The use of casts in the management of joint mobility and hypertonia following brain injury in adults: a systematic review. (Perspective).

Reduced joint mobility, hypertonia, and muscle contracture are among the many possible listed sequelae following injury to the central nervous system. (1,2) Hypertonia is the increase in joint resistance to passive movement and results from spasticity (ie, hyperactivity of the stretch reflex) and/or changes in the characteristics of the muscular and connective tissues. (3,4) The application of casts as a technique to manage hypertonia and contractures and to ultimately increase range of motion (ROM) has been described. (5-7) Casts offer a temporary, specific, and noninvasive intervention (5) as an alternative or complement to other interventions (eg, pharmacological agents). Casts made of either plaster or fiberglass are most commonly applied over the ankle or elbow joints for 3 to 7 days. (8-11) In some instances, the casts are changed sequentially (ie, serial casts) every few days or every week to incrementally increase joint ROM. (9)

The use of casts to control hypertonia was first described in the 1960s in the treatment of children with cerebral palsy. (12,13) Since that time, casting of the ankle joint has been examined and found to be effective in improving joint ROM and reducing hypertonia for children with cerebral palsy.(14,15) Despite the recent use of botulinum toxin for the management of hypertonia, there appears to be a consensus that other interventions, such as casting, are important in cerebral palsy. (16) Casting is also thought to be an important intervention for controlling contractures and hypertonia in adults with brain injury. (7) Although many different theories exist to support the use of casts in adults with brain injury, the effectiveness of casting for this population has not been firmly established. (17)

Proposed Theories

Spasticity has been defined as a velocity-dependent increase in the tonic stretch reflex (18) (ie, reflex hyperexcitability) and is one of the principal factors affecting rehabilitation following a lesion in the central nervous system (CNS). (3,4) More recently, the broader term "hypertonia" has been used to quantify the resistance to passive movement and results from spasticity (reflex hyperexcitability), in addition to altered mechanical properties of the muscles and connective tissues. (1) Scientists are only beginning to understand the contributions and interactions of reflex hyperexcitability and altered muscle mechanical properties and functions. (1,19,20) Reflex hyperexcitability and altered muscle mechanics can presumably result in the secondary effects of contractures, muscle weakness, reduced joint active ROM, pain, heterotopic ossification, and clonus. (4,21) There are, however, no clinical tests that examine the effect of altered mechanical properties of muscle. Hypertonia can negatively affect an individual's recovery, can interfere with normal motor function, and can cause discomfort; however, reflex hyperexcitability is also purported to have beneficial effects, including the prevention of muscle wasting and deep vein thrombosis. (22)

Theories have been proposed that offer different possible neurophysiological and mechanical mechanisms for the potential positive effects of casting on reflex hyperexcitability and mechanical changes. These mechanisms include the stretch imposed by the cast, the warmth provided by the enclosed structure, and the stability provided by the external structure. (5-7,17) Prolonged stretch has been shown to reduce reflex hyperexcitability. (23,24) Mechanical stretches applied to the elbow flexors for 30 minutes in adults with stroke have resulted in an increased stretch threshold (ie, the joint rotational speed at which the stretch reflex response occurs as measured by electromyography). (23) Similarly, Schmit et al (24) showed a reduction in the stretch reflex--induced torque of the elbow flexors, but not in that of the extensors, after 20 to 30 repetitive 10-second stretches in adults with brain injury. Recently, Tsai et al (25) measured reduced motoneuron excitability as measured by the H-reflex with 30 minutes of passive ankle dorsiflexion using a tilt table in adults with stroke who had increased resistance to passive ankle dorsiflexion as graded by the Ashworth scale. Thus, the prolonged stretch created by casting joints and muscles affected by CNS lesions has been hypothesized to stimulate the Golgi tendon organs and subsequently lb afferent fibers, which create an inhibitory response to the alpha motoneurons. (6,26) The type II muscle afferents also have been postulated to result in inhibition of the alpha motoneurons following a prolonged stretch. (25) Whether this occurs in patients, however, is a matter of conjecture.

The provision of neutral warmth and even total pressure from the circumferential cast may reduce cutaneous sensory input to the spinal cord, thereby reducing the overall level of excitability of interneurons and motoneurons; however, this theory has little supporting evidence. (27) Two research groups (28,29) have evaluated the effect of circumferential pressure through the use of an air splint over 5 minutes in subjects without lesions in the CNS and those with hypertonia secondary to spinal cord injury or cerebrovascular accident (CVA). They found a reduction of alpha motoneuron reflex excitability, as measured by H-reflex amplitude. This effect, however, did not persist when the application of pressure was removed.

Casting, presumably by promoting normal reflex patterns through external stability and opportunities for normalized sensory input, such as weight bearing, may aid the development of normal movement and CNS accommodation. (6,30) At this time, no evidence exists to support these hypotheses regarding casting. (5) Substantial evidence, however, exists that interventions that promote normal weight bearing and sensory input (eg, body support treadmill locomotion) can enhance postural and locomotor control in people with stroke or spinal cord injury. (31,32)

Casting immobilizes the joints and muscle for periods of time, and this can affect muscle and connective tissues mechanics in addition to neural mechanisms of spasticity. (33) In animal models, skeletal muscle has shown the ability to lengthen in response to being held in a lengthened position by casts. (34,35) Immobilization by casting for 4 weeks in a lengthened position resulted in the addition of sarcomeres in series, and conversely, muscles held in a shortened position showed a loss of sarcomeres in series. (34,35) Sarcomere numbers returned to normal within 4 weeks after the immobilization period. (35) According to Lieber and Friden, (36) remodeling in response to immobilization appears to alter the lengthtension properties so that the maximum isometric tension is found at the muscle length corresponding to the angle of immobilization. It is not known, however, whether this type of remodeling is similar to what occurs in the muscle of people with CNS lesions. Muscles in people who have hypertonia have extremely long sarcomere lengths (36) and increased intrinsic stiffness (37) compared with muscles in people without hypertonia. In addition, the extent to which muscles undergo sarcogenesis is likely dependent on the location of the muscles. (38)

Objective of the Systematic Review

Casting has been described as being a labor-intensive, lengthy (casts are worn over several weeks or even months), and relatively expensive method of treatment. (8,27,39) Therapists are believed to require additional skill to be proficient at casting. Moreover, pressure sores must be avoided, and therapists need to monitor the limb for circulatory problems. (7,10) In addition, many controversies exist regarding the theoretical premises for the use of casts in the management of people with CNS lesions and their changes in muscle as well as their joint hypomobility. The drive for evidence-based practice in rehabilitation, whereby therapists use "research evidence together with clinical knowledge and reasoning to make decisions about interventions," (40(p131)) is steadily increasing. The purpose of this article is to report on "best practice" for the use of casting during rehabilitation following brain injury. To achieve this, we sought to systematically examine the level of evidence supporting the use of casting.

Methods A systematic literature search and review was conducted to meet the objective of our systematic review. MEDLINE and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) databases were used to search the literature. These databases were accessed online through the local university's library system in October 2001, May 2002, and January 2003. The search was limited to articles written in English and was conducted for the time period of January 1982 to week 3 of January 2003.

"Casting," "brain injury," and "head injury" were used as key words for the search because we believe they reflect the intervention and condition in question. "Spasticity" also was added as a key word because spasticity management is often cited as the primary reason for using casts following CNS injury, even though authors frequently fail to define the term or they use it to represent different phenomena. (7) The search was conducted using the basic field index (ie, of titles, abstracts, and subject headings) and using the "AND" operator in 2 combinations: "spasticity AND brain injury AND casting" and "spasticity AND head injury AND casting." Using this method, the search combinations generated 8 and 3 articles, respectively, in MEDLINE and 23 and 3 articles, respectively, in CINAHL. After reviewing these articles, we believed that the CINAHL search results were more complete than the MEDLINE results for this particular topic. An additional search of CINAHL, therefore, was conducted using the combination of key words "casts OR serial casting," and 301 articles were identified. The titles and abstracts of these references were then examined, and articles that were not related to the proposed question (eg, those pertaining to orthopedic, burn, or cerebral palsy interventions) were removed. The remaining articles identified from this process were then -reviewed for additional references, and another 2 studies (41,42) were identified.

Following this screening process, 25 articles remained for further review for appropriateness and analysis. Articles were included in the subsequent analysis if: (1) they were experimental or quasi-experimental reports, (2) casting was identified as the independent variable or primary intervention, (3) subjects had what the authors called spasticity and a diagnosis of an acquired brain injury, and (4) subjects were adults. Articles in which subjects were children (under the age of 16 years) or the subjects had a diagnosis of cerebral palsy were excluded. The search was limited to adults because the adaptive response of muscles and tendons to imposed increases in length varies based on age. (43,44) Only 2 case reports (45,46) of casting for pediatric subjects with brain injury were found, which was not enough to conduct a separate review of evidence for this population. We decided to exclude cases of cerebral palsy because children with cerebral palsy have altered muscle properties and movement dynamics, in part, due to abnormally short muscles and long tendons compared with children without cerebral palsy. (47)

A qualitative review process was used to account for the variety of study designs, outcome measures, and analyses used. The rigor (quality) of the studies was identified using a modified version (48) of Sackett's 198149 critical appraisal criteria (random assignment, blinding, intervention monitoring, dropouts, reliability and validity of measurements, confounding factors). (48) When information within an article was not sufficient to ascertain if a criterion had been fulfilled, a "No" rating was given. The level of evidence (eg, level I=large randomized controlled trial, low error risk; level II=small randomized trial, moderate to high error risk; level III= nonrandomized design; level IV=case series, no control; level V= case report) supported by each study design and the grade of recommendation for identified outcomes (eg, A = supported by at least one level I study; B=supported by at least one level II study; C=supported by level III, IV, or V evidence) were then determined as described by Sackett (50) and Butler and Campbell?

Results

We found 13 articles (8,10,11,26,27,39,41,42,52-56) describing results following the use of casting in adult populations with brain injury (Tab. 1). The number of subjects in each study ranged from 1 to 105. With the exception of 4 studies, (8,20,27,56) the primary diagnosis for subjects in all studies was traumatic brain injury (TBI). Two studies (27,56) included subjects with either CVAs or TBIs. One study (8) included subjects with various neurological conditions causing spasticity (TBI, CVA, cerebral hypoxia, cerebral ischemia, and "other"). One article (26) was a case report of a woman who had a cerebral aneurysm. Casts were applied to the ankle plantar flexors in 10 studles, (8,10,11,27,39,41,42,52,55,56) the elbow flexors in 5 studies, (8,26,27,53,54) the knee flexors in 3 studies, (8,11,27) the wrist flexors in 2 studies, (8,53) and the combined ankle plantar flexors and knee flexors in 1 study. (11) Total wearing times varied from 1 to 4 days to a mean of 102 days in the study by Kent et al. (39) Of the 10 studies on ankle casts, 4 studies (41,42,52,56) had subjects bear weight (standing and walking) through the casts. The other 6 studies (8,10,11,27,39,55) did not explicitly state if any subjects were weight bearing while wearing the casts, but a number of conditions were present that suggested that subjects had limited weight bearing, including agitation, (55) a low level of consciousness, (10,11,55) and a low functional level. (10,55)

Ouality Review

The results of the quality review are presented in Table 2. Overall, the results were poor. Ten studies (8,10,11,26,27,39,42,54-56) fulfilled 3 or less of modified Sackett's 7 criteria demonstrating the rigor of the study. Two studies (41,53) met 4 of the criteria, and 1 study (52) met 5 of the criteria. Generally, the intervention was well monitored in the studies reviewed and adherence to the treatment was ensured. Two studies, (11,27) however, were retrospective chart reviews, and, because the issue of monitoring was not explicitly discussed, credit was not given for "intervention monitored" criteria. Finally, all but 2 studies (41,52) were rated poorly on the reporting of measurement reliability and validity. Of authors measuring changes in hypertonia, only Childers and colleagues (54) cited sources to support the reproducibility and consistency of their measurements with clinical scores (Ashworth scale and pendulum test) and made an argument for the use of their indirect measurement (the vibratory inhibition index as a measure of presynaptic inhibition). Ten studies (8,10,11,26,27,41,52,53,55,56) used passive range of motion (PROM) as an outcome measure, but most had little discussion of reliability or validity. The exceptions were Moseley's (41,52) use of standardized, torque-controlled PROM measures with reports of high reliability (intraclass correlation coefficient=.97).

Levels of Evidence

The level of evidence varied from level V (lowest evidence) to level II (Tab. 1). None of the studies reviewed used a true randomized control design. Two studies, (52,53) which were randomized trials with small sample sizes (fewer than 16 subjects in each group), were classified at level II. Three studies (8,10,39) had a cohort design with historic control groups (ie, data obtained in a retrospective review of patient charts or from 2 different groups in time) and were placed at level III. Of the remaining studies, 5 studies (11,27,41,54,55) were case series at level IV and 2 studies (26,42) were case reports at level V. In another study, (56) a single-subject research design was used, but the investigators did not report on the requisite minimum of 3 measures at each phase (baseline, intervention, and withdrawal) .57,58 Therefore, the study was classified as a case report at level V. The number of subjects used was small (<20) in the prospective studies, (41,52-55) but larger in the 2 retrospective reviews (11,27) (25 and 42 subjects), and largest in the study with the 2 historic groups (8) (105 subjects). Although casting was the independent variable in all of the studies, the type of casts used and the wearing time differed. In addition, the length of time of casting within a study was usually individualized, with only 5 studies (8,41,52-54) having a predetermined casting schedule. For one study, (8) however, the timing of cast changes was the independent variable--one historic group (1997-1998) had casts changed every 5 to 7 days, and the other group (1999-2001) had casts changed every 1 to 4 days.

Discussion

Although casting has been used since the 1960s for people with brain injuries, (7,17) only 13 studies relating to adult brain injury were found. Of these studies, only 4 studies (8,52,54,56) had been published in the last 6 years. In addition, all or most of the studies reviewed had problems with several quality criteria (eg, co-intervention, lack of random assignment, lack of blinding, poor measurement reliability and validity). This likely reflects difficulties in performing randomized and well-controlled studies in a clinical context.

All studies reviewed were conducted during the acute or subacute stage of recovery. Subjects therefore were receiving full rehabilitative services and, in some instances, medical and pharmaceutical treatment. As a result, contamination and co-intervention were impossible to avoid. Likewise, randomly assigning subjects to groups that received casting or groups that did not receive casting presents potential ethical and professional dilemmas for those therapists who use casting as a standard intervention for spasticity after brain injury. Moseley (52) and Hill (53) were able to avoid this problem by using a double crossover design, in which all subjects received some form of treatment at all times but in a randomized and controlled order. The strength of these prospective studies was compromised because there were small numbers of subjects. We believe, however, that this might be expected, given the limited number of eligible subjects likely to be available in any one institution during the course of a study on this topic. Unfortunately, the largest and most recent study (8) used historic comparison groups instead of a randomized design, which lowered the study's level of evidence, despite the large number (N = 105) of subjects studied. Finally, blinding subjects to the intervention would have been impossible. In some cases, however, subjects were at a low level of consciousness during the time of intervention.

With the exception of 2 studies, (8,27) a major limitation of all studies was the lack of any follow-up beyond the evaluation at the time of the final cast removal. Lehmkuhl and colleagues (27) reported that, for those subjects with follow-up data, the majority of subjects maintained their gains in PROM over a period ranging from 5 to 684 days after casting. Fifteen of the 19 people maintained or gained further PROM in their elbow joints due to casting, 3 of the 6 people maintained PROM in their ankle joints, and 2 of the 3 people maintained PROM in their knee joints. Pohl and colleagues (8) reported that, for both groups casted in their study, gains in PROM were maintained, with no differences between measurements taken at the end of treatment and 1 month following treatment (Tab. 2). In their case reports, King (26) and Zachazewski et al (42) commented that gains in PROM were maintained, but details on the time frame of these observations were not available.

Effect of Casting

When the 13 studies are grouped by area of measured outcome, 3 effect categories emerge: changes in spasticity, PROM, and "function." After compiling the information obtained from the quality and level of evidence reviews, we believe that comments can be made regarding quality issues and effects of casting for each of these categories. Grades of recommendation can then be given based on the level of evidence demonstrated by the studies, (50) and clinical guidelines can be created for what is "best practice" for the use of casts.

Spasticity. The reduction of spasticity is often cited as the primary goal of casting; however, only 5 of the 12 studies measured properties of spasticity as an outcome. In all 5 studies, there were reports of "improvement" to some extent in spasticity levels following casting. Measuring spasticity, however, presented a challenge. How spasticity was operationally defined and measured and how results were interpreted varied among the 5 studies. Two studies (26,55) relied on reports of clinically recognizable improvements. Hill (53) operationally defined spasticity as both the joint angle (measured by goniometry) at which a stretch reflex was elicited and the ability to perform rapid alternating motions. Before and during casting, Childers and colleagues (54) used electromyographic measurements of H-reflex amplitude, both with and without continuous vibration to the muscle tendon. They reported a reduction of the vibratory inhibition index with casting and attributed these changes to a decrease in motoneuron excitability resulting from the cast. This measure of reflex hyperexcitability (static H-reflex modulation), however, should be interpreted with caution because Levin and Hui-Chan (3) found that the severity of spasticity might not be fully described by reflex testing alone.

In the 5 studies (11,26,53-55) that measured components of spasticity, the rigor of measurements varied. In 2 studies, (53,54) changes reached statistical significance. At least one strong level II study is needed to award a grade B recommendation, (50) and therefore only a grade C recommendation can be supported by this review for the use of casts in reducing spasticity after brain injury.

PROM. In 10 studies, (8,10,11,26,27,41,52,53,55,56) changes in PROM as the outcome of casting were investigated, and all but one (56) reported improvements in PROM following casting. Jones (56) found no difference; however, the purpose of Jones' study was to investigate the effect of casting on gait variables and assistance required to promote external stability during ambulation. Consequently, all the subjects were ambulatory, and change in PROM was a secondary, not a primary, outcome. In contrast, subjects in the other 9 studies were either not ambulating or they had severe restrictions in PROM. The results of improved PROM following casting in these 9 studies were consistent and, in the 6 studies (8,27,41,52,53,55) in which statistical analysis was conducted, were demonstrated to be significant (Tab. 1).

Measurement issues, however, presented areas of weakness for 8 of the studies. The use of a PROM measure must be accompanied by evidence of the reliability and validity of the measurements produced. The reliability of PROM measurements of the ankle obtained by goniometry has been demonstrated to be poor (59,60) and is dependent on a variety of factors (eg, type of disease or injury, type of instrumentation). (61) We believe, therefore, that reliability must be established for each study, not assumed. In order for a measurement to be valid, it must first be reliable, and Moseley's studies (41,52) therefore were the only studies that met the quality criterion for validity.

The changes reported represented what are arguably clinically meaningful improvements for either the positive improvement in PROM or the prevention of loss of PROM. For example, the joint most commonly targeted for casting was the ankle (10 studies) and the reported mean change scores in PROM as assessed by goniometry ranged from a gain of 10.4 degrees to a gain of 26 degrees. Given that the normal dorsiflexion range is 20 degrees, this indicates that a number of subjects commenced treatment in plantar flexion. Changes in the elbow were investigated in 5 studies, (8,26,27,53,54) and reported improvements ranged from 35.2 to 75 degrees. Hill (53) did not report changes. Three groups of investigators (8,11,27) reported positive changes in knee PROM ranging from +15 degrees to +27 degrees. Of the 9 studies (8,10,11,26,27,41,52,53,55) reporting positive changes in PROM, one study (10) used a historical control group, and the authors reported an improvement in PROM in the casted group and also reported a loss of PROM in the historic group that had not received casting. Pohl and colleagues (8) examined the effect of duration of cast change intervals and found no difference on PROM gains between shorter and longer intervals, but they did find a lower complication rate with shorter change intervals. Finally, 2 studies (52,53) of the 9 studies that reported improvements in PROM met the criteria for Sackett's level II evidence (the highest in this review) and received the highest scores in the review of quality. Therefore, a grade B level of recommendation can be given for the use of casts in improving PROM or preventing the loss of PROM that results from the complications of brain injury and subsequent spasticity.

Function. Trends regarding the effect of casting on "function" could not be identified from the inconsistent results identified in this review, and no grade level of recommendation is given. Only 4 groups of investigators (39,52,53,56) measured aspects of "function" or ambulation. Kent et al, (39) Hill, (53) and Jones (56) were unable to find improvements using different functional classification scales. Jones, (56) however, reported improvements in walking speed using visual analysis of data, and Zachazewski et al (42) reported improvements in gait. Perhaps the scales used did not have the sensitivity to measure the impact of casting and associated improvements in increased PROM on function and ambulation. One study (8) used ratings from the Functional Independence Measure to establish similarity in functional level between historic groups, but did not gather rating scores after casting. Improvements in PROM, however, cannot be assumed to directly translate into improvements in function.

Despite recent conclusions that serial casting be considered as an adjunct to therapy to improve functional mobility, (17) therefore, the level of evidence available from research to date does not support this statement nor does it negate this claim. A recommendation for further research would be to include sensitive measures of function that measure levels of activity and participation according to the International Classification of Functioning, Disability, and Health classification. (62)

Conclusion

Of the 3 possible outcomes that have been suggested as a rationale for casting (improved PROM, decreased spasticity, and improved function), only the outcome of improved PROM has sufficient evidence to support the use of casts as current "best practice." Future research on casting for adults with brain injuries should use rigorous research designs and evaluate the effects on spasticity (a term that should use an operational definition shared by researchers), muscle function, functional independence, and activity and participation limitations. Additional research also is needed to determine the optimal timing and duration of cast applications as well as the longitudinal effects of casting on muscle force production and whether gains in PROM are accompanied by muscle atrophy as a result of the immobilization.

Table 1.
Levels of Evidence and Summary of Methods and Results (a)

                                            Level of
Study             Design                    Evidence  Subjects

Moseley, (52)     Randomized crossover; 7   II        9 subjects after
  1997              days each condition                 acute TBI

Hill, (53) 1994   Randomized double         II        15 subjects
                    crossover; 1 month                  after acute
                    each condition                      TBI

Pohl et al, (8)   Case comparison; 2        III       105 subjects
  2002              historic groups and 2               (172 limbs)
                    conditions of casting               with TBI, CVA,
                                                        or other
                                                        conditions
                                                        causing
                                                        spasticity

Kent et al, (39)  Cohort study with         III       18 subjects
  1990              historic control                    after acute
                    group                               TBI

Sullivan et       Cohort study with         III       10 subjects (18
  al, (10) 1988     historic control                    limbs) with
                    group                               acute TBI

Childers et       Prospective case series   IV        8 subjects after
  al, (54) 1999     without control group               acute TBI

Moseley, (41)     Prospective case series   IV        19 subjects (32
  1993              without control group               ankles) after
                                                        acute TBI

Conine et         Prospective case series   IV        10 subjects (18
  al, (55) 1990                                         limbs) with
                                                        acute TBI

Lehmkuhl et       Retrospective case        IV        25 subjects with
  al, (27) 1990     series with follow-up               TBI or CVA

Booth et          Retrospective case        IV        42 subjects
  al, (11) 1983     series                              after acute
                                                        TBI, grouped
                                                        by area of
                                                        lesion and
                                                        type of cast

Jones, (56) 1999  Single-subject research   V         4 subjects with
                    design (ABA); baseline              TBI or CVA,
                    not established                     all with
                                                        hemiparetic
                                                        gait pattern

King, (26) 1982   Case report               V         46-year-old
                                                        woman after
                                                        cerebral
                                                        aneurysm

Zachazewski et    Case report               V         25-year-old man
  al, (42) 1982                                         with acute TBI

Study             Intervention               Outcome Measures

Moseley, (52)     Ankle casts plus           Torque-controlled PROM
  1997              stretching versus no       (known torque of
                    casts                      passively dorsiflexed
                                               ankle and angle
                                               measured using
                                               photography)

Hill, (53) 1994   Serial elbow or wrist      (1) PROM using goniometer
                    casts (or both) 4-6      (2) 3-point scale of
                    casts over 1 month         function
                    versus traditional       (3) Spasticity measured
                    therapy                    as: (a) joint angle
                                               when stretch reflex
                                               elicited and (b) timed
                                               rapid alternating
                                               motions (eg, elbow
                                               flexion/extension)

Pohl et al, (8)   Serial elbow, wrist,       (1) % of normal maximum
  2002              knee, and ankle casts      PROM measured by
                    (duration of cast          goniometer
                    interval 5-7 days        (2) Number of
                    versus 1-4 days)           complications

Kent et al, (39)  Serial ankle casts worn    Holden Functional
  1990              for mean of 102 days       Ambulation
                                               Classification

Sullivan et       Serial ankle casts (mean   Goniometer measures of
  al, (10) 1988     of 5 casts worn for 5-7    PROM
                    days)

Childers et       Elbow casts spanning       Spasticity measured by
  al, (54) 1999     wrist to upper arm,        H-reflex amplitude with
                    worn 3 days                and without vibration
                                               (vibratory inhibition
                                               index)

Moseley, (41)     Ankle casts worn 7 days,   Standardized torque
  1993              plus periods of            controlled PROM
                    stretching into knee       measures (see
                    extension                  Moseley, (52) 1997)

Conine et         Serial ankle casts         (1) PROM using goniometry
  al, (55) 1990     Mean of 5 casts worn     (2) Subjective
                    over 4-64 days             impressions of
                                               spasticity (no scale
                                               used or analyzed)

Lehmkuhl et       21 elbow, 7 knee, and      PROM using pen markings
  al, (27) 1990     14 ankle casts worn a      and goniometry
                    mean 12, 15, and 22
                    days

Booth et          Serial ankle and or knee   (1) PROM using goniometry
  al, (11) 1983     casts worn for 7-92      (2) 4-point scale of
                    days                       spasticity measuring
                                               resistance to passive
                                               movement

Jones, (56) 1999  Bivalved ankle casts with  (1) Temporal distance
                    footplates worn during     gait variables (using
                    the day over 4 weeks       electrogoniometers)
                                             (2) Holden Functional
                                               Ambulation
                                               Classification
                                             (3) Ankle PROM (using
                                               electrogoniometers)

King, (26) 1982   Serial elbow casts over    (1) PROM (measurement
                    16 days and combined       instrument not
                    with traditional           reported)
                    therapy                  (2) Subjective report of
                                               spasticity (no scale
                                               used or analyzed)

Zachazewski et    Ankle casts over 4 weeks   Observational qualitative
  al, (42) 1982                                gait analysis

Study             Results

Moseley, (52)     Improved PROM for casted
  1997              condition (mean difference of
                    15.4[degrees], P <.05)

Hill, (53) 1994   Improved PROM in casted
                    condition (P=.014)
                  Improved spasticity only in
                    group casted first (P=.001)
                  No change in "function"
                    (P=.347)

Pohl et al, (8)   Improved PROM in both groups
  2002              (P <.001), no difference
                    between groups (P=.72)
                  Gains maintained at 1 month
                    follow-up (P=.54, P=.99)
                  Fewer complications with group
                    2 (P=.03)

Kent et al, (39)  No statistically significant
  1990              (P>.05) improvement in
                    ambulation score for casted
                    group

Sullivan et       Mean gain of 3[degrees] in casted
  al, (10) 1988     group versus loss of 15[degrees] for
                    noncasted group

Childers et       Decrease in vibratory inhibition
  al, (54) 1999     index during casting on day
                    3 (decrease of 27.2,
                    P=.033)

Moseley, (41)     Improved PROM (mean gain of
  1993              10.4[degrees], P <.001)

Conine et         Improved PROM (mean gain of
  al, (55) 1990     21[degrees], p <.05)
                  Subjective improvement in
                    spasticity

Lehmkuhl et       Improved PROM (P <.05) for
  al, (27) 1990     elbow (gain 35.2[degrees]); ankle
                    (gain 12.3[degrees]) and knee (gain
                    18.2[degrees])

Booth et          Improved ankle PROM for all
  al, (11) 1983     groups (mean gains from 17[degrees]
                    to 26[degrees])
                  Improved ankle spasticity in
                    37% of subjects
                  Improved knee extension in 2
                    groups (mean gain=27[degrees] and
                    15[degrees]) and knee spasticity in 1
                    group (mean change=0.2)

Jones, (56) 1999  Improved maximum walking
                    speed
                  Inconclusive results of
                    ambulation classification
                  No changes in double support
                    time, swing phase symmetry
                    and ankle PROM

King, (26) 1982   75[degrees] gain in PROM and
                    reported "improvement" in
                    spasticity

Zachazewski et    Reported general
  al, (42) 1982     "improvement" in gait pattern

(a) PROM = passive range of motion, TBI = traumatic brain injury,
CVA = cerebrovascular accident.

Table 2.
Quality Review: Criteria Demonstrating Rigor of Study

                              Quality Criteria

                              Avoided          Random
                              Contamination    Assignment
                              and              to          Blinded
                              Co-Intervention  Conditions  Assessment

Moseley, (52) 1997            No               Yes            No
Hill, (53) 1994               No               Yes            Yes
Pohl et al, (8) 2002          No               No             No
Kent et al, (39) 1990         No               No             No
Sullivan et al, (10) 1988     No               No             No
Childers et al, (54) 1999     No               No             No
Moseley, (41) 1993            No               No             No
Conine et al, (55) 1990       No               No             Yes
Lehmkuhl et al, (27) 1990     No               No             No
Booth et al, (11) 1983        No               No             No
Jones, (56) 1999              No               No             Yes
King, (26) 1982               No               No             No
Zachazewski et al, (42) 1982  No               No             No

                              Quality Criteria

                                                           Reported
                              Monitored     Accounted for  Reliability
                              Intervention  All Subjects   of Measures
                                                           Used

Moseley, (52) 1997            Yes           Yes            Yes (a)
Hill, (53) 1994               Yes           Yes            No
Pohl et al, (8) 2002          Yes           Yes            No
Kent et al, (39) 1990         Yes           Yes            No
Sullivan et al, (10) 1988     Yes           Yes            No
Childers et al, (54) 1999     Yes           Yes            No
Moseley, (41) 1993            Yes           Yes            Yes (a)
Conine et al, (55) 1990       Yes           Yes            No
Lehmkuhl et al, (27) 1990     No (d)        Yes            No
Booth et al, (11) 1983        No (d)        Yes            No
Jones, (56) 1999              Yes           Yes            No
King, (26) 1982               Yes           Yes            No
Zachazewski et al, (42) 1982  Yes           Yes            No

                              Quality Criteria

                              Reported      Total
                              Validity      Number of
                              Measures      Criteria
                              Used          Met

Moseley, (52) 1997            Yes (b)          5
Hill, (53) 1994               No               4
Pohl et al, (8) 2002          No               2
Kent et al, (39) 1990         No               2
Sullivan et al, (10) 1988     No               2
Childers et al, (54) 1999     No (c)           2
Moseley, (41) 1993            Yes (b)          4
Conine et al, (55) 1990       No               3
Lehmkuhl et al, (27) 1990     No               1
Booth et al, (11) 1983        No               1
Jones, (56) 1999              No               3
King, (26) 1982               No               2
Zachazewski et al, (42) 1982  No               2

(a) Interrater intraclass correlation coefficient=.97.

(b) Not discussed by authors, but methodology similar to
well-accepted protractor or goniometry measures.

(c) Not formally discussed, but methodology was based on the
H-reflex amplitude, which is a measure of reflex excitability.

(d) Information not provided.

References

(1) O'Dwyer NJ, Ada L, Neilson PD. Spasticity and muscle contracture following stroke. Brain. 1996;119:1737-1749.

(2) Yarkony GM, Sahgal V. Contractures: a major complication of craniocerebral trauma. Clin Orthop. 1987;219:93-96.

(3) Levin MF, Hui-Chan C. Are H and stretch reflexes in hemiparesis reproducible and correlated with spasticity? J Neurol. 1993;240:63-71.

(4) Umphred DA, ed. Neurological Rehabilitation. 4th ed. St Louis, Mo: Mosby; 2001.

(5) Carlson SJ. A neurophysiological analysis of inhibitive casting. Physical and Occupational Therapy in Pediatrics. 1984;4:31-41.

(6) Watkins CA. Mechanical and neurophysiological changes in spastic muscles: serial casting in spastic equinovarus following traumatic brain injury. Physiotherapy. 1999;85:603-609.

(7) Stoeckmann T. Casting for the person with spasticity. Topics in Stroke Rehabilitation. 2001;8:27-35.

(8) Pohl M, Ruckriem S, Mehrholz J, et al. Effectiveness of serial casting in patients with severe cerebral spasticity: a comparison study. Arch Phys Med Rehabil. 2002;83:784-790.

(9) Leahy P. Precasting work sheet: an assessment tool, a clinical report. Phys Ther. 1988;68:72-74.

(10) Sullivan T, Conine TA, Goodman M, Mackie T. Serial casting to prevent equinus in acute traumatic brain injury. Physiotherapy Canada. 1988;40:346-350.

(11) Booth BJ, Doyle M, Montgomery J. Serial casting for the management of spasticity in the head-injured adult. Phys Ther. 1983;63: 1960-1966.

(12) Groen JA, Domisse G. Plaster casts in the conservative treatment of cerebral palsy. S Afr Med J. 1964;38:502-505.

(13) Westin GW, Dye S. Conservative management of cerebral palsy in the growing child. Foot Ankle. 1983;4:160-163.

(14) Brouwer B, Davidson LK, Olney SJ. Serial casting in idiopathic toe-walkers and children with spastic cerebral palsy. J Pediatr Orthop. 2000;20:221-225.

(15) Brouwer B, Wheeldon RK, Stradiotto-Parker N, Allum J. Reflex excitability and isometric force production in cerebral palsy: the effect of serial casting. Der Med Child Neurol. 1998;40:168-175.

(16) Graham HK, Aoki KR, Autti-Ramo I, et al. Recommendations for the use ofbotulinum toxin type A in the management of cerebral palsy. Gait Posture. 2000; 11:67-79.

(17) Singer B, Singer KP, Allison G. Serial plaster casting to correct equino-varus deformity of the ankle following acquired brain injury in adults. Disabil Rehabil. 2001 ;23:829- 836.

(18) Lance JW. Symposium synopsis. In: Feldman RG, Young RR, Koella WP, eds. Spasticity, Disordered Motor Control. Miami, Fla: Symposium Specialists; 1980:487-489.

(19) Dietz V. Gait disorder in spasticity and Parkinson's disease. Adv Neurol. 2001 ;87:143-154.

(20) Kamper DG, Rymer WZ. Quantitative features of the stretch response of extrinsic finger muscles in hemiparetic stroke. Muscle Nerve. 2000;23:954-961.

(21) Doraisamy P. Spasticity: implications in rehabilitation. Ann Acad Med Singapore. 1992;21:414-420.

(22) Bontke CF. Medical advances in the treatment of brain injury. In: Kreutzer JS, Wehman P, eds. Community Integration Following Traumatic Brain Injury. Baltimore, Md: Paul H Brookes Publishing Co; 1990:3-13.

(23) Al-Zamil ZM, Hassan N, Hassan W. Reduction of elbow flexor and extensor spasticity following muscle stretch. J Neurol Rehabil. 1995;9: 161-165.

(24) Schmit BD, Dewald JP, Rymer WZ. Stretch reflex adaptation in elbow flexors during repeated passive movements in unilateral brain-injured patients. Arch Phys Med Rehabil. 2000;81:269-278.

(25) Tsai KH, Yeh CY, Chang HY, Chen JJ. Effects of a single session of prolonged muscle stretch on spastic muscle of stroke patients. Proc Natl Sci Counc Repub China B. 2001;25:76-81.

(26) King TI II. Plaster splinting as a means of reducing elbow flexor spasticity: a case study. Am J Occup Ther. 1982;36:671-673.

(27) Lehmkuhl LD, Thoi LL, Baize C, et al. Multimodality treatment of joint contractures in patients with severe brain injury: cost, effectiveness, and integration of therapies in the application of serial/inhibitive casts. J Head Trauma Rehabil. 1990;5:23-42.

(28) Robichaud JA, Agostinucci J, Vander Linden DW. Effect of air-splint application on soleus muscle motoneuron reflex excitability in non-disabled subjects and subjects with cerebrovascular accidents. Phys Ther. 1992;72:176-183.

(29) Robichaud JA, Agostinucci J. Air-splint pressure effect on soleus muscle alpha motoneuron reflex excitability in subjects with spinal cord injury. Arch Phys Med Rehabil. 1996;77:778-782.

(30) Bobath B. Adult Hemiplegia: Evaluation and Treatment. 3rd ed. Oxford, United Kingdom: Heinemann; 1990.

(31) Visintin M, Barbeau H, Korner-Bitensky N, Mayo NE. A new approach to retrain gait in stroke patients through body weight support and treadmill stimulation. Stroke. 1998;29:1122-1128.

(32) Harkema SJ, Hurley SL, Patel UK, et al. Human lumbosacral spinal cord interprets loading during stepping. J Neurophysiol. 1997;77: 797-811.

(33) Tardieu G, Tardieu C. Cerebral palsy: mechanical evaluation and conservative correction of limb joint contractures. Clin Orthop. 1987; 219:63-69.

(34) Shah SB, Peters D, Jordan KA, et al. Sarcomere number regulation maintained after immobilization in desmin-null mouse skeletal muscle. J Exp Biol. 2001 ;204:1703-1710.

(35) Tabary JC, Tabary C, Tardieu C, et al. Physiological and structural changes in the cat's soleus muscle due to immobilization at different lengths by plaster casts. J Physiol. 1972;224:231-244.

(36) Lieber RL, Friden J. Spasticity causes a fundamental rearrangement of muscle-joint interaction. Muscle Nerve. 2002;25:265-270.

(37) Sinkjaer T, Magnussen I. Passive, intrinsic and reflex-mediated stiffness in the ankle extensors of hemiparetic patients. Brain. 1994; 117:355-363.

(38) Spector SA, Simard CP, Fournier M, et al. Architectural alterations of rat hind-limb skeletal muscles immobilized at different lengths. Exp Neurol. 1982;76:94-110.

(39) Kent H, Hershler C, Conine TA, Hershler R. Case-control study of lower extremity serial casting in adult patients with head injury. Physiotherapy Canada. 1990;42:189-191.

(40) Law M, Baum C. Evidence-based occupational therapy. Can J Occup Ther. 1998;65:131-135.

(41) Moseley AM. The effect of a regimen of casting and prolonged stretching on passive ankle dorsiflexion in traumatic head-injured adults. Physiotherapy Theory and Practice. 1993;9:215-221.

(42) Zachazewski JE, Eberle ED, Jefferies M. Effect of tone-inhibiting casts and orthoses on gait: a case report. Phys Ther. 1982;62:453-455.

(43) Goldspink G. Sarcomere length during post-natal growth of mammalian muscle fibres. J Cell Sci. 1968;3:539-548.

(44) Tardieu C, TabaryJC, Tabary C, Huet de la Tour E. Comparison of the sarcomere number adaptation in young and adult animal: influence of tendon adaptation. J Physiol (Paris). 1977;73:1045-1055.

(45) Cherry DB, Weigand GM. Plaster drop-out casts as a dynamic means to reduce muscle contracture: a case report. Phys Ther. 1981;61: 1601-1603.

(46) Barnard P, Dill H, Eldredge P, et al. Reduction of hypertonicity by early casting in a comatose head-injured individual: a case report. Phys Ther. 1984;64:1540-1542.

(47) Tardieu C, Huet de la Tour E, Bret MD, Tardieu G. Muscle hypoextenisibility in children with cerebral palsy, 1: clinical and experimental observations. Arch Phys Med Rehabil. 1982;63:97-102.

(48) Roxborough L. Review of the efficacy and effectiveness of adaptive seating for children with cerebral palsy. Assist Technol. 1995;7:17-25.

(49) Sackett DL and Department of Clinical Epidemiology and Biostatistics, McMaster University Health Centre. How to read clinical journals, V: to distinguish useful from useless or even harmful therapy. Can Med Assoc J. 1981;124:1156-1162.

(50) Sackett DL. Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest. 1989;95(suppl 2):2S-3S.

(51) Butler C, Campbell S. Evidence of the effects of intrathecal baclofen for spastic and dystonic cerebral palsy: AACPDM Treatment Outcomes Committee Review Panel. Dev Med Child Neurol. 2000;42:634-645.

(52) Moseley AM. The effect of casting combined with stretching on passive ankle dorsiflexion in adults with traumatic head injuries. Phys Ther. 1997;77:240-247.

(53) Hill J. The effects of casting on upper extremity motor disorders after brain injury. Am J Occup Ther. 1994;48:219-224.

(54) Childers MK, Biswas SS, Petroski G, Merveille O. Inhibitory casting decreases a vibratory inhibition index of the H-reflex in the spastic upper limb. Arch Phys Med Rehabil. 1999;80:714-716.

(55) Conine TA, Sullivan T, Mackie T, Goodman M. Effect of serial casting for the prevention of equinus in patients with acute head injury. Arch Phys Med Rehabil. 1990;71:310-312.

(56) Jones CA. Effect of lower extremity serial casts on hemiparetic gait patterns in adults. Physical Therapy Case Reports. 1999;2:221-231.

(57) Hayes SC, Barlow DH, Helson-Gray R. The Scientist Practitioner: Research and Accountability in the Age of Managed Care. Boston, Mass: Allyn & Bacon; 1999:125.

(58) Backman CL, Harris SR, Chisholm JM, Monette AD. Single-subject research in rehabilitation: a review of studies using AB, withdrawal, multiple baselines, and alternating treatments designs. Arch Phys Med Rehabil. 1997;78:1145-1153.

(59) Elveru RA, Rothstein JM, Lamb RL. Goniometric reliability in a clinical setting: subtalar and ankle joint measurements. Phys Ther. 1988;68:672-677.

(60) Pandya S, Florence JM, King WM, et al. Reliability of goniometric measurements in patients with Duchenne muscular dystrophy. Phys Ther. 1985;65:1339-1342.

(61) Gajdosik RL, Bohannon RW. Clinical measurement of range of motion: review of goniometry emphasizing reliability and validity. Phys Ther. 1987;67:1867-1872.

(62) International Classification of Functioning, Disability, and Health. ICF Short Version. Geneva, Switzerland: World Health Organization; 2001.

PA Mortenson, BScOT, is Occupational Therapist, Children's and Women's Health Centre, Vancouver, British Columbia, Canada, and Graduate Student, School of Rehabilitation Sciences, University of British Columbia, Vancouver.

JJ Eng, PT/OT, PhD, is Associate Professor, School of Rehabilitation Sciences, T325-2211 Wesbrook Mall, Vancouver, British Columbia, Canada VtT 2B5 ([e-mail]), and Scientist, Rehabilitation Research Laboratory, GF Strong Rehab Centre. Address all correspondence to Dr Eng.

Both authors provided concept/research design, writing, and data collection and: analysis. Dr Eng provided project management. The authors thank the Canadian Institute of Health Research for a salary support award to Dr Eng.