Prostate volume ultrasonography: the influence of transabdominal versus transrectal approach, device type and...

We conduct a longitudinal non-invasive study of changes in urinary bladder contractility secondary to benign prostatic enlargement. In that study, the prostate volume is estimated by transabdominal ultrasonography. The accuracy of those measurements was verified by comparison of transabdominal to transrectal stepwise planimetric ultrasonography as the gold standard. Also, two different transabdominal devices used were compared, and the influence of different operators was studied.

Two series of measurements in 100 patients each were done. In the first series, transabdominal and transrectal sonography were pairwise compared in each patient. In the second series, transabdominal measurements were done with two devices (a hospital Aloka SSD-1700 and a portable Aloka SSD-900). Transrectal scannings were done by three investigators whilst all transabdominal scannings were done by one. Regression graphs, ratio plots and statistical analyses of the data quantified the reproducibility of different methods, observers and device types.

In the transrectal-transabdominal series of prostate volume measurements (in cm3), the Pearson correlation coefficient was 0.84 (p < 0.001), the mean of the means was 51.8 +/- 23.0 (mean +/- S.D.), and the mean of the differences was 1.0 +/- 1.4. In the series with two devices, the Pearson correlation coefficient was 0.73 (p < 0.001), the mean of the means was 31.0 +/- 10.9, and the mean of the differences was 1.0 +/- 1.3.

No statistically significant differences were found between the transabdominal- transrectal ultrasonography, two different transabdominal devices nor between different observers. However, for those using these measurements in everyday clinical practice, it is worth to point out that in our data a transabdominal scan and a transrectal scan in the same patient, on the same day, differed more than 30% in one fourth of the patients and that two transabdominal scans in the same patient (with two different devices, on two different days) differed more than 30% in every fifth patient.