Hello Nathan,
Thanks for reaching out! (How-to instructions are at the bottom)

Unfortunately, EPPI-Reviewer wasn't designed from the ground up in a way that would allow an external system (Archie, in this case) to control who has access to what. Thus, when implementing the Archie integration features we had to add functionalities on top of the systems we already had in place: this works, but requires to make some compromises.

In particular, it means that you can't add a role to a review member in Archie if the person in question has never opened the relevant review in EPPI-Reviewer. This behaviour is clearly not ideal, but considering that conducting complex reviews does require to collaborate closely, we thought it wouldn't be a disruptive shortcoming, for the time being.

We also discussed / thought about the possible ways of solving this problem: one option is to have all people join the review with the coding-only role, with the exception of who started the review in EPPI-Reviewer (is recognised as the "Review owner" by EPPI-Reviewer itself). This option is very easy to implement, but has obvious drawbacks: for the same reasons explained above, you can't assign a work allocation to people who have never opened the review in EPPI-Reviewer, so restricting people to read-only requires them to open the read-only interface for the first time,a requirement that guarantees they will see an empty interface the first time the log-on.

Another option is to give the "read-only" role to people when they first open an Archie review (with the same exception as above). It will then be the task of the review owner to assign roles to collaborators once they opened the review at least once. This approach looks slightly better to us, but still not ideal.

Finally, we could change this unwanted behaviour completely and make sure EPPI-Reviewer learns about all the people who can access Archie Reviews in real-time. This would be ideal, but is not trivial to achieve, and my guess is that being able to exchange review data (to check-back-in and/or to start an update) would be much more important, so takes precedence. Anyway, one important thing to remember is that we're open to suggestions, so please do let us know what you think about the above.

For now, the how-to instructions are:

1. Please instruct the other review authors. They should create an EPPI-Reviewer account, (optionally) log on EPPI-Reviewer, link their EPPI-Reviewer and Archie accounts, open the review in question and let you know. I guess you've seen the new Cochrane Training "get started" page already.
2. Once a given author has opened the review at least once, you will be able to assign roles to them via the Account Manager.
3. You will also gain the ability to create work allocations for them, so that they will be able to use the coding only interface to make their contribution.

I hope this helps,
Sergio