Video A Pharmacy vs Medical Benefits Perspective for Adopting Sensor-Based, Glucose Monitoring in the Managed Care Setting - Kenneth Schaecher, MD Play Pause Volume Quality 1080P 720P 576P Fullscreen Captions Transcript Chapters Slides A Pharmacy vs Medical Benefits Perspective for Adopting Sensor-Based, Glucose Monitoring in the Managed Care Setting - Kenneth Schaecher, MD Overview CONTINUE TO TEST Back to Symposium I hope you've enjoyed the two previous presentations and now we're going to get um and a little bit of the practical side of this. The uh focus of my presentation is um a practical roadmap for applying pharma co economic resource utilization and registered data to support and facilitate see Gm us and managed care setting. What I'm gonna do is run you through several areas. Um So the goals of my presentation include discussing the benefits and pitfalls of C. G. M. Is a pharmacy versus a medical benefit. Uh identified published literature demonstrating the cost effectiveness of C. G. M. And the diabetes in diabetes care populations for whom see GM may have shown clinical effectiveness, identify populations that had been found to be useful and cost effective and then also demonstrate a practical approach that we at our health plan took um in assessing whether we wanted it as a farm. Akhil are, I'm sorry, a pharmaceutical or a medical benefit. So what are the benefits and pitfalls of the C. G. M. Being placed on pharmacy or medical? Well I think one of the obvious benefits is for patients. It's a comedians. They're going to the pharmacy to get medications are going to the pharmacy to get their insulin. I'm going to the pharmacy is is very routine and it's easy it's local. Um Lower patient costs. Oftentimes you're able to put these supplies on a fixed copay versus a medical coinsurance which couldn't save the patient money. The process is simplified because it's just like getting a prescription filled. There is also perhaps better time than this because rather than having to think ahead send off into the mail or the ethernet um for um your refill you just go to local pharmacist if it's not ready. They'll call you also maybe some possible lower cost to the plans because the plans um may be able to reap some individual uh rebates or other discounts because now you can build very specifically what are some of the pitfalls? Uh I think one of the biggest pitfalls that you're going to see without going to the pharmacy benefit is that it moves the the the device from a medical benefit to a pharmacy benefit. And not all health plans Have both the pharmacy and medical benefit for each one of their clients. We for instance have one large about 30,000 member group that is we're just managing the pharmacy side for them. Um We're not managing the medical so sometimes that can create some disconnects in the management of their diabetes as you're not sure exactly necessarily what what's occurring not mentioned here. But it also can sometimes complicate the process for the provider's office where there used to sending things in one route. Now they have for some plans they have to go medical. For some plans they have to go pharmacy changing the plan can also cause confusion for patients when they're used to doing things a certain way. And um the other part of it may be that not all of the C. G. M. Maybe on pharmacy benefit. Some of them may only be offered as a durable medical equipment medical benefit and thus that may not um work out in that you can't consolidate things as well as you'd like that. Also then can cause some confusion for providers because they don't know which devices on which benefits as for medical coverage. The benefit I sort of it's the flip side of the pharmacy benefit. Um It allows uh um the airlines I should say with other medical supply acquisition processes so that um it's pretty consistent um And also keeps all the services with the same insurer rather than the potential as I've already mentioned, of one uh insurer covering pharmacy in a different group covering the medical sign. It does add some complexity to the acquisition. Uh You've got to go to a different site, It separates the insulin therapies and their supplies from C. G. M. Which could also be a negative. Um And also what we've seen is the DME manufacturers are very good at auto shipping. Um they send that 90 day supply whether you're using them or not because that's how they get paid. Um The other pitfall is less specific building building because you're using hick pick codes that are general instead of a specific NDC codes makes data analysis uh difficult so that you don't know which devices are being used by which patients necessarily. There are some ways to tell it. It just makes it more complex. So let's talk about the cost effectiveness literature. There are a lot of studies out there for both Type one and Type two diabetics. So most of them focus on Type one diabetics. Um There are limitations to this literature. However, many of the studies are european. If you're doing a cost effectiveness evaluation, it is very difficult to translate cost effectiveness in europe with the U. S. Because of the difference in the care delivery and the cost that is not translatable to how care is delivered in the U. S. Plus our costs tend to be higher. Many are registries or non randomized and thus they have more inherent biases. You don't see a lot of randomized controlled trials in this space. Some and actually many of them are are based upon modeling of data rather than the actual analysis of the data itself. And and modeling has inherent problems with it. It's can be useful at times, but it's hard to draw strong conclusions from some of that. And you'll see in some of the literature that I'm going to talk about how that can be a problem. Many studies are supported by device manufacturers and that's not in itself inherently evil. It's just that there is this question as to what biases may be present because uh unintentional considerations that the authors or others are giving unknowingly to the fact that their support is derived from a device manufacturer. So inherently, people tend to discount um studies supported by device manufacturers a little bit more than they do for those that are investigator led. Also, another um limitation is the fact that most of the cost effectiveness is defined by icer or qualities. Um and that's very useful in europe, particularly for equalities. Uh But it doesn't actually reflect cost savings or two areas where we in the US maybe um concerned about which is health care resource utilization because that's one of our issues. Lastly, the comparative effectiveness of the various diabetes management technologies is largely based on either the surrogate endpoints of hemoglobin a one c reduction or reduced hypoglycemic events. Um Those can be estimated or modeled costs associated with them. Um That can be both direct and indirect these sort of indirect comparisons and the lack of actually looking at health resource utilization and their associated costs. Um Get in the way of drawing strong conclusions. So let's talk about um One study that I think highlights the complexity of this sort of analysis of cost effectiveness. This is a study by peace um who uh they did a systematic review. Um You can see the inclusion criteria, their full or partial economic evaluations based on modeling or R. C. T. S. Of parallel and crossover study design six or more weeks in duration and included community dwelling adults with Type one. I think it's a nice inclusion criteria there They found 16,772 records that they could assess. However when they started to filter it They got down to 152 potential eligible publications out of that 16,000 plus. They then further because only 114 of them reported clinical outcomes and three of them. Um We reported economic evaluations not based on modelling R. C. T. S. They were left which is 35 articles out of the 16,772. So you can see the quality of literature is suffering a little bit and it makes it difficult to draw a strong conclusions. The other thing of note is that the majority of the cost effectiveness analyses were based upon Markov models of 93 and 20 of these used the Center of outcomes research evaluation or core diabetes model. So again, now we're looking at modeling instead of actual um studies with firm endpoint. The other thing to note was there were only um I'll start this part over. The other thing to note is only eight of the studies were based out of the US. are only 23%. The rest were from the UK Europe Canada Australia and the us again, because of the disparity in how health care is delivered and the pricing of services in these different areas. It's hard to draw um strong conclusions from economic evaluations done in other countries. It's also important to note that 26 of these 35 studies had manufacturer support again, causing one to pause a little bit and wonder now this systematic review did have some outcomes. And they did note that there was um a significant um range of Incredmental cost effectiveness ratios from $14,266 to $2,000,997,832. None of these were cost savings, but some of them didn't add a lot of costs. Um It was less clear when you look at in flint pumps when combined with glucose sensors if they were cost effective and that once again they were basing a lot of this on the impact on hemoglobin a one C or hypoglycemia rates rather than health resource utilization. It was noted in the study that that non integrated systems i. E. Multi dose insulin with um single measurement blood glucose Had an adjusted icer of $269,000. Whereas integrated systems with low glucose suspend feature um were noted to not feel be cost effective either. Um At 100 and 75 a range of 175 to 784 plus $1000 per quality. Um well above the thresholds that you're you're normally gonna see of 50 to $100,000. So again this was a large system systematic review and um shows you the struggles. So let's talk about the populations who may have potential benefit from uh C. G. M. I think the obvious are the Type one diabetics were uncontrolled despite evidence for strong efforts to manage their disease. That's sort of the standard population that we see it being used in. Also those with hypoglycemic unawareness. Again, they have a high health resource use and literature is supported pretty well that that um this is a good technology in that circumstance. Another group is pregnant women with gestational diabetes or type one diabetes. And um there was the concept trial Which was published in 2017 which limits um which looked at this now, it was notable that in this trial they didn't improve the hemoglobin A one C that much. Um but the time for these women to have the range uh Of their blood sugar within the normal range was 68 compared to 61 otherwise. And they also had less hyperglycemia. The consequence of that was not so much on the patient themselves but on the neonatal or the infant. Um What they saw is a lot fewer uh infants that were large for gestational age time in the I see you greater than 24 hours was reduced significantly and statistically significantly. And these these neonatal left the hospital sooner than those who didn't use a C. G. M. All of those have costs associated with them. And and show the value of of using a C. G. M. In this population. The other population are young, Children with diabetes type one, especially between the ages of 2-6. And most of this comes from the fact that they can't communicate well. So um you're trying to to monitor and use the C. G. M. As a hypoglycemic alarm because they don't understand what they're sensing and they cannot communicate it adequately. And that's another area that we see. A lot of you supported in the literature. A couple of less obvious areas include um adolescents And young adults. And in the study by lawful from 2020 they looked at this population and notably only 68 of this population, which was only 153 individuals, but 60 only 68 use their devices more than five days a week. This was the study that went on for six months and at the end of it, The A. one c. was reduced from 8.9 to 8.5. Still not well controlled. And it really calls into question whether there's value in this population, which tend to be a little rebellious, tends to uh not want to follow the rules. Um the other is the type one patients older than age 18 or older than age 60. Um And this was a study by Rudy. Uh and in in this study they looked at an older population again small, small number, but um they noted that this population, um 97 of the population used their device six days a week or more. And there was a one C reduction of .4. But more importantly than the A one C reduction was that the time that they had their sugar less than 2 50 was quite significant, with a p value of 0.6 which is quite significant. They also noted that these individuals had less glycemic variability um And with a p value of .02 suggesting that their control is just more even keeled. And it's felt that it's these excursions that lead to the complications more than the the elevation of the sugar itself. Again, pointing to patients who could benefit from this, particularly if they're on multi dose insulin. Now there are some populations that um we're not as certain about And one is newly diagnosed diabetics. And um in a study by patent um In 2019, they looked at Children ages 5-9 that um Were newly diagnosed within the previous six months with type one diabetes. Um the average age was 7.5. Um And they noted that actually um the hemoglobin a one c in this population was increased by 0.4 compared to the controls after six months, suggesting there may not be a benefit of starting a C. G. M. S quite as soon in this population. Now this study had a lot of limitations to it. One there was only 112 families involved, but also the demographics were very skewed. Um it was 88 White, almost completely suburban. And um and the families Also um were uh two parents. Um There were there were no single family households and stuff. So it's hard to draw a strong conclusion from a study like that. Another area of uncertainty is Children under two. And in the McKinley study what they did, they looked at me unaids and they noted that they're just the inherent nature of how see G. M. S. Works. Um the interstitial space. Um, there were limits with the technology. Now, this was a study from 2000 and 17 and that study um may not reflect current technology, but the literature suggests that the accuracy in this population is not ideal, in part because of the timeliness of the measures and the limitations to the changes in their interstitial reflecting the changes in the blood also. So, um the other thing to be no, to be noted in that circumstance is that um the uh Um none of the diabetic uh, see GM devices are FADA approved under age two. I don't believe so. The last group that's uncertain And I and I think this is being explored as we've heard. Um are those that are Type two diabetics. And I think what we have to understand is There's type two diabetics that are on multi dose insulin which are very much like type one diabetics. And that population probably do the literature does support but those on basal insulin by itself. Um The evidence is less clear. Um And Although it may improve hemoglobin. A one C the question there is whether it improves health resource utilization or or changes things in meaning meaningful fashion. Lastly, oral medications only. Uh the literature is scant in this area. Um There there have Carlson published to study. But Even in his study he doesn't define his Type two diabetics and who he looked at. And um One would argue that most type two diabetics aren't even doing finger stick glucose measurements to begin with. Um And as the drugs that are being used are moving away from the uh insulin secreted dogs like cell familiar areas um and moving to other mechanisms that don't have the potential for hypoglycemia. The need to monitor the sugars ask closely is lessened though certainly the importance of blood sugar excursions out and and great variation remain. So um sub populations demonstrating cost effectiveness. Let's talk about there for a moment. We do know that diabetic Type one patients. The diamond trial is one trial that has demonstrated the cost effectiveness in this population. We had 158 patients, the type one, all of whom have their A one c. Greater than or equal to 7.5. They were placed on C. G. M. and after six months they were looked at. They saw that the total six month costs were 11,000 for the c. g m and 7.2 but 70 200 for the control group. So the CG. M. was little more costly. But The lifetime complications from C. G. M. You increase your qualities by 0.54 which is not insignificant so half a year And the incremental cost effectiveness ratio is 98,000 per quality for the overall population. So below that $100,000 threshold, there's been another pop study recently published by Intermountain Healthcare that looked at type one diabetic patients and they um Look at 99 patients in a parallel randomized multi site prospective trial. So good trial design using C. G. M versus just standard blood glucose monitoring What they identified as an improvement in their hemoglobin, anyone saved to a P value of 0.01. But they also identified a reduction in health resource utilization because of improved control. So their total visits were down, emergency department encounters were down and the number of labs subsequently being done were also reduced. They identified that among non Medicare advantage patients per member per month savings and that's important to note savings. We're $417 per C. G. M. Compared to finger stick glucose. It costs a little bit more for Medicare that it did. Um For the commercial population, 70 of the C. G. M. users reported that the technology helped them better understand their daily disease and also Help them understand the effects of their diet on their sugars. Whereas only 16 of the finger stick glucose people had that um knowledge imparted upon them. So um definitely showing the benefit in that type one population. So other sub populations, We've already talked about the concept trial um which is probably one of the better trials. And we've talked about type two diabetics. There aren't many studies, there was this one Spanish study by Garcia Lorenzo um in 2018 that looked at real time glucose monitoring. Again, the challenges there, you're talking about different health care delivery and you can see they did not identify in Type two diabetics that um use of C. G. M. Was cost effective with uh qualities of 100 and €80,553 per quality for the Type two diabetics, Type one diabetics. They identified it to be even less. But again, this study was looking at different cost structures and different care delivery that may impact it Fonda in 2016 also modeled some outcomes. And they suggested in type two diabetic that there could be some cost effectiveness is they have um noted that um the icer was $9300 uh Or $13,000 annually. All right. Well $9300 annually. Or $13,030 per quality gained. What are the issues and barriers um with regarding moving um C. G. M. From medical to pharmacy? Well, see Gm. It is important to note that C. G. M. Is becoming a standard of care. It's replacing finger stick blood glucose is particularly as they have moved from having to do as many finger stick blood sugars to calibrate the C. G. M. Devices. Most of the literature, however, has not been able to demonstrate cost savings and that's because most of the literature focuses on the short term and not the long term costs, which is where most of the costs are with diabetes. Another barrier is that traditionally these devices are DME they're on the medical benefit and because of that, the codes used don't allow for cost production opportunities that exist on the pharmacy benefit by using NDC level billing. So it makes it harder to track products uh and and understand where those opportunities exist. Also. As I mentioned earlier in my presentation, not all members have a pharmacy benefit within the same plan. So for my plan to cover it on a pharmacy benefit may not impact members who are on my medical insurance. So pros and cons, I've gone through some of this earlier. There are no different than what we discussed. Um You could do NDC building, you can potentially lower the cost to members. You've got some contracting opportunities um and with the cons you do lose some coordination of care potentially. Um the auto fields may be an issue and uh in some instances to get those savings, you have to give up some of the controls you have in place on the medical side. And and that can be a challenge because that can increase utilization. So let's talk about some of the pricing. Now what I have here is a slide that shows the pricing for dex calm On both the medical and the pharmacy benefit. As an example. Now these prices were taken uh from the first data bank in April 2020. They may be changed a little bit and they also reflect some of our own internal contracted pricing. Um you can see that to cut to the chase that the annual cost for the decks calm on the pharmacy benefit is 53 91 On the medical benefit at 71 41. So without any other change other than moving it to pharmacy, do you end up um Saving approximately $1800. Now. If we go on to look at the decks. Com versus uh freestyle lee brae On two on the pharmacy benefit. You can see that the library offers even more cost savings. Now, I will say that these costs are not reflective of additional rebates that Dicks. Com is offering to move to the pharmacy benefits. So the disparity and costs that you see here is not um accurate. Um And I'm just not at liberty to share the actual cost differentials. But um let's just say that they're substantially more savings uh can be made by moving dicks. Com to the pharmacy benefit than what you see here. All right. So the situation's been presented to us as a health plan. We were we were approached to possibly look at um moving R. C. G. M. Uh to the particularly decks calm to the pharmacy benefit. The findings from the analysis for us. We were, we were presented several scenarios um, one of which had exclusivity, excuse me. One which had exclusivity involved another one um, which was less exclusive and allowed us to maintain some controls and, and, and some that that um allowed us to perhaps save even more, but we lost some of our controls that we had in place on utilization. So, um, we noted that there would be potential savings to the member by the reduced cost share from a, from a percentage uh, co insurance on the medical benefit to affix copayment In our situation. The member convenience would be greatly improved, possibly leading to better adherence and compliance. Um, significant potential cost savings by the plan could also be made by moving to an NDC billing model rather than the hit pick model. We also noted that there were additional rebates available to us by moving into pharmacy benefit. Um one of the considerations as I've already mentioned is that we could have saved more money by removing some of the coverage criteria but we were not ready to make that step and another is member disruption. They ban getting it through a certain supplier to so to make that change over, you have to make sure that you're you've got a good communication plan. So our determination at the end of the day Was that we were going to move the decks Com 6 to the pharmacy benefit. We would leave freestyle, library and library to are already on the pharmacy benefit and we're not um on the medical benefits. So we did not make a change there. We were continuing to apply our authorization criteria. Uh we gave up a little bit to do that, but we felt comfortable with that. But we also noted that we needed to maintain the Medtronic Guardian on the dems benefit as a non preferred product because of patients who are using um Medtronic insulin pumps. And this is a proprietary relationship that we did not want to um cause problems with or having to purchase a new insulin pump. So that's what came down in the long run. We looked at it, we maintained our coverage for those who we felt uh comfortable with. And we um went ahead and uh moved everything over to pharmacy. Beginning this last January. one. You've seen what one health plan has done. We made the decision January one. We're going to go ahead and cover uh some of the devices on the pharmacy benefit in the real world. What does that mean? Well, it appears, see GM is becoming the standard of care. We know that it can in certain circumstances impact health resource utilization, pregnant women. We know it reduces um the number of babies in the ICU. We know patients aren't tired of sticking their fingers um providers, I'm an internist. I still see patients providers are out there starting to use this earlier and more often because the patients don't have to poke their fingers anymore. So the reality is C. G. M. Is here to stay. We're going to see it move Um into the mainstream of diabetic management perhaps even in some type 2s. And what will also see is that um as the costs come down, you're going to see a reduction in management by health plans because it's just not it's not worth it. Um with the prices you're seeing for liberty to and with the rebates and how dex con is getting down to nearly the same pricing. Um It's just becoming a standard of care and at the end of the day I think that's where we're all going to be when that that is going to happen. I don't know. It may take a few years, but I think that's the way we're heading. So with that I want to thank you all for your time today and I'll turn it back over to our moderator. Published April 9, 2021 Created by Related Presenters Kenneth Schaecher, MD, FACP Associate Chief Medical OfficerUniversity of Utah Health PlansMurray, UT
