BHC-Lead Behavioral Health Clinic Measures – Part 2 of 2

Peggy O’Brien: Good afternoon, and welcome to the 5th Webinar in the Behavioral Health Clinic Data Collection and Quality
Reporting Webinar series, presented by the Substance Abuse and Mental Health Services Administration. Today’s topic is BHC-Lead Behavioral Health Clinic Measures
– Part 2 of 2. I’m Peggy O’Brien, a Senior Research Leader
at Truven Health Analytics, presenting on behalf of SAMHSA. There is a chat function where you can ask questions, and we encourage you to do so. I will pause at several points
to respond to questions that people have about things I’ve covered that may be confusing or about related matters that I have not addressed. Most webinars will have at least one poll question, which listeners will have time to answer and which we can discuss. A PDF of the slides for each webinar is posted as a
resource on the webinar site, and the webinars themselves will be available on demand for a year after the event on the webinar site where you are listening to this presentation. All you need to do to access the on-demand webinar on this webinar site is to register. So if you know someone who has not done so who would benefit from
accessing this webinar or earlier ones, encourage them to register. The webinar also will be posted on the SAMHSA website, probably within about a week or two at the most after it is presented. This is the webinar schedule with the intended audience identified in red. All are on Tuesday
from 2:00 to 3:30 Eastern Time. The next webinar is the first of the two Special Issues webinars covering general subjects related to data collection, analysis and reporting for these BHC quality measures. I will begin today by answering one logistical question that was raised
in the last webinar. As I mentioned, I also will pause for questions several times during the webinar, so feel free to ask questions by chat. Like last week, I may cut off questions at a certain point to try to get through all of the material, because we have a lot to cover
today, too. The main focus today, however, will be on the remaining five BHC-lead measures. So the one question or comment from last week that I want to address right now is that someone had difficulty with the links on the SAMHSA website to the specs and templates. And if you are having this
problem, there are two links, one for the specs, one for the data reporting templates. For the templates you click on the link on the SAMHSA Section 223 Quality Measures website and you’ll get a prompt to open or save the templates. If you click on open they will open. It’s an Excel document. You can save it for your use. For the
measure specs you click on the link and you will get a notice of copyright that you must accept. Once you accept it you will get a small message, likely at the bottom of your screen, as to whether you want to open or save. The main screen will look as though the acceptance button didn’t work and you will try to accept again, which you do not need to do. Just
look for the small message about open or save. If you select open, a ZIP folder will open with both volumes of the specs inside. You can extract them and save them if you wish. Additionally, someone indicated that because the specs are a PDF they cannot access the links. Links do work in
a PDF, and so I went to the specs on the 223 website, downloaded the specs, and was able to get the links to work. I will cover the final five BHC-lead measures today. The measures for discussion this week are listed on this slide. Some will be covered in greater
detail than others. This slide provides the age coverage and stratification requirements for the five measures we’re discussing today. As you can see, a variety of different ages are covered, 12 and older, 3 to 17, 6 to 17, and two of them cover 18 and older. All of these are stratified
by payer, and the first two are also stratified by age. The first measure I will cover is screening for clinical depression and follow-up plan. It begins on Page 91 of the specs, if anyone is interested. I will go through this measure in some detail. The denominator is consumers who
are 12 and older and seen during the measurement year at the BHC. Remember, the measurement year will be for the CCBHCs the demonstration year. The numerator is a subset of the denominator-eligible consumers who are screened for clinical depression using a standardized tool, and if positive for whom a
follow-up plan is documented on the same date as the positive screen. So if they’re screened and found to have no depression they’re counted in the numerator, and if they’re screened and found to have depression they’re only counted in the numerator if there is also a follow-up plan documented on the same day. Those two parts make up the
numerator. The measurement period for the numerator and the denominator is the measurement year only. Also as a reminder, the measurement period is the period of time covered by the data used. In this case, the measurement year and measurement period are identical. The measure is described
in Section A and is identified as a hybrid measure. It is stratified by age, 12 to 17, 18 to 64 and 65+, and by payer, with payer status determined on the date of the visit. Options for approach to the hybrid data and sampling requirements are discussed
in this section, and I will go over those in more detail on the next slide. This measure requires multiple approaches to data collection, which is the nature of a hybrid measure. Some parts of this are optional. To simplify it and provide a preview of the actual spec, I have boiled it
down to the denominator uses outpatient billing codes to identify the eligible population. The denominator exclusions use administrative data and medical records, that is, billing codes and medical record entries. And the numerator uses codes, or you have the option of sampling and doing a medical record review. I’ll
get into more detail in a little while, but you should indicate in Section E or F of the data reporting template if you sample. If you decide to sample you should review information in the introductory material to the spec manual related to sampling and hybrid measures. We also will discuss sampling in
greater detail in the Special Issues webinar next week. Section A goes on to identify that there are code changes from the source measure. These were designed to simplify the measure. Those changes, however, are actually also reflected in the Medicaid core measure. So if you were already calculating the Medicaid
core measure, this should be the same. It’s very important the encounter used to get the person into the eligible population and the screening and follow-up that gets them into the numerator must be on the same day, and if the person is seen multiple times during the measurement year, only the most recent such encounter
is used. Each person is only counted once. Also in Section A of the spec there’s a link to the source measure to access codes you will need. You’re encouraged to use all paid, suspended, pending and denied claims to the extent possible when you rely on billing data.
Section A also, as in all the measures, refers to the templates and appendices and provides the measurement period. The definitions include follow-up plan, which is defined as the proposed plan of treatment which must be directly linked to a positive depression screening and must contain at
least one of the listed components. The components are identified in the definitions. The provider entity is the BHC. This indicates the measure is aggregated at the BHC level, as are all the BHC measures. Screening is defined as a clinical or diagnostic tool used to identify people at
risk of developing or having a certain disease or condition, even in the absence of symptoms. Standardized tool is defined as one that is normalized and validated, age-appropriate. The name of the tool must be documented in the record. And examples of screening tools are included,
including for geriatric and cognitively impaired populations. This measure does not say the tool must be one of those that are listed, just that the tool used must be standardized and age-appropriate. Identifying the eligible population is a two-step process. First,
you use the outpatient encounter codes that are listed in the source measure. These are CPT or HCPCS codes. This tells you that the person was seen for one of those types of visits. From those people, you then identify those age 12 and older on the date that that eligible
encounter. Hearkening back to when I first started talking about this as a hybrid measure, this part of the eligible population, which is the precursor to the denominator, uses administrative data in the form of billing or encounter codes and demographic data for age,
which also is administrative. All of this can be pulled out of your EHR and/or billing data. Then Section D of the spec provides you with information on how to calculate the denominator, which, with the exception of exclusions, is the eligible population that I just explained
from Section C, and the numerator, which is, if you recall, screening and same-day follow-up. For the numerator you must identify first the number of consumers in the eligible population who were screened for clinical depression using a standardized tool, and, second, if they’re positive, who have a
follow-up plan documented on the date of the positive screen using one of the codes in Table CDF-A in the appendix in Volume 2 of the specs. When we discussed the hybrid nature of this measure I noted that the numerator uses codes, or you have the option of sampling and then doing a medical
record review for the numerator. So you can code it using the HCPCS G codes that are in the appendix and compute it electronically using the entire eligible population, or you can select an appropriate sample for the denominator, and for that sample then go through their medical records to ascertain screening and
follow-up for the numerator. I’ll talk a little bit about the codes in a minute. If you do use the hybrid sampling approach, you should indicate in the data reporting template that you’ve done so. You can see on this screen at the very bottom an asterisk referring to a hybrid flow document.
I’ll elaborate on that in a few slides. Finally, Section D includes the exclusions to the measure. The standard in writing measure specifications is to place exclusions below the numerator. Ideally, however, it would be better placed with the denominator, because they usually are exclusions
from the denominator and therefore also from the numerator. In any event, as I mentioned at the beginning of this measure discussion today, the denominator exclusions use administrative data and medical records, that is billing codes and medical record entries. The exclusions are active depression or bipolar
disorder diagnoses, refusal to participate, an urgent or emergent situation where delaying treatment would jeopardize health, or where functional capacity or motivation may impact results of the screen, for instance, some court-appointed cases or cases of delirium. For these exclusions what
is important is that they are documented. You will find in the appendix the codes you need to identify exclusions, including the ICD-10 codes that satisfy bipolar disorder and depression in this measure, and the two codes indicating that the patient was not appropriate for the screening or
follow-up One final note from Section E, a higher score on this measure is associated with better quality, as more screening and follow-up planning is desirable. This is a screenshot of Appendix CDF-BH.B from Volume 2 of the specs. As you can see, Table
A includes the codes indicating whether a clinical depression screening occurred, and Table B includes codes for exclusions, where a person was not eligible or appropriate for screening and follow-up. The latter includes the ICD-10 diagnosis codes and the
two HCPCS codes for where the person is not appropriate for screening and follow-up. You can use the Table B HCPCS codes to capture those who refuse the screening or follow-up plan development or who were in an urgent or emergent situation that precluded it or where functional capacity or motivation to
improve made them not appropriate. If the codes are not used, you will be stuck with having to do a sample and record review. In other words, you will need to review medical records unless you find a way to automate it using these codes. We have prepared a flowchart for how you follow this hybrid
measure, which is attached as a resource to this webinar, and it is designed to help guide you through the hybrid process for this measure. And as a last note here, Appendix CDF-BH-A in Volume 2 contains a sample calculation. These examples are
designed to help you visualize the measure calculation. Most of the process should actually be automated. For this measure of screening for clinical depression and follow-up plan, the encounter used for the denominator and screening
and follow-up planning for the numerator are all required to be on the same day, and the BHC can only count the most recent encounter during the measurement year. So what does this suggest for frequency screening? As I interpret this, if there is no
applicable exclusion such as existing depression or bipolar disorder or urgent or emergent, refusal to participate or functional capacity limitations, you will be conducting a formal depression screen on each visit to assure that it is done on the most recent
encounter. And how will BHCs ensure this is integrated into each visit? There are a few ideas for this on the next slide. Here are a few ideas for how to integrate routine formal depression screening and follow-up planning when it’s called for into consumer visits. It could be built
into the EHR if the EHR is used in the room with the consumer and if that seems comfortable with the consumer. Otherwise it could be built in as a reminder to the clinician before they enter the room and meet with the client. Also, it’s important to provide clinicians with age and developmentally appropriate screening tools so they
are available when they are needed. And for purposes of documentation beyond the use of the EHR, consider providing clinicians with the tool with the codes for screening, follow-up planning and rationales for not screening or follow-up planning to make sure it’s
properly coded. There are probably many other possibilities, as well. So I have a slide here that is the first poll question. Probably should be called rhetorical poll question, but I really do want to know what you
think. Is it easier to, Option 1, train providers to use all appropriate codes so you can use the EHR to gather the data, or, Option 2, sample 411 representative medical records to determine out of all eligible outpatient visits what percentage had formal screening for clinical
depression on the most recent visit, and if positive and there are no relevant exclusions, if a documented follow-up plan was developed on the same day. So I’m going to wait, and you can pick Option 1 or Option
2. And I want to see what the results are in a minute. Okay, let’s see what happens here. All right. That’s interesting. Eighty-three percent would train providers to use all
appropriate codes so you can use the EHR to gather the data, and 17 percent think it would be easier to sample the medical records and do a medical records review. Okay. I would love it if people could put comments in the chat box,
especially that selected the second option, as to what you see as the major impediments to using the codes and the EHR to gather the data. And I’ll – we’ll get back to those in a minute. I’m going to
move on, though. And actually what I’m moving on to is an opportunity for questions or comments. So, Amy, do we have any questions so far? Amy: Yes, the first question is from Maria Fernandez. The question refers to the definition
of active diagnosis for depression. What if you don’t know when the diagnosis was made or what depression screening tool was used? Peggy O’Brien: Okay. The term “active diagnosis of depression” or bipolar disorder is used as an
exclusion. So that means that the person has an existing diagnosis. That means that they presumably still satisfy the criteria for the diagnosis, and therefore you do not count them in either the numerator or the
denominator, because they’re already diagnosed. You don’t need to screen them fresh. So it really doesn’t matter what tool was used to diagnose them in the past. It’s just a way to eliminate those people, because you really only want to be counting
people who don’t have a diagnosis at this point. Amy: The next question is from Laurie Richardson. If the consumer is receiving therapy, how can they use the codes in conjunction with the therapy codes for screening and planning? Peggy O’Brien:
That’s a good question, and in a way it’s a technical question that I’m not equipped to answer. I do think that – if you have problems with the screening codes, the G codes, it would be helpful, I think, to
see if there is a way to incorporate them into your EHR, not necessarily as billing codes, but as codes that are designed for information purposes so that the EHR can gather the information and
calculate the measure using those codes. Amy: The next question is from Tracy Lieber. If we incorporate codes, can you use sampling from the baseline year since the codes are not currently being used? Peggy O’Brien: That’s one that I want
to think about. I don’t want to answer it off the cuff. I want to think about it and look at the measure in detail and respond to you separately. Amy: The next question is from Margaret Morris. Is sampling used? If so, is there a defined
methodology? Peggy O’Brien: Sampling can be used, and so clearly you have a choice of doing a sample and simply using the entire eligible population if there are codes that are used that allow you to capture it in your electronic health
record. None of these measures, the hybrid measures that rely on sampling or have the option of sampling, none of the source measures provide a definite methodology for the sampling. And so what I am going to do in the webinar next week is talk about sampling
for hybrid measures. That will give you some indication of what is available in the way of methodology. So I will be talking about that next week. Amy: The next question is from Donna Studlack. Just to clarify, if a client comes
in three times in a week, for example, once for an injection, another to see a therapist and a third to see your doctor, do we have to screen for depression three times? Peggy O’Brien: No. No. You use the most recent eligible encounter.
So let’s assume that all three of those encounters are eligible – well, maybe the answer is not yes. If all of those encounters or visits are eligible encounters and you don’t know that they are coming back, then theoretically you would want
to do screening each time, because you only use the most recent visit. Now, if any of those particular visits do not fall into what are considered eligible encounters, which are identified in the source codes, then you don’t have to do it during that
encounter. Amy: The next question is from Malik Manad. In reference to Slide 22, which talks about clinicians, who is considered a clinician? Is a clinician a doctor, a nurse practitioner, a PA, or does it include therapists and social workers? Peggy
O’Brien: There is nothing specific in this measure about the type of clinician or provider that can do this. And so what you’re left with is two things. You’re left with what are your state laws, your licensing laws in terms of who can do what? Are there any
restrictions? If there are restrictions then obviously that person can’t do it. The other thing is if you’re looking at the encounter codes that are used to define the eligible population, then it’s very possible that some of those encounter codes may be restricted in terms of
the types of providers that can code those. So that is the – those are the only two restrictions. There is nothing in this measure that says this has to be done by a certain kind of provider. Amy: The next question is by Ginger Bandine. Would you need to keep screening
at each visit even if the screening was positive already and/or it has already resulted in a diagnosis of depression? Peggy O’Brien: No. If the person is found to be depressed on a screen, they should get a diagnosis that reflects that, and they then become excluded
and they aren’t counted after that. It’s only up until the point that somebody gets screened and found to have depression. Amy: The next question is by Cathy Beveridge. GAO 431 is a
physical healthcare code which is not allowed by our behavioral health funders. Is the expectation that we use this code and just accept denials? Peggy O’Brien: That’s a good question
that I would need to look into, and I will look into it. And, Amy, I’m going to stop the questions at this point and move on so we can get through everything. The next measure we’re discussing today is the weight and BMI assessment
for children and adolescents, which begins on Page 50 of the specs. And there has been a change that I will point out later in the discussion about this measure. For the present I’m just going to go through the measure as it is discussed here on
the slides. The eligible population or the denominator are those that are ages 3 to 17 who have an outpatient visit with a PCP or OB/GYN during the measurement year. The numerator is those in the eligible population with evidence of BMI percentile documentation. This is designed to assure
assessment of BMI at least once a year, and the measurement period for both the numerator and the denominator is the measurement year, although for the hybrid spec you can look back to the prior year to determine if height, weight and BMI were documented. This measure
focuses on whether BMI is assessed rather than the BMI value, because norms vary by age and gender for children and adolescents. The measure is designed to be calculated using either administrative billing data or as a hybrid measure, and
it’s specified both ways. It is stratified by age, 3 to 11, 12 to 17, and by payer, and this version of the measure does not incorporate requirements related to physical activity and nutrition counseling. The eligible population must meet the continuous enrollment requirement
for Medicaid or for those dually eligible under Medicaid and Medicare, or they will be categorized as Other for purposes of stratification by payer. Numerator documentation must include a BMI percentile to count as satisfying the numerator. And if you use an EHR to gather any data it’s
requested you so indicate in the additional notes section on the data reporting template. Height, weight and BMI must be from the same source. Height and weight must be taken and documented during the measurement year. However, if you’re using the hybrid spec, documentation in
the medical record must include the height, weight and BMI value, but it can be dated either in the measurement year or the prior year. A sample can be used, and I’ll go into this in a little bit. But you should be sure to review the front matter of the specs and Appendix C
in Volume 2 of the specs. And, as I mentioned earlier, we’ll talk about sampling generally next week. There’s a link to the HEDIS value sets where you’ll find the applicable codes. Be sure to include all paid, suspended, pending and denied claims to the extent you can
if you’re using billing data, billing codes. And Appendix D includes definitions of PCP and OB/GYN that are eligible to include. Among the definitions that are included in this are BMI and BMI percentile. And I just want to stop
and mention that you will see a note at the bottom of Section A that mentions hypertension diagnosis. You should ignore this. It doesn’t belong there. Section C provides information on the eligible population, including age and
insurance enrollment requirements. The enrollment requirements are designed to help you determine if someone is Medicaid dually eligible or other by payer status for purposes of ratification. These are not commonly used in the BHC-lead measures, so I’m going to reiterate what these mean,
because we talked about it briefly in the very first webinar. Continuous enrollment is the period that they have to be enrolled for, and in this case it’s the measurement year. An allowable gap permits one gap of up to 45 days. However, if enrollment is verified monthly a one-month gap is all that’s allowed.
The anchor date is the last day of the measurement year, which means the person must be enrolled in Medicaid or as a dual on that day or they are treated as Other regardless of status the rest of the year. The anchor date is the one day that the allowable gap cannot encompass. So to
calculate the eligible population you must first make sure they are seen at the BHC once during the measurement year; second, make sure they are 3 to 17 as of the end of the measurement year; and third, from those, identify those who had an outpatient visit with a PCP or OB/GYN
during the measurement year. I’m going to stop here and explain how this has changed. It’s – we got additional information about this measure today, when it was too late to change the slides. So there will be something coming out in writing about it
probably in the questions and clarifications. We recognize that requiring that the BMI be done by a PCP or an OB/GYN is going to make it much more difficult for you to assure that it happens and to access the data. In some cases it may
be simple if you have somebody within the CCBHC who’s serving as a PCP or if you have a DCO that is associated with you that is an FQHC or something like that. But in general it’s going to be more difficult. The requirement for
PCP or OB/GYN screening comes from the source measure, and we’re not at liberty to change that measure. However, there is a way in the data reporting template for you to indicate deviations from the specification.
And it is possible that you might elect to have someone within the CCBHC who’s not a PCP or OB/GYN do the BMI screening, or your DCO might have somebody do it who’s not one of those providers, in which case you can go ahead and
do it that way and just make sure that you so indicate in the data reporting template. Okay? So the references you see in these slides about OB/GYNs and PCPs are what the measure calls for, but you may
find that it makes more sense for you to do it this way. Okay, so moving on, as I mentioned before you have the option of using an administrative spec or a hybrid spec for this measure. This slide focuses on
the administrative spec. The denominator is the eligible population we just discussed, seen at the BHC, age and some encounter, which could be at the BHC. It doesn’t have to be a PCP or OB/GYN. The numerator requires that the person was coded as having the height, weight and
BMI percentile documented. The codes are in the HEDIS value set, and there is a link to that in Section A of the spec. The exclusion is optional and involves consumers who have a diagnosis of pregnancy during the measurement year. Again,
the codes for that are in the HEDIS value set, and the HEDIS value set is linked in Section A. Section E contains the alternative hybrid spec should you elect to use it instead of the administrative spec. For the denominator, you
will take a systematic sample of the eligible population stratified by age and separately by payer. You’ll use the sample of 411 unless the BHC has a smaller population or there are special circumstances. An oversample is recommended in case you find out after the fact some were not
eligible. Appendix C in Volume 2 provides more information on using a reduced sample size, and I will discuss this more next week. For the numerator you can use administrative data or medical records. If you use administrative or billing data you use the administrative specification for the numerator, which
we just went over. Otherwise you use the hybrid sample and their medical records, which I discuss on the next slide. There is also an asterisk here regarding the hybrid flow document for this measure. And this one is not attached to the webinar as a resource like the
one for screening for clinical depression is today. So you can access the hybrid flow measure, the flowchart for the screening and clinical depression as a resource on the webinar website where you’re listening to this. You cannot access
it yet for the child and adolescent’s body mass index measure, because we haven’t finalized it. We were waiting to see if we could get rid of the OB/GYN and PCP requirement. So once it’s finalized it
will be posted. It should be in the resources for the webinar next week. For the numerator for the weight assessment BMI measure for children and adolescents, if you use medical
records you must make sure the record includes height, weight and BMI percentile for the same data source and that it’s in the measurement year or the year before. And it has to be the percentile or percentile plotted on an age-growth chart. If you rely on the medical
record for evidence of the pregnancy exclusion there must be a note indicating a diagnosis of pregnancy during the measurement year. Otherwise the record or EHR must include the appropriate pregnancy diagnosis codes. Additional notes for this measure include that you need documentation of all
the components of BMI, including height, weight and BMI percentile. Better quality is associated with a higher score, so more consistent documentation of BMI means more consistent assessment for this measure of relative height and weight, and it’s critical to monitoring
the health of children and adolescents. And I’m going to briefly touch on the data reporting template just to go over the part where you identify the data source for this measure, because we’re doing a hybrid measure. Actually we’re doing a measure that allows you a choice between
administrative and hybrid. That’s why I picked it. So in Section B of the data reporting template, that allows you to identify the data source. So in Row 7 for this measure you have an option of indicating administrative, hybrid or other data source. So if you pick administrative in Row 7, in Row 8 you
select MMIS or other, and if other you specify. So that’s for the administrative spec. If you pick hybrid in Row 7, in Row 9 you select the administrative source, which again is MMIS or other. If you pick hybrid, then in Row 10 you also have to select a medical record source,
so it could be EHR or paper or both. And if you do pick hybrid in Row 7, in Row 11 you have to indicate your sample size, in Row 12 you input the measure-eligible population, and in Row 13 you specify the source if you picked other than hybrid or administrative.
Because this measure provides you with a choice of administrative or hybrid there should be no reason to select other. Okay, so we have another poll question, and actually I’m going to
skip this one because I resolved it. Okay, so I’m going to stop again and see what kind of questions there are about the WCC measure for BMI for children and adolescents. Amy:
The first question is by Margaret Morris. If BMI percentile is present, why are height and weight required? Peggy O’Brien: I think that that’s their way of making sure that whoever calculated the percentile actually captured the height and
weight and did it accurately. I don’t really know. It’s a good question, but I don’t know the answer to it. It’s what the source measure requires, so that’s what this measure requires. Amy: The next question is from Nan
Guenther. Are psychiatric visits 99213, 99214 and 99215 going to qualify as PCP visits? As with PQRS, the plurality of visits are delivered by the psychiatrist. If not, how will PCP visits be apparent if it is
not delivered at the CCBHC? Peggy O’Brien: Okay, so we took care of that by giving you the option of not using a PCP or OB/GYN and just indicating in the data reporting template that it is a deviation
from the specification. Amy: The next question is by Ginger Bandine. Since the only options are administrative or hybrid, which includes administrative, it sounds like you’re saying that if we don’t have access to the administrative
data we can’t use our EHR for this data. Peggy O’Brien: You can use your EHR for this data, because your EHR would have the billing codes that are – or hopefully would have the billing codes that are going to show up in the administrative
data. Amy: There are no more questions, Peggy. Peggy O’Brien: Thank you. The next measure is the child and adolescent measure that looks at whether, out of all the visits during the measurement year by consumers in the pertinent
age group with a diagnosis of major depressive disorder, there is an assessment of suicide risk at the visit. This measure begins on Page 74 of the specs. The adult measure is the next one I will cover. I won’t go over it in any detail, as it is very similar to this child measure. For this measure the age group
is 6 to 17, and for adults it is 18 and older. And the measurement period for both the denominator and numerator is the measurement year for both measures. This measure is specified as collected from an EHR, and stratification is by payer, which is determined
on the date of the visit that’s being counted. The measure is based on the number of visits rather than the number of consumers. So one person could be counted any number of times during the measurement year. And more detail about this
measure is available at the ECQM Library and Resource Center, and links and access information are available in the spec in Section A. Value sets for this measure are found at the U.S. National Library of Medicine Value Set Authority Center, or VSAC. And there is a link to that
site in the spec. When you open the link you will see that a license is required to use the value set. You should be sure to use the value set that corresponds to the version of the measure. And we did receive a question about these value sets for both this and the adult measure. The question was for
the two suicide risk assessment measures do you have copies of the value sets? Apparently the Value Set Authority Center may have this input, but access to that information requires a license. The response to that is that the VSAC is part of the National Institutes of Health and the VSAC license is free. The
link is in the measures and is easy to access. As I mentioned on the last slide, the specs also refer to the CMS website for more information about the specs. This last information can be helpful as a supplement to the BHC specs, but generally what the user needs is to obtain the free VSAC
license and the value sets from the VSAC site and to use those in conjunction with the BHC specs. I also should note that Appendix SRA-BH-C in Volume 2 of the specs contains the eMeasure flows for this measure, which may be very helpful in
understanding it. So, skipping to Section C, the steps to calculate the eligible population are as follows. One, were they seen at the BHC and were they age 6 to 17 at the beginning of the measurement year? Two, from those, did they have at least two qualifying visits
during the measurement year? And three, from those, did they have an active diagnosis of major depressive disorder at the time of the encounter being measured? Section D provides the logic to calculate that eligible population using the
EHR. The EHR denominator logic includes language for age, certain types of outpatient visits – you’ll need two of those, and an active major depressive disorder diagnosis. So if a consumer comes into the clinic who’s 12 years old at the beginning
of the measurement year and on their first visit they do not have a diagnosis of depression but they are given this diagnosis on the second visit and they keep that diagnosis on the third through tenth visits, they have nine out of 10 visits that qualify, and each such visit is counted
separately in the denominator, as well as for the numerator, because this measure looks at visits rather than consumers, per se. The numerator section of the spec for the suicide risk assessment measure includes information about the type and magnitude of the risk assessment
being at the discretion of the clinician to meet the needs of the consumer. You will see later, though, that it does need to be a standardized tool. The numerator guidance requires the assessment to be performed at each visit when there is a major depressive disorder diagnosis. If it’s an episode
of care measure, it requires two initial visits to establish a relationship. Once that is clear, each visit with a major depressive disorder diagnosis counts as one where there should be a suicide risk assessment. And the suicide risk assessment must use a standardized
tool. The numerator logic provides language for the risk assessment and encounter criteria, and there are no exclusions. As is often the case, a higher rate on this measure is associated with higher quality. Consistent use of a suicide risk assessment when
a client has major depressive disorder is the objective that the measure seeks to further. Listed in Section E are the data elements, which are really the value sets for the measure that are available in the value set authority center. And the appendix in
Volume 2 provides the eCQM Flow reporting the measure. Some may find this more intuitive than the logic in the spec itself. So there were some questions about this measure. One was does this assessment need to be performed at every visit for every adolescent diagnosed
with major depressive disorder. And in response to that, yes, it does. The idea is that the child or adolescent has major depressive disorder and at each visit he or she is assessed for suicide risk. There was also a question asked that said for the
following measures can the state use MMIS administrative data rather than the data sources called for in the BHC quality measure specifications? And the person asking the question mentioned the clinical depression, which is administrative or hybrid, the child
or adolescent body mass index, which is hybrid, and the child or adolescent major depressive disorder suicide risk assessment, which uses an EHR. So, as I interpret it, this question’s really asking two things. First, can the state report some BHC-lead measures instead of requiring
the BHC to do it? The response to this is that when creating the quality measure tables and tech specs SAMHSA divided the measures into two groups based on who they thought would have the information and on the level of effort they felt it would take for them to report it.
However, the state is permitted to report data on behalf of CCBHCs, including data for the clinic or BHC-lead measures. Secondly, I think this question is asking if the state in those circumstances can use its reporting methodology for the Medicaid core set but distinguished by
BHC. The response to that is that states or BHCs should source data from the EHR or utilize hybrid data from the medical chart and claims as required by the respective BHC specs. If this is not possible, then administrative data derived from the claims may be used. If a state is not
able to report a measure as shown in the technical specifications, then it should provide a detailed plan in its demonstration application that outlines how it will move towards reporting the measure as specified.
Okay, I’m going to stop and see if there are any questions at this point. Amy: The first question is from Ashley Fir. Is there a reason why children need two
qualifying visits whereas adults only need one? Peggy O’Brien: I do not know what the source measure creators were thinking in terms of treating one population different than the other. However, the rationale is
provided in the child measure that you need to make sure that there is a relationship established, and that’s why two visits are required for the child measure. Amy: There is a comment by Ed Silva. Michigan Public Mental Health does not allow for
the G codes in encounter data. Peggy O’Brien: Okay. All right. So, again, that’s another question or actually comment that needs to be taken under advisement. But thank you for making it. Amy: The next question
is from Scott Girds. Do these measures align with the 2016 PQRS quality measures? Peggy O’Brien: I believe that these measure are Medicaid core measure and they align
with those. Amy: Those are the only questions, Peggy. Peggy O’Brien: Okay. All right. So the next measure is the adult major depressive disorder suicide risk assessment measure.
Begins on Page 82 of the specs. It’s the adult version of the child measure. And I won’t go through this in detail, just address key points that differ from the child measure. For this measure, unlike the child measure
that’s only specified as EHR, you’ll see specifications for both EHR and for medical or administrative data, so you have a choice. This measure looks at whether the person has a new diagnosis or a newly identified recurrent episode of major depressive disorder. So each time there is
a new diagnosis or newly identified recurrent episode during the measurement year it’s included as a separate count in the denominator. The flow for the eMeasure is in the Appendix SRA-A.A, and a calculation example for the non-eMeasure is in Appendix
SRA-A.B. Those are both in Volume 2 of the specs. The last of these measures is depression remission at 12 months. And this is on Page 95 of the specs. That’s where it begins. I’m going to go through the highlights and the
things that make it distinctive. This measure looks at adult consumers 18 and older who have major depressive disorder or dysthymia. The index visit is the first diagnosis of major depressive disorder or dysthymia during the measurement year with a PHQ-9
greater than nine. The numerator measures those in the eligible population with a PHQ-9 less than five at a point 12 months plus or minus 30 days after an index visit. This is considered remission at 12 months. The denominator measurement period is the measurement year
starting with the index visit. The numerator measurement period is the measurement year and 12 months after the index visit plus or minus 30 days. So, for example, and pretending that the measurement year begins on January 1, if a person is newly diagnosed with major
depressive disorder on March 1, 2017, the measurement period for the numerator goes from March 1, 2017 through February 2018 plus or minus 30 days. This lets you look at data for a full year after the new diagnosis to assess remission
at 12 months. This measure applies to newly diagnosed and existing depression or dysthymia with a PHQ-9 greater than nine. It would be the first such score during the measurement year. It is only reported once per measurement year for
any one consumer. And the value sets for this are on the steward’s website and the link is provided in the specs in Section A. The definitions explain the index visit in some detail. There must be all of the following at the same time. You have to
have a PHQ-9 that’s greater than nine. You need an active diagnosis of major depression or dysthymia. And the consumer cannot already be in a prior index period. So you have to have a clean period. The index period begins with an index visit where these criteria are
satisfied and is 13 months long. So if a person had a prior diagnosis of major depressive disorder 15 months before the new recurring diagnosis in the measurement year the new recurring episode counts for a new index period in the measurement year and is included in the denominator. However,
if the person had a prior diagnosis of major depressive disorder seven months before the first new reoccurring diagnosis in the measurement year, that new reoccurring diagnosis does not qualify the person for the eligible population. So you have to have that clean period. Also, if the provider
is a behavioral health provider the depression or dysthymia diagnosis must be the primary diagnosis. It is not necessary to be the primary diagnosis if the provider is a non-behavioral health provider. This means that if the BHC has a non-behavioral health physician or nurse that performs the
screening where there’s a PHQ-9 greater than nine it can be in any position as a diagnosis. If it’s a psychiatrist or other behavioral health provider it must be the first diagnosis. The definitions also define PHQ-9. It’s a
freely available tool and there is a link in the specs. Remission is defined as a PHQ-9 of less than five, and 12 months is defined as 13 months to allow for a grace period
for remission. The steps for determining if someone is in the eligible population are, first, they’re seen at the BHC during the measurement year; second, they have a major depression or dysthymia diagnosis during an outpatient encounter during the
measurement year; third, they have a PHQ-9 score greater than nine during that encounter, and you should note that there is a HCPCS code for this that permits you to automate the calculation of the eligible population; and, fourth, they are 18 or
older at the index visit. As usual, the denominator is the eligible population. The numerator includes coding for both a PHQ-9 less than five at 12 months or an equivalent record of a score in the medical record. And there is a code if the PHQ-9 was not
assessed or the score was greater than or equal to five at 12 months, so they were not in remission. So one code gets the person counted in the numerator and one gets them not counted. The not counted means they either were not assessed at 12 months, give or take 30 days, or they were not in remission with a
score less than five. It should be noted that not assessing at 12 months, give or take a month, counts against you, unlike some other measure we’ve discussed. Different measure stewards handle this differently. There are both required and optional exclusions. And I do want to
note that the exclusions apply to the denominator and numerator. So you omit people from the denominator if they fit an exclusion. This automatically gets them out of the numerator, too. So the required exclusions include an active diagnosis of bipolar disorder
or an active diagnosis of personality disorder, and both could be in any diagnostic position. Optional exclusions include being a permanent nursing home resident in the measurement year, hospice or palliative care during the measurement year, or they died during the measurement year. In this measure,
better quality is associated with higher care. More remission at 12 months is a good thing. So I’m going to stop here and recap something about data source. The denominator is capable of being fully automated. The numerator is capable of being fully automated. In other words, both can
rely on administrative or billing records. The exclusions, however, require forethought about how to approach them. If you only use the required exclusions, you can rely on administrative data because they are diagnosed. If you use the optional exclusions there is not a specified
code that can be used. This will require medical record review. So you would have two choices. You can either not use the optional exclusions or you can make sure that there are fields in your EHR that allow you to pull them out easily, even
though there are no specific codes. Okay, this is the last poll question, and if only the BHCs in attendance could participate that would help. So states, please don’t. It’s set up to allow multiple answers, although I won’t – I hope
you don’t select both yes and no. The first, do you presently use the HCPCS G codes that are referenced for the PHQ-9 results, yes or no? And I realize that asking you this out of the blue may mean that you don’t know the answer. I should’ve added that as an option. So if
you don’t know, don’t answer. And if the answer to that question is no, you do not use them, please select one of the following options, or actually multiple. Option 1: We will implement them and train providers to use them. Option 2: We will have another
electronic method of determining results. Option 3: We will undertake medical records review to determine results. So I suppose it’s possible that you
could have both Option 1 and Option 2. Okay, let’s see what we get. All right. Very few people currently use these codes, 3 percent, 4
percent. Okay. Forty-six percent indicate that they will implement them and train providers to use them. Forty-six percent, we will have another electronic method of determining results. Okay. And
almost 15 percent no, we will undertake medical records review to determine results. Again, if people have time feel free to put in the questions and answers or chat
box what the barriers are to you using these codes. And I’m going to move on and give people a chance to ask questions also. Amy: The first question is from Kristen
Kristenson. So the SRA-A captures the number of people, whereas the SRA-BH-C captures the number of visits? Peggy O’Brien: Okay, so the children is number of visits. Yes.
That is correct. I believe that they’re both number of visits. And if I’m wrong I will clarify that, but I think they’re both number of visits. Amy: The
next question is from Patricia. What if the child, adolescent or adult has depression and a co-occurring substance use disorder? Is their primary diagnosis of both mental health and substance use
disorder? Peggy O’Brien: I do not know if you can have two primary diagnoses, which could be a problem for – you could have conflicting measures here requiring multiple things, because this one would ask that you have the
depression in the primary position, and I think the follow-up after ED visit for alcohol or other drug use would ask that the substance use be in the primary position. So that’s problematic. Let me think about it
a bit more. Amy: The next question is from Ginger Bandine. For depression remission, can you please help me understand if it would count in the numerator as a remission if the PHQ-9 score of less than five was achieved and measured after
five months of treatment instead of a score of 12 plus or minus 30 days? Peggy O’Brien: Okay. There are two versions of this measure. There’s one at six months, which is probably also plus or minus 30 days, and then there’s this one at 12 months. And the one that was selected
is the 12-month one, and it’s designed to assure longer term follow-up and screening to determine if there was remission. So my instinct is that the answer would be no, that it has to be at 12 months and it has to be in that
30-day window on each side of 12 months. Amy: There are no more questions, Peggy. Peggy O’Brien: Okay. All right. So we have three more webinars to go. This is the remaining webinar
schedule, again, every Tuesday at 2:00 Eastern Time, not the last week of August. And this is a preview of the two Special Issues webinars. We’re going to discuss the topics listed here, but not necessarily
in this order. At this point I expect to cover the following next week: continuous quality improvement and the role of data; sampling for hybrid measures; differences regarding age coverage and stratification between the BHC measures, the HEDIS measures
and the Medicaid core measures; and when someone is a CCBHC consumer. I also will address a lot of unanswered questions from earlier webinars. So at the end of the webinar next week there is going to be a period of time where I will be addressing
things that were either left open during webinars or which required some clarification or a more definite answer, and I will also do that the following week. So in Webinar 7 we’ll be covering the quality bonus measures and
payments; data for dually eligible beneficiaries; lessons learned in state visits; and some more unanswered questions. So this and the next two slides are charts showing the measures by abbreviation, whether they’re state or BHC-lead, whether they’re
required as part of the CCBHC demonstration program, and in which webinar they are discussed, so you can find them if you want to refer back. Please submit any additional questions to the SAMHSA Data TA mailbox about what we covered
today, material that’s scheduled for next week or any other questions related to data collection, analysis or reporting. Also I wanted to mention that there’s still time to sign up for office hours if you have questions and would
like to have them addressed in a more individualized setting. So I’m going to stop and see, Amy, are there any other questions that have come in? Amy: There’s just a comment by Scott Girds, and it says our system does not allow a way to add
reportable codes for services. We use SQL to pull data and assign the appropriate HCPCS code. Peggy O’Brien: Okay. So, again, there have been a number of comments like this, which we will review. Okay, thank you.
I hope you’re able to attend the two Special Issues webinars over the next two weeks. We’ll be covering a combination of technical and less technical materials. So I hope it will be informative and somewhat
less dry than having to go through these specs in great detail. So, thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *