Adventures in Serial Casting Part II: Review of the Evidence

The term "serial casting" refers to the use of plaster casts applied over time to gradually increase PROM, decrease abnormal tone, and hopefully therefore increase functioning. Typical population for this intervention is either for children with cerebral palsy or adults post stroke or TBI. This (really long) entry aims to examine the evidence behind the intervention by answering multiple questions regarding the intervention.

For this review, I looked at 3 systematic reviews and 1 prospective uncontrolled intervention that was not included in the prior reviews. Of the reviews, one was from an OT publication, one from a PT publication, and one from a physiatry publication, so I feel that all relevant parties were represented. I focused only on articles that emphasized adults, and preferably the lower extremity since that was most relevant to my case. Annotated bibliography at bottom of page.

There are multiple theories for the effectiveness of serial casting and as per Lannin, Novak & Cusick, there is no strong evidence to state clearly which is the correct reasoning. (Categories by Lannin, Novak & Cusick though 1 or more are expressed in each article, further references as noted)
1. Neurophysiological (includes NDT)- casting prevents changes in muscle length, which eliminates excitatory input of muscle spindles and decreases spasticity. The concepts of neutral warmth, proprioceptive input to the limb, and even/constant pressure are also considered to play a part (Saracco Preissner). Per Mortenson & Eng, little evidence exists for the concept of neutral warmth providing the decreased spasticity.
2. Biomechanical- a low-load, long-duration stretch can prevent or correct contractures. By stretching, the Golgi Tendon Organs are stimulated, which stimulate the Ib afferent fibers and then inhibit the alpha motor neurons (Mortenson & Eng).
3. Motor Learning- support proximal joints until control is gained distally. Per Mortenson & Eng, no evidence to support this. I question how this can be applied to LE casts, since they are predominantly applied to the ankle and toe ROM is not a desired outcome. Also, though I have never casted a knee, it seems that if you were working off of this principle to give support proximally until distal control is given, then it might be easier to apply a bledsoe brace locked in position than a serial cast.

Mortenson & Eng outlined these well as: reducing abnormal tone, increasing ROM/reducing contracture (usually PROM is what is measured), and function. Measurement of these effects has been inconsistent across studies. Some studies give a very subjective therapist rating of tone, others use tools such as the Ashworth Scale. ROM was typically measured using standard goniometry, though Mortenson & Eng bemoaned the reliability of ankle goniometry. Mortenson & Eng also disucssed that increases in ROM do not necessarily correlate with increases in function, similarly, Saracco Preissner mentioned that abnormal tone did not necessarily indicate lack of function. "Function" is defined very loosely between studies and various outcome measures are used. Singer et al used the Transfer Dependency Scale, and other studies referenced the FIM. From my own limited experience in research, it would make sense that you would need an adequately sensitive measure and control for confounding factors (concurrent therapy, practice effect, time) to truly indicate whether improvement would be due to casting. I can think of several appropriate measures for UE functional improvement but don't know what has been researched in this direction arleady.

As indicated in the systematic review articles, there is no consistent protocol for serial casting. Lannin, Novak & Cusick identifies a key problem in a consistent protocol- namely that your background rationale will affect your decisions regarding casting time and limb positions. This article listed known indicators and contraindications to tx along with the level of evidence for each, however, a confounding variable to this information is that some factors that were exclusion criteria for some studies were inclusion criteria for others. Given the wide variability, comparison of RCTs was unable to be performed in this review.

Timing is a decision that has wide variability in each study. Saracco Preissner states (but does not reference) that there is no indication how long after injury casting is effective or when a person is too far removed from injury to benefit. It is stated that "most" advocate casting sooner for increased effectiveness, but again this is unreferenced. Length of time wearing the cast was highly variable as well, with the most relevant results being from a study by Pohl in 2002 (referenced in the following section) that showed no difference in results when casts were worn 1-4 days vs 5-7 days.

Protocol was specified for the Singer et al study, and stated to be "standard guideline." Briefly, casts were applied by 2 therapists w/ pt. in prone and knee flexed to 90* after gel pads were applied to bony prominences at risk for breakdown. They were able to insert a custom molded support for metatarsals if clawing of the toes was present. Casting was postponed if pressure areas or skin breakdown was present. Casts were discontinued when no change in PROM was seen over 3 casts, skin breakdown present, or if there was a need to emphasize other treatment prior to discharge.

*Singer et al performed a prospective uncontrolled intervention with 16 adults after aquired brain injury. Statistically significant increases in PROM were noted, and 13/16 improved their transfer ability. However, transfer skills were measured by "4 randomly selected scores" not admission/discharge scores, and could have been affected by concurrent cognitive increases. Of note, 3/4 patients who had limited response had brainstem dysfunction and decerebrate positioning. I believe it is clinically accepted that decerebrate and decorticate positioning is an indicator of poor prognosis medically and with therapy, but I can understand including these patients in this study to just add to that evidence. The authors suggested that the severity of injury rather than the severity of the ankle deformity was the more important predictor of success.

(The following articles were referenced in one or more of the main articles, but I did not follow up to read the full article. Slightly irresponsible, I know, so just take this info at that level)
*Pohl 2002- this study compared 2 groups who were casted for different lengths of time between cast changes- 1-4 days vs 5-7 days. All groups showed an increase in PROM with no difference between groups. Gains were maintained 1 month after cast removal. The Lannin, Novak & Cusick review stated that since there were slightly fewer complications with the group who got casts changed more frequently, there might be an advantage to more frequent changes, however, cost does not appear to have been considered as a factor.

*Mosley 2006- this was an RCT focused on adults with serial elbow casts vs PROM for 1 hr/day. The casting group decreased contracture 22* compared to stretchers when casts were removed, but this decreased to 11* the next day, and the improved effect had almost disappeared by 42 days.

*Booth 1983- This was a retrospective study of 39 patients who had casts s/p head injury. 37% showed and increase in ROM and decreased tone. They observed that pts with brainstem lesions got their casts longer out from injury and took longer to show progress than those who had cortical lesions. Per this study, traditional treatments (PROM, splinting, weight bearing, and PAMs) did not do enough to make an impact on spasticity compared to casting. I'm not sure why this study was given such prominence in the Saracco Preissner article since it was a retrospective study, a bit dated at this time, and its rather unclear how they drew such sweeping consclusions, but again, I did not review the actual article

*Hill 1994- a double-crossover design between traditional tx (PROM, static stretching, splinting) and casting. Improvements were seen for ROM in 14/15 participants and spasticity in 11/15. Stated conclusion is that casting was more effective than the traditional tx, but the gains did not translate into functional gains.

Per Lannin, Novak & Cusick, "There is insufficient evidence to either support or refute the effectiveness of upper limb casting ... There is no evidence of long-term benefits or long-term adverse effects." But this article also stated that there is Level Ib evidence that casting an adult's elbow s/p brain injury increases available extension 1 day after cast removal.

Per Sarraco, since immobilization (such as with spasticity) can cause physiological changes that would impair ADLs, and these changes are reversible, we should treat as able for spasticity. Serial casting has shown some effectiveness in improving ROM and spasticity.

Mortenson & Eng issued "grade" ratings for practice, which I am not familiar with. They say that there are inconsistent measures of "function," so no recommendation can be made on that front. A "Grade C" rating is given to using casts to reduce spasticity secondary to decreased rigor in measurement tools. A "Grade B" rating is given to using casting to improve or prevent loss of PROM, and they state that this is the only outcome with enough evidence to be considered a "best practice." Their studies showed gains of 10.4-26* improvement in ankle ROM, which is statistically significant, however they cautioned that ankle goniometry is not always reliable.

As with nearly all therapy research, there are many questions that need to be addressed. A summary from all articles would include the following, but it is not an exhaustive list:
How does casting work? Is it a biomechanical effect, neurological effect, or both?
What is the best protocol for serial casting? Included in this would be inclusion criteria, positioning of casts, length of time worn, concurrent therapy, post-casting program (including splints and exercises).
What is the comparison to other treatments or lack thereof (especially no stretching)?
How long are gains maintained?
How do improvements after casting translate into function? What specific functinoal gains are seen? Are these gains cost effective and best practice?
Randomized controlled trials comparing the above are also needed.

FULL DISCLOSURE: I am not a professor or professional researcher and do not claim that this is an exhaustive review of the literature surrounding this topic, but a review that I undertook relevant to a specific case. I am not an expert clinician. I do not intend to diminish the efforts or quality of research produced by any of the referenced articles. I would encourage you to do your own review and get necessary training prior to performing this intervention, which may not constitute entry-level practice for all practitioners. Please feel free to comment on additonal relevant research.

Singer, B. J., Jegasothy, G. M., Singer, K. P., & Allison, G. T. (2003). Evaluation of Serial Casting to Correct Equinovarus Deformity of the Ankle After Acquired Brain Injury in Adults. Archives of Physical Medicine and Rehabilitation, 84, 483-491.
A single study in Australia looking specifically at casted ankles in an adult brain injury unit.

Lannin, N. A., Novak, I., & Cusick, C. (2007). A systematic review of upper extremity casting for children and adults with central nervous system motor disorders. Clincal Rehabilitation, 21, 963-976.
A review focusing on UE casts but somewhat confounding as it includes many studies on children with CP

Mortenson, P. A. & Eng, J. J. (2003). The use of casts in the management of joint mobility and hypertonia following brain injury in adults: a systematic review. Physical Therapy, 83(7), 648-658.
Looks at adults after TBI and CVA only but includes studies involving wrist and elbow casting as well.

Preissner, K. S. (2001). The effecs of serial casting on spasticity: a literature review. Occupational Therapy in Health Care, 14(2), 99-106.
This review focuses mainly on management of spasticity, less on ROM gains.

No comments: