An interesting study for those considering surgery regarding erectile rehabilitation

[Nik]

I was quite surprised with this study, how many men who have bi lateral nerve sparing prostate removal, do actually succeed with erections within a year or two when using low dose sildenafil. They emphasise the importance of early rehabilitation, and not just to wait until things return by themselves as the penile tissue can suffer detrimental effect through lack of blood supply. I also read another article about how the vaccuum pump can also help keep the blood supply healthy in the penis while waiting for natural erections to return. Perhaps both sildenafil and the pump used together would be even more effective in speeding up recovery?

[Arnon Krongrad, MD]

1) The study is randomized but not blinded: The control group was not given a placebo. This means that the treatment (sildenafil) group knew it was being "boosted" for erections. This, combined with surgical patients' known bias of wanting to please their surgeons with favorable reports, opens the possibility of great motivation and some reporting bias in the treatment arm relative to the control arm. We cannot distinguish physiological treatment effect from psychological motivation and reporting bias because the control arm knew it was not getting treated.

 

2) We randomize to reduce biases between study arms. For example, to even out age distributions, cancer stage distributions, and the like between study arms. In this case, be it due to poor randomization technique or just bad luck, the study arms retained biases: The non-treatment arm had worse baseline function going into the study (28% potency) in comparison to the sildenafil arm (47% potency). Is this relevant to the observations about post-op erections? You bet! Patients with good erections before surgery are more likely than patients with bad erections before surgery to have good erections after surgery. So to an extent, the study started off with oranges compared with apples.

 

3) Another technique used to reduce bias from mal-distribution of patients between study arms is multivariate analysis. In this case, there was a univariate analysis (sildenafil yes or no), which is what's reported. However, there was not a multivariate analysis (sildenefil yes or no; baseline potency), which means we cannot distinguish if the observed differences in post-op function are associated with physiological effects of sildenafil or baseline function. (and forget about information relating to other known correlates of erections: Illness, smoking, obesity, depression ...) 

 

This is a lousy study. Badly designed and badly analyzed. What it does show is that untreated patients with poor and good baseline function have a hope of regaining function after surgery, not that this is news.