You should be able to import this but keeping in mind HL7 can have multiple delimiters e.g. |, ^ and & as well as carriage returns and there is a hierarchy to how these are applied with the pipe (|) being the primary delimiter then the others to identify sub components within fields.
What you will need to do is bring the data in once, separate on the pipe and then using either a resident load or QVD export perform each of the subsequent processing steps until you have broken everything down to the level you require.
Do you happen to know which HL7 schema is in use as this will determine which fields you have in your extract?
Hope this helps.