Turns out that some of the data loaded from the demo HCRS database (hcengs.db) has keys that are case sensitive.
If SQL server is set up with the Latin1_General collation but the case sensitive flag is not set, then certain records will not import properly and throw errors during the demo database creation step (unique index errors). When you look at the keys causing the error you will find that the only difference is case of the key (e.g. TEST, Test, test). If your SQL server installation doesn't have a case sensitive collation set, these sorts of keys will cause an error as they won't be unique.
You can check case sensitivity of your collation using the sp_helpsort stored procedure.
This is an example of what is returned:
| Server default collation |
|---|
| Latin1-General, case-sensitive, accent-sensitive, kanatype-insesitive, width-insensitive |
