Live 1057

Free Internet Radio Stations
Management of MRI Utilization for Low Back Pain (July 2015 Radiology Podcast)

Management of MRI Utilization for Low Back Pain (July 2015 Radiology Podcast)


Herbert Y. Kressel, MD: Hi. This is Herb Kressel
and welcome to the Radiology podcast. Today we have an interesting discussion on the management
of MRI utilization for low back pain. And we’ll start. Our two discussants are Dr. Cecilia
Ganduglia, Assistant Professor from the University of Texas School of Public Health in Houston.
Welcome Dr. Ganduglia. Cecilia M. Ganduglia, MD, DrPH: Thank you. H.Y.K.: Dr. Ganduglia and colleagues wrote
the article entitled Effective Public Reporting on MRI Use for Low Back Pain and she’ll
be joined by Dr. Brook Martin, the Dartmouth Institute for Health Policy and Clinical Practice
in the Department of Orthopedics. Welcome, Dr. Martin. Dr. Martin and Dr. Jarvik authored an editorial on this fascinating paper. So let’s get started. Dr. Ganduglia, tell us about the CMS hospital outpatient quality report. What’s the idea? C.M.G. So this initiative has been around
for some time now, actually. CMS requires hospitals to report a specific number of measures
that evaluate the quality of care that they’re providing. In this case particular to the
outpatient population. It was in 2009 that they incorporated a set of imaging efficiency
measures in today’s outpatient quality reporting program that measured some specific, they’ve sort of been increasing the number, they started with eight and then they remove a couple and they’ve added some other measures in later years and what they are requesting is hospitals to report some specific proportions of imaging studies that they are performing and the overall
requirement is they have to report these measures in order to quality for their annual payment
updates, but there is no particular penalty right now set in place for specific values
of these measures. H.Y.K. So what’s the general idea then? What’s
the concept? How is the reporting supposed to affect the utilization? C.M.G. So the reporting is right now used
in the hospital compare website that’s also supported by CMS and in theory this website
provides consumers or patients the ability to look at how hospitals are performing and
choose providers according to these measures. H.Y.K. Okay so aside from low back pain can
you give a few other examples of what else has been reported? C.M.G. So among the imaging efficiency measures,
they had others like CT scan proportion. For example they were looking at abdominal CT
scans and thoracic CT scans and the proportion of them that were performed with and without
contrast over the overall number of CT scans specifically for that side. There were also
CT scans for headaches if I’m not mistaken and there were a couple that were removed
and are still under consideration and they’re going back and forth. H.Y.K. Okay, now let’s go back to your study.
So what did you all decide to look at and how did you go about it? C.M.G. So we decided to focus our study on
the OP-8 measure which is the one that’s specifically looking at how MRIs of the lumbar spine are
ordered for patients with low back pain. And of the overall number of these MRIs how many
of them actually do not have a history of conservative management in the 60 days prior
to the MRI. We chose this measure because actually MRIs for low back pain have been
identified by many professional associations as a test or a study potentially being overused
and it’s also been identified by the Choosing Wisely campaign as something that should be
focus for a risk of overutilization. So we decided to see…the first thing we were measuring
was how Texas was doing on this measure and we wanted to look at geographic variation
throughout the state. And since we had data from before the measure was implemented and
after the measure we decided to look and see if this measure was in anything efficient
or had some effect on modifying the practice patterns in our state. So in order to do this
we used claims data which is the type of data the measure was originally designed to be
used with and we have data from a commercial insurance group here, Blue Cross Blue Shield,
that covers one-third of the commercial insurance population; and we also have the Medicare
data for the same time period. We measured 2008 through 2011 and we limited our analysis
to those that were covered in the fee for service type of program for which we have
the full information of those patients. So we actually measured…our study was focused
at the hospital referred region and we compared before and after implementation of the public
reporting measure. This measure, as Dr. Martin has actually pointed in his letter, later
on was changed and there were some additions to actually the exclusion criteria of the
measure. So what we included at the time was the specific measure that was active and valid
during the time frame we were measuring. And what we actually found was that there were
no significant differences. So we measured 2008 values for Medicare and Blue Cross and
then 2009, 10 and 11 and what we found was that there were no statistical significant
differences and if anything there was a slight increase in the Medicare overall rate for
the state. We did find, however, that there was a lot of geographic variations. So certain
areas in Texas were performing a lot better than other areas and this was really different
across payers so Medicare and Blue Cross were not behaving really in the same way, and we
did find that overall the Medicare rate was slightly lower so the proportion of MRIs without
conservative management were smaller in Medicare than in the Blue Cross Blue Shield population. H.Y.K. Thank you. Now Dr. Martin, in your
editorial sort of one of the first points was sort of concerns about the quality and
the definition that was used to define the target population for this initiative. Can
you kind of talk a little bit more about that? Brook I. Martin, PhD, MPH Yes, so I’ve been
paying attention to this measure since Medicare first announced it starting in 2009. I think
it began implementation in 2010. This is a claims based measure where they identify MRIs
for low back pain based on claims that are submitted to Medicare and then they linked
that to previous claims for the same beneficiary in order to identify whether these patients
who have an MRI for low back pain had anteceded conservative care and so that’s what the measure
is trying to get at is the use of MRI in the absence of anteceded conservative care. As
a claims based measure, we’ve tried to examine this measure both how it’s designed in claims
and one of my most fundamental concerns about the measure initially is that while it’s intended
to capture non-specific low back pain, it actually includes, they don’t actively exclude
people who have more specific spinal pathologies like stenosis and spondylolisthesis and spinal
fractures, postsurgical MRIs; and so when you add all of those in, we found in the 2010
data that 58% of the patients who were captured by this measure in the Medicare data actually
had some pathology in addition to a diagnosis code for low back pain, they had some code
that indicated more specific spinal pathology for which MRI is not the contraindication.
So in my mind it raised alarms that whether this measure is actually measuring non-specific
low back pain. Most patients with low back pain, I guess I shouldn’t say most patients,
but I should say many patients who have codes for low back submitted to Medicare also have
additional diagnoses for more specific pathology, and they largely in the initial roll out of
this measure didn’t exclude those with specific pathology and so the question to be raised
is whether the measure is really reflecting what the large base of literature is trying
to measure. I think the lack of movement that the nicely conducting study shows is there
wasn’t movement on this measure. I think there’s a lot of potential reasons we can
talk about but one of them that we’re asking is whether the measure is reflecting what
it’s intended to measure. H.Y.K. What about the demographics Dr. Ganduglia,
the population ages and the two groups are not comparable and so you would expect some
utilization, certainly some prevalence differences and resultant utilization differences. C.M.G. Yes, so we did find that there was
a prevalence difference definitely in low back pain because the first thing we measured
was the total number of claims we had with a principle diagnosis for low back pain or
related to low back pain; and it was much more common among the Medicare population
than among the Blue Cross population, the commercial insured one. We also found that
approximately the same percentage of them had an order of an imaging test performed
so that was not different among the populations, however when we applied the specific OP-8
measure, the specific measure from our quality reporting program, we did find differences
on how this was applied. There could be several reasons for these differences and our first
objective of including these two very different populations was first, we’re not able to
bring up a comparison group because this is something that was implemented nationally
and all physicians and hospitals were subject to it, so it’s not easy to find a comparative
group and we’re doing a before and after; but we wanted to at least look at two different
payer populations that were sort of affected slightly different in the sense that CMS was
the one requiring the measure. So we did not adjust it for age because it was logical to
assume that everyone on Medicare was older than people commercially insured and there
are significant differences. There’s also differences on how physicians may behave depending
on who the payer is and what are the managing regulations behind ordering a specific test.
However, we really weren’t…we cannot really explain that 20% difference we found
among Medicare and Blue Cross for the specific rate. H.Y.K. I see. Now one of the points that you
make in analyzing the data is that perhaps the problem is targeting the service provider
rather than the physician requesting the service and I think perhaps Dr. Martin you want to
comment on this. Where do you think the best leverage actually is in utilization management? B.I.M. Well I agree that the imaging for low
back pain patients targeting a hospital is pretty far downstream, a point that was really
I thought brought up well in this paper. I think we really need to focus on a primary
care population and ordering the MRI, because by the time it’s ordered and scheduled there’s
I think little opportunity to go back and revise that plan by the time the patient’s
already in the hospital suite or the free standing imaging center, they’re not likely
to change the plan at that point. So I do think it needs to go more upstream if we’re
really trying to change who receives these MRIs. So I think it’s a very good point.
I think integrating the referral patterns into electronic medical records is probably
the best approach to intervening to promote guideline concordant care in MRIs. H.Y.K. So I guess you’re talking there about
these decision support software that is integrated into the ordering system. B.I.M. I think that has a lot of potential
for improving guideline concordant care. H.Y.K. Dr. Ganduglia, any thoughts about where
we should go from here if we want to manage this behavior? C.M.G. I think that Dr. Martin actually made
a good point in importance in really identifying and defining good measures for quality of
care regardless of whether we’re looking at low back pain, MRIs or any other imaging
study. We really need to have measures that reflect the current knowledge and the science
behind it and I think a very important thing is to actually test these measures before
we go and implement them everywhere, nationwide, and ask everyone to report them and make sure
that we’re measuring work with that to see if we have an affect or not. H.Y.K. You raise a really important point,
in this whole quality of care effort, people understand that to follow something you really
need to measure it. But a lot of these things are very, very hard to get a finely granular
handle on it so people take what’s available and then you wind up with this kind of fairly
diffuse kind of situation where you don’t really, you have sort of apples and oranges
in there and you really are trying to modulate behavior but you can’t even sort out what’s
going on in your group. I think the lesson of this is I’m sure not limited to low back
pain. This is something that we all encounter as we’re grasping for measures based on
what we have, but they may not actually be telling us what we want to learn. And the
other thing I think for me and I’ll be interested in your comment, I mean the notion that by
reporting this that the disclosure of it in a database that patients may or may not look
at, I just can’t imagine someone the middle of low back pain that’s going to go consult
that before they choose their doctor. It seems like a stretch. One would hope that you would
kind of have a targeted approach that will sort of be more closely leveraged on the action
arm rather than on the awareness arm. So any closing comments Dr. Ganduglia? C.M.G. I think you’ve wrapped up the issue
pretty well and yes just what I was thinking, I think that it’s a good thing that this
measure has been highlighted and people are looking into this. I do think that overutilization
of imaging studies as well as other type of studies is something to be aware of and we
need to take some action into it, but I definitely think it’s something that requires a lot
of work ahead. H.Y.K. Dr. Martin, any closing comments? B.I.M. Well I think that these types of performance
measures based on claims actually do offer a lot of opportunities to improve care and
we can learn from them, but we need to not let policy get ahead of the science that validates
their use, that scientifically examines how well they measure what they’re supposed
to measure and whether they move and provide the right information that improves patient
care. And so I think there’s a lot of work to be done in validating this particular measure
and demonstrating that it’s a successful measure. Ultimately we’d like to have these
things linked to the outcomes and examine how much MRI imaging contributed to downstream
events like decision to have surgery or injections or other types of care. H.Y.K. Well I want to thank you both for a
stimulating discussion and some very nicely done articles. B.I.M. Thank you. C.M.G. Thank you. H.Y.K. So thank you very much.

Leave a Reply

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