| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters and Responses |
 | mCIT: The Whole Is More Than the Sum of Its Parts |
|---|
We agree that the question of why and how home-based efforts drive recovery deserves greater attention. The limitations of practicing tasks in the clinic (rather than in the patient's home environment), and trends toward diminishing in-clinic contact time, argue for protocols that place greater emphasis on structured, home-based practice scenarios. Consistently, the mCIT home component features highly structured "homework" sessions occurring for 5 hours/day in patients’ real-world environments. These sessions are a purposeful extension of practice provided in the mCIT clinical sessions. This is a stark contrast to the comparatively unstructured CIT restraint component. The careful structuring of the mCIT home program might influence the comparatively larger MAL outcomes observed after mCIT, as suggested by Wolf.
However, there are additional, favorable aspects of the mCIT protocol that also likely conspire to produce comparatively larger outcomes than CIT. These include:
MAL Maladies: A Caution About Wolf's Caution
Wolf also reported that the MAL Amount of Use section might not be reliable,9 suggesting that this MAL scale "should be used with caution." This could be a potential concern for our study results, because we used this version of the MAL as a measure of affected arm use. However, as in pilot mCIT work leading to the current trial,10 we corroborated MAL amount-of-use findings with activity monitor data. These devices have been shown to provide valid, objective measurements of affected arm use.11 Our activity monitor analyses confirmed that only subjects receiving mCIT showed significant affected arm use changes; they will be reported separately. Given the converging nature of our activity monitor findings and their established track record as a measure of affected arm use, we are confident in our MAL data trends. This finding should address Wolf's stated concern.
We also wish to point out that this "unreliable" version of the MAL was the sole affected-arm–use measure in the principal EXCITE publication.1 This MAL version also has been used in other CIT trials.12 Unlike our study, none of these CIT publications used surrogate measures of affected arm use (eg, activity monitors). Based on Wolf's concern about suboptimal MAL reliability, and in the absence of surrogate affected-arm–use measures in these reports, one could reason that the amount of use findings reported in these CIT studies warrant real caution.
Similarly, this version of the MAL was used in a recently published EXCITE follow-up study by Wolf and colleagues.13 This study reported sustained affected-arm–use patterns associated with CIT participation, as measured 24 months after treatment. The validity of conclusions made by such long-term studies hinges on the use of measures with high test-retest reliability. In light of Wolf's caution about the MAL's suboptimal reliability, one could reasonably question the validity of conclusions rendered by his own EXCITE follow-up study.
It is likely that conclusions based on MAL findings are valid in our study, as well as in Wolf's above-mentioned efforts. The above shortfalls from EXCITE work illustrate the importance of exercising caution, and of garnering more empirical support, before taking Wolf's remarks to heart.
Limitations of CIT: Still There
One of Wolf's most perplexing remarks was his assertion that therapists' largest CIT concern is its reimbursement. It is almost inconceivable that a practicing therapist's principal concern with a 6-hour/day therapy would be whether his or her facility gets reimbursed for it. What about the time that therapists have to invest? Patient adherence and "buy in" with such an intensive protocol? Patient fatigue? As noted earlier, Wolf himself concedes that "participants simply did not possess the endurance to train for the entire 6 hours [of CIT]." He also noted that patients in the EXCITE study were able to handle only about 3 hours/day of clinical therapy—a finding confirmed by a recent EXCITE publication.14 Thus, Wolf's own data suggest that CIT concerns transcend its mere reimbursement.
Several groups15 also have raised concerns with the high-duration CIT clinical and home-based components and their practical implementation. For example, our survey of subjects’ and therapists’ opinions about CIT16 found that the majority of patients would not want to participate in CIT, and more than 80% of therapists reported that their facilities could not administer such an intensive protocol. Fatigue, frustration, adherence, reimbursement, and loss of independence were mentioned as concerns by patients and therapists in this survey. Other CIT researchers have cited: (1) poor compliance with unsupervised constraint17; (2) burns, muscle soreness, and discomfort in the affected arm18,19; and (3) pain increase in 4 of the 5 patients in the constraint group.17 These data further suggest that CIT concerns transcend mere reimbursement.
Wolf acknowledges that CIT's components are high in duration. He then attempts to defray this limitation by citing an exit interview among EXCITE participants, in which subjects reported that the high-duration nature of the CIT components was "helpful." Although interesting, that study has an inherent selection bias, stemming from the fact that subjects who responded that the CIT duration was "helpful" had already been exposed to that intervention. Unless Wolf and his team also told subjects that other, efficacious durations were available, CIT, as administered in EXCITE, would constitute their only frame of reference from which to draw their opinions.
Wolf also correctly observes that we have previously included only part of a quote from his 1999 case study.20 He tries to "set the record straight" by including the rest of the passage. In our estimation, the fact that an apparently motivated patient has consented to a CIT study, yet does not adhere to one of its signature parameters, is striking. This is why we have historically not included the remainder of the quote.
In summary, Wolf's arguments regarding CIT limitations contradict existing data, including his own. They also do little to address the fundamental limitations with regard to CIT's high duration or its practical implementation, as cited by several groups. The fundamental CIT shortfalls remain.
Blinding the Therapists and the Subjects to Group Assignment
Wolf also questions why we would blind therapists to study hypotheses, citing 2 reasons not to blind therapists: (1) "clinicians’ knowledge about the intent of a study is important for fostering proactive involvement," and (2) "participating clinicians’ inherent intellect and training would allow them to easily deduce the hypotheses." With regard to point 1, it has been our experience that therapists are excited about the opportunity to learn the literature and techniques surrounding mCIT. As one might expect, neither therapists’ involvement nor their enthusiasm have been diminished by whether they know the specific study hypotheses, as Wolf suggests.
With regard to point 2, the main purpose of blinding therapists to study hypotheses was to diminish possible subjective biases or expectations as they administered treatment. Importantly, this technique has been used in clinical trial design for decades. It is certainly possible that study clinicians could form opinions about study hypotheses or anticipated patient group responses; this concern is present in virtually any therapy trial. However, such possibilities should not diminish the impetus to rigorously conduct scientific trials under the most controlled conditions that can be constructed.
The same also can be said for subjects: they could certainly gather information about the various therapy groups. From this information, they could then form expectations about their responses to the treatment condition to which they are assigned. In the interest of scientific rigor, we prefer to attempt to control this possibility to the best of our ability by limiting the hypothesis-related information that we provide to them.
The Clinical Portion of mCIT Has Been Reimbursed, Using Existing CPT Codes, for Several Years
Wolf correctly notes that mCIT does not have its own CPT code. He then suggests ways that our wording could be adjusted to be adequately descriptive of mCIT's reimbursement. In this spirit, we offer the phrasing used in the heading of this section as a more precise (albeit awkward) semantic alternative.
More importantly, Wolf is inaccurate in his claim that each CPT code is "defined for a specific intervention." This is an important point, because Wolf uses the above argument as one of his bases for his assertion that the CIT intervention needs its own CPT code. As a point of fact, neither specific techniques (eg, proprioceptive neuromuscular facilitation, neurodevelopmental treatment) nor general theoretical approaches to interventions (eg, motor relearning) have their own CPT codes. Instead, various techniques are grouped into a single CPT code according to the larger, common functional goals that the techniques serve. For example, a variety of activities intended to develop strength and endurance, range of motion, and flexibility are covered under the therapeutic exercise CPT code (97110).
In the case of mCIT clinical sessions, the signature clinical ingredient is motor learning–based, task-specific practice. During these sessions, patients repetitively engage in deficit components of motor skills that they want to relearn in a systematic fashion, under the guidance of the therapist (ie, part-task practice). Such motor learning/motor control–based techniques are commonly used by therapists working with patients with stroke,21,22 and they are regularly reimbursed using CPT codes listed in Wolf's letter.
Given the above, although the CIT family of therapies and its accompanying techniques (eg, shaping) are a novel advance, we would disagree that there is a need for creation of a separate CPT code for CIT, which Wolf has been pursuing for several years. We agree with Wolf's argument that the existence of such a code would stimulate important information regarding CIT's true clinical utilization and effectiveness (ie, how well it works in real-world environments). However, neither the absence of a code, nor the possibility of exciting research questions that could emanate from the existence of such a code, is reason enough to create a separate CPT code for CIT. The EXCITE team's own writings23,24 also note that large gaps remain in our knowledge about CIT. Such significant gaps—and the amount of evidence showing that a variety of practice schedules are efficacious (including availability of several, efficacious, lower-duration CIT variants)—would seem to preclude development of a CPT code for CIT at the intensity that Wolf advocates. With regard to the latter point, we are hard-pressed to imagine why payers would reimburse a therapy requiring 6-hour/day clinical sessions when efficacious, lower-duration alternatives are being validated. In light of the innumerable factors contradicting inception of such a code, one might say that Wolf's advocacy of a CPT code for CIT is "putting the cart before the horse."
We agree with Wolf's assertion that research questions should not be molded to fit payer parameters. However, we also wish to respond to Wolf's implicit speculation that the mCIT clinical component was "molded" to fit parameters of the health care system. Not only does this statement lack factual bases, but Wolf also attempts to use this remark as a foundation for several of his points. The formation of the mCIT clinical component was based on decades of motor-learning research examining the nature of practice (described earlier) and on neurophysiologic evidence showing that neural and motor changes can occur in as little as 15 to 30 minutes. Interestingly, around the same time that we were piloting mCIT, one of Wolf's eventual EXCITE colleagues wrote that "any technique that induces a patient to use an affected limb...should be considered therapeutically efficacious. This factor is likely to produce the use-dependent cortical reorganization...."26(p243) This quote suggested to us that it is not just the intensity with which the techniques are provided, but also the nature of the techniques themselves (and, specifically, whether they induce repetitive, task-specific arm use) that influences neural and motor change.
Collectively, data from these converging sources suggested to us that functional changes can be observed without intensive clinical practice. Thus, although it is fortuitous that the clinical portion of mCIT is reimbursed, a wide array of scientific literature was consulted in conceiving mCIT's relatively shorter time parameters.
Future Directions: Moving Both the Cart and the Horse Forward
For decades, drug and device studies have conformed to a "tried and true," phased clinical trial process. Through this process, the most efficacious and safe dosing level of a therapy is first identified, followed by larger efficacy and safety studies, and finally a multicenter, controlled efficacy trial (phase III). Unfortunately, rehabilitation researchers, including those involved with CIT, largely have not followed this validated route, especially as it relates to solidifying dose-response issues, a shortcoming expressed as a concern in a recent NIH report.27 We, therefore, agree with Wolf that more research on optimal dosing, timing, adherence, neurophysiologic impact, and other variables must still be examined in relation to CIT and mCIT. In other words, we agree that Wolf's metaphorical "horse" should advance.
However, significantly more information is now available about this family of therapies than when the EXCITE Trial was conducted. This is complemented by a wealth of data showing that mCIT is efficacious in all stages poststroke,10,28–32 with a treatment effect that is comparable to CIT.33 As Wolf noted, we also have shown that a variety of therapies (eg, mental practice, electromyographic-triggered electrical stimulation) can act as "bridges" to mCIT participation for patients who are more impaired. Other therapies (eg, telerehabilitative approaches, botulinum toxin A) can be co-administered with mCIT to optimize its efficacy and accessibility. The neural mechanisms underlying mCIT also have been identified.34 Collectively, these factors support advancement toward a phase III mCIT trial, especially given that these same sorts of issues were resolved at a comparable level when the EXCITE trial moved forward. The "cart," therefore, also should advance, not at the expense of the "horse," but in addition to it.
S Page, PhD, is Associate Professor, Department of Rehabilitation Sciences and Department of Physical Medicine and Rehabilitation, and Neuro-sciences Scholar, Institute for the Study of Health, University of Cincinnati Academic Medical Center, Cincinnati, Ohio
P Levine, PTA, is Research Associate, Department of Rehabilitation Sciences, University of Cincinnati Academic Medical Center
Footnotes
This letter was posted as a Rapid Response on April 9, 2008, at www.ptjournal.org.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
