New Login IDs have to be unique to the system to be accepted. In other words, there can’t be two “jimsmith” IDs. Also, they have to be at least four characters long and only have letters, numbers, and the underscore (_), dash (-) and “at sign” (@).
When you enter a login ID for a new patient, the system will look to see if the ID you have entered is available. If it is, you can use it. If it isn’t you will get an error message and have to try again. While this seems at first glance to be limiting, there are some strategies that you can use to be very likely to find an acceptable login ID for a patient.
First, if the patient is going to log in when they don’t have access to their email, the ID needs to be fairly memorable. You can use their name, some part of their name or some piece of information like a house address or birthday in combination to make an easy to remember login ID. Some examples for Jim Smith might be
jsmith (not likely to be available)
jims (not likely to be available)
jims1324 (name + house number (very likely to be available))
Or, you can add some kind of suffix that they will be able to remember but is unlikely to be used anywhere else. For example, our fictional Anytown Rehabilitation could use the patient’s name in combination with “AR”:
jmsar (the patient’s initials + “ar”)
The added initials (or “salt” as it is know in the security world) should come last so that sorting and finding patients via login ID is still possible. If they all begin with “AR” then searching for “jsmith” will return nothing.
An almost guaranteed unique name is the patient’s email address. “Almost” because members of the same family might have the same address. It is easy for the patient to remember, too, but it is also known by a lot of other people so some security is forfeit for the ease of remembering.
With login ID’s of four characters, there are more than one million possible combinations. With 6 characters, there are almost 2 billion. It should be pretty easy to find a way to create unique and memorable IDs.