| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Evidence In Practice |
Julie D Ries, PT, MA, GCS, is Assistant Professor, Physical Therapy Program, Marymount University, Arlington, Va
Rebecca Leonard, PT, is Adjunct Faculty, Physical Therapy Program, Marymount University
Submitted August 27, 2005;
Accepted January 11, 2006
| The purpose of "Evidence in Practice" is to illustrate how evidence is gathered and used to guide clinical decision making. This article is not a case report. The examination, evaluation, and intervention sections are purposely abbreviated.
|
A 28-month-old girl came to our physical therapy department with left-sided spastic hemiparesis, the result of a right cerebrovascular accident (CVA) in utero. Although she had delays in gross and fine motor development because of her hemiparesis, her cognitive and language development were typical for her age. Spontaneous use of her left upper extremity (UE) was markedly impaired in both quantity and quality of movement; however, she was able to use the left UE with prompting when she was given a 2-handed task (eg, holding a "sippy" cup with handles, holding a big ball with 2 hands) or when her right UE was therapeutically constrained.
The child’s parents were well educated and well versed in therapeutic opportunities for improving left UE function, and they approached our university through their primary physical therapist to investigate the opportunity of involving their daughter in a student research project on pediatric constraint-induced therapy (PCIT). Through their support and information network, they had heard many anecdotal success stories related to the use of PCIT with children with sensory and motor impairments in the UE, and they were excited at the prospect of their daughter receiving this intervention. They contacted researchers conducting controlled clinical trials on PCIT in Alabama and New York. They put their child on a waiting list for a program at a university conducting PCIT clinical trials, but they recognized many drawbacks to attending this program, including the financial burden and the need to leave their home.
 | Examination and evaluation |
|---|
She lived in a single-family, 2-level home with her parents. Her mother was 7 months pregnant. The girl was followed closely by a pediatric physiatrist who was in favor of PCIT intervention. The child had received physical therapy twice a week since the age of 6 months, once a week through county services and once a week with a private therapist who was an adjunct faculty member of our physical therapy program. She also received occupational therapy once per week through county services. Based on an examination, she did not need speech pathology services.
She received botulinum toxin type A (Botox
) motor point block injections to muscles of the left upper and lower extremity on 5 separate occasions, the first of which occurred in August 2003 when she was 11 months of age, with subsequent procedures at spans of 4 to 6 months. The most recent Botox treatment was 6 months before she came to our department. The left UE muscles treated with Botox included: teres major, biceps brachii, pronator teres, flexor digitorum superficialis and profundus, flexor pollicis longus, and flexor digitorum longus. The left gastrocnemius muscle was the only lower-extremity muscle that was injected with Botox. Total Botox dose for any single treatment procedure was 140 to 200 units. With the exception of the shoulder girdle injections that preceded improvement in functional creeping, her private physical therapist and physiatrist felt that the effects of Botox for distal UE function were somewhat equivocal. Although the Botox relaxed her hand, there was no noticeable improvement in hand strength, active range of motion (AROM), or function despite the therapeutic focus on these goals. They agreed to defer further Botox injections until after a PCIT intervention.
The child was independently ambulatory with a dynamic molded ankle-foot orthosis with dorsiflexion assist on the left foot. Observational gait analysis revealed the following gait deviations: (1) her left pelvis was retracted throughout the gait cycle, (2) she had a shorter step length on her left side (secondary to increased hamstring spasticity) and a decreased stance time on her left, and (3) she had recurvatum of left knee in mid-stance to terminal stance. Although admittedly not ideal,1,2 observational gait analysis is an acceptable method of documenting pediatric gait parameters according to the Guide to Physical Therapist Practice.3 She experienced more falls than a typical child of her age. Posture of her left UE during gait activities was with scapular retraction and elevation, shoulder extension, elbow flexion, forearm pronation, and wrist and finger flexion, with thumb adduction (ie, her hand was fisted, with her thumb positioned between second and third digits).
Her preferred floor sitting posture was side-sitting, with the left leg in internal rotation and the right leg in external rotation and with the right side of the trunk elongated and left side of the trunk remaining shortened. She was able to transition from the floor to standing and return to the floor independently through the halfkneeling position using the right UE for support. She did not use the left UE spontaneously for support or weight-bearing during transitional movements. After much work with her primary physical therapist, she could creep on her hands and knees, and her left UE posture on the floor during creeping activities was often open. When she was sitting on the floor focusing on a right UE activity or attempting a gross mobility activity, however, her left UE posture was similar to that seen when she was ambulating.
Passive range of motion (PROM) of all joints of the left UE, assessed using techniques described by Norkin and White,4 was within functional limits, with some effort required to passively achieve full elbow extension and full forearm supination because of end-range tightness in the elbow flexors and forearm pronators. We also noted hypoextensibility in shoulder adductors and extensors, making full overhead PROM difficult. Spasticity, defined as a velocity-dependent increase in resistance to PROM, was moderate in the following muscle groups of the left UE: shoulder adductors, shoulder internal rotators, elbow flexors, forearm pronators, wrist flexors, finger flexors, and thumb adductors. This was evidenced by a catch in the range of motion with rapid passive movement (often denoted "dynamic muscle length" or "R1"), felt at one half to three quarters of the available slow PROM ("static muscle length" or "R2") at the given joint. This clinical relationship of R1<R2 has been used to document spasticity in children with cerebral palsy by Boyd and colleagues,5–7 as a modification of a scheme presented by Tardieu and Tardieu.8
The AROM, strength, and motor control of young children with neurological impairment is often assessed through observation of movement in a functional context, rather than with specific tests and measures.3,9 Functional examination of AROM revealed that this child had strong and isolated movement throughout the right UE. She was able to move the left scapula and shoulder actively through space, if the arm was held in front of the body or to the left side, and below 90 degrees of shoulder flexion or abduction. The child was unable to independently get her left arm behind the body (eg, for posterior support) or to the extremes of her overhead flexion and abduction range of motion without assistance. She was able to cross midline. She was able to flex the left elbow fully, and could extend the elbow, although not to neutral position. Although she could not actively supinate the forearm, she could pronate when the arm was placed in a supinated position. She was able to actively extend the wrist and fingers through partial range of motion, having the most amount of difficulty extending and abducting the thumb.
The quality of her left UE movement was impaired, with poor precision and accuracy of movement and significant movement decomposition, with the inability to isolate movements distally. She could not point or poke with her index finger, and she could not move her fingers independently of one another. When attempting distal movements of the left UE, she mirrored the activity with her right UE. She had difficulty grasping and releasing objects: she was successful in approximately 50% of attempts at picking up a 1-inch cube, but dropped the cube before purposeful release. She was unable to pick up shaped puzzle pieces or a magic marker with her left hand despite multiple attempts.
A child’s temperament during the examination process is relevant to understanding how he or she might tolerate an intervention.9 Throughout the long examination process, the girl was attentive and engaged. She was very tolerant of therapeutic constraint of her right UE during assessment of left UE function, because she had been introduced to the general concept of constraint in short stints with her primary physical therapist.
When asked what might constitute "success" of PCIT in their minds, the child’s parents responded that any improvement in the use of the left UE would be considered a success. The mother stated that, if this improved ability led to any increase in spontaneous use of the arm or enhanced the child’s independence in any of her daily skills, especially dressing, it would be a major accomplishment. The parents’ stated goals for the PCIT intervention for their child were: (1) increased spontaneous use of the left arm and hand; (2) improved ability to grasp, hold, release, and manipulate objects with the left hand; and (3) less posturing of the left arm and hand. Interestingly, the parents were in favor of a constraint device that would not be easily removed (ie, a cast). They were concerned that if the constraint was removed regularly within a treatment protocol or if the child knew that the constraint could be easily removed, she would constantly request its removal or, quite likely, remove it herself.
Because of the parents’ interest in having their daughter receive PCIT and because we were aware of some success with constraintinduced therapy (CIT) interventions in adults with a CVA,10–12 we decided to search the literature to find evidence to support the use of PCIT to improve the quality or quantity of the girl’s UE function. We also wanted to determine the optimal parameters of the intervention, if the evidence supported its use.
| Databases searched |
|---|
 TopExamination and evaluation Databases searched KeywordsClinical decisionLiterature follow-upReferences |
|---|
| Keywords |
|---|
|
TopExamination and evaluationDatabases searchedKeywordsClinical decisionLiterature follow-upReferences |
|---|
because it allowed for simultaneous searching of the 3 databases. The databases were selected to represent both medical and allied health literature, as well as to access any related Cochrane Reviews. The search, using the final search string listed above, elicited 19 hits. After elimination of titles that seemed peripheral to PCIT (eg, dorsal rhizotomy, orthotics, seating systems, and respiratory function), 10 of the 19 hits seemed relevant to this case and worthy of further review.13–22 According to the abstracts, the majority of these publications were single13,14 or multiple15–17 case reports. A systematic review on management of UE dysfunction in cerebral palsy18 initially looked to be a promising resource; however, when we evaluated it further, it became clear that the only published evidence on PCIT at the time of publication was a case report,14 and the majority of the discussion in the systematic review on the use of CIT with children was based on speculation from adult studies. The best evidence for PCIT appeared to be in 2 experimental design studies: a small randomized controlled clinical trial of PCIT19 and a crossover study of the forced use component of PCIT,20 which was the subject of 2 critical appraisals.21,22 The strengths and limitations of these studies19,20 and their relevance to our clinical question are discussed below. Taub E, Ramey SL, DeLuca S, Echols K. Efficacy of constraint-induced movement therapy for children with cerebral palsy with asymmetric motor impairment. Pediatrics. 2004;113:305-312.
This study was the only randomized controlled trial that we found that examined the PCIT protocol of constraint, massed practice, and intensive training in the form of shaping activities (progressing level of difficulty, incremental task advancement, breaking tasks into component parts before progressing to completion of the whole task) on young children with cerebral palsy. It was a small study with 9 children each in the experimental and control groups, with ages ranging from 7 months to 8 years. The experimental group children were casted from the upper arm to finger tips with the elbow flexed for 3 weeks, and they were provided with massed practice and shaping opportunities in the context of age-appropriate play activities for 6 hours a day, 7 days a week.
The experimental design with random group assignment and "blinded scoring" were inherent strengths to the study; however, upon reading the text of the article, it became clear that only one of the outcome measures (the Toddler Arm Use Test [TAUT]) was actually scored by blinded raters. The other outcome measures used, the Emerging Behaviors Scale (EBS) and Pediatric Motor Activity Log (PMAL), were tools developed by the research team, administered or scored by nonblinded researchers, and presented without information on their validity or reliability.
The PMAL was the only one of the outcome measurements reported for long-term (3 and 6 months) follow-up testing. It is a structured interview that asks for parental perspective on the quantity and quality of the child’s UE movement, and was adapted from the Motor Activity Log (MAL). The article contained some discussion about validity and reliability data for the MAL; however, a closer look at the literature on the MAL revealed questions about its usefulness in monitoring change over time23 and some doubts about its use as a primary outcome measure because of its subjectivity.24,25 The research group that created the MAL26 conceded that the test has drawbacks typical of all self-report instruments, but contended that the doubt about the MAL’s ability to measure change over time was derived from a flawed interpretation in the study by van der Lee.23 Regardless, there was no information available concerning the psychometric properties of the PMAL, the sole outcome measure reported for long-term follow-up. In addition, the PMAL was not administered to the control group at the 3- and 6-month follow-up dates for comparison with the experimental group. Both of these factors—the lack of information about the PMAL and the elimination of the control group for follow-up comparisons—were threats to the internal validity of the study.
A potential confounding variable in this study was the contrast in the amount of time spent in intervention between the 2 groups. The control group, which participated in a mean of 2.2 hours of physical therapy or occupational therapy per week, was compared with the experimental group, which participated in a rigorous schedule of 42 hours of therapy per week. It begged the question: Was the improved UE functional performance in the children in the experimental group a product of the PCIT intervention, or simply a result of the increased time spent in a therapeutic environment? The authors addressed this question in their discussion and suggested that, because the defining feature of PCIT is "concentrated practice for an extended period,"19(p309) any intervention provided with this intensity and aimed at improving UE function could be potentially characterized as PCIT. They alluded to studies of CIT with adult patients from their own laboratory where they have demonstrated that the factor of time spent in treatment is not the critical component of CIT; however, they did not address the presence of a larger-scale adult study from outside of their lab that compared CIT with a neurodevelopmental treatment (NDT) intervention of equal time intensity and duration. This study revealed a much more modest difference between control and experimental groups.27
Despite establishing that there were no pre-existing differences on any of the testing measures in comparing pretest scores between groups, the authors used a repeated-measures analysis of covariance with pretest scores as the covariates to analyze EBS and PMAL scores. Statistically significant gains in the performance of the experimental group compared with the control group were evident with both of these outcome measures. Authors reported "large and statistically significant between-group differences" in TAUT scores after treatment,19(p308) but they did not present their statistical analysis, only the percent changes in scores for each group.
The conclusions drawn by the authors—that the PCIT "can produce large, sustained gains" and "leads to rapid and large changes in motoric function"19(p310) —were bold statements given the identified issues with the measurement tools, but the findings of the study suggested that the experimental PCIT group did benefit from the intervention in terms of ability to use the UE in the clinical setting and parental perception of improvement in quantity and quality of UE movement. We felt that this study represented the most rigorous evidence available for PCIT and anticipated that, if we proceeded with the intervention, we might model our treatment after this protocol.
Willis JK, Morello A, Davie A, et al. Forced use treatment of childhood hemiparesis. Pediatrics. 2002;110:94-96.
This randomized, crossover study evaluated the forced use component of PCIT. Although Willis et al seemed to use the terms "forced use" and "constraint induced" inconsistently and interchangeably in the text of the article, they suggested that forced use is a component of CIT, with the other major component being intensive training or practice. The experimental group (n=12) was casted below the elbow to the fingertips of the unaffected UE for a 1-month period, thereby eliminating the children’s ability to use the hand and fingers of the unaffected UE functionally; there was no training of the affected UE as part of the intervention. The children in both the experimental group and the control group (n=13) underwent their regular routine of activity and physical therapy and occupational therapy for the duration of the study; the control group actually received more hours of therapy than the experimental group.
The 25 children in the study, ranging in age from 1 to 8 years, had "chronic hemiparesis," but there was no indication of the number of children with acquired hemiparesis versus congenital hemiparesis. This information may be relevant because the response to treatment for those children with learned non-use (ie, once had functional use of the UE) versus developmental disregard (ie, never used the UE functionally) could be different. Neither the severity of impairment of UE function in the children nor the inclusion criteria beyond "chronic hemiparesis" were presented. These omissions of information make generalization of results of study findings difficult.
Of the 25 children initially involved in the study, only 17 returned for the 6-month follow-up, crossover component of the study (68%): 7 of the 12 original members of the treatment group (58%) and 10 of the 13 original members of the control group (77%). This was a fairly large number of "dropouts" for a study. Also, although no medical complications related to casting were reported, the authors said "several" parents withdrew their chil-dren from the study because of their children’s complaints and irritability associated with wearing the cast.
Outcome measures used in this study included the Fine Motor scale of the Peabody Developmental Motor Scales (PDMS) for the involved UE, with the omission of all 2-handed items, and a 2-question yes/no phone interview with parents. The interview questions asked parents whether they believed the cast produced improved UE function in their child and, if so, whether the improvement persisted after cast removal. All 22 parents (100%) reported improvement with casting, and 21 of 22 (95%) felt that the improvement persisted. Although the lack of sophistication of this outcome measure was not necessarily a problem, the subjective nature of the parental report and the lack of comment on validity or reliability of the measurements were potential problems. Use of the PDMS Fine Motor score for the paretic arm was a reasonable choice of outcome measure for the purposes of this study, and the authors appropriately justified their use of this tool. The PDMS Fine Motor scoring was performed by 2 of the researchers who were not blinded to subject grouping; clearly, the study would have been strengthened if the testing had been performed by blinded scorers.
Random group assignments resulted in 2 groups that were apparently different in terms of mean age and mean PDMS Fine Motor pretest scores, but the authors statistically analyzed the group differences appropriately. Differences between PDMS Fine Motor pretest scores between groups were determined to be statistically insignificant, although within-group variability was large. Age was determined not to be a confounding variable of treatment effects when examined with several different statistical methods. The statistical analysis of the outcome data with a 2-factor, repeatedmeasures analysis of variance was appropriate; however, the reporting of the statistical findings was not comprehensive (eg, no variance or standard deviation was provided for PDMS mean scores for the first leg of study).
This article was the subject of 2 critical appraisals,21,22 which made for an interesting comparison of the perceptions of the strengths and limitations of the study. Both critical appraisals identified the study design as an inherent strength and concluded that there is some potential value to the forced-use intervention as it was represented in the study. Hoare21 contended that Willis et al20 provide "weak evidence" that forced use can improve fine motor function in children with hemiparesis. Fetters et al22 concluded that CIT "has the potential to improve" UE skill in children with hemiplegia. Both Hoare and Fetters et al, however, warned readers to view the study results and authors’ conclusions with caution, given the multiple methodological limitations, most of which we identified.
As mentioned earlier, this was not a study of PCIT per se, but of the forced-use component of PCIT, which makes the very specific question posed in one of the critical appraisal papers22— "Is constraint-induced therapy an effective intervention for the treatment of upper extremity dysfunction in children with spastic hemiplegic cerebral palsy?"—somewhat misleading. Perhaps the study by Willis et al20 was the best evidence available at the time to answer this question, but this critical appraisal made no attempt to clarify the limitations of this particular study to answer the question as posed (the study examined forced use with children with congenital or acquired chronic hemiplegia versus CIT with children with hemiplegic cerebral palsy).
There are several case reports in the literature that describe assorted variations of PCIT, or components thereof, for the purposes of increasing UE function. Many of them were identified in the original literature search,13–17 others were found by searching the reference lists of articles from the original literature search.28,29 Full-text articles of each publication were procured and read. These reports varied in many ways, including subject age and diagnosis, specific type of intervention, outcome assessment tools used, and quality of publication. A similarity of all these reports was that they described some level of improvement in UE function of the children following the intervention.
| Clinical decision |
|---|
|
TopExamination and evaluationDatabases searchedKeywordsClinical decisionLiterature follow-upReferences |
|---|
Our clinical experience and understanding of motor learning suggested that an intervention that integrates therapeutic training of the involved UE in conjunction with forced use (casting of the unaffected UE) would be superior to forced use alone. The level of commitment and motivation of the parents, their seemingly realistic hopes for outcomes from PCIT, and their clear understanding of and support for the more intensive protocol were considered. The support of the physiatrist and the primary physical therapist and the temperament of the child also led us to believe that the more intensive protocol, as used by Taub et al,19 would be optimal. In addition, we had the benefit of sharing the burden of time and effort for this more rigorous intervention by involving several physical therapist students in the project. It was anticipated that, as constant faculty supervision was decreased, no one individual would be overwhelmed by the rigor of the intensive therapy schedule. The decision was made to proceed with an intensive protocol of casting for 3 weeks and 6 hours of training per day, following the model proposed by Taub et al.19
For practical reasons, some variations were made to the protocol: intervention was provided 5 days per week, in contrast with 7 days per week in the published protocol; intervention was provided by pairs of students consistently supervised on-site by a faculty member for the first week and a half, with less consistent (every other day on-site) supervision for the final week and a half; and intervention was provided in the child’s home.
| Literature follow-up |
|---|
|
TopExamination and evaluationDatabases searchedKeywordsClinical decisionLiterature follow-upReferences |
|---|
Given the excellent tolerance for the more rigorous protocol demonstrated by the 2 1/2-year-old girl, we felt confident in our decision to use the intensive PCIT protocol. If the literature on less intensive protocols had been available earlier, however, it may have made our decision more difficult. A more recent controlled trial31 using a less intensive intervention is discussed below. Ultimately, we believed that this evidence for a less intensive protocol would not have affected our choice of intervention.
Eliasson AC, Krumlinde-Sundholm L, Shaw K, Wang C. Effects of constraint-induced movement therapy in young children with hemiplegic cerebral palsy: an adapted model. Dev Med Child Neurol. 2005;47:266-275.
This controlled clinical trial was not randomized, a weakness that the authors conceded but felt was necessary for practical issues concerning subject recruitment. The 41 children with hemiplegic cerebral palsy included in the study ranged in age from 18 months to 4 years. The 21 children in the experimental group had 2 hours of focused play activities per day, in which the unaffected UE was constrained. This adapted model of PCIT involved a removable glove with a volar splint that limited the children’s ability to grasp, flex fingers, or oppose the thumb of the uninvolved hand. The play periods were supervised by parents or preschool teachers (after attending an introductory seminar), and these parents and teachers received weekly supervision from a therapist. These sessions were intended to be enjoyable for the children so that they would be motivated to participate, and they were designed to provide challenges for skill development and opportunity for repetition of newly developing motor skills. The 2-hour play periods were carried out 7 days per week over a period of 2 months.
The authors shared demographic data of all subjects effectively in a table, and the experimental and control groups were matched for age and degree of hand function (the pretest score of the outcome measure Assisting Hand Assessment [AHA]). There was great variability in hand function within both groups. All children in both groups continued with their regular therapy regimes for the duration of the research protocol, but there was no indication of whether the 2 groups were similar in the amount of therapy services they received outside of the study.
The intended goal of this study was to assess the impact of 1-hand treatment on performance of 2-handed tasks, and the AHA, a new test of 2-handed function, was used as the outcome measure. The AHA is a 15-minute play session where the child is provided with the opportunity to perform 22 test items; each item is scored with a 4-point rating scale. The sessions for this study were videotaped, and an inherent strength of the study design was the blinding (to subject group) of the individual scoring the AHA. The AHA tests were administered for a baseline measurement before intervention, after intervention (at 2 months), and at 6 months. The authors briefly addressed "evidence of" validity and "adequate" reliability of the AHA but did not provide any supporting data.
A repeated-measures analysis of variance was used appropriately to analyze AHA scores, and effect size (ES) calculations were used to determine overall effect of treatment. The modified PCIT group showed statistically significant increases in AHA scores between pretest and posttest measurements (2-month assessment), but the control group did not. The control group, however, did show a statistically significant increase in AHA scores between baseline and the 6-month assessment, indicating perhaps that maturation affected AHA scores, although the authors of the study did not address this question. The authors determined that the ES for the modified PCIT treatment was large at 2 months (ES=1.16) and medium at 6 months (ES=0.72). Interestingly, no correlation was found between duration of intervention and improvement in AHA score. Although all children in the experimental group were expected to wear the constraint device and engage in practice for 120 hours, this rarely occurred, as evidenced in diaries kept to document actual intervention time. Another point of interest is that older children and children with more marked UE impairment demonstrated the greatest improvements in AHA scores.
This study demonstrated significant improvement in 2-handed function in young children who were treated with an adapted and much less intensive model of PCIT than the Taub et al19 model that we used with our patient. The strengths of the study included the use of matched controls and a blinded rater. Limitations included the lack of randomization and the authors’ apparent disregard of the improvement noted in the control group, which threatens the internal validity of the study. The findings of this study were not persuasive enough that we would have altered our treatment approach had we had access to this information prior to initiating treatment of our patient. We felt justified in concluding this for the following reasons: (1) the parents of the child were concerned that a removable constraint would be more difficult emotionally for the child (and for them if the child were to continually request the removal of the device); (2) the parents were in favor of an intensive intervention provided by individuals with professional knowledge of motor learning and skill acquisition; (3) the authors of the study stated that the older children and those with more marked UE impairment showed the greatest benefit from this intervention, and, although the child we were treating would have met the inclusion criteria for this study, she would not have been considered "older" and likely would not have been considered "markedly impaired"; (4) although the daily intervention time was less than the intensive protocol we chose, the 2-month duration of the intervention was a longer commitment than we were able to make; and (5) the rationale for this adapted model was presumably to make the protocol more tolerable for the child and family, but our patient tolerated the more intensive protocol without any ill effects. Had the resource of the physical therapist students not been available to help provide the more intensive intervention, we might have given this modified protocol more serious consideration.
| Footnotes |
|---|
A case report of the PCIT intervention with the young girl described in this article was presented as a platform presentation at the Combined Sections Meeting of the American Physical Therapy Association; February 2, 2006; San Diego, Calif.
To view this content online, visit www.ptjournal.org
* Upsher-Smith Laboratories Inc, 6701 Evanstad Dr, Minneapolis, MN 55369. 
Allergan Inc, 2525 Dupont Dr, PO Box 19534, Irvine, CA 92623-9534.
EBSCO Information Services, PO Box 1943, Birmingham, AL 35201-1943.
| References |
|---|
|
TopExamination and evaluationDatabases searchedKeywordsClinical decisionLiterature follow-upReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
