ABSTRACT
OBJECTIVE: The aim of this study was to evaluate the predictive accuracy of the Pittsburgh Impairment Testing Tool (PITT). It was hypothesized that PITT would have a good overall accuracy (>80%) for predicting both ambulation and transfer ability and that overall accuracy of PITT would be higher than that of other scales. DESIGN: A retrospective chart review was used to classify 409 adults with spina bifida according to seven neurological scales. A Naïve Bayes classifier was used to obtain accuracy estimates for predicting both ambulation and transfer ability as a function of each scale. RESULTS: PITT was the only scale demonstrating >80% overall accuracy for predicting both ambulation and transfer ability. While several scales demonstrated >80% overall accuracy in predicting transfer ability, none were useful in predicting inability to transfer. Inability to transfer was difficult for all tools to predict. CONCLUSION: PITT demonstrated good overall accuracy for predicting both ambulation and transfer ability. Sensory and anatomic levels were less useful than motor level in predicting functional ability.
ABSTRACT
BACKGROUND: Gait monitoring is important for rehabilitation but clinic-based measurements provide a brief snapshot of gait patterns. Many consumer-based activity monitors facilitate continuous gait monitoring in daily life but may be subject to measurement errors. This study examines a novel consumer-based activity monitor that is worn on the feet, and therefore not susceptible to inaccurately detecting upper body movements as gait. OBJECTIVE: To evaluate the clinical validity of Sensoria smart socks, a commercially available wireless gait monitoring technology. DESIGN: A method comparison study between the GAITRite, a criterion standard for clinical gait monitoring, and Sensoria smart socks. SETTING: Outpatient rehabilitation clinic. PARTICIPANTS: Thirty individuals capable of supervised ambulation, with or without assistive devices, were recruited through the use of flyers and emails. Data were analyzed for 29 participants. Fifteen participants had no neurologic diagnosis. Fourteen participants had a neurologic diagnosis that could result in gait impairments. Diagnoses included Parkinson disease, stroke, brain injury, developmental delay, and acoustic neuroma. METHODS: Participants completed three gait trials with simultaneous measurements by the smart socks and the GAITRite. MAIN OUTCOME MEASURES: Measurements of step count, cadence, and velocity were compared between the two gait monitoring systems. RESULTS: There was no significant difference in step count measurements between the two systems. Although there was a significant difference in cadence measurements for the total sample group, the mean difference fell within the GAITRite Standard Error of Measurement. There was no significant difference in velocity measurements for the total sample group and the mean difference fell within the GAITRite Standard Error of Measurement. CONCLUSION: These results support the clinical validity of the smart socks for measuring step number and velocity. Further investigation is warranted to determine the efficacy and clinical value of the smart socks system for measuring cadence and for monitoring gait over longer distances outside the clinic. LEVEL OF EVIDENCE: III.
