The Reliability of Using a Handheld Device to Measure 2D Kinematics: A Systematic Review.

Only eight studies were found, and their agreement statistics were described in words rather than pooled, so there is no single estimate of how reliable these measures are. Almost everyone measured was free of symptoms: of 189 participants, only 12 had had an anterior cruciate ligament reconstruction and 40 had anterior knee pain. There is little here about people with other conditions or more severe disability. Most studies did not say whether raters were blinded to each other's measurements or to their own earlier ones, and four did not describe the raters' experience, so some of the agreement reported may be inflated. Only two of the eight studies reported the camera's resolution and set-ups varied widely, so the review could not work out which device or app feature actually drives reliability. The review measured whether raters agree with each other, not whether the angles they record are correct. Camera alignment error was not tested in these studies. Whether holding the device by hand rather than mounting it matters was never tested directly; it can only be inferred from comparing studies that did it differently.

The authors declare no conflicts of interest. For the ethics statement and consent they state only that they have nothing to report. No funding source is reported anywhere in the text supplied.

Do not read this as showing that phone video gives accurate joint angles. The review only checked whether different raters agreed with each other, and it warns that tilting a camera by 2 degrees can shift a measured limb angle by as much as 1 degree. Agreement between raters is not the same thing as measuring the right angle.