It is a transcript of my HIMSS interview with Bevey Miner, EVP Healthcare Technique & Coverage at Consensus Cloud Options. Normally I’d present the video however on this case my fancy new microphone didn’t work so that you’d solely hear a one sided dialog. Fortunately Youtube’s transcript considerably got here to the rescue–Matthew Holt
Matthew: One other THCB Highlight, I’m right here with Bevey Minor who a yr in the past I interviewed as Consensus Cloud Options and now your signal says eFax. So, what the hell occurred?
Bevey: Attention-grabbing query, Matthew. The corporate is Consensus Cloud Options. And the corporate’s at all times been Consensus Cloud Options since we spun off and went public ourselves. You’ll discover at our sales space we’ve received the eFax model — it’s eFax by Consensus Cloud Options. The rationale we’re exhibiting up as eFax is as a result of this yr at HIMSS we actually wished to set the report straight: digital cloud faxing will not be the issue with interoperability. Paper faxes are, however digital cloud faxing will not be the issue.
The issue is all this unstructured information — all of the unstructured information that occurs with faxes, with scanned pictures, with TIFF pictures. All that unstructured information can’t be queried. It could’t be a part of TEFCA. You’ll be able to’t question what you possibly can’t discover.
Cloud faxing is ship and obtain all day lengthy, and we do this very effectively and have been doing it for 27 years. About three years in the past, we launched an clever extraction answer. That answer doesn’t even have to begin with the fax, nevertheless it permits the “discover” piece to really turn out to be the important factor that we have to do. CMS defines interoperability as ship, obtain, discover, and combine. Fax expertise handles ship and obtain all day lengthy, however can’t discover. So as soon as we launched a “discover and clever extraction” answer, we will hearth up TEFCA.
I’ve talked to a whole lot of regulators, together with Dr. Thomas Keane and Amy Gleason with the CMS Align networks. You’ll be able to’t ignore this pile of unstructured information and simply assume the business goes to go magically every part’s on FHIR. We’re all utilizing FHIR as a result of all of these things has actually necessary affected person data in it.
What we need to clear up within the business is: don’t say we’ve to axe the digital cloud fax. Let’s axe the paper fax machine. Digital cloud faxing isn’t going away — in actual fact, it’s rising, particularly as we get rural well being off of paper fax machines. The following degree of maturity is digital cloud faxing. From there, as soon as it’s digital, now you are able to do all types of issues with it.
After we launched digital well being data throughout significant use — I used to be at Allscripts on the time — our dream was that we might take this paper report and remodel it into an digital well being report, so we might simply do away with the paper. As soon as we did that and there have been discrete information components in that EHR, we might do inhabitants well being, medical choice assist, efficacy, all types of issues — as a result of there are discrete information components now inside that digital well being report. That’s what a digital fax will do with the potential to do intelligence on prime of it.
So we need to make the business perceive that the fax will not be the issue. Extracting it and eliminating all that unstructured information is the answer.
Matthew: Okay. So since we final talked, I’ve had some journeys across the well being care system. I used to be despatched a referral from my pretty major care doc at One Medical. One way or the other the referral from One Medical received into the Blue Protect system and created a previous authorization. So I’ve an echocardiogram — I’ve written this up on my weblog — and what’s coming subsequent.
I name the quantity on the referral that One Medical needs me to name for the Marin Well being imaging middle, and so they go, “We don’t have any referral.” I say, “No, no, no, they despatched it to you, I see the PDF.” I name again two weeks later and so they go, “I don’t know, we don’t have the referral.” So lastly I name and say, “Can I electronic mail you the referral?” I get somebody’s electronic mail and electronic mail them the PDF and that individual goes, “Oh, this can be a cardiology imaging referral, not an everyday imaging referral.” It seems it goes to cardiology, which is after all in the identical constructing.
However right here’s the place it will get fascinating. They are saying, “We will’t take an electronic mail” — they advised it was as a result of HIPAA wouldn’t permit it.. So I downloaded some fax product — I feel it might have been eFax — a trial that later I needed to bear in mind to cancel. So I fax every part. Now they’ve received a referral which was typed on a PDF, and will or might not have a bit of paper someplace with the prior auth as effectively. It’s arrived, and now they’ll really do one thing with it.
What’s incorrect with that image, and what can eFax do in that image to repair it and get the information out that’s required to assist?
Bevey: Quite a lot of issues are incorrect. Primary, you’re going to get delayed therapy. You’re going to get annoyed as a affected person, and your well being outcomes — it’s possible you’ll even abandon care.
Matthew: That’s really what virtually occurred. I used to be about to desert it. I did hold going — I really do have an issue.
Bevey: However your well being outcomes might have been horrific. However you deliver up a very fascinating use case that we’re doing at the moment with imaging facilities. Hospitals that personal their imaging facilities, or IDNs that personal imaging facilities —
Matthew: this imaging middle, by the way in which, has each cardiology imaging and common imaging in the identical constructing, however these are completely different fax numbers.
Bevey: So for example — and that is an precise actual use case we’re doing — we intercept the fax, we pull out all of the structured information: all of the affected person demographic data, we pull out what the order is for. Let’s say it’s a mammogram with distinction. We map it into the radiology system to allow them to see the order with out anybody data-entering something. If it had come as a fax order, someone nonetheless has to data-enter it, or it’s sitting in a pile of all these orders they’ve acquired.
As soon as it’s extracted and mapped into the imaging middle’s RIS system, they’ll see — as a result of we learn that this isn’t only a mammogram, this can be a mammogram with distinction. Now they’ll see: do we’ve a machine able to a mammogram with distinction? Is that machine out there on Monday, Tuesday, Wednesday? And as soon as all that’s achieved, no human has really touched this. It’s getting checked out and mapped into their scheduling system, so you possibly can ship a textual content message on to the affected person inside 24 hours: “We’ve received your order and we’ve alternatives so that you can are available Monday, Tuesday, or Wednesday.”
All of that’s as a result of we pulled out the structured information from that fax for that picture. And for daily {that a} affected person anticipated to get the subsequent degree of care however is delayed — a mammogram or some form of imaging — 25% of belief within the system is eroded. That imaging middle must get that referral as a result of it’s their lifeblood. The affected person goes to go someplace else, or they might simply say to the physician, “They haven’t referred to as me — discover one other place.”
Matthew: So if they’d plugged in eFax someplace in the midst of that workflow, the referral would have been intercepted, learn, and mechanically routed.
Bevey: The product that does that is named Readability — that’s our AI information extraction piece. And when you’ve created all these discrete information components, if the imaging middle wanted it in FHIR so it could possibly be mapped, we will ship that as a FHIR message. Now we have a semantic interface spine. When you’ve received all this extracted information, now you are able to do prompt-based queries on it — for instance, “Present me all sufferers who want imaging with distinction as a result of these machines are sitting unused and we have to get these sufferers in quick.” After which you need to use immediate questions like on any generative AI answer
Matthew: I’ll inform you one thing amusing that results in one other query. Ultimately after the fax goes via, someone calls me again — and naturally I don’t reply, it goes to voicemail. I see the transcription of the voicemail and so they’ve organized for me the very subsequent out there appointment, which occurs to be in Santa Rosa — an hour from my home. So I’m going to MyChart pondering possibly I can change it. MyChart reveals me the appointment however doesn’t give me the choice to alter it — solely to cancel it. Ultimately I received it modified to a location 10 minutes from my home.
However this results in a query. You might have faxes going forwards and backwards between these departments, and this one continues to be very manually handled — guide appointment scheduling.However there’s a complete lot of knowledge within the EMR (Epic, Oracle, et al) There’s a whole lot of data sitting in there, with folks attempting to get it out and in by completely different strategies, not all digital. How does what you’re doing play into that?
Bevey: What’s fascinating is that there’s a very massive effort round affected person entry to their information. MyChart is an instance. This administration — Trump’s promise and marketing campaign was that each American would be capable to entry their medical data. And now we’ve a bunch of pledges taking place round affected person entry to their information, with QR codes and issues like that. However what do you do with all of the unstructured information? A affected person might have lab outcomes, medical data — the USCDI information set that offers the affected person their whole historical past and abstract — sitting in unstructured information that may’t be queried.
So the QHINs are solely working on a small subset of knowledge that’s FHIR-enabled structured information. It’s going to be biased, it’s going to have lacking items. Sufferers are going to assume, “Right here’s all the data I would like” — till they drill down and say, “Two years in the past my hemoglobin A1C was pre-diabetic, I need to discover that.” And it’s an unstructured PDF someplace.
Once you take a look at an digital well being report, there’s a whole lot of PDFs hooked up to the affected person’s chart.
Matthew: I used to be simply taking a look at mine. Each One Medical and Epic items have completely different buildings. A few of them they processed internally and present me my lab numbers with a pleasant little indicator, however I’m undecided they’re multi function place.
Bevey: Now we have a whole lot of work to do to get a typical report for sufferers. I’ve two MyChart accounts. It’s straightforward to hyperlink them, however I additionally see one other system that makes use of FollowMyHealth or one thing like that, and it’s not integrating because it ought to.
I do know the promise is nice — I bear in mind the times, and also you most likely do too, Matthew, the place each affected person was going to hold round a USB drive or a magical sensible card.
Matthew: You are able to do that in France. Most likely not out there in America.
Bevey: However I do assume there’s promise, and the promise is to actually acknowledge that there’s an entire bunch of knowledge. I go searching at a whole lot of these firms right here at HIMSS and so they boast they do workflow enhancements, prior authorizations, inhabitants well being — however once I ask them what they’re doing with all of the unstructured information, they are saying, “Now we have an OCR answer.” At any time when I hear that, it’s probably not extracting information at an clever degree. OCR is aware of that one thing is a B, a V, and a Y — my first identify — however doesn’t even realize it’s a reputation. In case you use clever extraction, it does contextual work: “Go discover me the identify of this affected person on this kind,” which could possibly be within the higher left-hand nook or the underside. It actually understands the best way to ask the correct of questions. OCR is a niche.
So we’re enjoying a task the place we will create very subtle intelligence off of all this unstructured information, so many of those workflow options can have higher information getting into.
I’ve talked to Dr. Thomas Keane, head of ONC, and I stated: “Do you actually need to make TEFCA profitable? You’ve received to speak about how we handle getting all of the unstructured information right into a FHIR-enabled TEFCA framework.” And rural well being — well being fairness has been a ardour of mine — Rural well being can’t rip and substitute. They’ll’t assist all of the FHIR-enabled options they’re speculated to assist. The CDC of their information modernization initiative is saying all digital case reporting must be achieved by way of FHIR — effectively, the neighborhood clinic undoubtedly must populate these repositories, in any other case these databases are going to be utterly biased.
And should you take a look at the prior authorization rule 57 going stay January 1st for Medicare Benefit sufferers — what’s that substance abuse clinic that should ask for an authorization so a affected person can keep in therapy for an additional 90 days? They’ll’t ship a FHIR-enabled prior authorization to get that permitted. We intercept their fax. We are saying to them — and we do loads in rural well being — “Simply hold doing what you’re doing. We’ll intercept it within the center and ship that as a FHIR prior authorization.” The business wants to know that not each supplier has tech fairness.
Matthew: Alright, give me your forecast. There are all these completely different holes — unstructured information, ambient AI, folks asking CharGPT questions. Do you assume the buyer, affected person, smaller supplier (or any supplier) expertise will get loads higher in two years, or are we nonetheless preventing this? Is Amy Gleason’s staff going to succeed, or are we nonetheless going to be in a struggle in 5 years’ time?
Bevey: I nonetheless assume healthcare is transactional. If everyone was in a closed system, everyone talked Epic, it could be nice. However we don’t have that sort of system. And that is an administration of deregulation — open for no matter you assume you are able to do from an innovation perspective. When you could have tight regulation, you see options emerge that truly assist suppliers and expertise firms fulfill that regulation — which is going on with prior authorization, which is why we’re actually useful for people who don’t have FHIR assets.
However with deregulation and pledges, I feel in two years should you gave me the identical interview, we might see some child steps. And in addition as a result of there’s a rising consciousness amongst sufferers — “I can really do loads with AI. I can push arduous on my docs. I can say that is the most effective therapy.” And never at all times use Western medication to try this. Sufferers are going to carry a ton of energy, and with that energy they’re going to say, “I’ve a duty to maintain my well being.” I feel that’s going to maneuver sooner than we even thought, particularly with ChatGPT Well being, Gemini — sufferers are utilizing these instruments. They’re studying their reviews and asking questions.
Matthew: I’ve uploaded a whole lot of my reviews from my imaging and I now use it on a regular basis. However , there are individuals who have been dragging their ft, and a few years into this, it’s beginning to really feel like fantasy being repeated three or 4 years in a row. So I’m hopeful, however I’ll acknowledge the joke of the second. I do assume there’s a whole lot of consensus that we’ve received to get some stuff achieved.
You and I’ve been doing this lengthy sufficient to recollect when you may go anyplace together with your information, and now everybody else is preventing over data blocking.
Bevey: The sufferers don’t care what the pipes seem like on the again finish. The docs don’t even care. The docs are saying, “I don’t care what you do as a expertise vendor. I don’t must see all of the stuff flowing via. I simply want data so I could make a extra knowledgeable choice, deal with my sufferers one of the best ways, and earn money doing it.”
We’re going to have some headwinds the place a whole lot of hospital programs are going to must take care of uncompensated care. HIMSS is somewhat gentle this yr, and I feel hospitals are taking a look at budgets and saying, “We will’t provide the similar IT price range to buy groceries. We’re going to must shut down service traces — behavioral well being, ladies’s well being — simply to handle the uncompensated care.” That’s a brand new world we by no means thought we’d see. However I feel we’ve received to climate via that too.
Matthew: I’ve been talking with Bevey Miner from Consensus Cloud Options, Thanks on your time
