While at Harvard Medical School, Senan Ebrahim was deeply affected by witnessing a still birth. Convinced better data could have avoided the tragedy, he founded Delfina Care with Priyanka Vaidya from EMOTIV, the EEG platform, and his brother Ali Ebrahim, PhD, a former Tech Lead at Google. Senan and Priyanka joined me for a great chat about the exciting things happening in digital health.
Highlights:
Sal Daher Introduces Senan Ebrahim and Priyanka Vaidya of Delfina
Senan Ebrahim Was Interviewed in a Prior Episode Titled “Hikma Health”
Obstetricians Are Flooded with Data from Pregnant Patients Leading to Poor Outcomes
Priyanka Vaidya’s Experience with Gestational Diabetes Highlighted the Need to Coordinate Care
“...I realized discontinuity of care is a major problem.”
“Delfina wants to be that one place that integrates all of these discrete sources of data so that the mom stays in the center of the care...”
Integrating Health Data Shows Up in Interviews with Jeremy Wiygul, MD of Pela Health and Ryan Hess of Connective Health
“...how are we going to take better care of the four million pregnant moms in 2023 than we did in 2022?”
Senan Reached Out to Priyanka for Product Savvy and to His Brother Ali for Experience Building Sophisticated Data Models
The Patient Journey at Delfina
“This data is being enriched as the patient travels along the journey.”
Machine Learning Is Just Starting to Do Useful Things in Medicine Like Identifying a Stroke Quickly or Predicting Preeclampsia
Incentives Are Aligned in Pregnancy Care; Nobody Wants Babies to End Up in Intensive Care
What Delfina Means by “Closing the Loop” in Patient Care
Delfina’s Platform Is Actually Live at a Practice in California
Commercialization Will Focus Initially on Under-Served Populations
How the Name “Delfina” Came About
The Founding Story of Delfina Care
ANGEL INVEST BOSTON IS SPONSORED BY:
Transcript of, “Delfina: Safer Pregnancies”
Guests: Co-founders Senan Ebrahim, MD and Priyanka Vaidya
Sal Daher: I'm really proud to say that the Angel Invest Boston podcast is sponsored by Purdue University entrepreneurship and Peter Fasse, patent attorney at Fish & Richardson. Purdue is exceptional in its support of its faculty for its top-five engineering school in helping them get their technology from the lab out to the market, out to industry, out to the clinic.
Peter Fasse is also a great support to entrepreneurs. He is a patent attorney specializing in microfluidics and has been tremendously helpful. Some of the startups, which I'm involved, including a startup that came out of Purdue, Savran Technologies. I'm proud to have these two sponsors for my podcast.
Sal Daher Introduces Senan Ebrahim and Priyanka Vaidya of Delfina
Welcome to Angel Invest Boston. Conversations with Boston's most interesting angels and founders. I am Sal Daher, an angel who is tremendously curious on how great startups are built. Today I'm really excited to have with me, Senan Ebrahim and Priyanka Vaidya of Delfina. Welcome.
Priyanka Vaidya: Thank you.
Senan Ebrahim: It's a pleasure to be here. Thank you, Sal.
Sal Daher: Senan and Priyanka for Senan and Pri, call me Sal. Nobody says, Saleh. I became Sal when I heard that people with one-syllable names have higher earnings than people with multiple syllable names.
Priyanka Vaidya: That is fantastic.
Sal Daher: Sounds like an irreproducible result from an experiment somewhere but anyway, Senan and Pri, tell me what problem Delfina is solving.
Senan Ebrahim: It's so great to be back on, Sal. Thank you so much for having us and congrats on the 100 episodes by the way.
Senan Ebrahim Was Interviewed in a Prior Episode Titled “Hikma Health”
Sal Daher: Oh, duh, I should have said, Senan is an alumnus of this podcast. The previous episode was regarding Hikma Health and he's still involved in a less executive way. Look up Hikma Health [read or listen here: Hikma Health Episode], it's a very interesting podcast talking about applying a lightweight EHR [electronic health records, medical records] to solve problems of a population of refugees and people who are in difficult circumstances. It's really remarkable. Anyway, today, we're going to talk about Delfina [Health], which is solving a different problem but also using data. Take it away, guys.
Senan Ebrahim: Thank you so much, Sal. It's such a pleasure to be back. Congratulations on the 100 episodes and really enjoyed chatting with you last time about Hikma Health. Today, really excited to share with you a new problem that I'm working on with my co-founder Pri, which is the maternal health crisis in this country and around the world. The problem is truly massive. It's here in the United States with tens of thousands of pregnant patients and their babies suffering from the complications of pregnancy. That number gets to millions when you consider worldwide populations.
Obstetricians Are Flooded with Data from Pregnant Patients Leading to Poor Outcomes
It is particularly notable that here in the United States, our numbers are worse by a factor of two or three at times compared to most developed nations. Distilling it, observing obstetricians in practice, we learned that the bottleneck of care, particularly in this country, is that OBs [obstetricians, ie pregnancy doctors] are typically trying to manually collect and aggregate and interpret patient data in order to make what are usually fairly time-sensitive care decisions throughout the nine months of pregnancy and beyond.
Sal Daher: This is phenomenal. It's so funny because, before the podcast, we touched on another startup that I'm invested in, that is integrating data, that case of mental health patients but it's the same thing. Data from different sources, from EHR, data from prescription providers, and all sorts of different places to make it available to the practitioner to help improve results for the patient because we have all this data that's out there but it's not usable. It's siloed. It doesn't go to the right places. People can't interpret it properly. I understand that you're assembling data from three different sources. Explain to me what those sources are and how you are integrating that data.
Senan Ebrahim: Happy to dive into that but just to add a little more color to the problem, I think you're right. The data has to be in the right place at the right time in the right form and the complexities of our healthcare system as you know well Sal, make it difficult but Pri has a really novel perspective on that having worked in a variety of environments where sometimes the data is more streamlined or less available. I think as we build for the future, whereas we tell you about our product today, we're thinking about all those different environments.
Priyanka Vaidya’s Experience with Gestational Diabetes Highlighted the Need to Coordinate Care
Priyanka Vaidya: Thanks, Senan. I would like to add onto the data silo problem. The problem is that pregnancy, which is such a finite condition as we call it, is riddled with anxiety. Anything can go wrong at any moment, which just makes the mother feel anxious throughout the course of the pregnancy. If you happen to have a challenging pregnancy, which unfortunately I faced five years ago, you are directed to several different practitioners. I was unfortunately diagnosed with gestational diabetes.
Sal Daher: Oh boy.
Priyanka Vaidya: That came as a bit of a shock to me because I had no family history. It was an unexpected event in my life. I was then passed on to a nutritionist. I had someone look at my diet diary. I was also taught how to do the prick and stick blood glucose analysis. I was doing these recordings four times a day. I was checking my blood glucose levels, I was putting it in a spreadsheet. Sometimes my nutritionist would check in with it. My OB had no time to look at the data and no one really was clued into what was going on with me a day-to-day basis. That was a scary place to be.
Sal Daher: That is so moving and very touching because pregnancy is such a difficult and jarring thing. Every day, things are a little different and your body is going through all these amazing gyrations [laughs] such as it's not itself, and it is oriented towards the baby. The mother doesn't have enough nutrition, the mother starves the baby so that the baby can be well. Nature's designed things up that way. It is just such a difficult time. Anything that we can do to alleviate the anxiety and the tension and the problems?
Gestational diabetes is something which is very surprising. People are not aware of it. It's people who are not diabetic, don't have any predisposition that can happen. Preeclampsia, they can have a very high blood pressure during pregnancy. Your body is not working normally. It is a fantastically stressful thing, amazing and magical thing, miraculous thing, but at the same time, extremely stressful. Please continue with presenting how you're addressing that lack of coordination that exists with data sources.
“...I realized discontinuity of care is a major problem.”
Priyanka Vaidya: Absolutely, building on my personal experience, but also talking to friends and lots of patients, and now doing research, I realized that discontinuity of care is a major problem. Some of the data sets in the EHR in the OB system, some is sitting with the nutritionist, some moms do need mental health support so they might be seeing a therapist on the side, her data about her mental health is sitting on some other cloud somewhere.
“Delfina wants to be that one place that integrates all of these discrete sources of data so that the mom stays in the center of the care...”
Delfina wants to be that one place that integrates all of these discrete sources of data so that the mom stays in the center of the care, but builds a care team around her that has access to all her details and take care of her in a holistic way.
Sal Daher: This is extremely promising. This brings to mind another startup that was on the podcast recently, a company that is addressing... pregnancy can cause pelvic floor problems. The founder, Jeremy Wiygul is a practicing urologist in New York City. You're smiling Senan, do you know Jeremy?
Priyanka Vaidya: We actually listened to his podcast. [Listen or read here: Jeremy Wiygul, MD Interview]
Integrating Health Data Shows Up in Interviews with Jeremy Wiygul, MD of Pela Health and Ryan Hess of Connective Health
Sal Daher: Oh, you listened to the podcast. He's a practicing urologist. He explained this in the podcast, his wife had a problem after giving birth. This is what he treats. She was having trouble coordinating the care. What he's doing is creating a platform with some connected devices, type of a lightweight EHR, shades of Hikma Health here to help create this patient-centered-- I think what you're saying, Pri, about centering this on the mother. This is centered on the person that has the pelvic health problems.
In that case, it's not so much data as devices and different types of practitioners, and so forth, that are helpful, that have to be integrated. I don't know if he mentioned on the podcast, but there's also Connective Health, which is doing that for mental health patients and try to integrate the information for them and to center around the mental health patient. This is extremely promising, very interesting what you're doing, extremely powerful because it's not as if you're building out any one of these legs of these things. To a certain extent you are, but it's not a huge reach. Please continue.
Priyanka Vaidya: It's funny you mentioned that because devices and tele-health providers also play into this ecosystem. We are also thinking of a remote patient monitoring system, where the pregnant mom walks home with a connected smart scale so she can measure whether her weight gain is within defined limits. If someone has gestational diabetes, the doctor then can give her a connected glucometer and for blood pressure issues, we send the patient home with blood pressure cuffs.
The Delfina app is also looking at all this data in real-time. It's continuously triaging the patient in the backend. We are making use of devices and we are also adding in tele-health practitioners. Perinatal mental health therapists, registered dietitians who guide the patient all along.
Sal Daher: Very powerful and very promising.
Priyanka Vaidya: Fantastic. Thanks.
Senan Ebrahim: Thank you, Sal. I think there's a lot of parallels. You know a little bit about my background. I previously was a neuroscientist and I've seen the power of when you can pull the right data at the right data at right time and the right form. How it's transformed neurologic care, functional neurologic care from epilepsy to sleep, and how it's transforming cardiovascular care that we were just talking about earlier, Sal. It's transforming disciplines across medicine and it's really inspiring to see we really enjoyed listening to the podcast that you did with Dr. Wiygul.
It's great to see other innovators in this space in women's health and in particular in pregnancy as historically, the NICHD is one of the most underfunded of the NIH bodies. In general, this has been even looking at private investment from investors like yourself and venture capital, it has historically over decades been underfunded. Right now it's finally having a moment in the sun and we're really excited and grateful to be part of that. The problems are extremely complex from a scientific standpoint, a clinical standpoint, an operational standpoint, and bringing together the right stakeholders.
“...how are we going to take better care of the four million pregnant moms in 2023 than we did in 2022?”
We've already done this quite a bit on our own team, which I'm very grateful for, but also as a field of obstetrics, how are we going to take better care of the four million pregnant moms in 2023 than we did in 2022? It's going to take a village, so we're grateful folks like Dr. Wiygul out there. We listened to his podcast, if he's listening to yours right now, you can drop him a line and we're building up a system. A system into which various other offerings, digital health offerings, hardware offerings we hope will be interoperable.
I saw that at Hikma Health where if you have systems that are interoperable and they work for a refugee in one place, and they show up in another place and you can access that data in the right time, you could save their lives. We believe the same is true in pregnancy care, not just in refugee camps where certainly I saw that as a need, but even right in Boston. I was in medical school at Harvard and I was fortunate to learn with the very best in some of the best hospitals on earth. Unfortunately, while I was there, I witnessed a stillbirth on the OB-GYN ward I was in med school.
It struck me that this poor patient had lost her baby, that we could actually do so much better for her and millions of pregnant patients like her all around the world if we could just get the data in the right place at the right time in the right hands. I dove into researching the space and building machine learning models that could predict these preventable complications of pregnancy. I quickly realized that indeed it had scientific legs, but at the end of the day, in order to actually execute on this vision, I would need the right people to tackle this problem with me.
Senan Reached Out to Priyanka for Product Savvy and to His Brother Ali for Experience Building Sophisticated Data Models
My first two calls when founding the company was to Pri who was the product visionary behind iMotions which you may know, Sal, is within the EEG [electroencephalogram, measuring brain waves] platform and I knew well from my PhD work in neuroscience. Ali who happens to also be my brother and also just the most brilliant engineer who specializes in machine learning, as you know.
Also happens to have a PhD in computational biology, they set the bar pretty high in our family.
[laughter]
Senan Ebrahim: He's at work coding as we speak right now. It's just been such a pleasure building the company with the talented team. Looking to the future, we're really excited to continue building the right kind of team to execute on that mission. Pri?
Priyanka Vaidya: Thanks, Senan. I was so ecstatic when I got the call. We knew each other from the EEG world. I've always been wanting to be a part of a really early growth company, and it became such a full-circle moment when Senan shared that he was thinking of implementing AIML [artificial intelligence/machine learning, ie making sense of data that is unstructured or not in broken up into fields like in a database] models in the pregnancy space, which as you now know, was so very personal. I was delighted to join the Delfina team and bring this machine learning model to the real world, which is very challenging but with the goal of saving a million babies, which really excites me as a product person.
Sal Daher: A very, very lofty goal. I hope that you get there. Let's get a little bit to exactly how this integration is being done. Let's review. You're getting data from devices, from electronic health records and from the patient herself. If you can walk through those three and explain what is it you're doing to capture that data and to integrate.
The Patient Journey for Delfina
Priyanka Vaidya: As a product person, I like to talk about journeys and so the patient journey in this instance is when a person finds out she is pregnant and sees her OB for the first time, she will be enrolled in part of the Delfina care system. What that means is the patient then downloads an app, an app designed by Delfina, and also goes home with a smart scale. In the app, we ask the patient to input her daily mood so we can start tracking her mental health. We start asking her to weigh herself daily so she knows how her weight is trending, whether it's in the right direction, she's gaining enough or not.
We also start tracking symptoms. Then, if necessary, blood pressure and blood glucose levels. These are the data sources that start flowing in from the pregnancy app. In the meantime, Delfina, the platform is also acquiring data from the OB's [pregnancy doc] EHR system. We have much more information about the patient's past history, her clinical details, her vitals that are not part of the pregnancy app. Then to add to it, we're also thinking of adding in this feature where a patient can opt in to a software that lets us, Delfina, fetch her clinical records from every other medical provider that she has visited, giving us a very data-rich source to start triaging her on the pregnancy risk calculator.
Sal Daher: Very interesting. This brings to mind a little bit what Connected Health is doing. Because they are integrating data, not just from the EHR, but also from the prescription medicine management system. In the case of mental health patients, it's particularly important, so that they can give a sense of how much compliance is occurring to prescriptions. If the prescription has been filled, if there's a record somewhere that the patient has been using the medication, and so forth, so that the primary care physician, when she looks at the data, she's seeing a much more complete picture.
In this case, I wonder if having prescription drug information could also be helpful to help complete the picture for the OBGYN.
Priyanka Vaidya: Yes, absolutely. Something on our roadmap that we want to start building in and adding in more rich data sources as we need. Moving further along in the pregnant person's journey, we then track this mental health data or the weight data to make it a closed-loop system. If the patient is trending mentally low for two or three days, the OB is notified in real-time to say, "You need to give this person a call. She has a history of a mental health disorder that we know of and she is trending low. This is a patient that you should track furthermore."
“This data is being enriched as the patient travels along the journey.”
This data is being enriched as the patient travels along the journey. I'll pass it to Senan to talk more about how the AIML models also work in the back end to make use of this data.
Senan Ebrahim: Thank you, Pri. As you heard from that walkthrough of, for example, for a mental health patient, there are some pretty clear indicators like the ones that Pri was sharing. If we can see the mood trending low, those are things that clinicians and folks, in general, they intuit that they are going to be predicted. From my past work in machine learning applied to different indications across different sectors in healthcare, I've seen this time and time again where there are so many factors that we may or I may not immediately go there, but they're still highly predictive.
Long-term, that's Delfina's real goal is after we have aggregated all this data, enable clinicians to use their best clinical judgment, seeing the data and its visualization, but also give them insights that are derived from us being able to crunch millions of data points and synthesize them in an actionable way. As an example of that, I'm really proud of our data science team led by my brother Ali, that really dug in. It's easy when it's your brother--
Sal Daher: Your partner in crime for the Hikma Health episode. We want to hear about the crimes You're up to in their backyard.
Senan Ebrahim: This is actually the third brother.
Sal Daher: Oh, the other brother. Oh, how much it takes.
Senan Ebrahim: There's three of us. Pri has gotten to know them both, as well as my grandma. They are very much part of the honorary Ebrahim family.
Basically, then in pregnancy, it's always been a challenge to predict a lot of these outcomes because there's so much that goes into it. Ali's team with this NICHD challenge, where they basically got this data set that has 9,000 pregnant patients. Our team was able to predict hypertensive disorders in pregnancy earlier and with higher predicted performance than any other model that had been published to date.
The real secret to that performance, everyone had had the same data set. We were recognized among 100 plus groups for innovation and addressing racial disparities in care, because our model had taken a lot of input from our clinicians who thoughtfully went through the data, and like you said, Sal, they made it usable. They made it interpretable. They made it useful to a machine-learning paradigm that doesn't necessarily optimize around the feature inputs of a bunch of ICD codes and other kinds of billing procedural code that came out of an EHR but it's translating that using clinical knowledge into useful features.
That's what can power machine-learning-based predictions well in advance and with higher performance and greater critical interpretability for clinicians who are looking to make the very best decision they can on a clinical action timeline for their primary patients. The future iterations with the Delfina product as we do more and more clinical research as well as validation of these models, we aim to provide those insights as well to all of our clinicians.
Sal Daher: This is interesting because EHR, Electronic Health Records, were created originally for billing purposes. The hospitals wanted them so they could bill for procedures. They're not geared towards providing information on patients but software had grown enough so that you can now use machine learning to get around the legacy, shall we say, original sin of EHRs, that they're designed as billing systems instead of health information systems.
They really are fulfilling the promise of becoming health information systems because you can pick up stuff that even though they're still just doing their mundane job of compiling all these codes, assigning things through the various billing codes, and all that...good stuff is coming out of it, slowly. It's taken way too long, just way too long but it's very promising now.
Machine Learning Is Just Starting to Do Useful Things in Medicine Like Identifying a Stroke Quickly or Predicting Preeclampsia
Senan Ebrahim: Absolutely and we believe that the promises, we're only just scratching the surface. There's been a whole generation of AI machine-learning tech companies. Some of which have been very successful and we have friends at some of them. You can see all of these neuro-radiology companies like Viz.ai congrats to our friends there who have just achieved a ton creating an algorithm that can basically identify stroke like a large volume inclusion in the brain very quickly and it can time and frame it. That they can help really save folks' lives and improve their outcomes in real-time.
That's been done and we've seen that and it's been successful. Then when you come to space like pregnancy where the machine-learning-based question is not quite as cut and dry and there are so many other inputs besides just the image, the second generation of machine-learning-based companies in healthcare, I truly believe now, I know you've seen this in investors in other companies. This generation of companies is not going to be about a data dump of a raw image or raw scientific stream. It's going to be heavily clinically curated and like you said, making the most of the data that is available.
A great example of that is a colleague of our Dr. Melissa Wong at Cedars [Cedars Sinai Medical Center] demonstrated recently that if you use natural language processing [artificial intelligence for understanding text] through birth models as applied to EHR records, you can actually much better predict preeclampsia, one of the most costly and devastating complications in pregnancy. It gives us an opportunity and a window as Delfina. When we think about in future how we can utilize various data sources, we're no longer just talking about those billing codes, we're also talking about digging into that note and potentially being able to extricate certain elements that could be predictive.
Sal Daher: The unstructured fields in EHR where the practitioner is writing notes and we still find it difficult to understand it. With natural language processing, you can understand it and make some sense and be able to predict preeclampsia from practitioner notes, the doctor's notes.
Senan Ebrahim: Exactly.
Sal Daher: Very promising. Let's think a little bit about what's the next big step for Delfina.
Senan Ebrahim: I'd be happy to share from a company perspective and I'll turn it over to Pri to talk about her 2.0 vision for the product. When we think about Delfina as a company, the main work that we've been doing this year is developing a proof of concept and validating it. That means doing clinical research with academic partners to ensure that we're able to predict outcomes with high performance. It means we're rolling out beta and getting user feedback from our clinician partners and our private patients and seeing it as users we're meeting them where we need to be.
Then it means that we're laying the groundwork for a future commercialization of these products where we're able to not just be able to demonstrate to the world that this is the very best use of a system for pregnancy care and delivering that care but it actually measurably moves in clinical outcomes thereby reducing cost for our customers who are on ultimately the risk-bearing entity in healthcare. From payers to employers to risk-bearing healthcare systems.
Those are the folks that are getting really excited about Delfina care because, besides the tragic human cost of all of stillbirths and babies born with permanent brain damage and preterm deliveries and maternal complications from hypertension to gestational diabetes to mood disorders, beyond the human cost of all of that, there is a tremendous financial cost on our already stretched healthcare system. Mainly through NICU [neonatal intensive care unit] admissions and long NICU stay and the maternal hospitalizations and related care costs.
Ultimately, our goal as Delfina next year is to demonstrate to the world, "Look, we've already shown you all we can predict these previously so-called unpredictable complications of pregnancy. Now let's demonstrate that when we roll that all out in a system, we can actually measurably improve the outcomes and save on costs for our system."
Sal Daher: That's what's going to drive adoption. If you can save on the costs, the payers are going to be all over you and that is going to pay off in terms of better healthcare for the patients because if a baby doesn't get admitted to the NICU, that's a huge victory.
Incentives Are Aligned in Pregnancy Care, Nobody Wants Babies to End Up in Intensive Care
Senan Ebrahim: Exactly. It's crucially what I love about this space Sal, is that the incentives are aligned. This isn't always true, unfortunately, in neurology or a lot of surgical specialties, sometimes incentives are misaligned, but in pregnancy, the payer who's on the hook financially wants nothing more than to see fewer babies having to go to the NICU and fewer moms having to be hospitalized. To our North Star, serving the patients first and foremost, that's what we want. We want healthier outcomes for moms and babies and the providers who are essential. They're the ones taking the 2:00 AM call on this.
They want to be able to take better care of their patients, spending less time on aggregating and trying to collate all this data, just getting the key insights that they need to be able to take care of that patient. That perfect alignment between the so-called three P's in healthcare doesn't always happen and while there are a lot of challenges in our space and I'm grateful for Pri's partnership really envisioning the product that can thread the needle. At the end of the day, the incentives from a financial standpoint, from a business strategic standpoint are incredibly well aligned in a way that it's not always true in healthcare. Pri?
Priyanka Vaidya: To build on what Senan said, the product strategy very much wants to leverage this AI and machine learning models which I see are competitive differentiators and so my vision for the product is really to bake these AIML models in the product as much as possible. Let me give you an example. There's thousands of pregnancy apps out there, but all of them are what I call open-loop. You get a lot of patient information articles that you read but think about a poor mother who has twins through IVF. Her situation is very different. She now has to trust Dr. Google to go look up resources for twins.
What Delfina Means by “Closing the Loop” in Patient Care
Sal Daher: Let's recapitulate a little bit. Open-loop means the patient is out on her own trying to interpret data. From my personal experience, I read something and my fingers are tingling and so I talk to my primary care physician and say, "Do I have diabetes?" He says, "No, you don't have diabetes. There's a virus going around. Your blood sugar is perfectly normal. There's a virus going around that makes your fingers tingle." I was like, "Oh no. My God. Oh, no." The patient has no idea.
It's really important for that loop to be closed for what the patient is feeling and so it has to be communicated to the practitioner. The practitioner then has to be able to interact with the patient, do it in a low-burden way, in a way that's effective and happens at the right time, and so forth. When you say close the loop, now I know what you mean. You see that? It wasn't clear to me before what you meant by close the loop. How is it to improve the closing of the loop, if you can flesh it out a little bit?
Priyanka Vaidya: That's a fantastic question, increasing the closing of the loop.
Sal Daher: No. Just give it more detail.
Priyanka Vaidya: More color and more detail, happy to. As part of the Delfina app, we have a short onboarding questionnaire. The onboarding questionnaire asks specifically about your past mental health or clinical history. It takes a snapshot of things like, what kind of diet are you on? Are you a vegetarian because that does make a difference compared to from a nutritional standpoint. It takes a look at your physical health. What is your BMI [Body Mass Index] going into the pregnancy? Are you exercising enough?
We are using all these data points to build a story about the patient, to educate her, and meet her where she is at. Also in the backend, use the machine learning models to triage her in such a way that the OB knows on a risk score where this patient fits. Is she a high-risk patient that I have to check in with 10 times whereas is this a low-risk patient that my midwife can handle now in a telehealth consult? We want to also focus on practice efficiency and reduce the burden on OBs, giving them noise, not noise, but signal as I like to call it. Give them data that is important to them and not just a host of patients' vitals.
Sal Daher: What's the timeline that you expect to be actually having customer adoption?
Delfina’s Platform Is Actually Live at a Practice in California
Priyanka Vaidya: I am very happy to announce this week is when our data actually launched. We are in the process of rolling out Delfina Care platform to a small practice in Southern California.
Sal Daher: Golf tournament clap. Golf tournament clap, clap, clap, clap, clap.
Priyanka Vaidya: Thank you so much. Yes, it was very stressful, but we did it and we are now live in Southern California. We will start seeing patients enrolling through the app, using it, and giving us real-time feedback.
Sal Daher: When is it that you expect for that to go into a commercial phase?
Commercialization Will Focus on Under-Served Populations
Senan Ebrahim: The commercialization process is, as you may have seen with other companies, Sal, is a longer journey. We decided basically, understanding the landscape here, who needs this? Who needs Delfina Care? Who really believe every single pregnant mother on earth can benefit at some level from Delfina care? Looking at our own country, it's really Medicaid patients that most need this. Many folks have probably heard the statistic that Black women in this country overall are 2X to 3X more likely to suffer from most complications in pregnancy, including perinatal mortality.
What I recently learned is that Native women are up to 8X as likely to suffer from that. If we think about making first of all a product that works for all these constituencies, it's a real credit to Pri and her team about how much time and deliberation they took with different users in diverse backgrounds. Even, for example, as such an early-stage company, making the first iteration of the app that's now in data, available in Spanish as well as English, I think speaks to our vision as a company of being able to serve all those patients.
To boot what we realized is if we are commercializing this technology direct consumer, most patients that can benefit from this are just not going to be able to afford it. Whereas as I mentioned earlier, the real payers that are covering largely the costs of pregnancy in this country, where if you look nationally, every year, they're spending $40 billion on pregnancy care, and roughly another $80 billion on the complications of pregnancy care.
There's a huge and historically under-appreciated market and opportunity for technologists like us, folks in the system space and device space, and elsewhere to really dig in and support lowering cost of care with better outcomes, and capturing some of that value as credit to you Sal our other early investors and the other investors, for companies like us in the space that see that massive and growing opportunity.
The way we actually get there is that we demonstrate off of these pilots that we've been doing, that look at the outcomes, look at how we actually deliver fewer of these complications of pregnancy. That is a very compelling narrative to an insurer who, typically year over year, that's one of the top three costs that they're trying to mitigate in their NICU loss rate. That's the commercial motion and us going with a credible set of data on the work that we've already done, both predicting retrospectively showing them that we with independent researchers are able to reliably predict at scale.
Then in our population that we've already served, being able to demonstrate to them that we're actually taking better care with lower costs overall.
Sal Daher: Amazing. I'm still processing the complication rate is three times in African-American deliveries, and eight times for Native Americans. That is staggering. So much suffering that's going on, and it's needless suffering because a lot of that can be addressed before it gets-- If it's happening at 8X and 3X, it has probably a lot to do with care, the delivery of care, access to care. There's probably a lot of low-hanging fruit to be picked off in terms of improving the delivery of care in those populations. Staggering.
Senan Ebrahim: We've already started addressing this, Sal, and there's often a lot of questions about this in the space. We're really lucky to have been working with a lot of thought leaders not just within the pregnancy space, but folks that have been looking at racial disparities in care. You're right, it's absolutely a question of access to care and a lot of social determinants. As an example, in that, an NIHD challenge that we participated in earlier this year, and we had the winning model for correcting hypertension, of the top 20 features, there were actually 200 features in the model. Just in the top 20, there were two that are actually social determinants of health.
One of which was the income level as a fraction of the federal poverty level, and the second was education status. This isn't a data set, mind you, where we weren't even using any kind of racial-ethnic data as a predictor. What we did was we looked between different racial and ethnic groups at the performance, our ability to predict, and we saw that it was actually lower in basically non-White ethnic and racial minorities.
What we did is we used a machine learning technique called SMOTE, and this has all been published on our website at hypertension.delfina.com. Using that technique, we were actually able to deliver higher levels of performance in these groups such that it closed the gap. It was still lower because these populations are just underrepresented in that data set as in so many other data sets. Looking to the future for us as a company, when you think about those kinds of statistics, how shockingly high they are, 2X to 3X for Black women, and 3X to 8X for Native women, we have to make sure that those folks are represented. If that means going to different practices that have that kind of representation, that's what we got to invest in early as a company with the right kinds of models that will perform at nationals.
Sal Daher: With those kinds of disparities, I wouldn't be surprised. If you're working with those populations, you're going to find that even though your platform is created for lowering the risk of pregnancies, you're probably going to be having ancillary effects and lowering the risk of cardiovascular problems and so forth. You might end up drifting in that direction because the signal is going to be so strong from the interventions that you do because if you're helping someone control blood pressure during pregnancy, you're helping them avoid kidney failure later on and so forth. Maybe that translates, so...astonishing. That is shocking.
Senan Ebrahim: It is truly shocking. We're grateful to work in this space. I know, Pri has been talking to a lot of pregnant patients, so I'll let her speak to the uniqueness of this experience. We see it as both a challenge and an opportunity. This is a time in someone's life where they have graciously partnered with us with their trust. What that means is delivering a high-quality product that's going to work for them, and being honest and transparent with them and with their care team with what we can do and cannot do. It's a time though, where they usually come with a higher level of engagement. Pregnancy is typically a time where folks take great interest in their health.
Sal Daher: You've got to [crosstalk]
[laughter]
Senan Ebrahim: Exactly. We have folks who have had chronic challenges throughout their life, whether it's hypertension, diabetes, substance abuse. We are at a time where often folks are coming to us and if we can provision the right resources, it can actually be a catalytic time. While we as a company are laser-focused on pregnancy care right now, you're absolutely right, Sal, one of the things we're tracking longer-term is with the interventions that we're enabling providers to do earlier in pregnancy, are there longer-term, positive, metabolic, cardiovascular, psychiatric ramifications of those interventions? Pri?
Priyanka Vaidya: Thanks, Senan. I would like to add that it was so eye-opening speaking with the population that ANA Women's Health serves in this pilot location. A majority of them are Spanish-speaking. I also found out that some of them did not have access to smartphones, which really, as a digital product manager, opened my eyes to the accessibility of care. What also became obvious was that I was, on the one hand, on paper saying, "Go see a registered dietician," but I wasn't thinking about the cultural nuances when it comes to-- [laughs]
Sal Daher: If they don't have smartphones, saying to someone, "Go see a--" There's so much to unpack there. It's crazy.
Priyanka Vaidya: Absolutely.
Sal Daher: Actually, in the interview with Senan on Delfina [Hikma] Health, we talked about this, the fact that they are creating an app, but does everybody have a smartphone to run the app on? I think we touched on that. Very good. Very interesting. Is there anything else that you want to touch on regarding the product? Because I wanted to get a little bit more on the entrepreneurial journey that you guys are on and how you came together as a team. Unless you have something else that you want to focus on right now, we can proceed to that.
Priyanka Vaidya: I think I'm good with the product discussion. It was wonderful to share with you, Sal. Thank you. Insightful questions.
How the Name “Delfina” Came About
Sal Daher: I'm going to listen to this very carefully again. Oh, one thing., delfina.com. That is such a cool name. If you're addressing a Hispanic population, Delfina is a perfect name. That is such a cool name and your little dolphin logo. How did you get that? How did you get delfina.com?
Senan Ebrahim: Thanks, Sal. We love the name too.
Sal Daher: Did you hire a multimillion-dollar naming consultant?
[laughter]
Senan Ebrahim: Well, if you consider, I was surfing with one of my buddies, so I guess, he'd charge me multiple millions of dollars but I did buy him lunch after we had this little experience, which resulted in the name. Shout out to my friend, DJ, who I went surfing with back in the spring.
Sal Daher: DJ is a genius. This is a genius name.
Senan Ebrahim: Well, I'll tell you what happened, Sal. Basically, we were out surfing in North Carolina, and we saw a little fin pop out and I was already thinking, "What's the right name for this company with this vision, transforming pregnancy care?" I see this fin pop out and what he thought, DJ, if you're listening to this, he thought it was a shark, for the record. He started making a beeline for shore. I am a slower swimmer than him by a lot. I was like, "All right, if someone's going to get eaten, it's probably me. See you later, DJ. Nice knowing you."
Then I actually stayed a second and another little fin popped up next to it and it was a mom dolphin and a baby dolphin doing their dolphin thing. Then it ended up being the whole pod. When I got to shore, I started reading about dolphin moms and it turns out of all the aquatic creatures, they are the most invested in terms of the pregnancy itself and the postpartum care that they delivered to their young. That was pretty inspiring to me. I figured it's better than getting eaten by a shark. In honor of that lady dolphin, we named the company Delfina.
My mom actually soon thereafter, told me, I should have probably known that having gone to medical school, but the Greek word delphi actually means womb. That's where actually dolphin comes from. It's the fish with a womb. I had already been thinking about Delphi, it's like Oracle of Delphi prediction.
Sal Daher: That's remarkable. I didn't know that. Jeez, a little bit of ancient culture [crosstalk] That's right because they are cetaceans, dolphins and so they give birth to live young.
Senan Ebrahim: Yes. Exactly. We basically, in thinking about the kind of company that we want to build and Pri really has been in nature force in our company for really always centering the patient. Where we realized the importance of the entire system of the provider who is really catalyzing the change that we want to see happen in pregnancy care of the payer, who's ultimately the customer we're serving. At the end of the day, it's really about in this pregnancy experience centering mom, centering the private patient, and that's what we aim to do by naming our company Delfina in honor of the pregnant lady dolphin.
Sal Daher: I'm going to make a joke for our Spanish-speaking listeners. Gracias a Diós que vieron un delfín y no un tiburón. ¡Si non, el nombre de la compania hubiera sido tiburona.com!
Senan Ebrahim: [Spanish language]
Sal Daher: [Spanish language] Not good. [laughs]
Senan Ebrahim: Claro, delfina es mejor.
Sal Daher: [Spanish language]
Senan Ebrahim: That is for Agustín if Agustín is out there listening.
Sal Daher: Agustín, you got to listen to this. You got to see [Spanish language] Genius. This is amazing. Shout out to your mom for this.
Senan Ebrahim: Absolutely. Thanks, mama.
Sal Daher: I learn so much from these podcasts. The connection to the womb.
The Founding Story of Delfina Care
Tell me the founding story and how you guys came to-- you hinted a little bit that you called Pri and others. I don't know who wants to tell the story. Go ahead.
Senan Ebrahim: Yes. I'm happy to kick things off. I'll let Pri share her version of the story.
Sal Daher: Describe the dramatis personae in the story, the characters. Senan, Harvard Medical School, completing your Harvard Medical School in the PhD program, MIT and Harvard Medical School. Please, continue.
Senan Ebrahim: Thanks, Sal. I was doing my MD, PhD, I was finishing it up, I always thought I was going to be in the neurosciences research unit for 10 years. As you know, I had founded Hikma Health, which we talked about at length on Angel invest Boston in 2020. I've always been passionate about technology as a means to better our society. I really credit my parents for that. I was born and raised in Santa Fe and Silicon Valley. I also lived in Boston for 13 years, of course, I have a lot of love for the Boston tech ecosystem that Sal you've been part of for a while, and I miss it, I have recently moved. I do miss Boston.
When I was in these hospitals, and going to all these events and MIT about tech innovation, a lot of that was thanks to Hikma Health, we've been through a bunch of MIT accelerators, and Harvard accelerators. I was pretty keyed in now to this being a space where we can actually start creating some real products. It's not just about innovation in a lab environment, in a sandbox, where we can publish a paper about it, we can actually start creating real technologies and tools for our patients and physicians to take better care of patients.
I'd already been thinking about a variety of different-- I had this little notebook that I went through when I was going through med school, I filled it up with 50 different ideas where machine learning technology could improve our practice of medicine. Ultimately, when I witnessed that stillbirth, I had to put the book down because it just dawned on me how big the gap was between what we could do for our patients in this space, and what we were actually doing for them.
That was the moment where I decided not to submit on that residency application. I knew I had to go find the right team to get on this mission. Pri was basically a legend in the EEG world for her work on iMotion where I'm just going to embarrass her for a second, where it's very difficult, especially in the neurosciences to create a product that resonates with not just a bunch of EEG epileptologist people that are going to look at the EEG signal, but with the patients themselves, with the person who's actually putting on that EEG headwear for you making it accessible to them.
Then with other B2B stakeholders who, no one had even thought might derive value from this. That was where it really struck me that I need someone like Pri, and then, of course, Ali. Without a very, very strong machine learning-based technologists in the mix, this would all be still a dream. We're really grateful for his partnership, as well his partnership currently being him being head down coding, while we had this fun conversation. Pri?
Priyanka Vaidya: Thanks, Senan. Like I mentioned, I was so happy to get Senan's call because as a product person working in the health tech industry for 15 years, I feel my mission, and I feel much more impactful on a daily basis when I work designing and delivering health tech products. Knowing that something I do makes someone's lives a little bit better is what really gets me up in the morning. To really design a pregnancy app that keeps the mom in the center of the care was a dream come true from a professional standpoint, but also very much a personal standpoint.
It's been a year, and it has been an incredible year that we've learned so much, I've learned so much personally along the way, and just looking forward to build the Delfina machine even more.
Senan Ebrahim: I can chime in a little bit about Ali, in absentia since among the three people here, I think I share the most genetic background with him.
Sal Daher: You know all the secrets.
Senan Ebrahim: Don't worry.
Sal Daher: He has kompromat, compromising material yeah.
Senan Ebrahim: He has more on me than I do on him, him being the older brother because I don't remember the earliest years that were probably the most embarrassing for me. In our family, I'm very lucky, my whole family I've learned a lot from different people, but Ali, literally he taught me physics, he taught me how to code in Python when I was 15. He's really always been the brains of the outfit and he and I have worked on a variety of things over the years. When I was embarking on this journey and Pri, me and Ali were really literally getting into this rowboat together, it's not even a sailboat yet. We couldn't even put up the mast yet.
Thinking of someone who would just be in it with us through thick and thin. The technologists usually being the one who really-- if there's a hole in the boat, usually the technologist is the one furiously duct taping. Thinking about someone who had not just the technical genius and vision, but also the grit I've seen his work at Google, where he had been a leading software engineer on the team that built the Gmail spam filter at the counter abuse team that failed to be more infamous for their work on the YouTube comments section, as he likes to say.
He's done a lot of fantastic work building machine learning solutions at scale at a top-tier enterprise tech company. That's exactly what we wanted to build from the ground up. Our vision for Delfina is delivering the solution not for a few hundred or a few thousand patients, but eventually for millions around the world. Bringing in someone with both the grit, to execute the technical vision to build the kind of product at scale that will work for millions of people. You can count the number of people on earth that can do that on one hand, and I'm just really grateful that one of them happens to be my brother and he dove in on this company with me and Pri.
Sal Daher: That is so tremendous. I think your team, let me tell you, there's a lot of energy. I'm going to make Senan blush right now, but Senan takes no prisoners. The guy is on top of things, he gets stuff done. I can just imagine that bringing together the Blues Brothers here and the Blues Sister-
[laughter]
Sal Daher: -with a blue shirt. It sounds extremely promising. That's great. One comment, we mentioned Agustin. Just for context, Agustin is a friend of ours. He connected Senan, me, Agustín López Márquez, shout out to him. He's at Nference and he's a great guy. I met Agustin when he was president of SQZ Biotech, which is now public. Thank you, Agustín, and I hope you're listening. I hope you like my lame joke.
Senan Ebrahim: I saw him just a little while ago for dinner. I'm glad we're doing the dinner and then the joke, but I'll let you handle him. We had a really nice time.
Sal Daher: By the way, tell him I'd love to have him on the podcast.
Senan Ebrahim: Will do.
Sal Daher: Very good. We talked about the entrepreneurial journey, we talked about the product journey or the patient journey. Any other topics that you want to touch on?
Senan Ebrahim: Yes. I think one thing we haven't really talked about is this pandemic that we're just coming out of. I think Pri and I, we've been reflecting a lot on I'd say, hopefully coming out of [crosstalk]
Sal Daher: Inshallah.
Senan Ebrahim: Inshallah, exactly.
Sal Daher: Or as they say in Spanish, ojalá.
Senan Ebrahim: You're pacing us here, Sal, with the languages. We really hope it's coming to a close here, but at the end of the day, it's fundamentally transformed healthcare. I don't need to tell you, Sal. You've seen it in all your investments. You've seen it in the Boston [crosstalk] out there.
Sal Daher: Oh yes. Absolutely.
Senan Ebrahim: MGH is doing things differently. They've gone PI. They're all adapting. It's great to see because in some ways like for us as physicians, it was the kick in the pants we needed. It was like there were decades in which we could have done certain elements of what has now come to pass. It was really only during the pandemic that I was volunteering at Brigham actually and I would see pregnant patients come in being wheeled in by one volunteer with no support partner having missed their last two or three appointments that they were supposed to come in for high risk.
Just seeing the tears stream and thinking, "We did not have our act together as the healthcare system for this to be happening like this right now where--" It was eye-opening for me, and it's just still a lot of what is important for us not just as Delfina but as the health innovation community to future-proof our inventions, and future proof the way we work with our provider partners, as well as the patients that we serve.
I think it's really a credit to the kind of product that Pri has been developing. We anticipate that whatever level of at-home care and/or in-person care is needed in the future, based on where we are with pandemics, with the kind of patient demands which generation things are like private care. We are building for that future for decades from now that those pregnant patients will continue to benefit from Delfine care the way I wish they could have in the spring of 2020 when the pandemic hit.
Sal Daher: The silver lining in the horrible disaster of COVID is that a lot of these barriers have been broken. The barrier to delivery of telehealth and so many levels. I think the next time that something like that comes around and will definitely come around. These kinds of pandemics do occur from time to time, we'll be better prepared.
Senan Ebrahim: Pri, I know you have some thoughts as well from the patient perspective on engagement.
Priyanka Vaidya: Yes. As Senan was saying, we've been thinking about telehealth in general and how it affects the OB space and the pregnancy space. There are some things that have to be done in person. The pregnant patient has to go in for an ultrasound exam. We respect that and we understand that, but we also want to take into account a mom who is high-risk, who is in her third trimester of pregnancy, and has to come into the hospitals three times a week, which if she's a working mom is simply impossible.
Trying to navigate and support patients like these, who could do much better with the care with daily health check-ins instead of seeing the OB and taking time off their work to go make these appointments, is something that we're definitely thinking of adding to the roadmap and supporting these high-risk patients much more.
Sal Daher: This very much reminds me of a startup in which I've invested. It's also a startup that I had the founder on, Imago Rehab. People who are rehabbing from stroke used to have to go to see a therapist. They used to have to go to the therapist's office. What Imago is making possible with a connected device, a robotic glove, and telehealth... is amplifying the reach of the physical therapist through this device and through the telehealth platform.
There's an explosion going on of interesting directions that we can go in with this connection of delivery of health through telehealth, through connected devices, integration of databases and EHR, and things like that. I'm so excited about this space.
Senan Ebrahim: Thanks, Sal. We know Chrissy from-- we're together in the LX Program at the Harvard i-lab, so it's a great--
Sal Daher: Oh, cool. Chrissy Clover. Shout-out to Chrissy Glover.
Senan Ebrahim: Yes. Chrissy Glover is amazing. I think what Imago Rehab is doing is just one of, you said it, Sal, a Cambrian explosion. There's so many peer companies that we're learning from, in so many other spaces from rehab to mental health, to diabetes care, you name it. We have lessons to learn from each other. We think this is from a financial standpoint, such a huge multitrillion-dollar problem at the overall healthcare level that-- every problem is now just so complex in healthcare that it's going to take in a village.
We're really excited to keep learning from peer companies like that and partnering, because again, our goal is really for that patient that we met that night three years ago on the hospital floor and to help her have the best pregnancy. For every patient that comes to our healthcare system in the future, to help them have healthy outcomes.
Sal Daher: Oh, that is tremendous.
Priyanka Vaidya: Thank you. Can I just say, this was my first podcast and you made me feel so comfortable and it felt like I'm sitting there right in front of you and talking about Delfina, so thank you.
Sal Daher: Pri, the reality of podcasts is that it's very low pressure because it's going to get edited. It's not live. You can always do a retake, and so it's low stakes in that sense. I end up sounding a lot smarter. Raul gives me 10 IQ points or maybe sometimes 15 if I haven't had too much coffee. Taking out all the ums and uhs, so I speak in perfectly complete sentences. [laughs] It's amazing. It's like you sound so much smarter in a podcast, but somebody who sees me in a video, I'm like, "Uh, mm, uh, uh." That's great.
I hope that the podcast will also help to make people aware of the great work that Delfina is doing, delfina.com. I'm very grateful to Senan Ebrahim and to Priyanka Vaidya for making time to be on the podcast. I would not be surprised if there's somebody out there who's working at some adjacent space who just makes a connection because of the work that you've done. This is very valuable. Thank you both.
Priyanka Vaidya: Thank you for having us.
Senan Ebrahim: Thank you so much, Sal.
Sal Daher: This is Angel Invest Boston. I'm Sal Daher. I'm glad you were able to join us. Our engineer is Raul Rosa. Our theme was composed by John McKusick. Our graphic design is by Katharine Woodman-Maynard. Our host is coached by Grace Daher.