September 17, 2009: Afternoon Session (continued)
Female Voice: Only by age. Age is the denominator.
Female Voice: Otherwise, the 22 and 22a are considered essentially equivalent, and so then it becomes 22 or 22a versus 42.
Female Voice: So 22a is the community health centers; 22 is the traditional NCQA-HEDIS ((Health Plan Employer Data and Information
Set-National Committee for Quality Assurance) type population health plans, the actual plans, and then 42 is Medicaid [indiscernible].
Female Voice: Alright, that's fine. So we have three different sources.
Female Voice: Three different sources. Clarification is 22, I thought I heard that the difference between 22 and 22a was just age and—
Female Voice: And data source.
Female Voice: And data source. It is different then.
Female Voice: Yes.
Male Voice: Right. Let's vote on these this way, and if 22 and 22a are—if we don't have a majority in any of these we'll have an instant runoff.
Male Voice: This question of denominator seems to me to be pertinent to all the measures we're looking at, not just this measure. So voting on this measure based on the difference in denominator doesn't seem to make sense to me because, I mean, it seems to me the difference, the real difference here is whether we want to look at whether somebody gets any dental visit or whether they get a preventive dental visit. And the denominator issue could be solved by saying this is how we're going to want to look our measures. So it seems to me making the distinction around the denominators on this set of measures—
Female Voice: But detailed specifications were submitted for each measure, and if we're going to say this is the measure in the core set and we want everybody to measure it the same way, then you have to use the specifications. I can't say we're going to use EPSDT, but we're going to make sure that you've been enrolled for 6 months. We don't have the authority to change what the specifications are for these measures and, frankly, the denominators I think do matter, not in all the measures. It's important to look at them in all the measures, but I think in some of the other ones we've been making choices between, the denominators are not very different. Here they're quite different.
James Crall: But from a process standpoint then, can I go back to the issue I was raising before—with CMS (Centers for Medicare & Medicaid Services) and the 416, we actually have a measure that looks at whether they got any service, whether they got preventive services and whether they got treatment services. In NCQA, we have one, and the issue is I mean, as a group, what we want to recommend that if the NCQA happened to be the preferred approach for gathering or reporting on whether a child got any service, would we extend that approach to the assessment of preventing services and treatment services? It seems to me like that.
Male Voice: So I think, I think what you're saying—
Female Voice: That's [indiscernible] measured at all.
Male Voice: Jim is really, as you vote on this, consider what are the dental measures.
Female Voice: No, we shouldn't. I don't think we should because they're really different. They're really different. This is just did you get a visit period, right? And the other ones look at what type of visit you got, so I think that's a very important distinction. Jim, is what you're asking about would we then advocate that the same sort of NCQA-type denominator be used in those other measures, or would we accept them as the—you know what I mean.
Female Voice: Thank you, measure denominators as they're specified.
James Crall: I mean that was the question I was raising.
Female Voice: Yes, I think those are two different measures that we need to consider separately from this discussion and stand alone from this.
Male Voice: But to be clear, we're not going to create a new measure with an NCQA format for the type of service in this group.
Female Voice: In the other two EPSDT measures, the 12c and 12d I think it is, right?
James Crall: If we decided we want to have reporting on whether it's only getting preventive services, preventive dental services, or dental treatment services, the recommendations will only come out currently using the CMS program.
Female Voice: That's all that's available right now.
Male Voice: So we'll be voting for different format.
Female Voice: I think deciding to, if we were to go with the NCQA measure here, that does not negate us using the other EPSDT measures, right?
Male Voice: That's just—right.
Male Voice: From a procedural standpoint, the States are going to have to determine whether or not the kid got any service period, anyway. I mean that's part of the analysis.
Female Voice: But later on we're voting on PHP43, which is the preventive services, so it doesn't have to be [cross-talking].
Female Voice: I only wanted to say that with respect to the EPSDT data, even if you can't change exactly how it's calculated right now, the analog that Jim is talking about, a full-time equivalent which is a kind of the outburst of some survival analysis kind of thing, is absolutely a reasonable statistical analytic technique to take, and it might require a tad of programming of EPSDT data. But if in fact it's a much better, broader measure of who is included, and then you do that kind of calculation he's talking about, I would have said that sounds like a preferable measure to me, particularly thinking ahead to 43.
Male Voice: Okay, Doreen?
Doreen Cavanaugh: Some point of clarification. When we're talking about the definition of continuously enrolled, I believe it's 11 out of 12 months, is that enrolled in Medicaid or enrolled in a plan? You used the word plan, Rita, and I was thinking it was enrolled in Medicaid, but I need a clarification on that.
Rita Mangione-Smith: So for the NCQA measure, I think is a plan-based measure, is that correct? So it would be Medicaid managed care that does it right now.
Doreen Cavanaugh: What do we do with fee for service?
Rita Mangione-Smith: Again, you would have to adapt to NCQA's specifications with fee-for-service data, which I am right there with you is not an easy thing to do. It's not impossible, but it's not an easy thing to do.
Mary McIntyre: It's not impossible to be done. It can be done, and the reality is when I spoke up about the manager with the 22 versus the 42, the reality is that even if this group decides that it's going to do the NCQA measure, that until CMS tells us that it's going to be changed as far as what we are going to have to report, we're going to have to do both of them. So I just wanted to point that out because we haven't gotten anything from CMS to say that they're going to adapt or change the measure requirements. We would hope that they would do that in light of what happens with this group, but the reality is that if you all decide on the NCQA measure, we're going to still have to report the CMS-416 EPSDT measure, and I don't see that as being a big issue because we currently, if we run CMS data and our DSS system now, it's really just pulling the information using the DSS system, putting the logic in. So it's not a big deal.
Jeffrey Schiff: Okay. I think we should vote. We have minutes to go. I think we need to do a runoff between 22 and 42, so we need to eliminate.
Cathy Caldwell: Can I comment just really quickly?
Male Voice: Sure.
Cathy Caldwell: Just a reminder that CHIP does not report EPSDT standalone, and I think Mary made a good point. The data, the way they're currently reported will continue, and I really advocate for dental measures, so I'm not trying to knock anything, but I do want some good quality dental measures in here. I, as a CHIP director, struggle constantly with what denominator to use when I'm reporting dental because right now you could have a 6-month-old with no teeth who's in the denominator. You can have somebody enrolled as little as 1 month. So they could be completely compliant with their routine check-ups but they're not going, that will not be reflective.
So a measure that truly has the appropriate denominator I think is desperately needed. Now whether that—is it only go with that one or go with—I don't know but there is a real need to have a measure with the appropriate denominator.
Jeffrey Schiff: Okay, okay 22 and 42. Run again. Okay, so 22 actually is the winner in the past.
I guess it's time for public comments. So we want to do that, but I think there's a theme going here and that we need some denominator enrollment calculation. It's a pervasive issue here, and when we put up our list of recommendations going forward, I think that I'm going to put that up there, and we can talk about it if we have time, but I'm getting a lot of nods. Okay, only really it's a dental measure. Your comment? We have one more.
Okay, so we have one more now, then we have to really get—we have a lot more work. We have public comment, and now we have a lot more work to do, so if there's something related to these identical measures, let's not talk about it otherwise. I'll take our public comments.
Return to Contents
Denise Dougherty: Okay. So I guess the first person would be Juliana Belelieu from TeenScreen National.
Juliana Belelieu: So good afternoon. I want to thank you for the opportunity to comment today on the establishment of a core set of pediatric quality measures of care provided by Medicaid and CHIP. This is an important undertaking as you all know with the potential to improve the quality of care provided to the majority, well, to a significant portion of all children in the United States.
Again, my name is Julie Belelieu, and I'm the deputy director of health policy at the TeenScreen National Center for Mental Health Checkups at Columbia University. Our mission is to expand access to mental health screening for adolescents and to improve outcomes through their early identification and treatment of mental illness.
As you consider the establishment of core pediatric quality measures, we urge you to include a measure that will allow us to assess the rate of mental health screening performed using a standardized evidence-based tool during adolescent well-visits. Quality measures should reflect current professional knowledge, and there is now an overwhelming consensus among medical professional groups and scientific review panels that comprehensive preventive services for adolescents should include mental health screening.
Groups endorsing mental health screenings for adolescents include the United States Preventive Services Task Force, which recently recommended screening adolescents aged 12 to 18 for depression; the Institute of Medicine, which highlighted the screening as a preventive measure in its report this spring; the American Academy of Pediatrics; the American Medical Association; the American Academy of Family Physicians; the Society for Adolescent Medicine, the American Academy of Child and Adolescent Psychiatry; and the American Psychological Association.
Unfortunately, a significant gap exists between the recognition of a value of providing mental health screenings and the translation of this recommendation into clinical practice. Epidemiological studies have estimated that approximately 20 percent of adolescents suffer from a diagnosable mental disorder, with approximately 10 percent suffering from serious functional impairment. However, right now, despite the existence of scientifically proven screening methods and effective treatment options, only about one-fifth of all mentally ill youths are identified and receive services, resulting in a missed opportunity to improve outcomes. This jarring disconnect between the consensus on the need to screen and the low rate at which adolescent mental health screening actually occurs highlights a need for quality improvement and provides a compelling rationale for including this measure in the core set.
To be totally effective, a quality measure also requires the ability to collect and report valid data. Fortunately, the Child and Adolescent Mental Health Measurement Initiative has developed detailed guidance on how to best measure the rate of developmental or mental health screening performed using a standardized tool, and a number of State and local programs are already collecting this data.
For example, the Massachusetts Medicaid Program, MassHealth, is currently measuring the rate of mental health screening using the 96110 billing code, which denotes limited developmental screening. I should note that that's also the same code used for younger children, and smaller programs have measured screening rates using chart review and patient surveys. Not surprisingly, these measurement efforts have helped to improve quality and have demonstrated increased rates of adolescent mental health screening over time.
In conclusion, the TeenScreen National Center for Mental Health Checkups urges you to include a quality measure assessing the rate of adolescent mental health screening performed using a standardized evidence-based tool in the core set of pediatric quality measures for Medicaid and CHIP. Doing so will improve quality and reduce the burden of untreated mental illness by reinforcing the importance of adolescent mental health screening to clinicians, public programs, private health plans, and others involved in the provision of care.
Jeffrey Schiff: Thank you very much. Our next comment is from Mark Antman.
Mark Antman: Good afternoon and thank you for the opportunity to comment. My name is Mark Antman. I'm a senior policy analyst at the American Medical Association and on staff to the Physician Consortium for Performance Improvement (PCPI), which is one of the measure developers from whom you received a number of measures for consideration. And I'm very pleased to say that a number of the measures from PCPI are included in the list that passed the most recent scoring.
I'll be brief, just a couple of comments about measures that are included or not. On the past grid, one of the care transition measures is listed, the measure for the transition record with specified elements received by discharged patients for inpatient discharges. As Dr. Schiff noted this morning, one of the companion measures in that group, AC21, is incorrectly listed and did not pass the Delphi scoring.
But there is another measure in that measure set, a measure for the timely transmission of the transition record, which was part of the measure submitted by the PCPI, which, frankly, I have not found on either the passed or the failed list. And so if that was an oversight, I do encourage the subcommittee to take another look at that measure, in large part because the PCPI—and in particular the workgroup that developed these measures—certainly felt that the provision of the detailed information listed in the numerator specs for the measure that is on the list, AC20, the transition record.
Although the group felt that that certainly is very important, they felt that it was equally important for the information to then be transmitted promptly to the next provider of care. And so that companion measure, the timely transmission of the transition record, does specify that the information in the transition record should be transmitted to the next provider of care, whether that be a primary care physician or another inpatient facility within 24 hours.
So again, unless I have missed it on the failed list, I encourage you to look that up.
Female Voice: So we did look at that. So the smaller of the documents we distributed has a failed set and a passed set. We did look at the whole set, and the one you're referring to is AC22 care transitions, timely transmission of transition record from inpatient. That is AC22, and it is on our failed list.
Mark Antman: Okay, thank you. One other quick comment then if I may. Moving to the next page of the long grid under depression care, again I was very pleased to see that the suicide risk assessment under the category of child and adolescent NDD measures did pass the scoring and was included in the list. I would submit to the group that if that is the only measure, that it moves forward for depression care. It does provide the somewhat limited view of care for child and adolescent patients with major depressive disorder. There are, and I recognize that the other measures in that set are on the failed list, so I recognize that for one reason or another they did not meet the criteria to be on the passed list. But I would encourage the group to consider that the addition of perhaps one of the therapeutic measures would provide a somewhat more complete picture of the care for these patients. Thank you very much.
Jeffrey Schiff: Our final in-house, onsite public comment is Deborah is it Fillin?
Female Voice: Wellspolinger [phonetic].
Jeffrey Schiff: Wellspolinger. I apologize.
Deborah Wellspolinger: Thank you and I will consider this more of a technical assistance or whatever, but I just wanted to thank the group for your hard work. This is a very complex project really in terms of all the different levels of consideration that you have to, the different ways you have to consider these measures, and I just want to thank you for this work, and it's an exciting opportunity.
I did want to take you back to the 22a I think it was, for which there was a question earlier about 22a, the community health center example of are there any dental visit measures. And just to think of it as just affirmation that at the community provider level, this work can be done, and if you think of it that way, if it just supports your decision you pick 22, and to let you know that there are clinicians on the ground, providers on the ground who have actually tried this measure and have done it successfully. So it's just an additional bit of data for you to consider.
I'd like to go back to PHP26 just for a moment also, and I'm really excited that there was a measure of low birth weight selected, and we settled on the 2500. Just to go back to an earlier statement that once you're looking at the birth certificate data, and I think that's what at least at this point you've decided to do, there are lots of opportunities that then present themselves. So if you look at the data, and you choose to only record those that are under 2500, that's very—that is a measure.
But in the measure development world, we often combine different bits of data into one measure, so it is possible that the entity that goes to the record could also pick, choose to provide additional information beyond just the 25. They could tell you about singletons. They could tell you different weights easily. So they do measure very different things, the very low birth weight and the low birth weight measures say something different about the care that's been provided and about the mother, and the singleton births versus the multiple births also tell you something very different because of the wonderful fertility success that's occurred, we're getting multiple births and lots of low birth weights from very different populations. So I'm sure you're all aware of that, but I just wanted to, for the record, indicate those things. We're available for any other questions that you may have. So thank you.
End of Public Comments
Return to Contents
Afternoon Session Resumed
Rita Mangione-Smith: Thank you. So we do have a couple of more measures that are close to each other that we definitely want to do some voting on, and then we're moving to the arduous task of going through this in a more comprehensive way. Yes? Alright, that's fine either way you want to do it.
Jeffrey Schiff: Okay. Our next issue is we thought we needed to talk a little bit about the eight acute care inpatient measures. There is a bundle of inpatient measures, including a couple of bloodstream ones that could be considered the same, and then there's an inpatient pediatric quality composite and safety indicator. So I'm looking at AC4 through 26 on the passed grid. I think what would be most expeditious as far as time is we should have a conservation about whether we wanted to consider these individually and as a composite or just as a composite, and I think we'll do that first before we look at whether these bloodstream infection ones are the same because the blood side is composite, then we're going to make that easier. Does that make sense?
Rita Mangione-Smith: Okay, so the composite measure is AC26, and I do want to point out that there are things included in the composite that we did not rate individually and that some of the things that we did rate individually are not in the composite. So I think everybody just needs to know that the composite does not capture all of the individual measures but does include some things that are not among the individual measures. And we'll try to track which things those are, but it will become obvious as we go over the measures.
Female Voice: Rita, excuse me, does the composite include measures that individually we failed?
Rita Mangione-Smith: That is an excellent question. Let me see. So the pain assessment is not in the composite, care transitions. I'll answer that in just a second. No, none of the measures included in the composite were let go as individual measures when we rated it.
Female Voice: Okay, one other point then. So you're saying that only the measures that are on here are rolled up into the composite?
Rita Mangione-Smith: No, the composite includes several individual measures that are rolled together. Some of those measures that are in the composite we rated as individual measures, like an example of the accidental puncture or laceration during a procedure. That was rated both as an individual measure, AC24, and it's part of the composite, AC26. However, the composite also includes post-op hemorrhage or hematoma. We did not rate that as an individual measure. So the composite includes some of our individually rated measures plus some that we didn't, and there are some individually rated ones that it doesn't include, and I'll try to sort that out for everybody. It's just not voting for them individually versus as a group. I just wanted to make that clear, and I'm sorry I'm confusing everyone.
Female Voice: Can I just have one final clarification on the order that you suggested that we consider composite versus single versus considering the two [indiscernible] measures. The data that we put in our one-pager on PDI12, the AHRQ, what is labeled then, it's an erroneous label actually. Central line infections was all wrong, and so I don't think we assessed it. If you simply look in our one-pagers, the CDC definitions, which is the measure that is AC4, were applied to the administrative-based AHRQ measure, as was all the information about the validity and reliability of it in our one-pager. So I'm not sure people assessed it accurately. And the last point I'll say is that it is actually infections due to medical care; it's not just central line. So it's all mixed up in one-
Female Voice: Actually, I want to respond to that because I was largely responsible for doing a lot of the one-page activity, and we did not use the stuff that you have submitted to do AC23. So it was based on information that came from the age AHRQ PDI measure.
Female Voice: No but I'm looking at the AHRQ Web site. AHRQ measures are built off of administrative ICD9 data. Yes, and so when you look at the one-pagers on AC23, you have put the CDC definition. There's no way to extract these from administrative data. Number of central line days? There is no administrative data that takes that denominator. So all the information, and I'll even take it further, the published evidence that measuring this changes outcome is based on the CDC definition, not the AHRQ administrative database measure.
Rita Mangione-Smith: So according to the specs I looked at for the AHRQ measure, it said medical records data. It didn't say admin data.
Female Voice: It's admin.
Female Voice: So then do you want to just take it off the table?
Female Voice: Well, no. I mean let me explain. Well, first of all, what's written on the one pager for AC23 as the numerator and the denominator is an error. So it should be whatever the PDI 12 definitions are, which you won't mind a little—
Female Voice: I actually have them. I pulled them up as we were thinking about this.
Female Voice: Okay, but so—but just to explain the PDI 12 and this is what I was told about why the name of it has been changed from selected infections due to medical care to CLABSI. So the name has been changed by the AHRQ quality indicator team, okay? So it's now known as the CLABSI measure. And the reason for that is it's all—there's a whole bunch of stuff about all the PDIs—I mean IQIs and stuff on the WebEx but the reason I was told—and I'm skeptical, too, even though I work for AHRQ, very skeptical, is that of the 80 percent of true cases that have been identified as to the validity of the selective infection measure, 60 percent of them were attributable to central line infections, and that's why they changed the name. I mean that's their decision on our team.
Female Voice: Having done this in my own institution of about 30 cases, there was an overlap of one. And more so, the measure that all the households across the country are using, the measure that the Joint Commission requires is AC4, not the PDI. It is just—I would lodge my vote it's a misnomer to label the AHRQ measure a CLABSI measure because there's a whole CDC-defined CLABSI measure.
But I would even take it further. I guess my main point when people were scoring it, and the reason I'm saying it is because it's rolled up in a composite, and if I think it's erroneous, you need to know that as you're thinking about the composites. But the data that's out there that says measuring CLABSI and improving it changes the outcomes of children's lives and adult lives is based on the CDC definition.
And on here, all that was missed was displayed as if it was valid in the AHRQ measure, and none of that validation work exists for the AHRQ measure.
Female Voice: Okay.
Rita Mangione-Smith: The true numerator and denominator for this—the AHRQ measure, the AC23, the denominator is all surgical and medical discharges under age 18 defined by DRGs, exclude cases with ICD9CM codes 999.3 or 996.6 in the principal diagnosis field, newborns, neonates with a birth weight less than 500 grams, length of stay less than 2 days, MDC14, pregnancy and childbirth. So those are the exclusions. Numerator is listed as discharges among cases meeting the inclusion and exclusion rules for the denominator with the ICD9CM code in any secondary diagnosis field, and I guess those must be the codes for infection, hospital acquired infection. I mean—you would know this, Marlene, 999.3.31 and 996.62.
Female Voice: Well, just to be very clear, the measures that the hospitals use are based on infection control reviews of charts. They are not based on any administrative data codes. These codes are, when you match them up, and I told you not my sample, it happened—they completely do not match up with what infection control practitioners across the country defined as CLABSI cases.
Rita Mangione-Smith: Okay, is everybody clear on the differences between the two measures? Do you want to deal with the vote on those tests now?
Male Voice: I have a question.
Rita Mangione-Smith: Talked a lot about them. Yes.
Male Voice: Are we lumping in all of the other measures above AC26 as the alternative to AC26, or are we dealing with AC20 separately?
Rita Mangione-Smith: No, we're not. So AC26 only includes some of the measures above it, not all of the measures above it. But I think the point that Marlene was trying to make is it does include this measure that we were just discussing, this AHRQ measure, that looks like it's not such a hot measure.
Male Voice: No, my point is that if we're lumping, I would like to call AC20 out of that discussion because I think that it is a different—
Rita Mangione-Smith: Oh yes. No, we won't be including that.
Male Voice:—a different kind of measure and that we should be talking about it separately.
Rita Mangione-Smith: Right, yes. They'll be talked about separately.
Jeffrey Schiff: Okay, okay. We're going to need to have a vote on AC4 versus—
Rita Mangione-Smith: AC23.
Jeffrey Schiff:—versus AC23 first.
Rita Mangione-Smith: Can we do that?
Female Voice: But wait a minute. The patient safety composite with the selective infections was NQF-endorsed. So I'd like to understand—
Rita Mangione-Smith: Would that affect—?
Female Voice:—your CLABSI measure was NQF-endorsed. The patient safety composite from AHRQ was NQF-endorsed, so how could—they're very different then as you're telling us, and how could the patient safety composite get endorsed when your measure got endorsed?
Female Voice: It's not my measure. It's the CDC's measure that's been out there for 15 years, and it's just that indictment of the approval process, if you will, that it's not necessarily a rigorous issue we'd like it to be perhaps or believe that it may be, and I'll leave it at that.
Rita Mangione-Smith: Let's just put that on the table.
Jeffrey Schiff: Okay. Okay, let's keep going and start—we have to start moving, so let's do this right now. We're going to vote on these CLABSI versus the CDC versus the AHRQ measures and respond one or two, so we're on.
Female Voice: Could you clarify what is an AHRQ measure and what is a CDC measure?
Female Voice: And what's PDI while you're at it?
Female Voice: The PDI could probably be the indicators.
Jeffrey Schiff: Okay, well, the vote on AC4 we're not going to pass so—
Return to Contents
Proceed to Next Section