Table 4.

Inter-reader agreement

	PRECISE score (1 to 5)		PRECISE 1–3 vs PRECISE 4–5
Per-patient
	κ value	Percent agreement (%)	κ value	Percent agreement (%)
UCL (n = 40)	0.81 [0.49–1]	80	0.95 [0.86–1]	97
Sapienza (n = 40)	0.55 [0.07–1]	78	0.66 [0.42–0.88]	90
Overall (n = 80)	0.71 [0.37–1]	79	0.83 [0.71–0.94]	90
Per-scan
	κ value	Percent agreement (%)	κ value	Percent agreement (%)
UCL (n = 111)	0.70 [0.31–1]	86	0.74 [0.61–0.87]	93
Sapienza (n = 68)	0.48 [0.07–0.89]	75	0.56 [0.35–0.77]	88
Overall (n = 179)	0.61 [0.30–0.93]	81	0.67 [0.56–0.79]	91

0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; 0.81–0.90, strong agreement and > 0.90 almost perfect agreement; interquartile ranges in brackets [IQR]. At UCL, three different scanners were used: two 1.5-T (Symphony or Avanto, Siemens) and one 3-T system (Achieva, Philips), with a pelvic phased-array coil. At La Sapienza, all exams were performed on a 3-T scanner (Discovery MR750, GE Healthcare) using a 32-multichannel surface phased-array body coil, but in some of the earlier scans, an endorectal coil was also used

UCL University College London