Webinar: Functional Outcome Measures in IRF & SNF: How is CMS Measuring Them?

Disclaimer: This transcript is intended to provide an overview of the main topics discussed in the webinar. Because it has been auto-generated, it might contain errors (including to proper names, industry terminology, and punctuation that result in altered meaning). To hear the webinar in full, please listen to the archived recording.

Watch Webinar Recording

Learn more about Outcomes Improvement.

Webinar Transcript

Welcome everybody. Thanks for taking the time to listen to this presentation on what I believe is a very important subject for us: The functional outcome measures currently in place for skilled nursing facilities and inpatient rehabilitation facilities.

These are our objectives for the hour: We’re going to describe what the measures are, look at how they’re being used, understand how to retrieve and interpret the outcome reports, and then examine competencies and processes so that you can ensure the data is reliable and accurate.

I like to start these live presentations with getting an idea of who is on the call. So this first poll is about you and what is your role at your facility or in your organization. Are you a PT, a PTA, an OT, an OTA, an SOP, an MBS, an assessment coordinator, in nursing leadership, or do you case manage? And lastly, are you an administrator?

We’ve got about 70 percent of people responding, so we’ll go ahead and show what the poll results are. Looks like most of us are physical therapists or physical therapist assistants, but we’ve got a fair number of administrators and occupational therapists.

So as a backdrop to this topic, let’s remember how this move towards measuring functional outcomes has happened. Attendees of this course, you might be very familiar with the fact that for the last decade, Medicare and policymakers have raised concerns that the prospective payment systems encourage some providers to favor patients who required more therapy services than those who are medically complex.

Now over the same period, Medicare became increasingly interested in care transitions. Medicare does believe that hospital readmissions are directly impacted by quality of care. Another trend that has been documented is that spending on similar patients is very different depending on what region of the country a person lives in, like east versus west or urban versus rural. And then spending differs across states for some of the same types of patients.

Now over time, all of these trends culminated in policymakers making firm decisions that the payment systems for care needed to change and that the focus needed to shift from what services were being provided to what was the outcome of those services.

So as trends became increasingly recognized, they became a concern for legislators. And in 2005, the deficit reduction, DRA, was passed. And this directed the Centers for Medicare and Medicaid Services to develop a post-acute care payment reform demonstration with the goal of creating one unified payment system for all post-acute care settings. That includes long-term care hospitals, skilled nursing facilities, inpatient rehab facilities, and home health agencies.

To do this, Medicare hired a contractor to develop a unified assessment tool, since currently each of these settings had their own individual assessment tool. So in 2006, industry stakeholders were gathered to assemble what is known as the Continuity Assessment Record and Evaluation Tool, or The Care. The idea with this was to develop a tool that took the best parts of each setting’s assessment tools and merge them into one universal tool, which was named The Care.

They then tested that tool between 2008 and 2010. Over 150 providers used the tool alongside their setting specific tool, and the data was then submitted to the Medicare contractor for analysis. After analyzing that data, they then submitted their report to Congress. The results of the study were compelling to congressional representatives. They saw an opportunity to both improve patient care and perhaps decrease the variability in spending.

So the Improving Medicare Post-Acute Care Transformation or Impact Act was drafted with input from industry stakeholders and in record time it was passed and signed into law in October of 2014.

So the Impact Law required the collection of standardized patient assessment data across the post-acute care settings, something that did not exist and had never been successful before. Congress and policymakers were convinced that if there were standardized data across the settings then a couple of goals might be achieved: Care would be improved because providers would all be talking the same language and Medicare would better understand the differences in patients treated in these settings because the same patient characteristics would be collected. It would be easier to share information between settings, perhaps making interoperability easier, and it might lead to the development of one payment system for all four of the post-acute care settings, rather than having to maintain separate payment systems for each setting.

Now, as I said earlier, many of the standardized items actually come from the current assessment tools used in the four post-acute care settings: the Altec Care B tool, the Skilled Nursing Facility Minimum Dataset, the Inpatient Rehab Facility Patient Assessment Instrument, and the Home Health Outcome and Assessment Information Set.

So if you notice the title of the slide, it says standardized patient assessment data element. That is the name CMS has given to these items that will be sprinkled through the four settings. Now we all know Medicare loves acronyms and this one is conveniently nicknamed SPADES. So once an item is selected to be a standardized patient assessment data element or SPADE, then it is proposed for each of the settings during rulemaking and becomes a required item for providers to collect. This slide shows the quality measure of domains that were included in the impact legislation, and they are functional status, cognitive function and changes in function and cognitive function, skin integrity and changes in skin integrity, medication reconciliation, incidents of major falls, and communicating the existence of and providing for the transfer of health information and care preferences. Now, of course today we are only going to be focusing on those items that measure the functional status, cognitive function, and changes therein.

It’s important to understand where quality measures come from. The truth is that they are years in the making and they must undergo scrutiny and testing to become a part of the CMS measure system. When the Affordable Care Act put into motion the idea of measuring value in the Medicare system, it had to have a vehicle to determine what those measures would be. After all Medicare is not in the business of doing research in measure development. That’s where the National Quality Forum comes in. NQF or the National Quality Forum is a not-for-profit, non-partisan, membership-based organization, focused on improvements in health care. They were created in 1999 by a coalition of public and private sector leaders based on a report that said an organization was needed to promote healthcare quality through measurement and public reporting. So think of them as kind of a neutral third party.

Think of them as a bank of measures. Measures can be developed by specialty societies or universities or companies that are invested in research. In fact, Medicare has hired contractors to develop measures for some of these impact areas and the contractor shepherds the measure through the process. Medicare may select a measure from this bank and decide to propose it to use in one or more settings. And that’s done in the rulemaking cycle annually. Medicare would define its process for selecting the measure, their rationale in selecting it, and how they intend to use it. We saw this in the 2016 rulemaking cycle where Medicare proposed to include four functional outcome measures for skilled nursing facilities. Stakeholders are allowed to submit comments for and against these measure proposals. Medicare then finalizes their decision in the final rule. Ours started collecting this information on October 1, 2017. A year later, these measures were finalized for the SNFs and the FY 18 final rule, which meant that skilled nursing facilities would start collecting them on 10/1/18.

After that measure is approved, then the technical specifications are posted. These are written by a contractor for Medicare. The specifications tell us what the co-variates are, how each is weighted, and where and if regression analysis is used. So those four measures are change in self-care for medical rehabilitation patients, change in mobility score for medical rehabilitation patients, discharged self-care score, and discharge mobility score. And again, the application of the functional outcome measure simply means that this measure was originally approved and endorsed by the National Quality Forum for inpatient rehab facilities. Medicare then is applying this same measure methodology to the other settings. And that’s why the term application of is used.

So, why the focus on function? Well, with over 150 of you being therapists or working in the therapy industry, you probably could answer that question very easily, but the reality is that the majority of patients admitted to an inpatient rehab facility or skilled nursing facility have functional limitations. And many of those patients and residents are at risk for further decline in function.

Just as importantly, functional status is important to residents, patients, and their family members. Functional outcomes, though, are different than clinical outcomes. Clinical outcomes are usually focused on clinical symptoms, such as what the patient’s breathing status is, whether their heart is in rhythm, whether there are signs of infection or delirium. Functional outcomes focus more on areas related to physical ability, such as mobility or areas related to vocation, such as returning to their roles at home and in the community. Moreover, functional outcomes and clinical outcomes don’t always parallel each other. A person’s disease process can resolve or stabilize while functional recovery remains stagnant. The opposite is true too: Functional recovery can surpass clinical recovery at least for awhile, especially as it relates to chronic progressive diseases.

So a resident’s functional status is related to or predictive of many things: healthcare spending, hospitalization of older adults living in the community, risk for nursing home placement, frailty, and even mortality. So in the skilled nursing facility and inpatient rehab facility, therapists assess and document patient residents’ functional status at admission and discharge. And this is a process that we have extensive education in. The results from our evaluations tell us what specific impairments there are. And we then use our clinical evidence base to determine what the most effective plan of care is. But then we must reassess at regular intervals and compare a patient’s change in function. This informs us of the effectiveness of the rehabilitation care provided to the individuals in a residence. And that process should help us get better at getting our patients better.

And as the second bullet tells us, let’s not be naive about this. Medicare is intending to look at the data. They intend to monitor provider outcomes. They will use that information to identify providers who achieve good results and those who achieve great outcomes. They will attempt to draw conclusions about provider types, patient characteristics that might be predictive of functional outcomes, and lots of other quality metrics. So let’s dive in, do the change in function measures. Now we’re going to discuss the change in self-care and change in mobility measures together since they are very similar. But first of all, question for you. How do you currently use the change in function measures in your practice? Do you use them to help determine length of stay, to determine appropriate discharge destination? Do you not use them or maybe you just don’t know how they’re currently being used?

So take a few minutes to respond to that question. There probably could be other ways to use the change in function measure, but these are some global areas. Certainly there are a lot of other areas that we could have listed underneath each one. But the main goal here is just to see how many providers, especially those on the call here today, are actually looking at these numbers and using it to improve your care processes. And that’s what we’re going to focus on more towards the end of the webinar.

Okay. So we’ve got close to 60 percent of the audience responding. So it’s really kind of even. We have about a quarter of you who are using a determined length of stay and a quarter of you are honest and saying, look, I don’t know how we use them. That might be something that you put on your to-do list for after the webinar, to get with your colleagues and team members and figure out a way to use them. About 33 percent are already using them to help determine discharge disposition. And then there are 17 percent who again, admit that hey, we aren’t using them, but again, you could have an action step after the webinar.

So the change in function measures are intended to capture risk-adjusted change in function from admission to discharge, for residents admitted with an expectation of functional improvement. So that’s a key piece. It’s not going to be used on everybody, only if a patient resident has an expectation of functional improvement. So that means that patients or residents with certain conditions would be excluded since they would not typically be expected to improve. So we’ll look at some of those.

First is a slide to remind us of what the section GG items are included in the change in self-care measure. Note the wash upper body item. You may not even be aware of that one unless you have practiced in a long-term care hospital. So all seven of these self-care items are included in the calculation of the change in self-care measure. These are some of the section GG 0170 items that are included in the change in mobility measure. The skilled nursing facility and IRF both collect the same ones. The LTCF of course collects fewer.

These are the remainder of the section GG 0170 items. Now a special note about wheeling. It is collected by both the IRF and the skilled nursing facility, however it is not currently included in the skilled nursing facility calculation of the measure at this time. Like last year, Medicare announced that they are going to start including it in the IRF calculation for those patients who do not walk. So that’s a change coming up that providers will want to take into consideration.

And this is the rating scale used by both the IRF and the skilled nursing facility for coding section GG. It’s imperative that all clinicians involved in coding the IRF PI and the SNF MDS really understand these definitions. These are the items used in scoring. If you didn’t assess a particular area, if the patient is assessed you want to take care to ensure you are choosing the right reason as the scores do say a lot about your assessment procedures. Did the resident really refuse to complete the activity, meaning that you were going to assess the patient in that area, but they refused to participate in the assessment. Not applicable means that they never did perform this activity in the period prior to the current illness, exacerbation, or injury. And therefore it’s not an appropriate area to address under your care. Not attempted due to environmental limitations, it’s pretty self-explanatory.

And then 88, not attempted due to medical condition or safety concerns. Again, you want to make sure that that is used judiciously, meaning that it wasn’t that it wasn’t that it was not attempted because a clinician or nurse didn’t have time to assess it, but that it was not assessed due to concerns about safety.

So what are the measure specifications for these two change-in-function measures? The patient population for which it is required is slightly different in each setting. The measure is only going to use data collected on traditional Medicare Part A, fee-for-service patients, while the IRF results are going to be calculated for all part A and C or Medicare Advantage patients. Now, you may remember that in last year’s skilled nursing facility proposed rule they proposed to begin collecting the quality measures on all patient populations, but in the final rule they declined to move forward with that. So it could happen in the future, but it’s not right now.

They use section GG items as we discussed and look for the coding down at admission and discharge collected over the first three days and the last three days. And then each type of provider submits the information through their own quality improvement and evaluation system or Qs system. So this is a representation of the calculation. The numerator is the risk-adjusted change in score between admission and discharge. And the denominator is the number of Med A-covered resident stays minus any exclusions. So what are the exclusions? They’re listed here. They’re excluded if the stay is incomplete, if the resident or patient is independent with all, meaning all, of the activities at the time of admission. If the following conditions would exclude a patient, and they’re all mashed to a list of ICD 10 codes.

So in the SNF that would be found in B 0-100 or the section I, and in the IRF, they would be found in items 44 D or the ICD 10 codes. If a patient’s less than 21 years old or discharged to hospice, they would also be excluded. Now since any patient admitted to an IRF must require rehab services, then all patients admitted to the IRF are used in the calculation. But since not all residents admitted to a SNF received skilled therapy, then the only patients included in the calculation are those that receive PT and or ST services.

I mentioned incomplete stay in the previous slide. How would Medicare determine that or know that? Well, they will know based on how the DC assessments are coded. So if a patient or resident discharges due to a medical emergency, leaves AMA, has a length of stay less than three days, or dies while in the SNF or IRF, then they would be considered an incomplete stay.

On the next two slides I’ve listed some of the risk adjusters for the change in self-care measure and the change in mobility measure. For the purposes of today’s webinar, they are displayed here in broader categories. Each one of these has specific assessment items associated with them or specific ICD 10 codes or specific claims data that is used to calculate the provider’s final score. Examples are co-morbidities using the hierarchical condition categories and primary medical condition category for which the patient is admitted, which comes from item I-20-B on the MDS and item 22 on the IRF PAI. Prior surgery comes from Section J, Prior functioning comes from GG 0110 and cognitive abilities is measured by staff assessment of cognitive function. So as you skim over this, hopefully you begin to appreciate that there’s a lot of information that goes into the risk adjustment, which means that there is a lot of information pulled from the MDS and PAI that are used in the adjustment.

That’s why it’s important to ensure that every section of the MDS and PAI are completed correctly, not just the ones that contribute to your payment.

So this slide shows the mobility risk adjusters. Many of them are the same, but there are a couple differences. The comparison to the admission mobility scores rather than self care. The fact that prior level of function is different. For mobility it uses indoor ambulation, stairs, and functional cognition where the self care uses prior self care and indoor ambulation only. And the mobility measure includes the history of one or more falls as a co-variant. Now in the third box on the first row are also some slight differences between the IRF and skilled nursing facility adjustment model. TBN use is included in the SNF while the IRF uses a modified diet and the patient’s body mass index. So now we’re going to look at how the measure is actually calculated, but from a high level.

So for each resident, the GDO 170 discharge assessment is scored. Now, one thing I want you to know here is that the scoring is done using the ratings that were entered into the respective assessments. To obtain the score you’re going to use the following. If the code is between 01 and 06, then the code is the score. But if the code is 07, 09, 10, or 88, then it is recoded to a zero one. And that is used as the score. If the item is dashed or missing then it is recoded to 01 as the score. Next, the GG 0170 admission assessment is scored and the admission score is subtracted from the discharge score. That total is then divided by the total number of patients and residents for the period. And that equals your facility observed change in mobility. Remember, that’s your raw score.

Then for each patient and resident the expected change score is calculated. In other words, based on all the characteristics represented in your assessment, what does the algorithm used by Medicare predict your change will be? And this is done on a patient-by-patient basis. And then again, divided by the total number of patients, residents in the period. Then the expected change scores subtracted from the change score in mobility. And that’s represented here. Now this exact same methodology is used for the change in self care measure, except for the fact that GG 0130 items are used and the risk adjustment is slightly different as I showed you. So you take the facility observed change and subtract it from the facility expected and you get a number. If it’s a negative number, then it means that your outcomes were worse than what was expected based on the resident’s characteristics. If it is a positive, then it’s better than what was expected. And if it is a zero, then it was equal to what was expected.

Then the provider’s difference value is added to the national average change in mobility or self-care score to obtain the provider’s risk-adjusted mean mobility or self-care change score. And the national average change in self-care score is the mean of the observed change in self-care or mobility for all patient stays using these same steps. So let’s look at an example: Four beneficiaries.

We have four beneficiaries admitted and discharged, and the observed and expected scores are calculated for each one. You add all the observed together, divide by four, add all the expected together and divide by four. So the mean observed change is 23.5 and the mean expected is 22.6, which is a 0.9 difference. Then that would be added to the national average change in self care score to get the facility’s final result. So if the national observed mean change was 0.5, then this provider’s final score would be a 1.4, which is great. But if the national observed change was 0.9, then that would bring this provider score back down to a zero, which means it was equal to what was expected.

Let’s move to the next chapter, which is the discharge functional measures. Now as with the change in function measures, these are designed to capture risk-adjusted discharge scores and function at discharge for patients or residents admitted with an expectation of functional improvement. It too was finalized to meet the impact act requirements. The same GG items for self care are used for this measure and the same mobility measures are used. But as I mentioned earlier, the IRF will begin including wheelchair items in their measure for those patients who cannot walk. The rating scale is the same, as is the rating scale for items not assessed. The measure specs for these differ slightly. The who is included as the same, but of course only the discharge items are used for the measure. So that means that the measurement of the patient’s function at discharge is the focus of the measure. And it is submitted the same way through the queue system.

The numerator of this measure is the number of Part A residents in a skilled nursing facility or Part A and C patients for the IRF with a DC score equal to or higher than the calculated expected DC score. And the denominator is the number of Part A residents in a skilled nursing facility or Part A and C patients in the IRF minus any exclusions.

So what are our exclusions? The same things that we indicated for the change measures, so it’s easy to remember. Incomplete stay, complete independence, presence of one or more of the conditions listed, less than 21 years old, discharged to hospice. And again, for the SNF, this measure is only going to be calculated for those residents who receive PT or OT services. The incomplete stay criteria are also the same.

The risk adjustment for both measures is the same as well. In fact, there are over 55 different clinical variables that are included in the risk adjustment model for all four of these measures and they are divided into seven domains: demographic, primary medical condition, functional status, cognition and communication, clinical conditions, services and treatments, and clinical diagnoses. So to take a look at this measure, we’re going to look closer at the self-care measure. So for each patient and resident the sum of GG 0130 discharge scores is tabulated to get the observed DC’s self-care score for any item that is coded as not assessed or has a dash in it, the item is going to receive a 01 for purposes of calculating the final score. Then for each patient and resident, Medicare calculates an expected discharge score for the patient using the intercept and regression coefficients we discussed earlier to get an expected discharge self-care score.

The expected DC self-care score is then subtracted from the observed DC self-care score. And essentially all the residents or patients with a score of greater than or equal to zero are put in one bucket. And all the patients and residents with a score less than zero are put into another bucket. Then each of these is then divided by the total number of patients or residents not meeting the exclusion criteria. And that is multiplied by a hundred. So the discharge mobility measure has done the same way. For each resident you determine, did they meet or exceed or did they not?

So let’s look at an example. We have our four beneficiaries again, and we calculate the observed and expected discharge self-care score. Patient one had an observed discharged self-care score at 32.4, but the expected was 33.1, which means patient one goes into the box with a lower than expected discharge score.

The same is done for patients two and three. The patient four had an observed discharged self-care score greater than the expected one. So that patient goes into the higher than expected box. Since the measure looks at the percentage of patients and residents who meet or exceed the expected amount, then we’re mostly concerned with the number of people who ended up in our bucket of greater than or equal to zero, which in this case was only one out of four. So this facility score for this period is going to be 25%, which essentially says, if a patient comes to this facility for rehab, then they have a 25% chance of meeting or exceeding their expected rehab outcomes. That’s not necessarily something that you want to be advertised.

Now let’s look at the report that shows these numbers. There are three that you want to be aware of. but we’re only going to focus on two today. So another poll question for you, how often do you pull your measures and review them as a team? Twice a month, once a month, once a quarter, once a year, or we don’t? And again, as a therapist, are you aware of what your facility does? The reason I’m asking this is because you may not. You probably aren’t the one to pull them, but do you have an opportunity to review them with your team on a regular basis? Are you aware of not just what you get in your physical or occupational or speech therapy department, but aware of what the facility’s scores are?

And of course, if you’re a manager, are you asking the MDS REI or the RFP PAI coordinator about what some of these reports are showing? So we’re going to wait for a few more people to respond. Definitely don’t be shy and don’t be afraid to say what the honest truth is. So these are new reports and that’s one of the reasons why we’re doing this webinar is to try to make more therapists aware of the types of reports that are out there and how they can get information. I’m also wanting to encourage people to be sure and submit questions in the box. So we want to be able to have them.
All right. So let’s look at the results. About a third of you are looking at them once a month, which is great. Another 30% about once a quarter, which is great. And a third of you say that you don’t. So again, that might be an action item for you following the webinar.

So let’s look at the three types of reports, the Review and Correct report on the left and the Provider Preview report on the right. Those are the ones that we’re going to be focused on today. The quality measure reports are certainly ones that you do want to be aware of and look at. But they are not the ones I’m going to focus on today. Now these are frequently referred to as CASPER reports. CASPER stands for Certification and Survey Provider Enhanced Reports. So the first one is the Review and Correct. And you can pull these whenever you want them. And they’re confidential to the facility. They are available the first day after a quarter ends, and they display quarterly data for the assessment-based quality measures only, not the claims-based one. So that’s why you want to look at these for the four we discussed today. And you can use them in conjunction with other CASPER reports to determine any errors that may affect your performance. They also display data correction deadlines, and whether the data correction period is open or closed. So if you make corrections to your data, then you can run it weekly to ensure that what you perceive the problem to be has been corrected. Providers are able to obtain aggregate performance for up to the past four full quarters as the data are available. And when a new reporting quarter begins, the oldest quarter is dropped. Now, keep in mind these Review and Correct reports are not risk-adjusted. They’re only the observed raw rates. So this is not the report that tells you if you met your annual percentage update threshold.

This shows the timelines for each of the four quarters. Data collection occurs on a calendar year basis, which is a little bit different from what we’re used to from the payment or fiscal year. Following the close of a quarter, you have approximately four and a half months to review your data and make any necessary corrections.

This is a summary slide showing the difference between this report and the compliance calculation that is done later and reported on the provider preview report. Now I’m going to walk through how these reports are retrieved, but keep in mind that generally therapists are not able to physically retrieve them from the queue system. That’s because Medicare only allows a provider to have a certain number of authorizations and to log in. Some people do use a proprietary product, however, that kind of hovers over your system and essentially calculates the data for you just as a queue system would do, but it does have its limitations due to the variable risk adjustment and national means that vary depending on the time period. But therapists, you should be asking your REI or MDS or PAI coordinators about the reports and looking at them together.

So this is a screenshot of the CMS queue system home page. You would then log in here. And this is the home page. And you can see a lot of reports and folders here for providers to utilize. The Review and Correct Reports are in the Reports tab that is circled in red. And the instructions from here differ just a little bit because the naming structure is different in the SNF than it is in the IRF. But essentially you look for your state, your IRF, and your facility ID. Of course, no one else’s should be in your folder. So that makes it a little bit easier. But if you are a regional administrator, you might have access to several facilities’ reports. On the nursing facility side, nursing home provider reports start with an LTC and hospital swing bed providers start with an SB.

And this is what you might see when you go in there. Again, these are blocked out to protect facility identity. You just download them and can save or print them, whatever you want to do. And then of course, hopefully you’re studying them. The provider preview report is the next report that I want to go over. And this one tells you whether or not you met the percentage compliance threshold for your annual payment update, or APU. These become available about five months after the end of each data collection quarter. And you only have 30 days to appeal the decision if you disagree. So, as we noted earlier, Q1 data for January 1 through March 31st must be submitted and any corrections made within four and a half months, which takes us to August 15th. Around September 1, you will receive your provider preview report for that quarter telling you if you met your threshold. You have 30 days from that date to make an appeal if you want to. And after 30 days has passed, it’s baked, it’s done and there’s no going back.

So this is an example of an IRF provider preview report. I know it’s rather blurry and I apologize about that but it is a screenshot. So hopefully in your handouts it might be a little bit easier to see. All the IRF information is at the top: name, address, type of ownership. And then this report shows you the results for the percent of residents with pressure ulcers that are new or worsened. So we know this is a report from last year, since that measure has been replaced. The measure ID is in the first cell, then the name of the measure, and the third cell over is the number of eligible patients discharged from your IRF. In other words, the number that qualified for the measure. In the fourth cell is the IRF’s risk-adjusted performance rate, which in this example is 3.1%.

And the last column is the US national rate, which is 0.8% at the time of the report. This is an example of what an IRF report for the four of five functional measures looks like: the first row is the admission and discharge functional assessment and a care plan that addresses function. The second row is the discharge self-care score. The second cell of the second row is the CMS measure ID. The third cell is the numerator. So for this period of 1/1/2018 to 12/31/18, 11 patients were included. That means 11 patients met or exceeded the expected discharge score for self care. The fourth cell at the row is the denominator, 19. So 19 patients did not mean did not meet exclusion criteria, meaning they had complete stays. That makes their facility percent 57.9% in the fifth cell. The last cell shows up the comparison group, 56.1. So for this IRF, they were performing better than the national average.

Now this slide is a little bit easier to see. This is a skilled nursing facility report. It looks very similar, just a little bit different at the top. Top left is the facility ID and on the right is the report period comparison group period, and the run date. The report also shows us percentage of residents who were patients with a pressure ulcer and the information is all the same here. So one patient was in the numerator 14 and the denominator. This report shows us the four functional outcome measures for skilled nursing facility. Column one is the measure name, column two is the report period. And then which for this report is 1/19 through 12/31/19. ID is in the column three, Medicare ID for the measure, and the ID discharge date is in column four.

The reason that there are two measure numbers here is because of the new fiscal year in the IRF. So the ID changes had changed from year over year. Column five shows us the average observed discharge score. And at the top, the observed score is a 24 and the expected is a 23.2. There were 18 persons then who met or exceeded the expected discharge score during this period out of a total of 31 complete stays. So the facility percentage was 54.5 and the national average was 52.8. So this facility did better than the national average. At the bottom of the slot is the change in mobility score. And if you move over to the fifth cell of that last row, there were 33 total patients. Their average observed admit score was 29.8. And their average observed DC was 40.6 giving an average raw change score of 10.8. When that is risk adjusted, it becomes a 21, which is greater than the national average of 19.2. So you might ask, well, how come your risk adjusted rate is so much higher? Well, it probably means they had a more frail or sick population. And so because of the adjustment, it actually makes their change look greater.

Now, these reports are important to monitor because failure to submit the data can cost your facility 2 percent of its annual revenue. IRFs must meet or exceed two thresholds. One for the quality measures set at 95 percent and the other for the infection control and other measures submitted via the national healthcare safety network. For the skilled nursing facilities, their data completeness threshold is still set at 80%. In other words, 80 percent of all items are required on every patient. They must be submitted in order to avoid the 2 percent penalty.

But it’s also also important because the data is going to be risk adjusted, meaning that your results will be out out there on IRF compare and national and nursing home compare. They’re going to be public. Providers and payers alike will be able to use it to make decisions about where to send their patients and, or their loved ones.

So lastly, let’s take a look at some of the competencies and processes you need to ensure great results. First, one more poll question. Is function discussed in section GG language during your weekly or daily meetings? So think of how often you use section GG. And are you talking in the rating scale language, in other words, supervision, touching, moderate partial assist, substantial maximal assist. Are you using their DP, the section GG dependent definition when you’re talking about a patient being dependent?

We’ve got about 40 percent of respondents. Go ahead and continue to submit your results. So at this point we have about 70 percent of people who are talking in that language, which is great. Okay. So if you’re not, then you do want to kind of ask yourself why you aren’t, right. This is the kind of language that we need to be speaking in so that we can ensure our documentation backs it up.

So think about who is involved in collecting your data. We’ve talked about the rating scale. We’ve talked about the importance of collecting the data accurately using guidelines Medicare has given us. But it still represents a challenge for us because we’re not used to collaborating very well. We know Medicare expects it to be interdisciplinary. Nurses and therapists should have been begun talking and collaborating on these for at least two years.

But if you know you aren’t doing this as effectively as you could have been, then now’s the time to start. Section GG matters for your payment in both the IRF and the SNF. So then following the rules for collecting the data becomes a real compliance risk. If there’s not documentation to support how the scores were determined, that becomes a problem. If only one discipline is scoring GG, that becomes a compliance problem too. It’s easy to incorporate the conversations about functional levels in your daily morning meetings or other huddles. Talk about what nursing is seeing in these functional levels and what therapy is seeing. Talk about why there are differences. Some differences would be expected, but there needs to be an agreement on what is the usual and what level best reflects the patient’s baseline abilities. It’s important to be aware of the definitions of each of the GG items, as well as the rating scale, especially the part about allowing the patient to perform the activity as independently as safely as possible.

Training is important. And repeating that training on a regular basis is even more important. I’ve met several therapists over the last year who were coding section GG incorrectly, and their facility was relying on them to know the definitions. I’ve done several reviews of medical records, where what is coded in GG makes absolutely no sense based on the documentation. I can’t stress enough how important this is to your outcomes and to proper payments. If you are still using different rating scales in your therapy documentation, ask yourself why. Could it be contributing to the problem? Is nursing using the GG language in their documentation, or are they still using extensive and limited assist? That could cause a problem since there isn’t a crosswalk between extensive assistance section GG, but it certainly makes a difference in your payment. So you want to close that gap.

We discussed how coding section GG items is important, but also coding other parts of the MDS. We discussed all the covariates used in the outcome measure calculations, and this data has to come from somewhere. So you want to make sure it’s accurate too. That’s why ensuring you have a complete picture of your patient’s medical history becomes important, and that you’re accurately reflecting that in your ICD 10 coding. So this slide has a few coding tips, remembering some of the unique attributes of the coding roles such as usual performance, not their best and not their worst.

One caution is to ensure you are not just taking the ratings from day one and transposing them into section GG. While it might likely be baseline it may not always be. And remember, the coding guidelines tell us that we are to assess the patient over the first three days. So you have to at least meet the intent of the regulation with assessing the patient.

You may find key differences across those three days. You also want to make sure everyone understands what the not-assessed code use. If you have a therapy evaluation lasting only 15 to 20 minutes, but every area has a performance code, is that likely or is it possible that perhaps the patient was just interviewed? Remember, GG is first and foremost an assessment and not a questionnaire.

Now, everyone is talking about clinical protocols these days, but what are they? The truth is there are very few black and white clear cut guidelines for caring for the types of patients seen in the IRF and the SNF, but it is important to have a way to monitor the outcomes along the way. So clinical protocols can help with this by attempting to standardize to some degree the care that’s delivered. And that’s the only way that we know what works and what doesn’t work, but how are you measuring that?

Are you reassessing Gigi at regular intervals? Are you estimating what a discharge score should be to ensure you’re getting there as quickly and as safely as possible? And are you using it to help you decide best location for the patient to transition to. While we don’t have published benchmarks for GG you could begin to integrate it into your conversations. Since we have to collect it, and since Medicare is going to be using it to measure us, you would probably be well-served to use it as a tool for you and your team. And lastly, truth is that people play a big role in what your outcomes are. Your patients trust that you have a high-level competent therapist and nurses that are knowledgeable about how to achieve functional outcomes in these settings. So do you, and how do you know? If you are displeased with your functional outcomes, it is a natural part of your SWOT analysis to look at your people.

What is your retention rate? Do they feel comfortable addressing all the areas in GG? How much time do they spend on each component? For instance, does OT spend more time with toileting than they do with putting on shoes and socks? Does PT regularly address car transfers or picking up objects from the floor? If not, why? These are all important areas to think about.

And last but not least, verify, look at your documentation. Make sure you have your interdisciplinary process outlined and that you are holding people accountable to it. Do you have your documentation? Is it clear how you arrived at your decisions to code GG one way or the other? Does nursing and therapy documentation both support those levels and how is it being collected at discharge? The same processes are just as important there too. And one last point here. As I hope you realize by now, the data that you put in column two of section GG on the MDS at admission has nothing to do with any of these four measures I’ve talked about today.

Note I’ve never mentioned it. That’s because it is the, it is only the admission and discharge scores that you code that you assess the patient to be that are used in the calculations. But if you are setting GG goals for every single item, then you better be sure to have them covered in your care plan somewhere. That’s where people could get into trouble with surveyors. If you’ve identified a functional area that you set a goal for, are you certain someone is working towards that goal? Remember the GAO came out with a report last year that said skilled nursing facilities don’t always address all the functional impairments identified in their assessments. So I’ll leave you with that thought and I’m going to turn it now to the question and answer session.

So let’s look at the first one.

Once rehab intervenes and starts the training and education process, I do not believe this is the resident’s usual performance afterwards. Your thoughts? Again, I think it depends on what area you’re talking about. So I do know of times where I might assess a patient late in the day and they were not appropriate to perhaps assess lower body dressing and putting on shoes and socks. Or from a PT perspective, they had just arrived and I could not do stairs or steps with them. So even though I might’ve started training and education on some aspects of mobility, I didn’t necessarily start on all. So I think that we have to be careful about trying to create bright lines around when things start and stop and think about it from a clinical perspective and a bigger picture standpoint.

So the next question is: If a facility codes in stylos, is that a problem for compliance? I would definitely think that it would be a risk for compliance. Again, there needs to be discussion. It’s very clear in the REI manual that this has to be an interdisciplinary process. And as I mentioned, I’ve done some audits with PDPM and and looked at some of the documentation and found some stark disagreement, even between the therapy evaluation and what was coded in GG in addition to nursing. So we really do need to kind of crack that process and make sure that we are following it so that we get the best result possible.

The next question is: Is the time period for calculating all of the scores one year? Yes it is. They look at the last four rolling quarters. So when a new quarter ends, then the earliest quarter falls off your report. So it’s not always a calendar year, it’s a rolling four quarters.

The next question is: If an IRF uses standardized outcome measures in addition to the QI (the quality indicators, which is what section GG is referred to as well), will Medicare take into account the results of set outcome measures capacity? I don’t know if they will or not. They have not published that they will be. So they are not collecting any of that other information anymore. So I do know that prior to last October, inpatient rehab facilities collected the FEM data. And so now they are using GG for their payment. Medicare hasn’t said if they’re going to use the FEM data in any capacity or not. So I can’t answer that for sure. It doesn’t look like they will at this point, but of course, that could change in an instant.

Another question: For FY 2021, rules for wheelchair items to be added to the discharge and change calculations. Oh, this is a long one. I don’t want to read it out loud. Okay. So she’s asking about the fact that the wheelchair items to be added to the discharge and change calculations to the question walk 10 feet. What if the answers are different between admission and discharge? I’m assuming that’s what this question is getting at.

The specifications on that have not been published yet, so it’s unclear how they would treat answers that are different at beginning to end. You know, if a person doesn’t walk, then technically walk 10 feet should be coded as 09, not applicable, versus someone for whom a goal is written. Then it might be 88 or someone who is a walking candidate. Then it would be scored 88. So that is likely to be how they are going to differentiate between when to look at the wheelchair scores and not, but again we will wait for that for the final specifications.

Debra asked, How was the expected discharge score determined? I hope I answered that in the webinar. There is a very complicated regression analysis that’s done looking at all of the different covariates that I showed on the slide and they calculate that based on your patient characteristics. So that’s why it’s so important that all sections of the MDS be filled out in its entirety.

And then last question is, Have you seen this being used at all in the acute care setting? What is the benefit of using outcome measures in acute care? These particular items were tested during the PAC PRD demonstration in the acute care setting, but it was determined not to add them at that time. So certainly I think it’s a benefit to use outcome measures in acute care in a general sense, just so that you can evaluate your care processes and make sure that you’re doing everything you can while someone is in acute care. I realize that acute care therapists have special special challenges due to short lengths of stay and sometimes very sick patients. But there can be and should be some measures of functional outcome in that setting, but GG has not been discussed to my knowledge in recent years for insertion into the acute care model.