Scientific Articles   |    
Functional Outcomes Following Single-Event Multilevel Surgery of the Upper Extremity for Children with Hemiplegic Cerebral Palsy
J. Adam Smitherman, MD1; Jon R. Davids, MD2; Stephanie Tanner, MS1; James W. Hardin, PhD3; Lisa V. Wagner, OTRL2; Laura C. Peace, OTRL2; Mary Ann Gidewall, OTRL2
1 Department of Orthopaedic Surgery, Greenville Hospital System University Medical Center, 701 Grove Road, Greenville, SC 29605
2 Motion Analysis Laboratory, Shriners Hospital for Children, 950 West Faris Road, Greenville, SC 29605. E-mail address for J.R. Davids: jdavids@shrinenet.org
3 Department of Epidemiology and Biostatistics, University of South Carolina, 800 Sumter Street, Columbia, SC 29208
View Disclosures and Other Information
Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor a member of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity.

Investigation performed at the Shriners Hospital for Children, Greenville, South Carolina, and the Arnold School of Public Health, The University of South Carolina, Columbia, South Carolina

Copyright © 2011 by The Journal of Bone and Joint Surgery, Inc.
J Bone Joint Surg Am, 2011 Apr 06;93(7):655-661. doi: 10.2106/JBJS.J.00295
5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case



Outcomes following single-event multilevel surgery of the upper extremity for children with cerebral palsy have not been well described in the literature. Since 1996, all children with hemiplegic cerebral palsy at our institution thought to be candidates for upper extremity surgery have had serial Shriners Hospital for Children Upper Extremity Evaluation performed for both clinical decision making and outcome assessment. The goal of the current study was to determine the functional outcomes, as described by the Shriners Hospital for Children Upper Extremity Evaluation, following single-event multilevel surgery of the upper extremity in children with hemiplegic cerebral palsy.


The study design was a retrospective, case-control series. The case group consisted of forty children with hemiplegic cerebral palsy who underwent upper-extremity single-event multilevel surgery. The control group consisted of twenty-six children with hemiplegic cerebral palsy who had not received any upper-extremity interventions. The spontaneous functional analysis, dynamic positional analysis, and grasp-release analysis sections of the Shriners Hospital for Children Upper Extremity Evaluation were compared between the two groups.


The operative and nonoperative groups were comparable with respect to age (p = 0.09), sex (p = 0.97), initial spontaneous functional analysis scores (p = 0.37), dynamic positional analysis scores (p = 0.73), and grasp-release analysis scores (p = 0.16). For the single-event multilevel surgery group, significant improvements were noted for the mean spontaneous functional analysis score (p < 0.0001) and the mean dynamic positional analysis score (p < 0.0001), but not the mean grasp-release analysis score (p = 0.75). For the nonoperative control group, no significant changes were noted for the mean spontaneous functional analysis score (p = 0.89), the mean dynamic positional analysis score (p = 0.98), or the mean grasp-release analysis score (p = 0.36). Significant differences were noted between the single-event multilevel surgery and nonoperative control groups for the mean changes in the spontaneous functional analysis score (p = 0.01) and the mean change in the dynamic positional analysis score (p < 0.0001), but not the mean changes in the grasp-release analysis score (p = 0.56).


Children with hemiplegic cerebral palsy showed significantly improved dynamic segmental alignment and, to a lesser degree, spontaneous use of the upper extremity following single-event multilevel surgery compared with a comparable nonoperative control group. However, the grasp-release ability did not significantly improve in either the operative or nonoperative group.

Level of Evidence: 

Therapeutic Level III. See Instructions to Authors for a complete description of levels of evidence.

Figures in this Article


    cerebral palsy ; arm
    Sign In to Your Personal ProfileSign In To Access Full Content
    Not a Subscriber?
    Get online access for 30 days for $35
    New to JBJS?
    Sign up for a full subscription to both the print and online editions
    Register for a FREE limited account to get full access to all CME activities, to comment on public articles, or to sign up for alerts.
    Register for a FREE limited account to get full access to all CME activities
    Have a subscription to the print edition?
    Current subscribers to The Journal of Bone & Joint Surgery in either the print or quarterly DVD formats receive free online access to JBJS.org.
    Forgot your password?
    Enter your username and email address. We'll send you a reminder to the email address on record.

    Forgot your username or need assistance? Please contact customer service at subs@jbjs.org. If your access is provided
    by your institution, please contact you librarian or administrator for username and password information. Institutional
    administrators, to reset your institution's master username or password, please contact subs@jbjs.org


    Accreditation Statement
    These activities have been planned and implemented in accordance with the Essential Areas and policies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint sponsorship of the American Academy of Orthopaedic Surgeons and The Journal of Bone and Joint Surgery, Inc. The American Academy of Orthopaedic Surgeons is accredited by the ACCME to provide continuing medical education for physicians.
    CME Activities Associated with This Article
    Submit a Comment
    Please read the other comments before you post yours. Contributors must reveal any conflict of interest.
    Comments are moderated and will appear on the site at the discretion of JBJS editorial staff.

    * = Required Field
    (if multiple authors, separate names by comma)
    Example: John Doe

    Jon R. Davids, MD
    Posted on May 20, 2011
    Drs. Davids and Hardin respond to Dr. Nichols
    Chief of Staff, Shriners Hospitals for Children, Greenville, South Carolina

    We appreciate Dr. Nichols’ careful review of our article. His letter to the editor raises three issues. First, what do you call the study on which we reported? In the strictest sense, he is correct to challenge the study designation as case-control. A case-control study design is one in which investigators look through old data to locate people with a disease or condition. These people are then compared to matched persons without the disease or condition. In our study we looked through old data to locate people (children) with a condition (hemiplegic type cerebral palsy). We noted two different treatments for that condition and compared the outcome of the two groups. We chose, and the journal agreed, to describe our study as “case-control” based upon the argument that we identified a group of patients treated a certain way and then matched them to patients that didn’t get treated that way. From a statistical science perspective a more accurate description of our study would be a “retrospective study” or a “non-randomized retrospective study” where the non-randomized refers to the fact that there was no study design in which patients were randomized to either of the two treatments. However, from a clinical medicine perspective, and certainly from an orthopaedic surgery perspective, comparison of a treated group to an untreated group (which reflects the natural history of the disease process) is of practical value in guiding clinical decision making. Rightly or wrongly, study designs such as ours are frequently referred to, and recognized by clinicians, as “case-control” in the orthopaedic literature. We believe that the goal of describing a study’s design should be to facilitate the readers’ understanding and analysis of both the data and the authors’ conclusions.

    Second, should we have chosen a cutpoint in our statistical analysis? There were multiple options for the examination of the data in our study. Turning the continuous measures of outcome into binary measures is favored among some investigators because, at some point, both investigators and readers must make a clinical (diagnostic/prognostic) decision. While such a black and white decision must be faced at some point, this certainly doesn’t mean that the entire analysis should be under such limitations. Adding analyses based on cutpoints necessarily requires certain choices about the utility function to be maximized. That is, do we choose a cutpoint that maximizes sensitivity, specificity, Youden’s index (sensitivity plus specificity), positive predictive value, or some other value? Unless there is a clear answer for a particular statistic based on clinical motivations, there are many possible ways to proceed. Properly applied statistical analysis brings order and clarity to data derived from the messy world of clinical medicine. When the outcome is as clear as life or death, a binary approach makes sense. Unfortunately the outcome(s) following single event multilevel surgery of the upper extremity in children with hemiplegic cerebral palsy is (are) multidimensional and therefore much more difficult to characterize. In this setting, is the subject with an outcome score one point below the cutpoint really that different than the subject with an outcome score one point above the cutpoint? We do not think so, and to imply this by performing a binary analysis would be misleading. Without clear clinical motivations, we would worry about the clinical utility of any “rule” that would be implied by the establishment of a cutpoint.

    Finally, should the journal/authors make data available to readers? We favor maximum transparency at all stages of the peer-review publication process in the medical literature. Open access to the data should promote critical analysis by readers and support subsequent analyses by future investigators. Electronic publication platforms can greatly facilitate access to data sets that could not be included in traditional paper-based print formats. Like Dr. Nichols, we encourage our profession’s journals to embrace this opportunity afforded by online publication.

    David P. Nichols, MD, PC
    Posted on May 20, 2011
    Case-Control vs. Means Comparison
    Orthopedic Surgeon

    To the Editor:

    I read with interest the article, "Functional Outcomes Following a Single-Unit Multilevel Surgery of the Upper Extremity for Children with Hemiplegic Cerebral Palsy”, by Smitherman et al. (2011;93:655-61). The study design was erroneously labeled as a “retrospective, case-control series”. The authors labeled the “control group” as comprised of the “26 children with hemiplegic cerebral palsy who had not received any upper extremity interventions”. Rather than being the control group, this group is the non-exposed group. Those patients exposed to the surgery were not the case group, they were the exposed group. Cases and controls were not defined by the authors.

    Rather than being a case-control study, this study is a means comparison. The authors select several outcomes including SHUEE score, dynamic positional analysis score, and grasp release score, then compare the means of change in score in the operative and nonoperative exposure groups. Unfortunately, the authors did not select a cut point which would determine a good or bad outcome (the cases and controls). Merely comparing means does not allow us to make an inference about which treatment would more likely lead to a good outcome. The authors should have taken it upon themselves to determine either from previous research or from their own rationale what the cut point for the cases and controls should have been in each of their selected outcomes. Then, an odds ratio could have been calculated along with its confidence interval which would guide the reader in selecting operative versus nonoperative treatment. This study sheds little light on the research question about which treatment would be most likely to lead to a good outcome for this class of patient. In probabilistic terms, a case-control study would answer the question “Which outcome group was more likely to have been exposed to operative or nonoperative treatment?”

    In failing to make this cutpoint determination for cases and controls, the authors of the article and the editors of the Journal of Bone and Joint Surgery have the responsibility to provide the raw data in spreadsheet form to the reader allowing them to analyze the data and make the determination about treatment selection on their own. It is my hope that JBJS will in the future provide electronic data sharing for its readers. This would assist them in making evidence-based decisions regarding treatment for their patients.

    Related Content
    The Journal of Bone & Joint Surgery
    JBJS Case Connector
    Topic Collections
    Related Audio and Videos
    PubMed Articles
    Clinical Trials
    Readers of This Also Read...
    JBJS Jobs
    New York - Icahn School of Medicine at Mount Sinai
    District of Columbia (DC) - Children's National Medical Center
    S. Carolina - Department of Orthopaedic Surgery Medical Univerity of South Carlonina
    W. Virginia - Charleston Area Medical Center