For the February #JHSJC discussion, we looked at two articles from Volume 47, Issue 2 of JHS: “Evaluation of a Comprehensive Telemedicine Pathway for Carpal Tunnel Syndrome: A Comparison of Virtual and In-Person Assessments” and “Minimal Clinically Important Difference for PROMIS Physical Function and Pain Interference in Patients Following Surgical Treatment of Distal Radius Fracture.” We thank all of those who joined us, especially our newcomers!
As a reminder, we are now using a new platform for our events, TwChat. Our former discussion platform, tchat.io, is no longer functional. Please review the information here on how to use TwChat and join us for the next event!
Feel free to continue the conversation below in the comments! And please check out previous #JHSJC transcripts (please scroll to the bottom of the page).
Brent Graham @BgTalkinman:
Welcome to the Feb session! Please use the #jhsjc hashtag in all of your comments. We are using that hashtag in the TwChat chatroom, which worked well last month. Instructions for using TwChat are at http://jhsfocus.org/jhsjc. Let’s start with the paper on #PROMIS
The authors used the #PROMIS PF and PI scales in patients who underwent surgical treatment for #DRF and used the familiar approach of an anchor question to estimate the #MCID for these scales. These values were close to previously reported values in other conditions. #jhsjc
To me the findings are not surprising because the #MCID should be considered a property of the scale, not of the condition to which the scale is being applied. Thoughts? #jhsjc
Chris Grandizio @ChrisGrandizio:
I have found PROMIS in general to be an easy scale to discuss with patients bc many are familiar the idea of 0-100. I think the MCIDs add context to those discussion in terms of what people can/should expect after surgery. #jhsjc
Brent Graham @BgTalkinman to @ChrisGrandizio:
The 0-100 metric is easily understood although in practice these scales are much more narrow. Most studies show a ceiling effect for the UE scale at 56. #jhsjc
Brent Graham @BgTalkinman:
The #MCID values for these scales are consistently in the same narrow range when they are used in a variety of upper extremity conditions, consistent with the #MCID being a property of the scale and that scale having varying responsiveness to these various health states. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @ChrisGrandizio:
That is interesting. You don’t find patients have trouble understanding T-scores? It’s not a traditional 0-100 metric where patients are expected to reach 0 or 100. This has been confusing from my experiences. #JHSJC
Brent Graham @BgTalkinman to @BrochaSternOT:
Dr. Stern, what is your take on the meaning of the ceiling effects of some of the #PROMIS scales? It feels to me like they actually range from 35-60, not 0-100 with a mean of 50 and a SD of 10. #jhsjc
Brent Graham @BgTalkinman:
I ask these general questions about the overall metric bc I sense that the scales are not really like they are described in much of the literature on the general topic of #PROMIS, at least in the hand. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
It depends if one is using CATs or short forms. CATs have a wider range. Even with CATs, most measures usually range from 20-80 (not 0-100) – sometimes a bit smaller or wider range. #JHSJC
Brent Graham @BgTalkinman:
The anchor question-based estimates were 5.2 and 6.8 for the PF and PI scales respectively and for distribution-based estimates were 3.8 and 3.7 respectively. What does that say about the idea of a mean of 50 and SD of 10 for both scales – and all the #PROMIS scales? #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
But I think the possible range of a measure is somewhat different than (although overlaps with) a “ceiling effect” which is more about the proportion of patients who are at the highest score, indicating that they could have scored higher. #JHSJC
Brent Graham @BgTalkinman to @BrochaSternOT:
Almost all our submissions use the CAT format and still have this ceiling effect for the UE at 56. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
I’ve noticed that too re. the UE CAT specifically. I think the UE measure in particular may need some more work. #JHSJC
Brent Graham @BgTalkinman to @BrochaSternOT:
Understood! But doesn’t it get at the idea of responsiveness? That is what the papers we’ve published are also showing when they estimate the #MCID for different conditions. #jhsjc
What I’m getting at here is whether we should be re-thinking the assumptions of the #PROMIS scales, at least in the upper extremity. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
Great point. Yes! I think the MCID piece though is a challenging one more generally as there’s really still no gold standard for estimation and we try to use these for individual decision making when we are basing them on group-level estimates. #JHSJC
Warren @whammert:
Late to the discussion- the biggest difference here is the increased PI MCID anchor likely based on previous level of function #jhsjc
Brent Graham @BgTalkinman to @whammert:
Thanks for weighing in! To me that observation is another reflection of varying responsiveness. #jhsjc
Warren @whammert to @BgTalkinman:
I agree with varied responsiveness – we realized PROMIS is not perfect for everything, as we noted in early changes after CTR. But I think PROMIS is here to stay and understanding the limitations as well as the benefits will help us in the fiture #jhsjc
Brent Graham @BgTalkinman to @whammert:
I agree with that! #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
Sorry – getting used to platform, so a little slow here. I am not sure if the MCID estimates really challenge the assumption of a mean of 50 and SD of 10. Is that the assumption you’re asking about? #JHSJC
Brent Graham @BgTalkinman to @BrochaSternOT:
I think the idea that some of our authors like @whammert have brought forward that anchor-based approaches to estimating the #MCID reflect patient experience are persuasive and I think those should be the gold standard. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
Agree with anchor-based but don’t think that maps onto things like SD then & don’t know then if that challenges PROMIS assumptions in any way (although there are still obvious limitations to these measures). But good food for thought for me. #JHSJC
Alex Hollenberg @alex_hollenberg:
For conditions that are not responsiveness, like early stages of CTS, what would be the solution to measure changes in PROs if PROMIS can’t capture change? Other legacy instruments? #jhsjc
Brent Graham @BgTalkinman to @alex_hollenberg:
The legacy ascales may also have issues with responsiveness so I’m not sure there is a ready answer to that. For #CTS, the #BCTQ seems a good scale for early change. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @whammert:
Have you looked at all about whether MCID using anchor-based estimates for PROMIS measures in UE vary for improvement vs. decline? Something someone just asked me recently. #JHSJC
Warren @whammert to @BrochaSternOT:
We try to look at MCID as change- up or down, indicating better or worse. But not always clear. We have calculated MCID using the somewhat better of statistically better, but then used that number to look at both directions #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @whammert:
Thanks! Appreciate the response! #jhsjc
Brent Graham @BgTalkinman to @whammert:
That is an extremely important detail that is often not noticed. #jhsjc
Brent Graham @BgTalkinman to @BrochaSternOT:
I am suggesting that it is a much more complex situation than one where the mean is always going to be 50 and the SD is 10. That just doesn’t ring true to me and some of these papers seem support that notion of more complexity. #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
Understood. I think the idea behind PROMIS is really population-based. And at that level, that statement of 50 and 10 may be true. Once we start to get into clinical groups, things are going to change. #JHSJC
Brent Graham @BgTalkinman to @BrochaSternOT:
Agreed! and recognizing that is an important insight. #jhsjc
Brent Graham @BgTalkinman:
Let’s move on to the other article #telemedicine for #CTS diagnosis! #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
I found this paper interesting as it starts to tap into that subgroup variation in score distributions: https://journals.lww.com/jbjsoa/Fulltext/2019/12000/Establishing__Normal__Patient_Reported_Outcomes.12.aspx #JHSJC
Brent Graham @BgTalkinman to @BrochaSternOT:
I will try and remember to check that out. Thanks for bringing it to my attention!
Brent Graham @BgTalkinman:
The authors assessed pts with the #CTS-6 in a #telemedicine consult and then reassessed in person when the patients came for surgery. There was generally very high agreement except for some physical exam components. #jhsjc
Given that only patients with symptoms and physical findings of a certain severity, in other words sufficient to meet the indications for a CTR, were included, does that inject any selection bias into the findings? #jhsjc
What I’m asking is whether the same high agreement would be likely if patients with lesser symptoms of #CTS were assessed? #jhsjc
Warren @whammert to @BgTalkinman:
I think if you are comfortable saying the score is > 12.5, you can do with telemedicine, but if looking for absolute number, not sure how to do 2pd virtually. But that has been my limitation when trying to look retrospectively as I don’t consistently do this. #jhsjc
Brent Graham @BgTalkinman to @whammert:
A score of 12.5 on the #CTS-6 corresponds to a probability that the diagnosis is #CTS of 80% – that is even without the physical exam. Is that high enough? #jhsjc
Brent Graham @BgTalkinman:
The sensory evaluation had the lowest agreement. The authors point out that, even without this component of the clinical evaluation, there was high agreement between the two assessments. Does that have any impact on the generalizability of the findings? #jhsjc
Warren @whammert to @BgTalkinman:
I think it is as I find history often more helpful the exam. But like more date to help with discussion #jhsjc
Brent Graham @BgTalkinman:
We know #CTS is common for most hand surgeons and maybe this approach can be used effectively in that condition. Are there other conditions where a #telemedicine approach might be effective? #jhsjc
Brent Graham @BgTalkinman to @whammert:
Sir Wm Osler said in the 19th Century that history should allow an accurate diagnosis 80% of the time. Are we going to go back to that in a #telemedicine world? #jhsjc
Warren @whammert to @BgTalkinman:
IT does eliminate the idea of laying hands on patients ? #jhsjc
Brent Graham @BgTalkinman to @whammert:
True! #jhsjc
Chris Grandizio @ChrisGrandizio:
It remains difficult to reproduce 2PD…the modification of the ten test did not perform well. May be hard to generalize for ulnar nerve compression. I agree with @whammert on the history component. #jhsjc
Brent Graham @BgTalkinman to @ChrisGrandizio:
So what is the immediate future for this use of #telemedicine as the pandemic may be winding down? #jhsjc
Brent Graham @BgTalkinman:
Even after the pandemic lessens the need for #telemedicine I think it will remain an important part of care, especially in rural settings like where this study was performed. #jhsjc
Perhaps what is needed is more research on diagnostic scales that emphasize hx. I think there’s a lot of room for innovation there. #jhsjc
Chris Grandizio @ChrisGrandizio to @BgTalkinman:
Patients with long travel distances, which we see frequently in our rural center. Early postoperative care. #jhsjc
Warren @whammert:
I think there will be a role and this will vary based on practice locations and types of patients. The more complex, the less the role. #jhsjc
Brent Graham @BgTalkinman to @whammert:
I agree and I will be curious to see how that gets integrated into practices. #jhsjc
Brent Graham @BgTalkinman to @ChrisGrandizio:
Do you think that will be the role of #telemedicine – mostly early post-op care? #jhsjc
Chris Grandizio @ChrisGrandizio:
There may be additional ways to utilize this technology but I think that some of that will be driven by patient demand for these services. If patients continue to want to use it, it will lead to innovation (hopefully for the exam) #jhsjc
Brocha Stern, PhD, OT @BrochaSternOT to @BgTalkinman:
Full circle to PROMs and other self-report measures that include hx, etc. We need higher-quality self-report measures than ever in virtual contexts. #JHSJC
Dr. Tara Packham @TaraLPackham to @BrochaSternOT and @BgTalkinman:
Really enjoyed hearing @ChaoLong talk recently about her work developing low literacy digital versions of hand #PROMs with animated video cues. Timely for virtual environments
Brocha Stern, PhD, OT @BrochaSternOT to @TaraLPackham:
Yes! Dr. Long and @theaviram’s work in that space is really groundbreaking from perspectives of equity and access.
Brent Graham @BgTalkinman to @BrochaSternOT:
Agreed. Diagnosis and diagnostic uncertainty are going to dictate what can and cannot be managed with #telemedicine. There might be important and unexpected ramifications for healthcare access. #jhsjc
Brent Graham @BgTalkinman:
Really great discussion tonight. Thanks to all for participating! Join in again tomorrow at 3 pm EST – or next month. We return with a new session March 8. #jhsjc
