Season 1, Episode 9: Transcript
Angela: Welcome everyone. Thank you for joining us. You’re listening to Atlanta Business Impact Radio with Veanne Smith, and I’m your co-host Angela Greenwell.
Atlanta Business Impact Radio is a podcast that showcases some of Atlanta’s most innovative and forward-thinking business and technology professionals. In our first season, we take a deeper look into the world of health care IT.
In today’s podcast, we talk about the challenges that hospitals face with patient identification, which leads to duplicate records and overlays, and how biometrics such as iris recognition are being used to protect our health records and allow hospitals to spend less time on manual data correction.
Both Veanne and I are very excited to welcome as our guest Michael Trader of Right Patient.
Veanne: Welcome everyone. Thanks for joining us. This is Veanne Smith with Atlanta Business Impact Radio. In this episode of our podcast, we will look at the use of biometrics as a method for patient identification to increase patient safety, protect patient data and to improve the patient experience. I’m excited to welcome Michael Trader from Right Patient as our guest today to talk on this topic.
Michael is the president and co-founder of Right Patient where he is responsible for overseeing business development, marketing activities and government outreach and for providing senior leadership on business and policy issues.
As a serial entrepreneur with a passion for innovation, Michael has been an early stage and founding member of several successful technology companies in the areas of strategic sourcing, biometrics and health care. He has advised dozens of fortune 1000 companies, traveled to more than 30 countries, is a frequent presenter at various health care symposiums and has been interviewed by media and research outlets such as CNN, Fortune magazine, Health Data Management and Frost & Sullivan.
Hello Michael, and welcome to Atlanta Business Impact Radio.
Michael: Thanks so much for having me, Veanne.
What are the approaches for patient identification and what problems exist today?
Veanne: All right. Great to have you here. So, I’m very excited about this topic around patient safety and patient identification. To get started, I guess for my own personal perspective and my own experience, the traditional approach to making sure that the medical professional is performing the right procedure on the right patient is to ask the patient his name and his date of birth, and then they match that to some data stored in a system somewhere, or they might look at a wristband or an armband.
Am I on target with that? Then I’d love to have you tell us about the approaches out there today for patient identification and what problems exist.
Michael: Sure. So typically a health care organization that does not have a platform like Right Patient, for example, they will authenticate patients or attempt during the registration process to figure out who the patient is and whether or not they have a medical record that exists within their electronic health record software by checking their identification.
That might be a driver’s license, looking at their insurance card, asking questions, trying to get a Social Security number or date of birth, a proper last name and using that information in order to locate the correct medical record.
If they find that a medical record does not exist, then at that point they would create a new medical record for the patient. It’s typically a completely manual process. Most health care organizations have policies and procedures and processes that they follow in order to avoid making any kind of mistakes, but the problem is that mistakes inevitably are made.
As we were talking about earlier, a big reason for that is because of the human element. We have human beings that are involved in that process. Another big challenge is standardization, so really there is a lack of standardization in terms of how the data, the different attributes that would help a health care professional to determine who we are, are not really standardized across organizations or oftentimes even within the same health care organization. So that makes it a big challenge.
There was a study that was done recently by the ONC, which is the Office of the National Coordinator of Health IT, which found that—and they surveyed I don’t even know the exact number but hundreds of different hospitals throughout the U.S., and they found that the best error rate in terms of patient matching was 7 percent, which is kind of frightening.
Veanne: Wow, that’s scary.
Michael: Yeah. So, it’s a really big challenge in the marketplace. You know, patients can be misidentified for all kinds of different reasons. You have patients who have very similar names, for examples. You have patients who come in who have no identification on them whatsoever. Maybe they’re incoherent. They can’t properly communicate with the health care professional.
You have registrars in hospitals who “fat finger” the registration information. Maybe instead of typing in, you know, 1, 2, 3 as the beginning of the Social Security number, they type 2, 1, 3. So there are all kinds of different causes for this problem, but the problem definitely exists on a pretty wide-scale basis.
Veanne: Yeah, I’ve heard from other folks I know in the industry. It’s not impractical or impossible to have two patients in the hospital at the same time with the same name on the same date of birth.
Michael: It’s very common actually.
Veanne: And where do you go to figure out which one is which, right?
Michael: Yes. It’s very common, especially if you look at certain parts of the country where you have a demographic that has very similar names. You know, you look at border states, for example, where you might have a large Hispanic population.
These kinds of issues are in some cases even more prevalent in those kinds of areas, and again, having similar names is one potential cause. There are a number of different causes. We have customers who had problems with what they refer to as frequent flyers who come through the ED, the emergency department.
They’ll use different aliases in order to get their hands on different pharmaceuticals, so fraud is another big, big problem in the marketplace as well.
Veanne: Well, I’m just kind of laughing thinking of my name. You know, I have Smith as a last name, but thank goodness I have Veanne, you know? I won’t be mistaken for someone else some day at the hospital as I get old.
Michael: That’s right.
What are the problems with patient identification?
Veanne: Any way, very interesting. Let’s talk about some of the approaches that are being taken now to eliminate the potential for errors in patient identification.
Michael: So, a couple of the big problems that exist in the market today are with what are known as duplicate medical records and overlay medical records. A duplicate medical record might be created when a patient arrives at let’s say the ED, for example.
Most of the time when a patient arrives at the ED the nurses that triage those patients, their primary focus is on providing treatment to the patient, so they really want to get the patient registered in the system as quickly as possible, and a lot of times they’ll do quick registrations where they get the patient into the system, especially if it’s an emergency situation, and then later they’ll try to figure out if that patient has an existing record and figure out what their actual identity is.
Veanne: And we’re thankful for that, that they do worry about treating them first.
Michael: Yes, absolutely. So what happens, though, sometimes is that if you have patients who arrive at the hospital who cannot be properly identified for whatever reason, maybe they don’t have any form of identification with them or they can’t communicate properly, a duplicate medical record will be created during that process. It’s sometimes referred to by the health care providers as a John or Jane Doe record.
So, when duplicate records are created, it’s a potential problem for a couple of different reasons. First of all, our health records or electronic health records contain a lot of very valuable information that is necessary to properly be evaluated by a health care professional.
Maybe it has allergies that we are allergic to or pre-existing conditions, so if that information is contained in a duplicate medical record and the patient is not identified correctly and matched to that record, then that information can be missed by the clinician during that evaluation.
So, it has the potential, you know, substantial impact on patient safety. That’s a big reason why providers are looking to invest in this kind of technology.
Now, an even scarier situation or scenario is when the health care provider creates what we refer to as an overlay medical record, and that would be where a patient is incorrectly identified, and their health information for that visit is actually applied to someone else’s medical record. These are very, very messy situations.
Most hospitals will have a department that’s called their HIM department where they have resources that have a number of different responsibilities, but one of them is to find and resolve duplicate and overlay medical records. So, typically they have tasks that are kicked off by a system internally that’s monitoring when duplicates or potential duplicates are being created.
On a regular basis, they have to go in and look at these medical records, try to manually compare all the information, determine whether or not it’s the right person. In the cases of overlay medical records, it can be very messy because you’ve got a lot of health information that you have to sift through and figure out what belongs to—
Veanne: It’s become intertwined.
Michael: Exactly. It was actually a lot easier when we had paper medical records, but with the move towards the digitization of this information, it makes it a lot more difficult. There are some statistics out there from studies that show that, first of all, the cost—the average cost to a health care provider in resolving duplicate medical records is about $60 per duplicate pair.
When you take into account that, again as I had mentioned, the best error rate in terms of patient matching is about 7 percent, and there are other studies from organizations like HIMSS and AHIMA that put the average duplicate rate at anywhere from 8 to 14 percent, so when you run those numbers in looking at the annual number of patient visits for a particular health care provider, it can add up to a considerable amount of money.
Overlays happen less frequently, but they’re much more expensive. It typically can cost anywhere—or it can take anywhere from 60 to 100 hours to resolve an overlay medical record.
Michael: And you’re talking about HIM resources, and you figure out what you’re paying them, and you can run the numbers to see what that potentially can cost.
Veanne: $10,000 or something along that line?
Michael: Yeah. An HIM resource, I think the study that showed the average hourly rate for an HIM resource was around $19 or $20 an hour. The other thing, though, to keep in mind is that this is not—while this is important to us as patients and our safety, you know, resolving duplicates and overlays, it’s not a really value-added activity for a health care provider.
One of the other activities that HIM resources spend their time on is coding, which is directly related to the reimbursement that the health care provider receives, so that’s a much more value-added activity.
So, one of the big benefits of technologies like ours is that we free up time that these HIM resources, who might otherwise be working on duplicates and overlays, can spend on things like coding, which goes directly to revenue cycle and the bottom line for the hospital.
What are the approaches to eliminate patient record errors?
Veanne: Yeah, that’s amazing. Thanks for all the explanation of overlays. It’s a terminology I’d not heard about before. I was not familiar with that and didn’t know that that was a problem, so thanks for that education. What are the approaches out there today that are being taken to eliminate the potential for errors in patient identification? Maybe you can talk about what’s going on, and that segues into what you’re doing.
Michael: Sure. So, the typical approach that is taken in the absence of a solution like Right Patient is manual processes that a hospital institutes where the registration professionals follow certain rules and guidelines and try to standardize the process as much as they possibly can, but inevitably, mistakes are made because, you know, there are humans beings who are involved in the process. Sometimes you just have situations where it’s impossible to know the identity of the patient.
I’ll give you a perfect example. We have a customer who’s based out of North Carolina called Novant Health. They are a pretty large system, a 14-hospital system. They have over 340 clinics in four states, I think. They actually published this story on their website, so any of the listeners can go on to Novant’s website and check this story out. They had a situation where a woman came into the emergency department. She was confused. She was incoherent. She couldn’t communicate properly. She didn’t have any identification on her.
Normally in those kinds of situations, the provider will do their best to try to figure out who the patient is. They’ll maybe call the local police department or fire department and use whatever resources are at their disposal in order to find out this person’s identity, which is important for a couple reasons. One, if the patient has an existing medical record within that hospital’s electronic health record system, then clinicians want to be able to access that so that they can look at their medical history and provide the best care. Two, you want to be able to notify this person’s family and loved ones, etc.
Because Novant was utilizing the Right Patient solution with what we call photo biometrics, which is iris recognition cameras, they just simply took a picture of the patient. That’s how the technology works.
Our software extracts the unique iris pattern data from the photo and uses that to identify the patient, and then we integrate with the hospital’s electronic health record software. In this particular case, they took a picture of the patient. We identified the patient and pulled up her medical record. She had one already in the system, so they knew immediately who she was, and they knew immediately what her entire health history was so that they could provide the highest level of care.
What are the different types of biometrics that can help with patient identification?
Veanne: That’s such a great success story. Congratulations on that. So, we’ve kind of segued into the area where processes are changing now in biometrics, and I guess that’s what you are specializing in, so we’ve kind of moved into biometrics. I’d be curious to hear. There are probably multiple forms of biometrics. You’ve talked about iris scans. Can you tell us about others and explain how they work?
Michael: Sure. So, first of all with respect to iris recognition, we actually refer to it as photo biometrics. The reason for that is because—one of the reasons is because we’ve found that there is a lot of misinformation in the marketplace about iris recognition specifically. You tend to hear people refer to it as retinal scanning, for example, which is unfortunate because not only is retinal scanning a completely different type of biometrics altogether, but it’s not even utilized at all. There is no vendor in the market that exists who is providing—
Veanne: That’s interesting because that’s the term I always hear.
Michael: Yeah. Well, and I think it’s just a matter of information, false information that’s being put into the marketplace about how the technology actually works. We kind of take a different approach. We have a different background and approach compared to other vendors in this particular space because we have almost 14 years of experience in biometric technology, so we know how all these technologies work inside and out and what is going to best position our customers and the health care market for the future. With photo biometrics, as I mentioned, you take a picture of the patient with a camera that works just like any other digital camera.
Veanne: A normal camera?
Michael: A normal camera. It’s not like anything you’ve seen in a Tom Cruise movie, right? You know, there’s a red laser that’s shot into your eye or anything like that. All of that is false information. You don’t have to get really close to the camera. You basically just stand 18 inches away from it, and it just takes your picture. In fact, one of the big advantages of this technology is that we are simultaneously capturing a high-quality color photograph of the patient, which then becomes part of their medical record as well, so that any clinician who is retrieving that—
Veanne: Can look at the picture.
Michael: Can see the picture.
Veanne: That’s another—almost another safety point, right?
Michael: Exactly. A big safety point. On the clinical side, the clinicians really like that feature. So, when you take the photo, our software goes in and reads the unique iris data and uses that to authenticate the patient. One of the big benefits of iris in particular is that it is the most accurate form of biometrics in the marketplace. There was a study that just came out, actually last November, by Frost & Sullivan who confirmed through their analysis that iris is more accurate than any other form of biometrics.
Veanne: Did you evaluate other forms before going to iris? I was curious.
Michael: Yes, absolutely. Our software actually supports many different forms of biometrics.
Michael: We support, in addition to iris, we support fingerprint recognition. We support vascular recognition, palm vein and finger vein. These are technologies that read the vein pattern underneath your skin.
Veanne: Right. Okay.
Michael: And facial recognition as well. So, our software platform technically supports all these different forms of biometrics, and you can utilize Right Patient with any of these forms of biometrics.
Veanne: But the most accurate is the iris.
Michael: The most accurate is the iris. It has some other advantages over the other forms of biometrics as well, simultaneous photo capture. It also has a very low minimum enrollment age as well, so the minimum enrollment age for iris is 12 months in comparison to something like palm vein, for example, might be 12 years.
Veanne: Because you wait for it to develop?
Veanne: The iris is more developed at an earlier age?
Michael: That’s right. Yes. So, fingerprint, typically the minimum enrollment age is about 5. There are some advantages. Another big advantage that our customers see is that the iris cameras are completely contactless, especially in this day and age where everybody is now talking about the Zika virus.
Michael: Before it was Ebola. So hygiene and infection control clearly in a health care environment are extremely important to the provider, first of all, but also to us as patients.
So, when you’re looking at a device sitting on the counter that you know is meant for every patient to use, it’s a little different than like a door handle on a restroom door, let’s say for example, because not every patient is grabbing that door handle necessarily, but with the biometric system, you’re assuming that in order for the provider to get the most value from it, every patient should be using it. Sometimes hygiene is a concern that comes up in our discussions.
Veanne: It would seem to me that it’s less invasive or noninvasive compared to having everyone take a fingerprint. There’s almost a legal connotation of giving a fingerprint away, I would think, versus, hey, take my picture. I’m used to that.
Michael: Yeah. I mean, fingerprint definitely has that kind of stigma associated with it because of its longstanding link law enforcement. Some people, I think, are maybe a little bit more apprehensive about scanning their fingerprint comparative to other forms of biometrics. A fingerprint still is also a very accurate form of biometric. No question about it. Our first recommendation would be photo biometrics because it offers certain advantages over the other forms of biometrics.
Veanne: This is kind of a side tangent, but I just happened to read yesterday that there was a case in Tennessee where a hospital performed tongue surgery on the wrong baby. I guess now I’ve learned from you, so the iris recognition wouldn’t have been able to prevent that because the baby wouldn’t be old enough to have that on file. So I guess there are going to always be certain situations where it just can’t be prevented? There’s nobody to solve every problem yet, or what’s your perspective on that.
Michael: I think, as is normally the case with any form of technology, there are going to be outlying situations where the technology is not going to be able to necessarily add value, but I think that those situations are in the vast minority of cases. What we see from our customers is first of all very, very high enrollment rates. With the photo biometrics, for example, our customers report 99 percent plus enrollment rates from their patient population.
Veanne: That’s basically everybody.
Michael: Pretty much everybody. A lot of that also, I think, has to do with how the technology is explained to the patient prior to their enrollment because this technology, while it provides a tremendous amount of benefit for the provider, it also provides a lot of benefit to us as patients as well. We have peace of mind knowing that once you have enrolled into the system no one else can come behind you and steal your identity. That happens a lot. We’ve got another customer who is based in the South. They are in a very large metropolitan area with a very, very diverse patient demographic. They have a convenience store that’s located right around the corner from their main hospital where it’s well known that anybody can go in there and purchase a fake ID, and it’s relatively easy. Anybody, any of your listeners just Google “buy fake IDs.”
Veanne: For the young listeners, stop listening. [laughs]
Michael: Yeah, right. Now it’s time for the parents to turn off the program. No, if you Google “purchase fake ID,” I mean, you’ll see. There are places you can go online and just buy a fake ID these days.
Veanne: Unfortunately true.
Michael: Yeah. So, it’s not difficult to get your hands on one. We also know that our health information actually fetches 10 times more value on the black market than something like a credit card number.
Veanne: Ten times?
Michael: Ten times more valuable.
Veanne: I did not know that statistic.
Veanne: Is it because they’re trying to get prescriptions? Meds? What’s the play on that?
Michael: It’s because health care is expensive. If you are in a desperate situation, for example, and you are faced with a large health care bill, the prospect of paying $500, let’s say for example, to get somebody else’s information—
Veanne: Is better than the alternative.
Michael: That’s right. It’s very, very valuable information. Health care fraud is, unfortunately, on the rise. There’s a study—the last study that was done on this was at the end of 2014, and it showed that 2.3 million Americans, adult-aged Americans or family members, were prior victims of identity theft, health care identity theft.
So when you couple that with what we’ve seen happen in the health care marketplace over the past five or six years, which is mass digitization of health care data through electronic health records, increased access from the Affordable Care Act—you’ve got 17 plus million more people now who have access. Now with a push in the market towards interoperability, you have this situation where it’s potentially more prone to identity theft.
It comes at a great cost to the provider. I think the average cost for us as patients to—if you’re a victim of identity theft is about $18,000 in order to restore your identity. Providers face a lot of litigation expenses when these things happen because patients, the real patients will turn around and sue the provider in some cases, claiming that they didn’t do a good enough job. So these are all things that our solution helps to resolve because you’re preventing all this on the front end.
How do you test biometric identification technology?
Veanne: I’m going to switch gears just a little bit. My brain has been going to, you know, as health care organizations look to bring in biometric identification technology into their organization, are there concerns and things that you have to overcome in terms of testing the solution? Is that an easy process or difficult? Maybe can you talk about that a little bit?
Michael: The way our process works is we integrate with the provider’s electronic health record software. Of course, there’s a period that we go through—
Veanne: Does it integrate with most of the electronic health record software out there today?
Michael: Yes. We integrate with—there’s about—if you look at the hospital landscape in the U.S. market, there are really five players that own most of that space. It’s Cerner, Epic, McKesson, Meditech and Siemens. Now Siemens is owned by Cerner. So, most of the hospitals out there are using one of those products. There are other ones out there, of course. We’ve integrated to a lot of them, so we’ve integrated to all of the top five and a lot of the other ones out there as well.
So, there’s a process where we work through—we work with the provider organization, and we integrate our Right Patient platform into their EHR software, so of course, that needs to be tested. With our delivery model, we actually have a SaaS-based delivery model for the backend of our software where the entire biometric matching system and data storage system and application layer are hosted in our secure cloud, and that makes it a lot easier for the provider to implement.
They don’t have to deal with provisioning or maintaining servers on their side. So the implementation process is fairly straightforward. We’ve done it numerous times. In terms of testing the actual biometric technology itself, at this point—and we’ve been in this market for a long time—the actual technology doesn’t have to go through any kind of testing period.
Veanne: It’s already tested.
Veanne: That’s awesome. Well, we’ve talked a lot. We’ve covered a lot of ground, and I feel like I’ve gotten a good overview of Right Patient, but is there anything else? I would love you to give a chance to talk about the firm that you represent and the exciting things you’re doing. Is there anything about Right Patient that you’d like to share that we haven’t covered? I’ll just give you the opportunity to do that.
Michael: Sure. The new SaaS model actually that I had mentioned is a new model, delivery model, for us.
Veanne: How new is it?
Michael: We really launched it at the beginning of this year.
Veanne: Awesome. Real new.
Michael: Yes. Our previous model was a little bit different and more like an enterprise deployment model. We’re really excited about it because we think that it will allow for the easier adoption of the technology and more rapid adoption.
Michael: I think that one thing, just one final comment, one thing that we find very exciting and interesting about this industry is that with the right platform, and, of course, we believe that Right Patient is the right platform, we think about patient identification in terms of identifying a patient at a physical site, like a hospital for example.
There’s a whole new segment of the health care market know as patient engagement where we are encouraged to become more self-actualized with respect to our health care information, take on more ownership and responsibility of that information. It’s our data at the end of the day, so the prospect of then accessing that information remotely from something like a smartphone or from a tablet or a PC, how do you secure that access?
We hear all the time about data breaches in the health care industry and market, so I think that one thing that our customers will be doing in the future is leveraging our platform in those kinds of areas because we are unique in the fact that because we have a very versatile platform, we allow for our providers to utilize the technology in some unique ways.
So, we’re really excited about that. I think that we’ll have more information coming out from our customers who are leveraging the technology in the future. We’ve got also a new camera that we’re getting ready to release in a few months, so keep an eye out for that as well.
HeroX and CHIME Challenge
Veanne: Congratulations. I feel great success ahead for you based on what you’ve been sharing with us today, so I think exciting times are ahead, and I’m sure there’s a lot of success in store for you.
The last area I want to talk about here before we wrap up is I understand—I’ve done some research here—I understand that HeroX is an organization that was founded for the purpose of enabling anyone to create a challenge that addresses a problem by building a community around that challenge to lead to a breakthrough innovation. Can you tell us about the challenge that CHIME and HeroX have launched?
For our listeners, CHIME is the College of Healthcare Information Management Executives, but I think that you may know something about this topic.
Michael: Yes, absolutely. So, they put this challenge out to entrepreneurs like us, for example, to basically come up with a way to solve the patient identification challenge in the marketplace. CHIME is an organization that knows very well what this challenge is, and they also recognize that it can have a significant impact on not only cost in the health care industry but also on the safe and secure exchange of information, which is directly linked to our patient safety.
As I’d mentioned that earlier, one of the big pushes now in the marketplace is towards interoperability, so different health care organizations sharing our health information. That means that the level of data integrity is very important. You don’t want to be sharing dirty data among different health care providers. It’s not really doing anybody much good.
Data integrity, levels of data integrity can be linked all the way back to proper identification. The patient identification part is the very, very first step in the health care continuum. If you don’t get that part right, then you’ve got a lot of potential problems downstream, and that’s what CHIME recognizes and the ONC recognizes.
There has been a lot of chatter for the past year plus about this serious issue that we have with respect to patient identification, so we’re really excited that CHIME has put this challenge out there. We are definitely participating. We think we have a fantastic proposal. I mean our technology—
Veanne: Yes. I would think you’re well ahead of most people, right?
Michael: Yes. And we’ve got some secret sauce too that we’re going to be bringing to the table in our proposal that I think is going to set us apart even more so, so we’re very excited and looking forward to it.
Veanne: What does the winner, I guess, of this get? Is there is an award?
Michael: Yes. There’s awarding $1 million to the winner.
Michael: For the best proposal. I think the award is going to be given next year.
Michael: 2017, March I think is when they’re announcing the final winner. So, right now we’re just working on fine tuning our proposals, all the details, and we’re going to get that submitted. I think we’ve got a great shot. I mean, we already have the perfect platform, in my opinion, to address this problem. Like I said, we’re really looking forward to it.
Veanne: Well, good luck, and I am sure that if you guys are the fortunate winners, you will put that money to good use, right?
Michael: Absolutely. You can count on that, definitely.
Veanne: Well, great. If our listeners would like to reach out to you personally or learn more about Right Patient or what you do, what’s the best way for them to get in touch with you?
Michael: Website is www.rightpatient.com. You can find us on all the social channels as well, so Twitter @RightPatient. We’re on Facebook as well, LinkedIn. Please interact with us. We’d love to answer any questions that you have, or certainly, if you have any interest in the solutions, if any listeners are out there, please contact us. We’d love to talk to you.
Veanne: Yes. I know we all love to get engagement and talking about can technology replace the human error part. I think there’s a lot of dialogue out there, so I hope you get some of that from this podcast here. It’s been a great pleasure having you here today. I appreciate your time. Thanks so much.
Michael: Thank you so much for having me. I appreciate it.
Angela: You have been listening to Atlanta Business Impact Radio with Veanne Smith, and I’m your co-host Angela Greenwell. This program is brought to you by Soltech. For more information about the podcast, including other episodes, you can visit our website at Soltech.net or find us on iTunes. Thank you for listening, and we look forward to having you join us again.