Implementation of a biparametric prostate MRI protocol with a recall system for intravenous contrast: an analysis of recall rates and comparison of PI-RADS category distribution

Objectives The prostate magnetic resonance imaging (MRI) protocol changed from a contrast multiparametric MRI (mp-MRI) to a non-contrast biparametric (bp-MRI) protocol at a single centre. This study describes the recall rate, median time to recall, and PI-RADS distribution pre-/post-implementation. Methods This retrospective descriptive study involved a consecutive sample of all 3 Tesla (3T) prostate bp-MRIs following the protocol change (01April2023-23May2024) to determine the proportion of bp-MRIs recalled, the indications for recall, and the median time until recall MRI (Part 1). A subset of these prostate MRIs was compared to a consecutive sample immediately prior to the change in protocol to compare the distribution of PI-RADS categories (Part 2). The distribution of MRI reporting frequency by radiologists was compared pre-/post-implementation (Part 2). Inclusion criteria: 3T bp-MRI protocol, out-patient, male, and ≥ 18 years of age. Results A total of 554 prostate bp-MRIs took place during the study period, 547 (96.8%) were included. Of the 547 subjects, 36 (6.6%) were recalled for intravenous (IV) contrast. Most recalls were for PI-RADS 3 reassessment (33/36, 91.7%). PI-RADS categories were upgraded in 19/36 (52.8%) subjects. The median time between initial and recall MRI was 35 days. There was no statistically significant difference in the distribution of PI-RADS categories before (n = 225) and after (n = 225) implementation of the bp-MRI protocol (X ² , p = 0.09). There was no statistically significant difference in the distribution of MRI reporting frequency by radiologists pre-/post-implementation (X ² , p = 0.06). Conclusion A small proportion (6.6%) of bp-MRI subjects were recalled for contrast reassessment and the majority of those (52.8%) received PI-RADS category upgrade. There was no statistically significant difference in the distribution of PI-RADS categories following implementation of the bp-MRI protocol. Implementation of a bp-MRI protocol at a single centre reduced patient exposure to IV contrast while not changing PI-RADS distribution.