Crowdsourcing a mystery from the Markingson case

Here's a mystery from the case of Dan Markingson. Yesterday I posted some correspondence between Steven Olson and the University of Minnesota's IRB about Dan Markingson's "evaluation to consent" to the CAFE study.  Well, it turns out that there are actually two different versions of that "evaluation to consent." 

In one version, a space at the top labelled "subject initials" has no initials, only a large black mark.  At the bottom right-hand corner of this uninitialed form is the phrase "IRB."

In the other version, the space labelled "subject initials" has a different black mark with the initials "DRM" above it. At the bottom right-hand corner of that form is written the phrase "psy." 
 
So the mystery is: why would there be two versions of the same form? What is under the blacked-out spot on each form, and why are those spots slightly different? Why does one version have the initials of Dan Markingson while the other one does not?

Any ideas out there? 

Here are links to the two forms:

http://www.scribd.com/doc/123356784/Markingson-evaluation-to-consent-without-initials

http://www.scribd.com/doc/123357156/Markingson-evaluation-to-consent-initials-added