What Happens When You Replace the Fax Machine with Shana Palmieri

welcome to Mental Health Evolution, a podcast about what's changing in mental health and why it matters. I'm your host, Rachel Harrison, inviting you into honest conversations with people from all perspectives in the field. Clinicians, tech founders, investors, insurance companies, and all the folks in between. Let's explore what's working, what's not, and what's next. Welcome back everyone to the Mental Health Evolution Podcast where we talk about how the landscape is rapidly evolving in the mental health industry. So today we are joined by Shanna Palmieri, who is the Chief Clinical Officer and co founder of Transfer all, which if you'd like to Google it after our episode, is spelled X F E R A L. Shanna is a licensed clinical social worker who spent years on the front lines of behavioral health, including as the Director of Behavioral Health Services at George Washington University Hospital in Washington, DC. It was there that she watched something that would change the direction of her career patients in psychiatric crisis sitting in overwhelmed emergency departments for hours, sometimes days, while her team of highly trained clinical staff spent the majority of their time making phone calls and and sending faxes to try to find them a placement. She co founded TransferAll to solve that problem directly. TransForAll is a real time patient transfer and care coordination platform that replaces a manual fragmented process with a digital system that matches patients in crisis to available care in minutes. Today we are going to talk about how that technology works, what she has learned about where technology belongs in behavioral health and where she believes it should never go. As always before we talk to our guest, we would like to bring up some relative articles related to our topic today. These articles may be helpful for listeners who want to learn more and dive deeper. All articles will be linked in the show Notes for this episode. The first one here is From NPR in August 2025 titled Children in a Mental Health Crisis can spend Days in the ER waiting for Treatment. And this article talks about a new study that found nearly 1 in 10 children on Medicaid who visit an emergency department during a mental health crisis end up stuck there for days waiting for psychiatric placement. In states like North Carolina, Florida and Maine, as many as 25% of those visits led to children boarding in the emergency department for three to seven days. The study documents what emergency departments across the country have known for years that the mental health system is failing people at their most vulnerable moments and the emergency room is often the worst possible place to wait for care. The second article here also from NPR is titled AI in the Mental Health Care Workforce is Met with Fear, Pushback and enthusiasm from April 2026. And this piece looks at how artificial intelligence is moving into behavioral health settings and the real tension it is creating on the ground. It covers a one day strike by 2,400 Kaiser Permanente Mental health providers in California, many of whom felt the AI driven triage changes were removing licensed clinical judgment from the process. This piece captures both the genuine promise and the very real risks of deploying technology in settings where getting it wrong can cost someone their life. It sets up the conversation Shanna has been having for years about where AI belongs and where it does not. PR Newswire is the third article. TransferAll expands real time Patient Transfer Platform in Southern California. This is from February 2026. And this is Transferall's own announcement of its expansion into Southern California. And it includes data from the platform's impact in Central and Northern California, where IT launched in 2023. Healthcare facilities using the platform reported a median patient acceptance time of just over 60 minutes by the end of 2025, compared to the national averages of 7 to 8 hours or more for psychiatric patients awaiting placement. One county reported cutting their average wait time from more than two hours to just 17 minutes. The numbers illustrate what's possible when you replace fax machines and phone calls with real time technology. These three pieces tell the same story from different angles. The crisis is real, the technology exists to help, and the debate over how much we should trust technology with our most vulnerable patients is happening right now. Shanna has been living at the center of all three, so I am excited to begin this conversation. Thank you for being here.

Thank you so much for having me, Rachel. It's a pleasure to be having this conversation today.

Yeah. And I'm excited about this idea of wait times. I feel like it is a known situation in our field that anytime a patient goes to the hospital for psychiatric care, there are extended wait times. Like that is sort of just part of how the landscape has looked historically and even worse, I think, sometimes for kids and teens because that requires a specialized placement. So I'd love to kind of talk through your journey about how you thought about the need and the solution to this problem.

Absolutely. Now I'll start. When I was the director of Behavioral health for this hospital, it was an academic hospital downtown, District of Columbia. Very busy. We continued over the years to become more busy and more patients came. So we had a team. We expanded the team 247 added staff. And what was quickly apparent to me was that as clinicians, where I really wanted us to be spending our time was in front of patients. They're in crisis. That is a critical moment for a therapeutic crisis intervention with a patient and their family. We're trying to determine best place for them to go.

Yeah.

And also to have an environment so that they can move quickly to the next place. And where we ended up spending the most of our time was administrative coordination of finding a bed. It was incredibly cumbersome. Even when you're a large academic medical center and you have some of the advanced tools, you may have a very advanced electronic medical record, you can send documents out pretty efficiently. The challenge is that the system's not connected. So you have your system in the way that you do things. And then you have a all of your partners that are part of different health systems, they have different technologies, they do things a different way, different intake departments. And so because this was so fragmented, when I have traveled subsequently into many hospitals across the country, so many hospitals, they have a list of paper with the names of the hospitals on it, the fax numbers, and then cross em out when somebody says, oh, I have a new number to send it to or a new email or send it this way. And then they're trying to remember what all of these different facilities do. So what's your specialty unit? What age range do you take all these different clinical capabilities. And what was happening was I would have this team of social workers that would call the same group of hospitals and if they were full, then you just wait to the next shift and then you call again or you wait for hours, or maybe you send the fax through and then you call them and you say, did you get the fax though? It didn't come through, but in the meantime you went saw another patient. So now you're four hours later. Right. So you can see how these administrative, operational, underlying challenges were leading to this larger system level issue. And what oftentimes I read is that the problem is when I read articles on this, there's not enough beds. So maybe this is true, maybe there's just never enough beds. And as we launched it out, we found out there's a lot of beds available. There is an operational problem with how we efficiently link all of the services together in real time and find where the closest, best clinically matched bed is for the patient, where sometimes there's five. And then the family can have that choice. And instead of this process going on for hours before you get a call back and it's on your voicemail and then you collect this information, you go talk to your family. Now sometimes it's in a matter of five minutes. And our median response time on the platform across the nation to get a response back about whether there's a bed available is 15 seconds. So within seconds, you know, I sent this to 17 places, six of them have a bed open right now. They're going to take a review of this information. I can go have a conversation with this family. The impact of that patient and their families experience drastically changes.

Oh yeah.

And we've done this in so many other markets in the world. Right. Not I would say in healthcare, but you know, oftentimes I'll say to myself, like during the time that we would spend going to get the, the fax sent and get somebody to come back, I'm like, the ER staff has ordered Uber eats. We got an Uber for a patient home. We use all of these sort of business to consumer platforms to get things done where it uses this network approach, it uses this real time digital approach. But we haven't transformed these technologies, these systems into the healthcare space where they can be really valuable and save lives and drastically transform someone's experience.

Yeah, I mean that is incredibly revolutionary for the way that I've experienced the mental health care system as a clinician over the last 28 years. Like that is a game changer. Yeah, I just, I just am amazed. It's pretty awesome. So can you walk us through how it works? If I'm a clinician in an emergency department who needs to find psychiatric placement for a patient, what, what happens next?

Sure. So when we think about if you're in the emergency department, this could also be clinicians in the field that may be at a mental health agency, be a crisis team going out to do placement. So you know, patients are in crisis in all sorts of settings. Many of them land in an emergency department. But any clinician who's with a patient instead of that, you know, they realize their patient needs a different level of care and service than where they're at right now. And so they can have this app on their phone, they can have a web based platform and they basically answer some basic questions about their patient and it will geolocate either to where the patient lives or their current location. And then it finds clinically matching facilities along the continuum of care. This might be patient may need an inpatient unit stay, they may need a crisis stabil unit, they may be looking for the closest empath unit so that the patient has a different experience in terms of what their crisis assessment Environment is like they maybe they need residential substance use disorder, detox, php, iop. They can come in here and select what they're looking for. And then they select maybe this patient they want to find who provides TMS or ECT or dialectical Behavioral therapy. They want to make sure they have certain types of services. So they click these off and then it matches these facilities and they can start to just click the ones they'd like to get a real time alert. They can put the clinical documentation in there. Everything's HIPAA compliant, it's protected, we have contracts and privacy just like you would with an electronic medical record that you're using and then they can submit that request. And rather than today where maybe you make a phone call and get a voicemail or you send a fax, you get a fax confirmation, you're getting real time notifications that are coming back. Somebody looked at was read, so you get a read receipt. When somebody looks at it, you can have real time chat, messaging back and forth with the other providers. Maybe they have an additional question or they want an additional lab work sent over. And so all of this is real time and it's also team based. So maybe the shift changes to the next person who's coming on. And now they can see everywhere it was sent, all of the communication, what happened, what the next step was, and on both sides. So nothing's also getting lost in this process. Sometimes I'd be in the ER and I would have gotten my patient accepted. And then my colleague comes on and they call the intake department and that person's like, I don't, did we accept that patient? I don't know. And then they start the whole process over again, right? So you have these gaps, whereas when everything's really digitally documented in real time, you have this back and forth until you get to your final acceptance. They can say, okay, the patient's on the way over. You get your accepting doctor, or maybe it's a PHP program or outpatient. They say, this is your appointment date, this is where you send the patient, here's all the information and send you attachments back to give to the patient and the family with information and then the patient moves on to that level of care. Both sides also collect an incredible amount of information and data on their performance, their response times, how often are they, do they have consistency in how they're accepting patients or applying decline criteria across staff, across your physicians, across your providers. So you start to really understand the operations in a different way. And then from a higher level, we can start to see truly where are the system gaps in geographical areas. So we may see this region doesn't have the necessary medical psychiatric facilities, but three hours away they have four medical psychiatric facilities that take patients with dialysis or intense for care needs. Right. And so you can really start to see even things like how far do certain patients have to travel? How far are the adolescent and children traveling? Because maybe there is a bed capacity for a specific population, a specific type of patient.

Yeah.

But when we say overall, well, we just need more beds. Well, what type of beds do we need? Right. What type of clinical services do we need to really truly serve the population of behavioral health patients that are needed? And that's different by region, county. And so when we start to have some of this information, it really allows us to build out systems that can provide the needed services to patients. So there's different levels that I really look at it. One is that direct experience of the clinician and the patient in the moment. And then we start to be able to look at more system level opportunities of where do we need services, where do we have challenges, where are the delays? This way we're really engaged in this continuous quality improvement process from a system level, a hospital level, local level.

Right. How do organizations opt into this? You mentioned, I'm just wondering, are all hospitals in an area like is this something that the system goes and collects this data or is there a way that organizations opt in?

Sure. So I'm going to talk about this from two different sides. One is what we refer to as the placement side. And so you have healthcare providers, systems, crisis teams, counties, they're looking to place the patient maker referral. And sometimes people are both, they're trying to bring patients in and they have patient go out, Right? Sure. But on the placement side, they're joining the network or the platform or getting access to this. They're signing the necessary documents for, you know, we go through their cybersecurity and HIPAA compliance process because they are putting patients phi in the system. So we take that very seriously, just like any other technology that is utilized to protect this information. So once we do that contracting, we then train their teams how to use the platform. So that's how any provider or healthcare organization can join to find placement on the what we refer to as the intake side, which would be your intake department or anybody who's receiving referrals for behavioral health programs. When we go into these large markets, states and overstate lines, it is the entire national database of every single behavioral health facility in the Country.

Okay.

And so it is PHP programs, substance use disorder programs, mental health dual diagnosis, what their clinical capabilities are. So you can search the entire United States to find these programs. And then we link them all in to be able to get requests. So whether they are participating with us or not, if there is a transfer placement user, they can send to anybody on this network and they can send that digitally instead of using fax, which saves significant time in itself, because fax technology, no matter how you use it, the underlying technology, you're looking at at least 30 seconds a page for them to go through. And it can have high failure rates. There's a lot of challenges, and the healthcare system is still heavily dependent on facts. So we change this to a digital transmission of this information so that it's instantaneous, it's real time, and it can be viewed immediately. It also allows us to have a lot of documentation on who's viewing the information, making sure that it's going to the correct place. So we're really looking at enhancing the security of a patient's private healthcare documentation as well. And then you have another level for transfer all intake. And that is what we have. That's our intake suite. And so we provide an intake platform to intake departments that want to join and have a contract with us. And what we do with those is we take all of the methods that they could potentially get a referral. Maybe it comes in through a fax, maybe it comes in through an email, maybe the patient walks in through their front door, or they're getting a real time request from somebody using our platform. It all streams into the same platform and organizes all of these requests in one place for their entire team. And then we're helping lift information that's really necessary so they don't miss anything to make sure that the key clinical information is considered. One of the challenges that we see in general is you may be getting hundreds of referrals a day. You need to make sure your facility has the clinical capability and the staff that is necessary to take care of patients needs. Behavioral health, psychiatric patients many times have comorbid medical issues. And facilities need to make sure that they can take care of the entire patient, not just the fact that they're suicidal or that they have bipolar disorder. So they we need to take a look at all of this. And sometimes these clinical packets that can send over can be 50, 100, 250 pages long. So it's prone for missing key information if you're trying to quickly get through the packets. And so we are finding and lifting really key clinical information and data to ensure that the clinicians who. And this is a particular philosophy of our technology company, is that the clinicians make the decisions, the clinicians make the final judgments. We are providing a technology and using every advanced underlying innovation that we can to help ensure that nothing gets missed, that there is clinical accuracy, and that we're lifting exactly what they need to see to make the right decision. But it is our fundamental belief that the clinicians remain the person who makes those final decisions. Because there's nuance to it. There's a lot of nuance to behavioral health, human behavior. You have to take a look at what's happening in your unit. There's acuity issues. So that's how the platform and the different customers sort of come to be and use the service.

That's amazing. I'm wondering if this is ever something that could be accessed by patients themselves.

Yeah, it's definitely something that I would love to see this happen at some point. I think that it would be an incredible space to be able to eventually move to a consumer access. So the challenges, and I see this all the time, I'll have family members or friends and they call me and they say, where should we go? And they're in crisis and they are trying to find the best match and facilities and recommendations. And it is a complicated system for families. So if you're not sitting there with a clinician, and sometimes you may be seeing a therapist, but they're daily job is not to understand the network of resources. Their job is to be a clinician. And so when their client escalates it, that's a challenging space too. Right. So that is something that I think would just be incredible to be able to bring to patients and their families to have access to that as well. We started with clinicians because of the challenge of what happens Right. In these health systems. But I think in the future that's certainly a place to end going.

I love that idea because I think it's interesting that you started this with do we really have a problem with beds or is there an X, like knowing where the beds are kind of problem. Right. And solving for that. I think there is potentially a similar thing when it comes to access to care in general of that not knowing where to go, how to find it. And so helping people do that improves access to care for all levels. I love, I love the concept behind

this, Lee, because, you know, when I talk to a lot of the companies, they're trying to find access to who are the people who need care and they have these amazing programs. There's some fundamentally amazing work going on in this space where there's advanced treatments, advanced opportunities. It's through these traditional marketing methods that happened in the past for behavioral health. So now you're doing a search on Google or you're searching through websites and you're trying to find information. And meanwhile these programs are trying to figure out how to do outreach. Right. Where can we help people gain awareness? And it's a challenging and fragmented system on many different levels. I think in the last decade there's been incredible growth in the behavioral health space in all of the levels of

oh yeah, oh yeah.

A lot of services and programs are out there, but there's a massive gap between awareness of where they're at and what they do and an easy way to find them without sort of these traditional just constant search methods. And when you're in crisis, that's the last thing anybody wants to sit down and do is a, you know, eight to 12 hour day as a family member calling and researching. And that's not your specialty in life. Like it's very challenging. So certainly I think there are some huge opportunities.

I love that idea. And I just want to highlight that you are. I. You said this, but HIPAA compliant information is protected. That's not always a given in our, in our field right now. So I just want to highlight that. And is this something that AI is part of your technology?

Here it is. So when we look at artificial intelligence, I think with any new innovative technology, there's a space where we have to be very mindful and ethical with our use and really taking a look at what the best use cases are. I always feel like there is a very positive use case we don't want to miss out on because it can really add to the treatment of our patients. So I'll give some examples of where I think that it is really impactful. So one of the ways, and this is one of the things that I was mentioning was really lifting critical information to make sure that nothing is missed. So clinical accuracy is incredibly important. You know, I've seen in my career you have a patient come into the emergency department or you're trying to accept a patient and you just have so many records and somewhere in a note somewhere there's something critical that if you miss could be detrimental to your patients outcome and sometimes it could be fatal. You know, a patient has an allergy, they can't take a certain type of medication. There are many challenges that we experience in the health system like that today. And this technology absolutely helps with pulling critical information to the top and organizing information in a way that allows clinicians to have at their fingertips the information they need to make critical and oftentimes life saving decisions for patients. So in that way, when we're looking at how are decisions made, when we have lots of clinical data that need to be considered, we do use artificial intelligence to help pull this information so that we can really take a look at it. And when we really are thinking about artificial intelligence, what we're really looking at here is an incredible ability to look at pattern recognition and to pull information. And I think it's important to really sort of understand what it is that it's being used for. Because when we look at the treatment of behavioral health patients, we need clinicians, we need human beings to be at the center of that treatment. That's our best intervention point that we have, is being present as a human being with another person to help move them and help formulate a plan and an intervention to help that person move forward clinically and with their treatment. When we look at the operations of how the entire behavioral health system works and is connected, I think there's incredible opportunity there to connect resources, look at patterns, look at where are we missing things? Are there areas where we could make operational improvements? So I look at the use of artificial intelligence in our particular space. We're using this as an operational tool and we're using it to lift clinical information, to enhance the accuracy and to give clinicians what they need in order to make decisions and to be at the forefront and still that final decision maker and still that clinician doing that clinical work. And we also use it to reduce administrative burden on highly trained clinicians. Nobody wants to REType A field three different times in three different systems, right? That like over documentation, connecting technology systems together, making sure that data flows on the back end. This all allows us to do our jobs better and it removes clinicians from having to do administrative work and gets them back in front of their patients. So I think that operational use case for artificial intelligence is really helpful. And I think right now that's where, you know, I would say personally a lot of my energy is focused because I think there's an incredible operational. There's incredible operational challenges in the space and opportunities so that we can really connect this system together and allow patients to get where they need to go very quickly. And we address this issue where we could be missing key clinical information. I think there's another area where the pattern recognition, when we look at your diagnosis of a patient, your clinical formulation of a patient, your treatment options of a patient. The way that I look at this, and this is not something that my company is involved in at all, we are completely operational and working on efficiency and connecting patients. But I think it's a very interesting space to have a conversation about because I think this is where much more of the fear exists. And I think that when you look at diagnostic categories today, there's system, you know, there's symptom clusters in the DSM and we're basing our treatment off interventions off of those clusters that have been developed in the dsm. And so there's traditional challenges that behavioral health space has in comparison to where medical diagnosis has come. Like we can't. I can't do a blood test and tell somebody that they have what type of psychosis that they have.

Yeah.

And one of the things that I think artificial intelligence can help us with because it can engage in this really powerful pattern recognition, is not diagnosing our patients, but ensuring that clinicians are asking the right questions and that we don't miss a question to ask. Because if I have a patient who comes into my emergency room with psychotic symptoms, I have to do an entire differential about where those symptoms came from. And I want to make sure that this isn't delirium. Did they have a uti? Are they using substances? Maybe they. Some patients in rare cases have brain tumors. There's this entire differential. And so where I feel like in the future artificial intelligence can help us is making sure that clinicians are asking the right questions and not missing anything for their clinical diagnostic formulation. It gets us to a better treatment plan. We have to understand these underlying etiologies are in order to get to the right diagnosis, in order to get to the right treatment. And then the final space is that I think a lot of psychiatry can be based on risk prediction. Today we want to understand what a patient's future risk is. And what we're trying to do sometimes with that risk is prevent the person in a way where maybe we're saying, well, we need to hospitalize them. And what I would love to see in the future with behavioral health, if we have powerful tools, is how do we move from some of our current mindsets, which sometimes can be, how do I prevent this person from engaging in violence or this or that, through detaining the person and taking away their rights, to a resiliency based model where we can use massive pattern recognition to say, what are the interventions that are truly going to help people. Right. And so sometimes we're using risk prediction in a way of like the bad thing that could happen. But I'd love to see massive pattern recognition, say if you do these three things, your patient's likely to get better. You're likely to reduce your risk level if you address these three variables. And so when we start to look at it from a healing and a resiliency perspective of being able to use this information to mitigate information that we did, we forgot to ask the right question or maybe it was a newer clinician and they haven't had experience with some of the different underlying contributing factors to what first break psychosis could be from. These are areas where I think we could really advance the field. And I don't think that the future is humans versus AI. I think this is a tool that helps us provide a better level of care and outcome to our patients and create a better system and how it operates. And I think we have to be really careful along the way and very mindful of how we do it because what we don't want to do is say, well, we have a 71% risk that this patient may commit violence and so we're going to incarcerate or detain the person. That I think is a negative use of the technology in compared to there's these three things we can do to really reduce this person's risk and provide a healing environment that's going to heal and help somebody in the future. So that's where I sort of have my philosophy around. I don't think it's either or I think that we need to look at how we can constantly use whatever tool is at our fingertips and to improve the system of care and transition behavioral health and psychiatry into the future in a way that allows us to help people heal.

Yeah. Yeah. Well, I appreciate that well formulated thought process on AI. We are going to need to wrap up this episode but Shannon, I want to thank you for being here. I am excited about transfer all and the opportunities that that creates and for our listeners. All of the articles will be in the show Notes. We hope you enjoyed today's interview and we will be back next week exploring everything that is changing in the mental health landscape. Thanks for listening.