Past and current use of walking measures for children with spina bifida: a systematic review

Objectives: To describe walking measurement in children with spina bifida and to identify patterns in the use of walking measures in this population.

Data sources: Seven medical databases-Medline, PubMed, Embase, Scopus, Web of Science, CINAHL, and AMED-were searched from the earliest known record until March 11, 2014. Search terms encompassed 3 themes: (1) children; (2) spina bifida; and (3) walking.

Study selection: Articles were included if participants were children with spina bifida aged 1 to 17 years and if walking was measured. Articles were excluded if the assessment was restricted to kinematic, kinetic, or electromyographic analysis of walking. A total of 1751 abstracts were screened by 2 authors independently, and 109 articles were included in this review.

Data extraction: Data were extracted using standardized forms. Extracted data included study and participant characteristics and details about the walking measures used, including psychometric properties. Two authors evaluated the methodological quality of articles using a previously published framework that considers sampling method, study design, and psychometric properties of the measures used.

Data synthesis: Nineteen walking measures were identified. Ordinal-level rating scales (eg, Hoffer Functional Ambulation Scale) were most commonly used (57% of articles), followed by ratio-level, spatiotemporal measures, such as walking speed (18% of articles). Walking was measured for various reasons relevant to multiple health care disciplines. A machine learning analysis was used to identify patterns in the use of walking measures. The learned classifier predicted whether a spatiotemporal measure was used with 77.1% accuracy. A trend to use spatiotemporal measures in older children and those with lumbar and sacral spinal lesions was identified. Most articles were prospective studies that used samples of convenience and unblinded assessors. Few articles evaluated or considered the psychometric properties of the walking measures used.

Conclusions: Despite a demonstrated need to measure walking in children with spina bifida, few valid, reliable, and responsive measures have been established for this population.