sync with upstream SCILHS code

[dckc]

In particular...

• Adopt their Oct 10 run script solution to our request to split things up.
• figure out why they prefix their table names (pmnVITAL) and why we don't and whether to change ours or ask them to change theirs.
  • see trouble with aliases and SAS #66
• prune obsolete TODOs such as Also, Error(57,57): PL/SQL: ORA-00904: "MO"."PCORI_NDC": invalid identifier

[dckc]

We're 12 commits behind SCILHS:master. I wonder how substantive those are.

cc @njgraham @archaean @mprittie

[dckc]

Jeff responded to several of our issues lately; e.g.

ARCH-commons#4 (comment)

[dckc]

for diagnosis.admit_date, we're using observation_fact.start_date rather than visit_dimension.start_date.

This was fixed upstream in ARCH-commons@1254d76

We're now 60 commits behind.
cc @njgraham @archaean @mprittie

[dckc]

We're getting bit by not picking up the change to "Create table statements are now outside of the procedure ..." ARCH-commons@21f9680

We're now 78 commits behind SCILHS:master.

[dckc]

Perhaps upstream is more open to renaming tables than I thought...

... Let's also rename the tables with a pmn prefix to the correct name ...
-- jk Feb 6 ARCH-commons#50 (comment)

[dckc]

For performance reasons, we moved some stuff from runtime to design-time based on local knowledge.