Measurement Properties of a 2-Dimensional Movement Analysis System: A Systematic Review and Meta-analysis

Objectives: To critically appraise, compare, and summarize the quality of the measurement properties of the Dartfish software across various populations and motion tasks.

Data sources: Systematic electronic searches were performed in the PsychInfo, Embase, Medline@Ovid, CINAHL, and Google Scholar databases from January 1999 to January 2020.

Study selection: Prospective measurement studies published in English peer-reviewed journals that reported on at least 1 psychometric property (reliability, validity, measurement error) using Dartfish were included. An independent reviewer performed searches and identified studies.

Data extraction: We followed the COnsensus-based Standards for the selection of health Measurement INstruments 2018 guideline for abstracting and assessing data quality. Independent extraction was performed by 2 individual authors. The extracted data involved the author, year, study population, setting, sample size, and measurement properties, as well as information on camera positions, analyzed movement variables, and the corresponding strategy for addressing perspective error.

Data synthesis: In total, 23 studies were included in this review. Studies were pooled to examine inter-rater reliability estimates for different tasks: single-leg squat angle (2 studies, 115 participants; intra-class correlation coefficient [ICC], 0.94; 95% confidence interval [CI], 0.62-0.99), single-leg vertical drop jump angle (2 studies, 94 participants; ICC, 0.92; 95% CI, 0.20-0.99), and vertical drop jump angle (2 studies, 100 participants; ICC, 0.88; 95% CI, 0.83-0.92). Concurrent validity (2-dimensional Dartfish vs 3-dimensional Vicon) was established using the push and release task, single leg stance, and single leg stance with acutely induced dizziness in 45 healthy Parkinson patients. A correlation of 0.59 to 0.98 was reported. For tracking angles across various movements, a measurement error of approximately 10° retest variation was reported in 3 studies.

Conclusion: Dartfish is a reliable software for assessing a variety of tasks across multiple contexts of assessments. Evidence suggests that the estimates of motion obtained with Dartfish are valid for single plane movements.