AH069 - Poster Presentations: The Utility of ICD Codes, and How Text Messages & Pharmacist Outreach Aids Medication Adherence

Transcript

Lightly edited for clarity.

[00:27] Justin Venneri: Hello, and thank you for joining us for another episode of the Astonishing Healthcare Podcast. This is Justin Venneri, your host and Senior Director of Communications at Capital Rx, and today we're highlighting a couple of recent poster presentations. Zach Brunko and Nash Albadaran, both pharmacists here at Capital Rx, have presented recently at conferences, and I'll link those up in the show notes as they are on our Insights page. But I always feel like talking things through is valuable and can sometimes be more helpful in raising awareness of the work we're doing. Zach's presentation was at AMCP Annual in Houston, and Nash was at PQA, the Pharmacy Quality Alliance.

• Poster: Evaluating ICD Codes on Pharmacy Claim Submissions: Are They Reliable?
• Poster: The Impact of Text Messaging and Pharmacist Outreach Interventions on Medication Adherence Rates in a Commercial PBM Population

Zach and Nash, thanks for joining me today.

[01:06] Zachary Brunko, PharmD: Yeah, of course, Justin, thank you for having us.

[01:08] Nash Albadarin, PharmD: Yeah, thank you.

[01:09] Justin Venneri: So, starting with you, Zach, tell us what you were looking to study here on ICD codes. Evaluating ICD Codes on Pharmacy Claim Submissions: Are they reliable? Was the title of your poster, right?

[01:20] Zachary Brunko, PharmD: Yes, that is correct. So, starting off, for anyone unfamiliar with ICD, it refers to the International Classification of Diseases Codes. So to kind of provide some context on the why, the National Council for Prescription Drug Programs has a standard free text field that allows for these ICD codes to be submitted on pharmacy claims. So there's roughly over 70,000, give or take, different ICD codes. And these can range anywhere from something like a fracture on a leg to some kind of infection, headaches, to even encounters at the doctor's office.  

These ICD codes help track accurate cause of death, they can help track disease and injury, and these things can be used to make informed clinical decisions and overall accurately bill services that are provided to members. So these codes, however, are not required to be submitted during the time of medication claim submission. When we were looking at different previous studies, we kind of found somewhat of a mixed bag on how accurate ICD codes were based off of those studies. If they were to be submitted incorrectly, it could provide challenges in population health management. So, kind of circling back to your question, Justin, we wanted to see if the ICD codes that do have something populated in those free text fields on pharmacy claim submissions were valid. Additionally, we wanted to identify any patterns across different drug classes, disease states, different pharmacy networks, and across different states. Then we wanted to see if There were any potential applications of these findings to improve any processes in house.

[02:56] Justin Venneri: Okay, and what was the methodology you employed here to perform this research?

[03:01] Zachary Brunko, PharmD: Yeah, of course. We pulled claims from 19 different eligible capital RX clients over the course of a quarter. So, for this one we pulled from quarter four of 2024. Overall, this yielded about 525,000 claims.

[03:16] Justin Venneri: Okay.

[03:17] Zachary Brunko, PharmD: From that we wanted to boil that down to claims that actually had something submitted on the free text field for, for a diagnosis code. So, something that I mentioned a little bit previously in the background section. This brought about 244,000 claims. So in order to validate whether these claims are accurate, we used both an external source being from CMS or the Centers for Medicare and Medicaid Services, their ICD code that they have on the web, and then our own internal ICD code table.

Of those nearly 244,000 claims, about 230,000 of them had a valid text field that could be linked back to a diagnosis code. So about 14,000 had something in the text code, but not something that could be linked back to either CMSS or our own internal ICD code table.

[04:04] Justin Venneri: Is that normal?

[04:05] Zachary Brunko, PharmD: Yeah, it is normal to have like some sort of like Z codes, for example, that may just be something populated in there. But when we're doing that research, there was also things such as random jumble of letters or random numbers that could also be put into there. Now given that these are free text fields, like I was mentioning, you can enter anything into that. So if, say for example, you had a claim for metformin, in theory, someone could enter in an ICD code for hypertension.

So, we wanted to further determine if different medications and drug classes were indeed accurate and had the correct diagnosis. Now if I was to look through about 230,000 claims, that would probably age me about 10 years. We decided that we wanted to look at a much smaller subset of those claims. So, we ultimately went on looking at a total of 111 different medications that had ties back to prior authorization, whether it was a preventative medication, whether it was on like our Affordable Care act drug list. So in total we looked at about 27,000 claims to kind of like make that a little bit more digestible. We wanted to see of that validity. We determined what would be an appropriate diagnosis for that medication. So we use different clinical databases such as Micromedex and Lexicomp and clinical studies, and we wanted to look at both FDA-approved and “off label” indications.  

So, these ICD codes that were deemed inappropriate or inconsistent we wanted to further analyze to determine if they could be linked to another disease or medication in the member's prescription history.

[05:40] Justin Venneri: Okay. And then any other data massaging or things you had to do to be able to work with this subset of the larger group of claims you identified?

[05:48] Zachary Brunko, PharmD: Yeah, of course. So we use Capital Rx's adjudication platform, Judi®, to dig for more information on members' fill history to accurately categorize them - the member claim for Metformin and ICD code of hypertension, if for whatever reason they had, say, lisinopril on there, we still deem these as incorrect, but more so misaligned, since they had an adjacent medication with that claim.  

And then in other cases, for example, if they filled a Trulicity® and had a code for type 1 diabetes and no other legitimate claim to be tied back to that ICD code, we deem these to also be incorrect, but more so inappropriate.

[06:25] Justin Venneri: So, at the end of the day, what did you find? What were the key takeaways?

[06:29] Zachary Brunko, PharmD: So, I'd say the biggest takeaway is that of the submitted claims that we reviewed, a large majority, nearly 92% of the 27,000 claims that we will look back, were accurately tied back to an appropriate ICD code. About 6% were in an undefined category like we were talking about previously, like that 14,000 that just had something populated in the free tax field. So about 6% of that 27,000 were undefined. And then finally of those incorrect, about 2%, it was kind of split on what was inappropriate and misaligned.  

This was fairly consistent across the different trends - that around 90% correct claim submissions, when we were looking at some of those different categories. So, for example, like your CGRPs, your anti diabetics, opioids, et cetera, across different medications, whether it was, for example, like fluoxetine, tretinoin, or Wegovy®, just to name a few, and even across different volumes of claims submitted by clients, different pharmacy networks, whether it was like independent or retail. And then I would also say, like we were mentioning earlier in the show, another key takeaway is that almost 43% of the original 525,000 claims did not have anything populated in that free text field.  

If I were to use this as a jumping off point, that if an ICD code is submitted on a pharmacy claim, that we would almost nearly 9 out of 10 times justifiably rely on them, we could potentially use these ICD codes in some different ways to help improve our business. In my opinion, what feels the most obvious would be A smart or silent prior authorization (PA) system that could identify these ICD codes required to meet different criteria and automatically approve or deny based on what is being submitted. It would, in theory, hopefully shift some of the workload off of a PA pharmacist queue. Some potential other ideas could be using or connecting patients or members with services if they don't have any already based on the ICD code submitted. We could also utilize these ICD codes as extra data points for predictive models that measure adherence. And it could be for whatever reason that a certain ICD code or a broader definition of that ICD code may influence non-adherence or may it have some trends toward that. So, there's a couple different areas that could be worth exploring in the future.

[08:46] Justin Venneri: That's awesome. And I would think as we work toward this unified infrastructure where medical and pharmacy are running together on the same chassis, that that could be pretty helpful in terms of those workflows you have identified, that you could rely on better data to take further action, right?

[09:02] Zachary Brunko, PharmD: Definitely.

[09:03] Justin Venneri: All right, awesome. Zach, thank you so much. Nash, thanks for sticking with us here. And also, thanks for coming back on the show. You were our guest back on episode 43, and that was an explanation of the DMP and MTM programs. So, appreciate you coming back on.

Related Clinical Programs Content

• AH043 - Pharmacy Benefits 101: DMP & MTM, Explained, with Nash Albadarin, PharmD
• AH035 - Pharmacy Benefits 101: Clinical Programs, with Bonnie Hui-Callahan, PharmD
• AH006 - Pharmacy Benefits 101: Clinical Care Teams, with Amy Stockton, PharmD

‍

[09:18] Nash Albadarin, PharmD: Yeah, thanks, Justin. Appreciate you having me back on again.

[09:22] Justin Venneri: You're gonna become a household name here too. Just some of our other guests. So you looked at the impact of text messaging and pharmacist outreach interventions on medication adherence rates in a commercial PBM population. What was the thesis or the background on the clinical program, really?

[09:40] Nash Albadarin, PharmD: What we were looking at was wanting to evaluate the effectiveness of one of our clinical programs named RX Enhance. It's an adherence program and what we wanted to really look into was its effect on improving medication adherence among members with specifically diabetes, hypertension and high cholesterol. So that was kind of the focus of the study here. And I can give a little bit of background for those not familiar with RX Enhanced. So, this program really aims to improve member adherence and overall chronic condition management through behavioral nudges via motivational text messages. And we pair that with in person clinical pharmacist interventions by the member's local pharmacist. So, it's completed by a community-based pharmacist.

[10:17] Justin Venneri: Does our team interact with the pharmacist?

[10:19] Nash Albadarin, PharmD: Yeah, so we do coordinate, kind of help drive the interventions that get delivered to the pharmacists and then leverage the local pharmacists just because they have a relationship and a rapport with the members. So it really provides them kind of that personalized clinical support, and that's the background on the clinical program.

[10:33] Justin Venneri: Totally understand the adherence aspect. What was the methodology you employed for this study?

[10:39] Nash Albadarin, PharmD: Absolutely, yeah. And so just kind of talk through how we initially even target these members. So what we do is we identify these members that are at risk for non-adherence and then deploy these interventions like I mentioned to the community based pharmacies and the intervention include things like refill reminders, adherence check ins, trying to get them converted to a 90 day supply and then consulting on any new to therapies that they may be starting in terms of the actual study and how we went about this. So we ended up conducting a retrospective review of deidentified pharmacy claims and we used a rolling 12 month look back and the adherence was tracked over that 12-month period pre and post intervention.  

And so what we ended up doing was using the medication adherence PDC or proportion of days covered greater than 80% across the four drug classes. So statins for cholesterol, diabetes medications, and then hypertension medications - specifically calcium channel blockers or CCB's and then renin angiotensin system antagonists or RASAs. And ultimately what the study really compared here was the adherence rates over a 12 month period before this program was turned on and then post program implementation for six of our clients that are that were enrolled in the program and so ended up comparing kind of pre and post adherence rates before and after we turn the program on for these clients.

[11:55] Justin Venneri: What were the key takeaways or findings?

[11:57] Nash Albadarin, PharmD: So the program actually had really great results. Overall program led to an average of 9.5% improvement in the adherence across all the drug classes, which is really quite incredible because it's almost more than triple the industry standard of 3% increase.  

So, we saw a significant increase in the adherence rates in this population specifically with the hypertension medications, the calcium channel blockers, and the RASAs. They saw the highest improvement. We saw a 14% increase in adherence in this population. I think we saw a big improvement there because of the conversion in the 90-day supply intervention specifically. So we saw a lot of 90-day supply fills that were a result of the interventions that were deployed.

Some other notable takeaways that we kind of identified was in terms of the diabetes medication classes and the improvement in adherence. We saw a smaller improvement in adherence of 1%. But I will say it was interesting for this diabetes medication because our pre adherence rate was actually really high. It was sitting at an 80% adherence rate. And so it was already, you know, at the threshold of adherence. So we did actually get an increase of 1% in adherence in this timeframe. But notably, I think one thing that we kind of were thinking through of the impact there was this time period that we studied was during a lot of the GLP-1 drug shortages.

[13:12] Justin Venneri: Oh, interesting.

[13:13] Nash Albadarin, PharmD: Yeah. And so that was one of our thoughts is that I think that may have contributed to kind of that smaller increase in adherence rate for that class specifically.

[13:21] Justin Venneri: And so in within that class, it's the GLP1s and the other just standard diabetes medications.

[13:27] Nash Albadarin, PharmD: That's correct. Yeah. So that’s what we studied in the diabetes medications, kind of all. All the diabetes meds, but specifically, I think kind of notably was the GLP-1 drug shortages in this time period that we were studying that I think impacted improvement in adherence rate that we could achieve. And again, I think that's probably one area where we could do a little bit further research just to validate some of that.

[13:46] Justin Venneri: Okay, are there any specific nuances? Is it just having better data like an accurate phone number for the member to send a text message, or being able to communicate with the pharmacy quickly because of Judi? Like, are there any little nuances of the program that helps?

[14:01] Nash Albadarin, PharmD: No, no. It's a great question, actually. So I will say ultimately, I think what we found is when we look at the overall improvement adherence rate that we got of 9.5% just across the board, it is really a significant improvement in adherence that we were able to see with RX enhanced program. So in terms of things that we noticed and took away from it, I think that was a big takeaway is the effectiveness of this program in improving adherence across the drug classes that we measured. I think to me, it really signals the importance of both the integrated digital aspect of these interventions, so, you know, incorporating those motivational text messages, but also the key importance of the human intervention still, and kind of pairing those together can really significantly outperform your industry standard in terms of adherence improvements. So, you know, as much as we can kind of automate, I think that human touch is still really critical and something that I think has helped our program drive and show that improved adherence.

[14:50] Justin Venneri: Got it. I guess. Last question. Here we are. Anything surprising or astonishing that you think otherwise is worth following up on or maybe is there a particular takeaway for plan sponsors, something they should consider about a program like this for their populations?

[15:04] Nash Albadarin, PharmD: Yeah, no, absolutely. So I think one like we talked about just with the impact of those GLP-1 drug shortages kind of how that would have impacted our data a little bit further in the diabetes classes. But also really wanting to dig in a little bit further and kind of dissecting the different aspects of the program, the text messages versus the in person pharmacist interventions and specifically which of those, whether it be the adherence check ins, the 90 day conversion, the refill reminders, which one of those specifically drive the most impact, can really help optimize the program.  

I think down the road now looking at the benefit to plain sponsors, I think really what we take away from this and what plan sponsors can take away from this is that these types of programs and you know, Rx Enhance is, is one of these types of programs, but an adherence type program really can improve and reduce your overall healthcare costs and improve the members chronic disease outcomes and conditions. Specifically, what we do with Judi I think does a good job of allowing us the capabilities of targeting these high-risk populations and then combining that with the automated and personalized outreach as part of the Rx Enhance program.  

So I think the capabilities of Judi to really automate and help us identify these high-risk members and quickly get these interventions deployed to their local pharmacy helps us achieve the adherence results that we're seeing. And again, I think like we talked about, we noticed a big improvement for the hypertension class specifically because of the 90-day conversions and getting people to fill the 90-day supplies. So, I think that's a key takeaway the plain sponsors can look at.  

So I would definitely encourage, you know, if adherence is a concern, kind of discuss different options. But I think solutions like Rx Enhance that do the automation through technology but also have that personalized pharmacist touch really drive adherence further and can improve patient outcomes.

[16:42] Justin Venneri: Makes sense. And we've heard recently, as recently as the most recent episode - AH068 - The Future of Health Benefits: Integrating the Power of Pharmacogenomics (PGx), with Burns Blaxall, PhD, and Caitlin Munro, PharmD - how the pharmacist intervention or pharmacist participation on the care teams and their involvement in helping with managing patients with chronic conditions can really be beneficial to the system as a whole. So it's kind of like a nod to the pharmacist involvement in taking care of patients.  

I guess one thing before we go that I just want to push back on a little bit - and to either of you - it's interesting to think about the total cost and it's like medication adherence. So people are taking their meds and you would expect them, if they're adherent, to have better health overall and lower costs over time. Is this a situation where it's the qualitative aspect of it, you know, it's quantitative. They're taking their medication, they're more adherent, there’s an uptick, that's good.  

But the ROI on that, the lower overall costs over time, is that something that just takes time? But at least now we have a, a baseline you can measure from and say these patients didn't end up going to the hospital or they didn't, they were healthier over time. Is that kind of the next leg of this and we have to wait and see, or is there a way you would respond to that if someone kind of pushed back a little bit?

[17:46] Nash Albadarin, PharmD: Yeah, no, I, I think that's a really valid question, actually. And with a lot of these adherence programs, you know, what we, what we know and understand is that as you improve adherence, right. Inevitably the goal is to get these members to take the medications kind of as prescribed. Right. And improve kind of their ability to have access and take the medications and be compliant with that. And so what that results in, right. If you improve adherence is ultimately they're going to be filling and picking up the medication more frequently.  

And that may, to your point, increase the drug spend for a plain sponsor, but the benefits highly outweigh that because oftentimes, if we talk about these classes, typically, diabetes, hypertension, high cholesterol, the majority of these are generic drugs that will have incremental kind of increases to your drug spend. But the benefit to the medical cost avoidance, in terms of what you mentioned, like avoiding an ER visit, avoiding a hospital visit, those are really where you're seeing the savings and driving value for a plan sponsor.  

And so, yes, to your point, that's kind of how I push back of. Yeah, you know, increasing adherence inevitably. Right. You're getting the member to take the medication more frequently and as they need it. And so that may drive up your Rx spend marginally, but it does improve your medical cost avoidance significantly by improving their health, avoiding a heart attack, and avoiding any kind of cardiovascular events down the road. That's where you see the value and benefit of these types of programs.

[19:05] Justin Venneri: Makes sense. Makes sense. Okay. Well, Nash and Zach, thanks so much for taking the time out of your day today to join me on the show and discuss the posters you recently presented.

[19:15] Nash Albadarin, PharmD: No, I appreciate it.

[19:16] Justin Venneri: Hope you have a great rest of your day.

[19:17] Zachary Brunko, PharmD: Thanks, Justin.

‍

‍

‍

If you would like to learn more about Capital Rx’s full-service benefit administration solutions, including our clinical programs, CLICK HERE to get in touch with our team.

‍

Want to stay apprised of the latest Capital Rx news? Sign up for our monthly newsletter!

‍

‍