A Tape Made from a Strip of Cloth: Stretch Error in Homemade Mid-Upper-Arm Circumference Tapes and the Misclassification It Buys

Where a printed mid-upper-arm circumference (MUAC) tape is not available, a caregiver or community worker may cut a strip of cloth, plastic, or paper, mark it against a ruler, and screen children against the 115 mm and 125 mm cutoffs used in community practice. This paper computes the error budget of such a homemade tape from stated assumptions about elastic strain under the tension a caregiver applies, humidity and ageing drift, hysteresis, and reading parallax. At a reference circumference of 120 mm, a woven cloth strip strains 3.00% under the nominal tension and carries a total bias of -1.80 mm with a standard deviation of 1.82 mm; a plastic strip carries a bias of -1.56 mm with a standard deviation of 1.40 mm; a paper strip, which barely stretches, is instead dominated by humidity drift and non-repeatable set, with a bias of -2.88 mm and a standard deviation of 3.73 mm. The practical cost is misclassification near the cutoffs: a child whose true circumference lies 5 mm from a cutoff is placed in the wrong band with worst-side probability 3.9% by the cloth tape, 0.7% by the plastic tape, and 28.5% by the paper tape, and at 2 mm from a cutoff the paper tape is wrong on its worst side 59.3% of the time. A weekly calibration ritual that costs nothing — wrapping the strip around a food tin of diameter 73 mm and circumference 229.3 mm and re-drawing the cutoff marks — reduces the residual bias of every material to below 0.70 mm in magnitude and brings the cloth and plastic tapes to worst-side error rates of 0.5% and 0.0% at 5 mm. It does not rescue paper: the calibrated paper strip still misclassifies 12.4% of children 5 mm from a cutoff, because calibration removes scale bias and paper's problem is noise, not scale. The conclusions are that a homemade tape should be plastic or calibrated cloth, never paper, and that any reading within about 5 mm of a cutoff should be treated as undecided and re-measured or referred rather than acted on alone.