Table 3.

Measurement error of the three reliability datasets for Siena and for JI. Table shows standard deviations (SD) of the model residuals of the pooled reliability datasets (column 2). The other columns show the median and the 95th percentile of the absolute (unsigned) VL measurements of the pooled data and for each cohort individually.

	pooled			Maclaren		Biberacher		OASIS
	SD	median	95%	median	95%	median	95%	median	95%
BVL Siena	0.28	0.16	0.49	0.15	0.43	0.13	0.93	0.23	0.53
BVL JI	0.34	0.21	0.65	0.23	0.64	0.18	0.92	0.09	0.34
GMVL JI	0.29	0.16	0.51	0.19	0.49	0.19	0.93	0.10	0.33
deep GMVL JI	0.43	0.26	0.81	0.30	0.79	0.27	1.17	0.13	0.36
ThalaVL JI	0.50	0.33	1.02	0.35	0.96	0.34	1.41	0.18	0.46