Since the development of the first version of PI-RADS in 2012122 with subsequent versions in 2015 (v2.0)123 and 2019 (v2.1),124 the system has been widely adopted and has standardized the reporting of mpMRI. Multiple studies have confirmed PI-RADS score, either on a per lesion or per patient basis, correlates with likelihood of detecting any cancer and GG2+ cancer. Table 5 summarizes the detection prevalence for any prostate cancer and GG2+ prostate cancer based on the PI-RADS score when 23 studies125-147 identified by the systematic review were pooled. Of the 23 studies, 10 reported on a per lesion analysis126, 128, 129, 131, 133, 134, 137, 138, 142, 145 and 13 reported on a per patient analysis using an index lesion.125, 127, 130, 132, 135, 136, 139-141, 143, 144, 146, 147 While PI-RADS v2.1 provides a structured system for lesion-based scoring approach and has contributed to the wider use of prostate MRI over the last decade, some of the required evaluation criteria remain subjective. As a result, reader variability remains a challenge,145 especially for novice readers.148 Reported measures of interobserver agreement for PI-RADS v2.1 include a weighted kappa value of 0.700 for a study with 5 radiologists of varying experience149 and a Conger kappa value of 0.64 for a study with 6 radiologists of varying experience.150 While interpretative variability remains a limitation, there is evidence that agreement is greater for PI-RADS v2.1 compared to v2.0 and also greater for more experienced readers.151 Due to the paucity of data using PI-RADS version 2.1 specifically, pooled estimates in Table 5 reflect PI-RADS version 1.0 through 2.1. Reader variability is only one of multiple factors that may influence performance differences between sites, including heterogeneity in patient selection, technical factors (e.g., MRI manufacturer and field strength, and use of an endorectal coil), method of prostate biopsy used for pathological correlation, and pathologist expertise and variability. Minimum training requirements to establish reader experience have been proposed and are under investigation.152, 153 Continued evolution of training criteria and further iterative refinements of the PI-RADS should result in greater accuracy and reader agreement. In the interim, clinicians should interpret PI-RADS scores in the context of known local experience and expertise. This statement applies to both initial and repeat biopsy situations. <img src="images/Guidelines/Guideline%20Images/2026%20EDPC/2026%20EDPC-%20Table%205.png" alt="TABLE 5: Prevalence of Prostate Cancer Detection based on PI-RADS Score" title="TABLE 5: Prevalence of Prostate Cancer Detection based on PI-RADS Score" width="700" height="253" class="blockImg" />