Estimation of clinically significant prostate volumes by digital rectal examination: a comparative prospective study
Ahmad Sarfraz; Manecksha Pervez Rustom; Cullen Michael Ivor; Flynn Joseph Robert; McDermott Eugene Dermott Thomas; Grainger Ronald; Thornhill Alan John;
Department of Urology, The Adelaide and Meath Hospital incorporating the National Children's Hospital, Tallaght, Dublin, Ireland
INTRODUCTION: Reliable quantification of prostate volume is important to correctly select patients with benign prostatic hyperplasia (BPH) most likely to benefit from medical therapy [e.g. 5 alpha-reductase inhibitors (5-ARIs)] and in selecting appropriate surgical approach. We aim to determine the reliability of digital rectal examination (DRE) in estimation of prostate volume which may be helpful in patient selection for 5-ARIs therapy. MATERIALS AND METHODS: Patients requiring transrectal ultrasound (TRUS) guided prostate biopsy were recruited in this prospective study. DRE was performed twice for each patient. Clinicians categorized prostate volume on DRE into small, medium and large, and estimated prostate volume. Volume estimated by DRE at the first examination was intentionally unavailable at second DRE. TRUS volumes were measured using 2101 Falcon ultrasound machine. RESULTS: Comparative analysis of prostate volume (n = 248) by DRE and TRUS was performed. There was no significant difference between DRE-estimated prostate volume at the first and second examinations (p = 0.8). DRE-estimated volumes for prostates categorized as small, medium or large were underestimated in 59%, 58% and 53% of patients respectively. However, for clinical relevant volumes (> 30 cc), 94.5% patients were accurately estimated on DRE. CONCLUSIONS: We have shown that DRE had positive predictive value of 94% in identifying prostate above 30 cc. Hence, when considering treatment with 5-ARIs, DRE may be sufficient to identify suitable patients for 5-ARIs therapy. However, for prostate volumes between 25 cc-30 cc and above 80 cc, TRUS may be required.