Send Seer Zip Code with airSlate SignNow

Send seer zip code

good afternoon everyone and thank you for joining us for the webinar today fear and they show us data resource webinar highlighting newly available Medicare Part D claims data my name is Stephanie and I will be your WebEx host before moving into today's webinar I'd like to go over a few logistics all lines are in listen-only mode and if you need to view light-light closed captioning you can select this by using the media viewer panel on the right hand side of your screen should you have any questions at all during the webinar you can submit them at any time during the presentation just type it into the Q&A or chat panel and select host from the drop-down menu we will ask those questions on your behalf at the end of the webinar as a reminder this webinar is being recorded and will be posted online in the near future and I would like to pass it over to dr. Roxanne Jensen right hi everyone welcome thank you for joining us today my name is Roxanne Jensen I am a program director in the outcomes research branch at the National Cancer Institute and I'm the scientific leave for this year mhos data resource I am joined today with by dr. Amy Davidoff she is a senior research scientist at the Yale School of Public Health and affiliated with the Yale cancer outcomes Public Policy and effectiveness Research Center on this webinar today we will provide some background on seer Medicare and the Medicare health outcomes surveys discuss the new edition of Medicare Part D claims data to the resource then dr. Amy Davidoff will present her project examining pain interference and pain management using this year mhos data resource with Part C claims data and then I will discuss the application process with a focus on projects proposing the use of Part D claims and then we will take your question so the surveillance epidemiology and End results seer program provides information on cancer to statistics in an effort to reduce the cancer burden among the US population this here cancer registry data file is a comprehensive population-based source of cancer data in the US the cancer registries represent in this method in the sphere cover roughly 34 percent of the u.s. population seer contains demographic information and cancer information such as primary tumor site and morphology cancer stage of diagnosis and initial course of treatment radiation and surgery it's important to note that the seer data in serum HOF does have a vital status indicator and that is updated with each new data linkage this enables researchers performs survival analyses with the seer image with data resource the second linked data source are Medicare beneficiaries specifically those enrolled in Medicare Advantage Medicare provides insurance coverage for about 50 million beneficiaries these annuities are typically 65 and older younger people with disabilities people with end-stage renal disease Medicare Advantage is offered by private insurance companies approved by Medicare Medicare Advantage consists of managed care plans which paper enrollee versus fee-for-service this distinction is important because this means there are no Medicare claims available for this data resource over one-third of Medicare beneficiaries are currently enrolled in Medicare Advantage plans the Medicare health outcome survey is a survey of Medicaid Medicare beneficiaries enrolled in Medicare Advantage all Medicare Advantage plans with over 500 enrollees must participate and a random sample is drawn from each plan surveys are administered by mail or telephone and follow-up is occurs two years later with the same person so the survey captures a wide range of health outcomes such as health-related quality of life this can include physical health mental health pain interference a limitation to activities of daily living ADL's there are questions that are used to create heat as effectiveness of care measures quality indicators this includes risk fall management urinary incontinence and their additional variables of interest such as self-reported smoking histories of for protocol rapidity and BMI the response rate is 63% from baseline and 89% for follow-up survey so let's hear it I made to a data resource represents the overlay of these two data resources so where someone is diagnosed with cancer at any point since 1973 within a seer registry catchment area is then at some point later enrolled in Medicare Advantage plan and then was randomly selected to be part of the mhos sample cohort consisting of two surveys two years apart and also please note that the CR mhos data resource also includes non cancer cases to provide a comparison group to the cancer cases this map displays both cancer registries that are included in CRM h OS thos does not include Alaska Natives and Arizona American Indian registries and supplied represents numbers of the most prevalent cancer types available in the seer mhos data resource and when the surveys are occurring relative to the cancer diagnosis as you can see there are large numbers of prostate brass Clark along bronchial and this data is representing 1998 to 2017 about 6% of the patient or the patient's for each cancer completed a survey before and after diagnosis that means they were enrolled in Medicare Advantage completed a survey those here made to a survey and within the two-year period before they completed another survey were diagnosed with cancer it's a very unique cancer population and one of the true highlights of this data resource thank and you'll notice um while all of these cancer cases um you see they have surveys after diagnosis the rate for lung cancer is thirty percent versus between 66 and seventy seven percent amongst you to a lower life expectancy for lung cancer so um so people don't have as long as fill out a survey after a lung cancer diagnosis these numbers highlight some strength of this resource there's large sample size from many cancer sites it's one of the largest data resources on patient reported outcomes among cancer patients in the US and there's an ability to look change over a two year period and finally you can compare it to individuals who even to our individuals with and without cancer all right so for now now when we want to talk about Part D but for more information and more in-depth information about the data resource including a deeper dive into the questionnaires your mhos data I would encourage you to visit our website and review our introductory webinar for a little more information okay so on a little background on Part D Medicare Part D is the voluntary prescription drug benefits that started in 2006 Part D plans may be offered a standalone or link to Medicare Advantage plans met most Medicare Advantage plans offer a linked Part D plan and so as a result many of those enrolled and Medicare Advantage have a Part D plan and so in most relevant to cancer patients of Part D plans cover most oral chemotherapy and other systemic therapies supportive care medications and medications for the management of acute and chronic conditions Part D plans do include cost sharing such as deductibles and co-payments and low income beneficiaries may apply for cost-sharing reductions known as low income subsidy here's a list of some of the selected Part D data elements that are available um up at the top you see the patient or beneficiary which is a common identifier for linking claims to respondents in Sierre mhos to the date of prescription fill and in light blue there's a series of variables to identify the specific medication how its administered and quantity there is currently no linkage to Part D plant formularies or plan characters who files for CR mhos so the info that highlighted in green at the bottom here maximum step number quality limit prior authorization required um that use of this information would be limited as that it's used often in conjunction with that information so what's is my Medicare Part D Medicare key claims identify what prescribed of prescribe medications are filled and this concluded the daily dose duration of single fills duration over time for medication with chronic use medication swishy is switching those changes discontinuation and medication adherence claims based measures oh it's important to know that they reflect the medication availability but not if they were actually ingested so the most um this is a graphic detailing the data ability ability for mhos survey of the cancer diagnosis or care data and the party claims and you can see all of them start and stop at different dates and this is something to really keep in mind when you're setting up your research question using this rated source C or cancer information goes back to 1973 of 2015 mhos is updated cohorts every every linkage we have 1998 to 2017 and Part D claims we have from 2007 to 2016 each of these will be updated in future linkages however this is what we have available today and so finally I'd like to present a table that reflects here my Cho as with Part D D data this is very similar tables for the one I showed you before however this is the numbers based on party data that's linked and this data reflects 2007 to 2014 so there are a few additional cases that have been added in a most recent linkage as you can see here again you have a host of 50 percents with at least one or two surveys after diagnosis and you have a good number of people about 12% that have a survey pre and post diagnosis you do a lot of options when you're trying to think of research questions using Part D data in this population and now I will turn over to dr. Amy data okay well thank you very much Roxanne for of the great background and the introduction I'm going to be talking about one of several projects that team of us at Yale did as part of a demonstration of the feasibility of working with the linked data sets to address some important questions the work that I'm going to present today has been accepted for publication and supportive care and cancer and you can see the citation there and a link including my wonderful colleagues who helped me on this project so pain is highly prevalent among older adults about not 29% report persistent pain so there's been concerned about adequacy of pain management our patients who report pain receiving treatment are those who receive pain treatments still reporting pain in other words is pain management adequate and then is adequacy of pain management similar for older adults with and without cancer there is a perception that patients with cancer may be under or over treated relative to older adults without cancer there is a large literature that describes patterns of opioid use in particular for cancer compared to non cancer there are no large population-based studies that actually adjust for reported pain severity and since that may differ for adults with and without cancer it's important to adjust for that when assessing a pain management the study aim is to examine patterns of medication therapy for pain management in older adults with and without cancer our study design is cross-sectional we considered trying to look at only those observations that had to M Haas survey so we could see whether there was a change in pain management and pain report over time but we were concerned about sample size and so we decided to do a cross-sectional study the outcomes of interest were medication for pain management and opioids use detail on opioid use and then other types of pain medications the key independent variables of interest were pain interference and cancer status whether you are someone who is a cancer survivor or not we generated descriptive statistics and ultimately ran multi variable logistic regression models the dependent variables were either any pain medication or any opioid receipt and these are all measured within the 30 days prior to the M Haas survey and then we pulled our cancer and non-cancer cohorts and we interacted pain severity and cancer status in order to test whether there were differences in medication management for cancer versus non cancer across different levels of pain severity our covariance and the analysis were the usual demographic characteristics health conditions as reported in a chronic condition checklist on the M Haas we include an indicator for whether the individual was enrolled in an HMO versus a PPO and whether they had the Part D low income subsidy or extra help but just a little bit about our methods and these would be relevant for you if you're considering a study you need to think about sort of what the relationship is temporarily between the cancer diagnosis that's reported in fear and the Emhoff survey you want to have an imposter of a that perceives the cancer that that is after the cancer diagnosis and how far you are going to allow those a temporal gap and then you also need to think about how to refine us when there might be more than one cancer primary or more than one a maja state because we wanted to compare pain report and medication therapy for Medicare beneficiaries with and without cancer we wanted to make sure that the M Haas pain report is fairly soon after or as soon as possible after the cancer diagnosis date when when the patients still may be experiencing symptoms from the cancer or from the therapy and obviously some patients have very limited time frames when they're experiencing this and others have may have much longer time periods when they're experiencing pain after cancer so we decided to use a five-year window so for our sample selection we required that they have only one cancer of primary reported in the seer registry pull those for January 2003 through December 2012 they had to have completed at least one M ha survey between 2008 and 2012 and the first M ha state had to be within five years after the cancer diagnosis and you can see our little schematic there BM ha state between 2008 and 2012 and the cancer diagnosis less than five years prior to that M ha survey so other considerations is really the observation period that you want to use for the prescription drug use because this will determine how long you have to require continuous Medicare Advantage and Part D enrollment and not only that you need to consider if you're relatively new to using Part D claims that if you have an observation period you actually need to look at drug fields that might have occurred prior to that observation period because oftentimes a fill might be for a long period of time and might actually spill over into your observation period so we selected the time period of 12 months before and 12 months after now we actually didn't use that full time span for this particular project that I'm going to describe to you but we were doing other projects and wanted to make sure that we were sort of drawing in certain time periods that would be relevant for all of the studies that we undertook so we required that they have MA and Part D enrollment twelve months before and twelve months after the M ha state that we had selected so further sample selection we started out with forty eight thousand nine hundred and forty nine in the total M Haas cancer sample after we these are patients who have a cancer that's linked to M Haas known date of cancer diagnosis etc and then when we require that they be aged over 65 because we wanted to focus on older adults and an invasive tumor we come down to 32,000 when we put our timeframe in there of when they're surveyed in relation to the cancer diagnosis drops down to that 12,000 etc and you can see on the right-hand side of the slide that we have a similar sample selection process for a sample of individuals who did not have a seer cancer but had responded to the M Haas and lived in this year region so we ended up with the total of 90 100 with cancer and about 6500 without cancer to compare to for a total analytic sample of fifteen thousand six hundred and twenty-four so the Emhoff survey itself has some options in terms of how we could measure pain but the questions changed over time and we needed to carefully select questions that covered the majority of our timeframe between 2007 and 2012 there are questions about activity interference associated with pain in the past four weeks in addition there are several questions about chest pain low back pain etc but we felt these were not sufficiently Universal to incorporate so we focused on the activity interference with pain in four weeks now starting in 2013 there are new questions that focus on the last seven days and are a little different in terms of the wording related to day to day activities or social engagements and there's also a pain scale which will be in a very interesting addition as a resource for analysis so things to look forward to measurement for this study used the am Haas question during the past four weeks how much did pain interfere with your normal work including both work outside the home and housework and there's a five five level response and we lumped the top two and the bottom two so ended up with a severe moderate and mild or none you okay in order to identify our pain therapy we needed to identify the relevant medication classes which we did by looking at the literature looking at guidelines for pain management and then we had several wonderful Commission's in our team who are palliative peer experts and they gave us feedback throughout we identified once we had the classes we identified their relevant generic drugs within the class and then we selected the relevant records within the Part D claims that to select once you know that the type of drug you're looking for you can either use NDC codes or generic drug names to identify the actual claims we decided to use the latter we use a generic name field for the drugs we use string search to identify combination drugs and then our palliative care clinicians confirmed the relevance particularly when we had combination drugs relevance for pain management so we selected claims that had drug supply 12 months before and after our m ha state again to be useful for other studies that we were doing for this particular study we focused on the 30 days prior to the M ha survey date you the medication classes we included were opioids both short and long acting non-opioid analgesics NSAIDs and local anesthetics and adjuvant pain medication anti-epileptics and selected antidepressants that tend to be used more for opinion than for depression we do note that steroids are commonly used to manage pain but they're also used to manage nausea and as part of chemotherapy regimens because we only had access to Part D information we could not identify through other symptoms or chemotherapy regimens that might be ongoing where the steroid would be a component of that and because we were unable to assign therapeutic role to the steroids we did not include them in the category of pain medications once we had pulled our claims and particularly our claims for the 30 days prior to the M ha survey we created measures for any use the category of medications for opioids the specific agents used the duration the morphine equivalent daily dose and the initiation of opioids after the M ha survey we considered both sort measures of untreated and undertreated pain particularly patients who are reporting severe pain and not receiving opioids or patients receiving moderate pain without any pain medications the immortal until I notice that I mentioned is a way of creating equivalents across different types of opioids because they do differ in their strength and so they're all sort of there are conversion factors that are used so that there's a common metric so we identified all of the opioid prescriptions with supply on the M ha state or the closest state within that window going backwards we calculated the daily dose for the oral tablets we then converted the daily dose to this morphine equivalent daily dose using conversion factors that we pulled from the Centers for Disease Control and modified by the Centers for Medicare and Medicaid Services when we had more than one opioid that was used simultaneously which is very common then the morphine equivalent daily dose was summed to calculate the overall dose equivalent so some results from our study this slide just shows an overall description of our medication use for management of pain like cancer history and so you can see that we report for any medication use opioids non-opioid analgesic and adjuvant pain medication overall for cancer and for non cancer so overall 21.5% received any pain management medications with slightly higher 22.5% of cancer of population receiving on any medication and 20.1% for the non cancer comparison group the percent of receiving medications with somewhat higher for the adults with the cancer than non cancer for all of the different categories all the differences were not particularly large and remember this is not adjusting for pain severity slide shows similar results but this time looking at differences by means they're of a severity of of paying interference I apologize that we've got two blueish bars in there but the first one is the first somewhat blue or green one is the estimates for those with severe pain the middle group is for those with moderate and then the final group is for mild or no pain severity and you can see there's a monotonic association between the level of severity and the percentage of patients who are receiving medication and all of these differences across the severity category are significant this just gives you a list of the specific opioids that were used with the highest being hydrocodone and oxycodone names that are commonly heard in the news this just gives you a sense we didn't find that there were differences between the cancer and non-cancer samples in terms of the types of opioids that were being when we look in detail at the opioids first looking at the duration of opioid use prior to the M ha state we see a real mix with a fairly substantial group having fairly short term exposure to opioids and then almost 40% overall having had opioids for at least 30 days no difference between the cancer and non-cancer groups for the MACD we have fairly similar medians across groups but the mean for individuals with cancer was 58 point 3 milligrams compared to those without cancer 50.7 and that is a significant difference we also looked briefly at whether there is a higher very high M EDD above 90 milligrams and we found that overall almost 15% had that very high dose but that there were no differences between individuals with and without cancer the results of that our logistic regression analysis are presented here is adjusted relative risks you can see that the biggest factor is pain severity with significant effects for cancer I can give you the overall interpretation of these interactions with the interaction terms at two in two slides but other factors that were significant we did control for proximal deaths or whether the patient died within six months after taking the am house and that was a significant indicator or predictor of increased risk of having any pain medication other factors increasing age and Asian or other race ethnicity were associated with lower rates of pain medication Part D L is was associated with higher risk focus in on opioids the results are a bit more extreme in terms of say the risk associated with severe pain moderate of cancer proximal death and then a number of factors that were not significant in looking at any pain are in fact significant when we're looking just at opioids and you can see factors such as black rays Hispanic ethnicity marital status education poverty come up now as significant whereas they did not before for the any pain management so here is a slide showing sort of what happens when you pull together the separate results for cancer history and pain severity using predictive margins predictive margins provide the predicted probability when you're sort of turning on and off various dummy variables and so you can see here that for cancer patients with severe pain predicted 38.7% received any pain medication compared with adults without cancer 37.1% so that's you can see that the there's a higher predicted probabilities across all of the pain severity groups although it's only significant to the severe and mild non groups you can see that this pattern of results holds also for opioids where the individuals with cancer have higher probability of receiving an opioid compared to those without cancer across all three D severity categories and these are significant again for the severe and the mild non categories so just to discuss briefly this analysis looked at a snapshot in time 30-day period suggesting relatively low use of medication therapy for pain particularly opioids although we would have liked to have a sort of newly diagnosed cancer patient population the five-year window gave us somewhat more of a prevalent cancer population and I think that the medication use is probably consistent with that more prevalent cancer population appropriately we observe that medication use is higher with higher reported pain severity there seems to be evidence of under or untreated pain only 30% of those recording severe pain received in opioid and we would note that that actually taking the Emhoff survey they have somewhat of a Heisenberg effect in that it may help sort of alert patients to be thinking about their pain they may end up reporting pain to their physician and so there may be the addition of pain medications after the actual empath survey but that is not within our observation period but it is part of some of the other studies we're looking at so overall the receipt of pain medication was greater in older adults with cancer even after adjustment for different characteristics including severity of pain dur fear ins it may be that oncologists are better tuned to pain management issues or have less constraints on prescribing opioids we note that race ethnicity and socio-economic differences were present particularly with respect to opioid use we fell low over opioid use in race and ethnic minorities at higher opioid use with higher poverty rates and with the Part D low-income subsidy so there's clearly some issues there related to access to care that are likely affecting access to pain medication management so with that I will thank you and turn things over to Susan I believe okay um well before we get to do this I'm going to take us through very quickly the application process for C ramage OS with an emphasis on Part D requests so these are the pieces that investigator needs to apply first is your image OS project a cover letter application form IRB approval as required by your institution a signed data use agreement and a request for any restricted variable is necessary the process as you can see there's a lot of steps in this process for investigators focus on the purple boxes you're developing your proposal it is then reviewed there may be some new revision that's reviewed by C or PI then you need once we have approvals payment needs to be risky of Hoops revises necessary Green is a milestone of approval and then we move forward into finally the final green box at the bottom is the data is released to the investigator with a 5-year ziwei and after five years the DA can be renewed on a yearly basis as necessary you can see our website says it takes between three and four months to go through to getting the data released we encourage you to talk with us when you're starting the proposal process um you know and as you're thinking through your ideas some common application considerations uh accurate cohort collection for your research question is really important you should define clearly why you are selecting the cancer sites or types that you are you need to give good consideration to the timing of the cancer diagnosis in relationship to the survey and then also finally clearly defined why you need a non cancer comparison group if you are requesting one as part of your research questions um I think keep careful considerations about changes survey content over time the mhos survey does change there's three different versions from 1998 to the present and so you need to be mindful that the content you are considering is is available and always think about sample size and power um these things can get small very fast based on the available data and I think most relevant for this group is that you need to be thoughtful about how to justify they need for Part D data I'm just saying that you have it and want to use it it's not going to be enough um we'd like to know a little bit more about how it's being used and integrated in the research question and so um it's Part D justification think about how it's represented in your aims message variables analytic plan um you know I think thinking about sample size issues specific to the years and available party data is important documenting in your application availability and diffusability of getting the claims that you want and then also finally as part of your research team having someone with demonstrated expertise on Part D data is very helpful okay um and before we wrap up for questions Anna direct you to two helpful resources um the first link is to see ways that are public using this here may show us data these are applications that investigators have submitted and indicated that they are happy being public and - these are lists of research aims and current ongoing projects the second bullet are is a list of CMS publications so this can direct you to things that are out in the published literature and maybe also help you think about ways investigators are using this data resource alright without further ado I think we will end this over to Susan buckler to run the question in the answer session thank you so much Roxanne and thank you Amy I will now be starting the question and answers as a reminder you can and your questions into the chat box was mentioned at the beginning of this webinar and we will get to as many as possible that we can so I'm going to start with this first it's a two-part question I'm going to direct it to Roxanne the first part of the question is what types of claims are available and Medicare what's the claims are mobile and Medicare oh um so for this here at Mitchell s major resource the only claims that we have available are part date method and so please you know we do not have utilization data you do not have a or B thanks okay thank you it looks like another question is coming in asking if there are any in-depth trainings available on how to use the Part D claim I'm the graphics direct fan okay Susan um so there are two resources I would direct investigators to NCI periodically offers your Medicare training for researchers interested in learning more about these data there was one recently this fall this year and you can go to your Medicare website and there are the slides and videos from the ten of the most recent training and there is a section on Part C data and so while in general it is geared toward t4 service I think there's a lot to be taken about the party and the this year variables - for anybody interested in working with the seer mhos dataset resdac which is an external site also has a variety of workshops and online videos that describes Part D and the claims so I think well the party files on on the res excite they're slightly different than the ones we're presenting here it's not than those linked to your Medicare I think the underlying concepts and approaches using this data are the same and would help people investigators thinking through their research ideas and using this data resource Thank You Roxanne up next I'm gonna a question came in I'm going to direct towards Amy um so we'll let her come mute and the question is what type of brakes and coverage would show up and Medicare Part D claims for example if a patient switches plans for a period well it might actually not come up as a Part D claim but it would come up as a indicator in the enrollment file so there is an enrollment file that in indicates on a monthly basis whether the individual is enrolled in Part D and also enrolled in MA and if you see months that where that doesn't occur then that is a break in coverage typically we assume that if there's a one month delay break in coverage that there's some administrative snafu and we tend to ignore those and but if there's in any period longer than that we exclude the individual hold up sample selection process thank you Amy we have another question that came in I'm going to direct towards Roxanne it asks that they understand that Jeremy Joey has publicly available data through 2017 but how what is what are the different time periods for Part D and and MCA incident um so here image-wise data is available from 1998 to 2017 however the Part D claims that we have associated with the mhos data run from 2007 to 2016 so I think that's important to keep that in mind when you're considering your research questions and sample size um when we link again we linked spring of 2019 of this this year um and we linked every two years in our next linkage we will update the Part D will update the mhos survey and fear data enrollment data so everything will be updated in our next linkage but until then you know these are the this is the date ability that we have Thank You Roxanne the next question is I'm going to direct or do you Amy and it is what drug information is provided in Part D and also what is not included and Part D if you can go through go through of that differentiation sure so within the Part D claim there's information on the date that the prescription is filled the date it's to spend the amount dispensed the supply the decode the brand-name generic name drug plan code whether individual has low income subsidy prescriber code fill number dispensing status and information on the cost of ingredients and the fee for filling the prescription and then lots of information on what is the out-of-pocket requirement for the individual so that's Part D claims more generally the seer Part D linkage has a little bit more limited information in particular the but the date of fill and dispensing the amount the day supply the NDC code the brand and generic name are all part of the Part D linked data what's not included is any information on the medication class the pharmacy the information and what's not included also in the Part D linkage to information on the prescriber which is in the CCW of party files so a little bit different in terms of what's in the linked part d claims compared to the full of CCW Part D claims but a lot of really important information about the drugs that can be used to describe patterns of medication use great thank you that was very helpful I'm going to direct the next question to Roxanne it's about sample size and feasibility for a specific project what is the best way to get sample size or if you develop feasibility estimates is this something that can be provided to investigators before they apply for a dua and thanks then um there are a couple ways you can go about this one one great starting point is to get some sea air mhos sample estimates that can be generated through the seer status application um on our webpage there's more information about there's a specific serum each was flagged so you can understand by cancer type and by specific information view care most about you can understand who's in our data set linked to seer what kind of information we have especially if you're looking at rare cancers this can be very helpful it will not tell you whether we have Part D data on on the person that's there in the larger sample but I think it's a good place to start to see if you are in the ballpark I think another step that would be very important for investigators that want to use Part D data is to email our help email which is a seer MHO a seer - MH o s @ h CQ is back ward we have a serum HL s Medicare Part D linkage technical report it summarizes some of the information we discussed today and the challenges and also gives some preliminary data such as data table I put up there that's another place to look to understand feasibility and sample size of your study and then also I think please reach out to us um you know this is a new linkage it's just one on line this year and we're willing to work with people to a o2a degree to just see to assess feasibility before you go through the data do you a process does you do with process so we have a new question that just came in asking about after the dua is approved do we have an idea of how much the Part D data will call us yes um our website we have a cost calculator that will let you calculate the costs of your request and the name includes and be out on Oh Part C um the cost of spirit mhos theta varies by the number of cancer sites that you are selecting so I would encourage you to go there learn all the numbers that you may need to know next question that came in is there a way to get mhos data linked to Part D without the sphere cancer cases so this question is asking if you can use this data with party but not focus on a cancer related question research question and the answer is no um Sara Mitchell is a data resource that was designed to improve our understanding of cancer health outcomes for those enrolled in Medicare Advantage organizations and our our purpose is to focus on cancer and while we are interested in cancer non-cancer comparisons are you is a few ways that are submitted that do not focus on cancer populations will not be approved Thank You Roxanne we have time for a few more questions so I'm going to direct this next one to Amy how can you determine periods of coverage within the Part D data you can determine periods of coverage using the enrollment files which are sort of an auxiliary file that should be merged in that would really be the best way to assess yes in coverage is that what the question was or did I misinterpret nope I think that was I think that was exactly it okay I think we're going to pause to see if any more questions come in very briefly before we wrap up okay I think we did see one come in about the to repeat the URL for the technical report request can we post that in the chat so it's been posted in the chat but I will repeat it very quickly it's fear - mhos at HQ is 0rg but that should be posted in the chat as well for you to capture a final question before we're done I'm going to direct to Roxanne and that's are there any related fo A's or advice about where to seek funding for CR mhos Part D project great question Susan um so there are no specific fo A's particular um but I do investigators have reached out and written up grant applications such as our o 3s and r21s using this data set I think you know the things that we're always very happy to talk with investigators both to talk about what their research questions are and refine their projects before they submit to D way and I think anybody who is interested in this data set these questions we just encourage you to reach out to us um to see to check about feasibility and and what you are trying to do we're here for you guys perfect well thank you everyone for joining today and thank you for the great question as a reminder this webinar has been recorded and it will be posted online in the next couple of weeks so if you'd like to continue the conversation online or offline you can do so thank you again everyone and have a wonderful day