have a special population of patients who have enough embryos that you could put them into different systems and compare them, or potentially looking at, you know, larger populations, if you don't have those sibling embryos to look at, you could look at groups of individuals in those two different, you know, isolated, different ways of producing the embryo, for example. So it goes beyond what we're currently doing in the lab, which is observational, when we even when we look at time lapse imaging, we're looking at changes over time that those are very interesting markers. Because you could see slow development versus fast development versus abnormal development. And you can see all that in a time lapse imager, this is something that you could never see as a, the traditional way of analyzing embryos to pick for transfer is a, you know, a one, a particular time point. And looking at an individual, you know, time point is, is not as superior as looking at, you know, time time points throughout the developmental process over the five to six or seven day period, that we have them in culture. And what Eva's talking about is even more specific and more precise. And that is going after those molecular markers, where you look at gene regulation, you know, those kinds of subtleties are almost impossible to you may not see anything, but and they made the embryo may be developing perfectly well, you know, it's just looks like a normal embryo. But when you actually look at the molecular profile, and look at the genes that are upregulated or downregulated, compared to the perfect environment where you can't replace something like that, you know, and and in past times, some of the things that people have looked at are metabolomics. I don't know if you've ever heard that word, but it's okay, the embryo is growing, and we're looking at metabolites of growth, and you siphon off some of the culture fluid and you look to see oh, is it metabolizing? Well, but actually looking at gene regulation, and and looking at markers that are very fine detail of the health of an embryo could be a potential answer.
Griffin Jones 54:15
I appreciate you both giving these so much insight into the different obstacles that are inhibiting automation from fully taking the IVF lab by storm. How would you like to conclude with regard to what needs to happen in order for automation to take its full rightful place in the IVF lab?
Helena Russell 54:37
I think what we need to do are some very detailed studies, where we look at how the impact of these automations on you know, first adopters, you know, there's always going to be a group of people who say, I'm there with you, I want to go automation all the way I want to do these things that are going to assist us in in prevailing and thriving and And moving forward, those first adopters should be studied. And efficiency should be studied, we should study all aspects of, you know, their turnaround time for troubleshooting, they're, you know, catching things on the on the fly when there's a, you know, a detail that's out of place for their QC, their daily Qc is messed up and they get an automated announcement. And, you know, there are people who are malleable to this, you know, they will be early adopters. And so those are the folks that we really need to study we need to present at meetings, we need to maybe create the perfect training environment like we have here at Art Lab, where you can bring people in, expose them to this integration and say, Okay, this is how it could work in your lab. You show them something, and that barrier is may not be eliminated, but it's gonna come down a little bit.
Griffin Jones 55:55
Helena Russell. Eva Schenkman. Thank you both so much for coming on inside reproductive health.
Sponsor 56:01
You've been listening to the inside reproductive health podcast with Griffin Jones. If you are ready to take action to make sure that your practice thrives beyond the revolutionary changes that are happening in our field and in society. Visit fertility bridge.com To begin the first piece of the fertility marketing system, the goal and competitive diagnostic. Thank you for listening to inside reproductive health