Healthcare AI and Leadership Challenges with Medical Records | Aleida Lanza

Healthcare AI fails silently when the medical records it trains on are incomplete — Aleida Lanza on the AI strategy gaps leaders discover too late in production.

Healthcare AI fails silently when the medical records it trains on are incomplete, and most organizations do not feel the scope of that problem until production. Aleida Lanza, who spent 35 years as a medical malpractice paralegal before founding CaseDok, joins Chris Hutchins to examine why interoperability efforts that stop at the core clinical record leave every AI initiative exposed, and what data governance looks like when every image, claim, and note has to be in scope.

What We Cover

• Why patient portals only surface a small fraction of a patient's actual health record, and what is missing when AI models train on the rest
• How CaseDok enforces interoperability across the full record, clinical notes, imaging, itemized billing, and claims history, not just the fragment every vendor calls "interoperable"
• The economic argument: if a patient's insurance paid $113,000 in claims, that is $113,000 of health data the patient does not have
• Why the antecedent matters: how we got to the hospital is the missing story that reshapes treatment plans and malpractice exposure
• The regulatory path through Medicare.gov's Connected Apps Registry and what it takes to reach 65 million members

Key Takeaways

• AI readiness in healthcare is bounded by record completeness. A model trained on partial records inherits every blind spot in the documentation chain.
• Healthcare data analytics that exclude claims and imaging are incomplete by definition. The most consequential clinical patterns live outside the core EHR data most AI platforms touch.
• Static medicine ends when patients own their full record. Data governance that starts with patient access reshapes every downstream decision.

Frameworks & Tools Mentioned

• CaseDok full-record acquisition platform (clinical notes, imaging, itemized billing, claims history)
• Medicare.gov Connected Apps Registry (conditional approval path)
• United Healthcare, Aetna, and Florida Blue member data integrations
• Patient-owned health data models
• Antecedent-first clinical documentation framing

## Timestamps 00:00 Live from Data First Conference 01:20 Why interoperability is more than clinical data 03:40 Fragmentation, static medicine, and broken incentives 05:55 Why AI needs complete patient history 08:10 Missing data as invisible bias 10:55 Emergency care and inaccessible records 12:40 Patient ownership and transparency 14:30 Precision medicine and AI safety 16:10 Why patients should own what they paid for 18:30 How to connect with Aleida Lanza

About Aleida Lanza

Aleida Lanza is the founder of CaseDok, a platform that solves the acquisition and cost of complete medical records. She spent 35 years as a medical malpractice paralegal before leaving her career to build the interoperability infrastructure modern AI and clinical decision-making actually require.

Related Resources

• Episode: Caregivers as the Connective Tissue of Healthcare with Amanda Roser
• Episode: The Hidden Infrastructure of Trust with Amit Shivpuja
• Episode: The Scary Truth About AI in the ER and Why Clinical Judgment Still Wins with Dr. Natasha Dole
• Topic: Clinical AI & Patient Care<

  📘 Beneath the Signal, Chris's book on the human work behind trusted data and responsible AI in healthcare: Get it on Amazon

Transcript

Copy Transcript Copied

Aleida Lanza: 0:00

CaseDok, C-A-S-E-D-O-K. You can find us at CaseDok.com. We solved enforced interoperability, and we basically have solved the acquisition of medical records and the expense involved in the acquisition of medical records. Look at how we got to the hospital instead of what happened at the hospital. The antecedent is the story that we're missing to solve a lot of the issues in healthcare.

Christopher Hutchins: 0:40

I want to welcome everyone to the Signal Room. This looks different because we are in Las Vegas at Planet Hollywood at the Put Data First Conference. I'm really excited about our next guest. We've met over the last 24 hours or so. Welcome, Aleida, to the Signal Room podcast. It's been fun so far, but I'm most excited because you told me you just launched your business. Maybe we could talk a little bit, just tell me about your background and talk about the company and what your mission is. I also want to dig into the juicy stuff because this has been a fun event. I've never been to an event where there's so much excitement and passion. Welcome, and let's hear about you.

Aleida Lanza: First of all, it's such an honor to be here. Thank you so much for having me on the show. I am so excited to be here. And I wouldn't miss it for the world. About me, I've been a master medical malpractice paralegal for 35 years. I had to use a lot of tools and learn how to program to be able to manage the campaigns that I did at scale. I had to learn quickly. And in doing so, I definitely leveraged AI in a lot of the processing and rule connections and things like that. My love for AI, I'm still looking to potentially integrate AI safely into our just-launched product, which leads us to the most important thing: 1:28

tell us about the product.

Christopher Hutchins: 2:54

What is the name of your company? And by the way, for you listeners, she really wanted to be here because her company actually launched officially yesterday, and she's here with us. Unbelievable.

Aleida Lanza: 3:04

CaseDok, C-A-S-E-D-O-K. You can find us at CaseDok.com. We solved enforced interoperability, and we basically have solved the acquisition of medical records and the expense involved in the acquisition of medical records. When I talk about medical records, I'm talking about the whole corpus of medical records because currently every patient portal only reports a very small fraction of your actual record. We needed to solve that because at least from my perspective, I know health information is powerful and it can mitigate malpractice and certainly better inform our lives. So I set out to solve patient access and transparency more than anything. Absolute transparency in medical records. If you're not seeing exactly what your doctor sees, then come to CaseDok because we can fix that.

Christopher Hutchins: You are composing a masterpiece for the ears of the clinicians that are wanting to do AI, but they're scared because chief data officers like me really have not done a good job of proactively listening to where their pain points are. And you're talking about one of the things I hear most frequently: 4:13

interoperability. My friend Barry, I think you might have met yesterday, he's been a physician for a number of years, and one of his issues with us is that we talk about interoperability, but we really haven't solved it yet. And until we do, we're just continuing to keep people at risk.

Aleida Lanza: 4:48

Absolutely. And more importantly, when we talk about interoperability, currently anytime we see the word interoperability, they're really only having a conversation about interoperability of the core clinical record. Our goal is to solve interoperability for the entire corpus of the medical record. That's all medical records, all images, and all itemized bills. They come in different forms, especially because when you work with a hospital and they can afford very sophisticated software versus a practitioner in a rural area who may still be working with paper. We receive all of the information for the patient record in native format. If it's in HL7, we receive it in HL7. If it's in PDF, we receive it as PDF. What we are looking for is to collaborate in AI in a way that's safe because you can't put privileged information into certain systems. We're very interested in building a pipeline to make available the entire complete clinical record history of a patient so that AI can be unleashed on ending static medicine. Look at how we got to the hospital instead of what happened at the hospital. The antecedent is the story that we're missing to solve a lot of the issues in healthcare.

Christopher Hutchins: 6:25

You refer to it as people treating it as static. We've been measuring performance by the picture of an average person that no individual on earth actually fits the profile. We either penalize or incentivize a physician based on how they care for their patient against that profile. It doesn't work from an economic standpoint. Does it work from a quality standpoint in terms of care? How is it going to work if we treat AI with the same approach? Because they're still calling it a practice of medicine, which means it's an ever-evolving science and we're always learning. If our legislation and our policymaking does not have that same approach, we are going to exacerbate a problem that's already very disruptive. From your perspective, what are some of the things that you're seeing and how are you working to address this evolution and building the flexibility into what you do? Because there is change. I hope to God we don't ever have another COVID, but that was eye-opening for a lot of us because it was massively disruptive and we had to adapt so fast that we did make some mistakes. Unfortunately. But how do we mitigate that and how are you thinking about it?

Aleida Lanza: 7:46

I think that objective information is our best tool because in informing the informed consent, we have a duty to inform for our practitioners. They have a duty to inform the patients, but they rely on an accurate medical record. And we are all terrible historians of our health and our health history. The fragmentation that we face today in healthcare is causing us, if you want to do the math, we've been paying for the same x-rays our entire lives. Every time we go to a doctor and we have to report a medical history, we fill out another full high-level history. If you're thinking about the cost of the encounter, we pay for 15 minutes, and those 15 minutes are what we have to solve our challenge. And generally five of those minutes are spent trying to relay our medical history. We hope to solve it with precision, because we know that there's a complete record that we can curate. We want to unleash AI safely to be able to curate a precision medical history. AI can help, not in adding to fragmentation or hallucination, because there was recently a published article that showed that the reliability in emergency rooms was 33%, and that's a very alarming rate to be relying on in those instances. Whereas we can at least interject and inform providers on an emergency basis with a clear, precise medical history when the patient may not be available to report it themselves. That can save lives.

Christopher Hutchins: 10:06

You bring up another really interesting point because I don't know what everyone else's experience is like with their medical history, but I'm old enough that there's a lot of things I can't even remember. And if it's not in the record, how's a physician going to be able to help me with an informed prognosis or treatment?

Aleida Lanza: 10:30

And we aren't doctors. We don't know if that has relevance or clinical significance to the challenge that we have today that might change their clinical plan.

Christopher Hutchins: 10:44

We don't know what we don't know. And is it important?

Aleida Lanza: 10:46

And isn't it just time that we unleash AI to ask those questions?

Christopher Hutchins: 10:53

I think you're right. The challenge that we have to be mindful of and the flexibility we have to have in our models has to account for the fact that things are still evolving. And if you don't know what you don't know, AI is not going to bridge that gap for you completely. There's always going to be the possibility that there's missing information. And to me, that's the most dangerous kind of bias because you can't detect it, because there's nothing to detect.

Aleida Lanza: 11:20

And interestingly, healthcare is the only industry where we have to pay to get the receipts for what we paid for. That's another problem that we wanted to solve. Because in true accessibility, when you can look at your complete record history in one place, I've never been able to do that in my life. And I'm the CEO of this company. I haven't used it myself yet. Currently, we're only serving currently enrolled members for United Healthcare, Aetna, and Florida Blue. We are conditionally approved for Medicare.gov's Connected Apps Registry, but we have to wait for the government shutdown to hopefully end soon so that we can serve 65 million Medicare enrollees. They could, by scanning their license and health card, access their complete Medicare record history in one click.

Christopher Hutchins: 12:27

You seem to have this go big or go home approach because you're not starting small. You mentioned two of the biggest insurance companies on deployment.

Aleida Lanza: 12:36

I didn't leave a very successful career as a paralegal. I can assure you, I have about a trillion dollars in recovery. I have been privileged in being able to work with titans of law and educated by very amazing geniuses in law. And being able to support what they're advocating for gave me a great sense of reward. But I will share that I left this career because after 35 years in medical malpractice, I left. A family member was involved in an incident of medical malpractice, and I couldn't stop it, even though it was occurring, and I knew it was occurring as it was occurring. So I decided to leave my career and try to mitigate this by better informing and engaging patients so that they not only access their entire medical record history or claims history that exists in the carrier claim history, but to own it, download it, keep it. Get every CT scan that appears there, every MRI, whatever you paid, whatever the insurance company paid for that record. If it's $113,000 worth of claims, you're looking at $113,000 worth of data that you don't have. If I paid $113,000 for anything, you can be sure I'd have that in the safe.

Christopher Hutchins: 14:17

Yes.

Aleida Lanza: 14:18

I'm just saying, go get it while you can.

Christopher Hutchins: 14:21

This is exciting. I wish we could talk for two more hours. I would love to have you back if you'd like. I'm sure next time we have a conversation, there's going to be a lot more breaking news because what you're doing is so exciting. And I know your passion is going to get results. I am so grateful for you joining me today on the Signal Room. Let's stay in touch. How do people reach you if they want to get in touch?

Aleida Lanza: 14:50

I'm on LinkedIn, of course. Aleida Lanza, A-L-E-I-D-A L-A-N-Z-A. Our website is CaseDok.com, C-A-S-E-D-O-K.com. And my email is aleida@casedok.com. Anybody who's interested in bettering health, please reach out and collaborate.

Christopher Hutchins: 15:11

If you're trying to figure out how to get excited about AI, definitely follow up with this lady because she's got passion and she's excited. Thanks again, Aleida. It's been a pleasure. That's it for this episode of the Signal Room. If today's conversation sparks something in you, an idea, a challenge, or perspective worth amplifying, I'd love to hear from you. Message me on LinkedIn or visit SignalRoomPodcast.com to explore being a guest on an upcoming episode. Until next time, stay tuned, stay curious, and stay human.