| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
We want to thank Goodgold for her thoughtful and clinically relevant comments about our study. Our research team appreciates her acknowledgment of the ecological validity of monitoring ambulatory activity with a monitor such as the StepWatch. We believe that activities of "performance," or what a child really does, will be quite helpful in discerning the impact of activity interventions. Goodgold accurately noted large interquartile ranges for the median variable scores displayed in Figures 2 through 5, with much overlap of data between Gross Motor Function Classification System (GMFCS) levels. These observations raise questions regarding the normal distribution of the data and the true population values for each functional level.
In regard to the normal distribution of the raw data, the kurtosis statistic was <1.0 for all variables and functional levels except for the comparison group of children who were developing typically for the variables of (1) ratio of medium to low activity and (2) percentage of time in high activity (Figs. 4 and 5). The analysis outcomes reported at the bottom of the Figures 2 through 5 are based on parametric statistical tests (post hoc univariate analysis of variance with Tukey correction), whereas the box plots are nonparametric per limitation of the graphing software.
We concur with Goodgold that confidence intervals by functional levels would provide useful information regarding the true population values for the functional levels. The application of effect size to describe the degree of difference between functional levels is an interesting approach. In response to this suggestion, we computed the Cohen d for the variable of average steps per day. It ranged from 0.6 to 3.1 between functional levels, suggesting—at a minimum—a moderate to large effect size.1 Similarly, the Cohen d for the ratio of medium to low activity levels to high activity between functional levels ranged from 0.5 to 2.7, again suggesting a moderate effect size for the difference between some functional levels and a large effect size for differences between specific groups.
Goodgold expressed some concerns regarding monitor-wearing adherence. The StepWatch monitor was turned on from the moment it was calibrated to each child. Thus, if there were no step data, either the child was not wearing the monitor or the child was not walking. If a child was wearing the monitor upside down, the output displayed a distinctive pattern and was not included in the analysis. Data analyzed for this project included all minutes of the days that the monitors were worn by all participants when there were step activity data. Excluded from analysis was evening time, where there were no step data (ie, the child was assumed to be sleeping or not wearing the monitor), and days where the monitor was not worn throughout the full day (ie, the first day, when the monitor was calibrated, was an incomplete data collection day).
Despite participants being instructed to wear the monitors at all times except when they were in bed, we do not know directly whether each child put the monitor on as soon as he or she arose in the morning and took it off when he or she went to bed at night. In future studies, the addition of a daily report log of exact times that participants actually put the monitor on and take it off would improve our understanding of adherence. For the present investigation, additional activity data were collected with self-reported physical activity data (using the performance scale of the Activities Scale for Kids [ASKp]) for the 7 days that the participants wore the monitor. In future studies, we will be exploring the relationship of StepWatch and ASKp data across different functional levels.
Although the StepWatch system requires approximately $2,000 for start-up costs, the amount is comparable to the combined costs for a typical item of adapted pediatric equipment (ie, approximately $600 for a therapy roll) or standard pediatric assessments (ie, $950–$1,100 for the Bayley II Developmental Assessment, approximately $250 for the Peabody Developmental Motor Scales). The StepWatch conveys the added benefit of measuring the child's actual behavior, can be used over and over to monitor changes within and among children, and can be used for ambulatory monitoring of people with diagnoses other than cerebral palsy. Because of this, practicing clinicians or clinics may consider it a good investment. The printout of data from the monitor can be used to provide feedback to individuals to improve motivation and can be used for longitudinal tracking on an evaluative level.
Finally, we concur with Goodgold's cautionary note regarding the interpretation of a decrease in daily steps as related to a decline in function. We look forward to future investigations using the StepWatch to develop procedures for normalizing output for leg length as children grow and for normalizing functional ability. These data should improve interpretation relative to youth who are developing typically as well potentially define more clearly the natural ambulatory history of youth with cerebral palsy.
 | Reference |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
