Note that when looking at the first chart, the numbers at the top of the bars,
represent the
percentage of men in each group as discussed on the previous page.

The most successful group was the Xiaflex modeling group.  However,
what stands out are the results of those who only received a saline
injection (placebo) and no modeling.  Forty-four percent (44%) achieved
the desired curvature reduction goal and they actually had a slightly
better penile curvature reduction percent as compared to the Xiaflex no
modeling group.  They also did not report any significantly different
outcome on the PRO than those men who were in the Xiaflex
administered no modeling group.  Auxilium opines that the relatively
positive showing of the placebo group maybe due to participants who
had Peyronies for a shorter duration leading to a more spontaneous
resolution of the disease.  This is an interesting hypothesis, but I don’t
know if there is much evidence to support this conclusion. It it somewhat
hard to believe that saline solution injections would result in such a
significant improvement or that simple stretching and bending twice a
day for 60 seconds would cause such an improvement.

P value is an important statistical concept.  It means the probability that
an event or outcome will happen by chance.  A level of 0.05 is often
used.  Simply put, a value at that level or lower is good and higher than
0.05 is no good. The reporting on this study raises a number of potential
issues or it may be nothing more than incomplete reporting.  I really don't
know because all I have to go on are the public documents.

The no modeling group reporting on percentage of reduction in penile
curvature has a p=0.91. This means these findings are pretty much
useless. No p value was reported for the PRO questionnaire for the no
modeling group or the combined modeling/no modeling aggregate.  
There may be a simple explanation for this, but I don’t know.  Why was
the p value reported for some findings and not others?

In their report, the word "significant" is used interchangeably with
statistically significant.  Statistically significant has a specific meaning in
the realm of probability, whereas significant really has no meaning. This
may be just careless wording, but we will have to wait for further
reporting to find out.