Ob Hospitalist Group (OBHG) is the nation’s largest supplier of OB/GYNs with practically 2,000 physicians. Mark Simon, M.D., OBHG’s chief medical supply, not too long ago spoke with Healthcare Innovation in regards to the potential for synthetic intelligence to assist clinicians with administrative duties similar to coding.
Greenville, S.C.-based OBHG carried out a latest pilot with the Commure Autonomous Coding answer. Inside three months, OBHG clinicians decreased time spent getting into expenses by 83%, and AI now codes over 85% of all expenses. The subsequent step is rolling the answer out enterprise-wide. Becoming a member of the dialog was Commure Chief Working Officer Deepika Bodapati.
Healthcare Innovation: Dr. Simon, may you describe how your group companions with hospitals across the EHR and knowledge seize and sharing, after which perhaps how you have historically dealt with coding, billing and high quality reporting, in order that we will distinction how the brand new mode is impacting that?
Simon: We contract with hospitals to supply OB hospitalist protection. And the overwhelming majority of our work is completed within the hospital facility itself. And, after all, from a affected person security and affected person care perspective, our purpose is to doc the work that we do inside the hospital’s EHR as a result of it is best for the affected person, the knowledge is on the market for everybody to see, et cetera. After which we do our personal skilled payment billing for the providers that we offer. For numerous years, we’ve used PatientKeeper as a cost entry system. (Commure bought PatientKeeper from HCA Healthcare in 2021.) Our physicians and midwives would see the affected person, doc the observe within the hospital’s EHR, after which they might go into PatientKeeper and enter the cost for every billable go to.
HCI: And have been there some frequent coding workflow frustrations expressed by the clinicians in regards to the guide knowledge entry?
Simon: Completely. Clinicians, physicians particularly, do not like coding. It has its personal algorithm that should be adopted, and so they might not be as intuitive to a doctor as one may hope they might be. It is an additional piece of labor that they need to do. They’ve to enter a unique system after which take into consideration the go to another way and enter these expenses. Coding for physicians generally is a very burdensome exercise.
HCI: How did your group come to do a pilot of the autonomous coding answer from Commure?
Simon: Commure got here to us, and defined that they’d this providing the place they may begin to probably autonomously code our visits. That’s one thing that could be very interesting to us, by way of ease of doctor workflow and midwife workflow. If we may do one thing as an organization to make the physicians’ lives a bit bit simpler, we wished to analyze that and look into how this might work for us.
Bodapati: At Commure, we see ourselves as an AI companion to OBHG, asking how can we relieve stress and create extra efficiencies in present work streams? This was such a transparent place for optimization and uplift in relation to supplier time being spent, in relation to income leakage, and only a clear place the place automation might be launched to alleviate suppliers’ stress. It made numerous sense to deploy that expertise right here.
We have now an enormous doctor scarcity, so that is very high of thoughts for everybody, and certain will probably be for many years to return. We’re going to need to attempt to assist suppliers discover the aim again of their work, and that’s not by doing documentation and coding and going backwards and forwards with billers on why some cost obtained accepted and a few expenses did not get accepted. That isn’t why they went into the occupation, however the truth that these folks, who’re actually hard-working individuals are burnt out, signifies that there’s one thing very mistaken. They only do not need to work on this administrative nuisance. So we see our answer as a strategy to increase suppliers in that capability, to assist them code, to assist them doc and create a really seamless workflow to optimize income on the finish of the day.
HCI: In pilots, did you establish that it captures extra codes than the clinicians may themselves, after which they’ll simply approve the codes that have been captured?
Bodapati: We discovered that the period of time that clinicians spent coding decreased by 83% and we’re doing about 85% of all of the coding there.
Simon: From our perspective, we completely hear from clinicians that it makes their lives quite a bit simpler, and that there are specific encounters the place they may have simply forgotten the code. Clinicians need to do what’s proper, however they’re very busy, and so they can get simply distracted by the scientific work that they are offering. Generally they will write the observe or doc the go to, considering that they need to return and put that code in, however they neglect about it. However with the AI software, it makes it a lot simpler. Now the observe goes by, it will get coded, and so they have a chance to have a look at it and ensure it suits with their understanding of what it needs to be. After which it goes off to the income cycle workforce.
HCI: Does having this new AI companion require any coaching for the clinicians, as a result of it’s altering their workflow, regardless that they’re doing much less administrative work now?
Simon: What they should be refreshed on is that documentation is essential — how they doc the observe about what they did and why they did it, and for whom, and all that thought course of that goes into it, which is one thing they need to be doing anyway, proper? So this is not prefer it needs to be new, but it surely’s only a reminder to make it possible for their documentation goes to make the AI software work one of the best for them and the place it causes them the least complications. So, for instance, ensuring that their observe sort is right. In the event that they’re doing a supply, that it is actually a supply sort. That helps the AI software precisely code this, which implies they’ve much less work on the again finish. On the similar time, ensuring that their documentation covers all the suitable factors of what they did throughout that go to, in order that the code might be as correct as potential, particularly whenever you begin speaking about prognosis, ensuring that their evaluation consists of the suitable diagnoses which are related for that affected person. Due to this fact, it will likely be picked up by the AI. So it is a refresh flash coaching on documentation. A part of my position is danger mitigation and malpractice and all that, too. So I am a giant fan of higher documentation regardless.
Bodapati: That is actually meant to empower suppliers. There are new codes. Codes change comparatively usually. It isn’t one thing that is static, proper? So then everybody must get up to date on these adjustments. And what we’re discovering is the case is that always medical doctors really drop secondary diagnoses, and really do not precisely painting the complexity of the procedures they’re doing. For those who do not say that your affected person has COPD as a secondary prognosis and hypertension as a secondary prognosis, and there is a slight complication, that is really a better acuity process that you just’re not getting paid for, however you’re doing all the work of increased acuity process. Physicians needs to be compensated in a method that is completely commensurate with the quantity of labor and the complexity of the work that they are doing, and that’s not taking place right this moment.
Simon: I believe one of many different the explanation why AI as a software for coding particularly is helpful is that not solely do the codes change, however typically the foundations round the way you decide the suitable code on the CPT aspect, these get modified and up to date, and it is not high of thoughts of a doctor to remain abreast of precisely all of the nuances of particular codes guidelines. I believe it is an ideal use case for AI in healthcare, as a result of the AI can sustain with that, and the doctor does not really want to do this. It isn’t use of their time.
HCI: So Dr. Simon, are you now scaling this up throughout all of your footprint?
Simon: Sure, we’re trying to implement this answer at as many hospital companions as we will, as shortly as we will. When our IT chief introduced at our nationwide scientific management assembly within the fall that it’s OBHG’s intent to implement autonomously aided coding at as many websites as potential, as shortly as potential, he was the hero of all the convention, just about, as a result of they have been so excited that that is one thing that is going to make their workforce’s lives simpler.
