2. It Was A Very Good Year!
• This week continues the HIS-tory of Siemens Healthcare, today‟s
#3 vendor in annual revenue, whose HIS roots go back to the mid-
1960s when IBM first developed SHAS, featured last week.
• Thanks to the many HIS veterans who contributed to the origins of
SHAS, which automated patient accounting in thousands of US
hospitals who used it through local Blue Cross, state hospital
associations, and many proprietary firms like Gamut &SMS.
• This week, we cover the early days of Shared Medical Systems for
CIOs who may not have been born when it was founded in 1969.
I was fortunate to be one of SMS‟ early employees (#24, hired in
October of 1969), so I‟m going to relay the inside and human story
of SMS‟ amazing growth to eventually being the #1 HIS vendor.
3. SHAS Was Not Perfect!
• SMS started running IBM’s SHAS soon after its release, and like
all new HIS products (Millennium, Paragon, Soarian, etc.) it had
its share of bugs, design flaws, missing features, etc., all to be
corrected in the 4th quarter per the vendor (but in what year?).
• We touched on one last week which was the Scalar Date routine
IBM came up with to minimize storage requirements back in the
days of their 360 mainframe, whose disk drive had one (1) meg!
• An early SMS programming maven, Glen Marshall, tells the tale
(he‟s pictured on the right at our 2007 reunion in FL):
“In the mid-1980s I rewrote the old SHAS scalar date routine,
changing the base-date from 1/1/1900 to 1/1/1960. This
extended the range of dates until September 2049, well past
my 100th birthday. For the geeks among us: The original
scalar date calculation was done in packed decimal arithmetic:
year x 36525 / 100 (by lopping-off 2decimal places) then
calculations for month, day, and leap-year adjustments.”
(VC: pretty simple, huh?)
4. Y2K Pre-Cursor (still love that
pun!)“My rewrite was based on the date calculation formula used in
satellites, and that formula dealt with the Y2K problem as well. (I
saw it coming early...) In addition, the calculation was done in
binary register arithmetic, which cut the CPU time for date
calculation by 90%. This time-savings was significant.
The billing records were chock-full of dates that entered into the
insurance proration calculations. As I recall, the savings was
nearly a net 10% savings for the overnight billing program
runs. That is a major savings for a mainframe.
All the scalar-date using programs needed to be re-linked to pick-
up the new date calculation subroutine. A one-time conversion
program was run to change the date-base to 1960. Everything
works like a charm. Only one program was not re-linked, though,
due to an oversight. And that was the one that caused the
headache and headline in 1989.” – Glen Marshall
5. Start-Up Ups & Downs
• There was an amazing esprit de corps at SMS in those early days
– as I‟m sure there was at HIS new start-up: McAuto, HBO, SAI,
etc. Everyone knew we had to work hard just to survive, let alone
ever make the big times. The hours were long and hard too: I got
up one winter morning to a freezing rain at my home and
couldn‟t get the door to my „vette to open – the lock was frozen
solid! I tried heating the key with matches, to no avail. Waiting
an hour for the sun to do its job, the phone rang around 9AM – it
was President Jim Macaleer wondering why I wasn‟t there yet:
– we were supposed to start at 8:30!
• And I‟ll never forget the “Saturday Club” – a
small group of fools like me who got their dull
admin stuff done on Saturday mornings: “Big
Jim,” Harvey Wilson (Sr. VP Sales & Marketing),
Mike Mulhall (VP of Installations), Phil Jackson
(Terminals), Tony Sam (CSC)… you could tell
who was in by the cars parked in the near-
empty parking lot at 650 Park Avenue…
6. Inside Humor
• It wasn’t all just work during
those early 10-12 hour days
either – we goofed off a lot to
keep each other half sane...
• We IDs (Installation Directors)
received a stream of “ID Memos”
from K of P telling us of bugs
that were fixed and new features
or modules.
• I was an ID at SMS’ NJ office,
and wrote this mock memo to a
hot chick in King of Prussia HQ
trying to impress her with my
puny humor (she was an English
Major too).
• She laughed, but didn’t buy…
7. Outside Humor
• ID memos were pretty
technical, so they were
re-written in English (sort
of…) for clients to learn of
new enhancements by
our Customer Service
Center.
• They were called CSC
Memos and #531 went
out in 1977 that really
didn‟t do a good job of
explaining some changes
to our new Inventory
system...
• The next day, Big Jim
wrote this cover memo to
a re-written version of the
memo apologizing to our
100-odd (sic) clients!
8. New Product Break-throughs
• SMS had an amazingly talented team of programmers, and one
of their technological breakthroughs was called UNIFILE – Ken
Shumaker led the development of this powerful & precocious
1970‟s data base system, based on MRI‟s “System 2000.”.
• Unlike SHAS‟ batch processing, it processed transactions in real
time as soon as they were entered (like rival McAuto‟s HFC did),
and then passed them on to an on-line data base for inquiries.
• Needless to say, it sold well, but as more and more clients
jumped on board, things started to slow down as the water-
cooled IBM 370s of that era had trouble handling the many
census transactions, report writer requests, and db inquiries…
• It was later toned down to less-powerful but
more reliable versions called Focus &
Command, but at one of SMS‟ infamous Xmas
parties, I had a blast giving Big Jim, Harvey
Wilson and Ken Shumaker three T-shirts
labeled respectively Uni, Fiand Al!
9. Near Misses
• The earlier HIS-tory episode on SMS (#11 –see them all at
hispros.com) as a shared system pioneer covered two near misses
that might have put SMS out of business early in the 1970s:
– Regionalization – expanding from 1-digit to 3-charcater
hospital codes that brought SHAS down for days on June 30
– Cash Flow – turning the corner from red to black circa 1971
• Another close call was when SMS moved from its original rented
space at Ross & Royal Roads in Bridgeport to a former bank
building we owned at 650 Park Avenue in King of Prussia. Phil
Jackson, who was assigned a number of challenging tasks (like
ACTIon and the NYCHHC install) headed up moving the data
center, and he asked we IDs to go to client hospitals on three
Saturdays, the first 2 to test the move, the 3rd for the real thing.
• We all went to clients and dumped in batches of cards for the two
tests, with only a few problems switching the hundreds of phone
lines, etc. When it came time for the 3rd test we got the word: the
2nd one was the real thing – no need for #3! Few complaints…
10. Green IDs
• Another down side to start-up firms is the lack of
experience with the system by their “green” staff. Not
just green in terms of age, but practical experience.
• Most of we IDs at SMS in the early 70s were totally
new to computers, hospitals and accounting basics:
- I was an English major at my first “real” job
- Al College (eventual VP) was a former school teacher & coach
- Takis Petrakis (sadly deceased) set the record for ID novitiates:
he was the former captain of a submarine in the Greek navy!
• So what, you ask, doesn‟t every vendor hire rookies and train
them? We had a 3-week class that tried to teach us every aspect
of SHAS (several million lines of code!), accounting (debits vs
credits) and hospitals (what‟s the difference between an RN, LPN
and Aide?) – lots of luck! We learned as as much as we could
during those 3 weeks, then were sent out to the real world to
learn in the school of hard knocks, at our client hospitals‟ time &
expense.
• Aren‟t we so much smarter today? Every CIO insists on meeting the
11. Card Column 11 of the Header Card
• Al College & I were assigned to convert St. Vincent‟s Hospital in
Staten Island, which had been totally manual on NCR posting
cards. We started with AR, showing them how to fill out coding
sheets for their thousands of ledger cards for keypunching:
• The cards were then sorted into batches of ≈50 each for ease of
handling, and SHAS required each one to have a header & footer
card. On the header card went the hospital‟s code (St. V = “O”),
the batch type (new AR = 05), a batch number (001 to 999), etc.
• According to the SHAS OPS manual (our bible!), card column 11
indicated outpatients with a “6.” So Al & I dutifully sorted all the
hundreds of batches by IP & OP, entering a 6 in cc 11 for OP
ones. Wouldn‟t you do the same – it‟s what the book said!?
12. Catastrophe!
• I squeezed all the boxes of 5081 cards into my car on Friday
night, drove them down to K of P to load onto our mainframe over
the weekend. On Monday I went back to get the TCEs
(Transmission Control & Error report), and was dismayed to have
as many boxes of paper error printouts as we had submitted
keypunch cards! It seems what the SHAS OPS Manual meant to
say was that cc 11 separates OP vs IP charges (batch type 03):
new AR from cards was batch type 05. OyVe!!!!!!!!
• So I drove the boxes of error reports back to the poor folks at the
hospital, who started trying to correct the bewildering array of
duplicate errors that each batch had generated: some from the
AR program, some from the OP billing program. A nightmare!
• Precious days went flying by as all patient accounting activity
halted until we could correct all the errors and balance the AR –
we never did, and after a few weeks, the CFO just wrote off the
difference (6 figures…) before we proceed on to ADT & Billing…
13. Near-Death Experience
• We converted Census and Billing at St. V‟s much better, and the
hospital eventually benefitted enormously from automation – it is
still an SMS (Siemens) client to this day! But I must admit, I still
avoid driving over the Goethals bridge thru Staten Island, afraid
the CFO might still be gunning for me somewhere out there…
• I probably almost got fired for the screw-up – I
remember trying to explain to Steve Macaleer my
ID Manager about the error in the SHAS OPS
manual, but he told me to not screw-up again…
• The real irony is that I learned from my mistakes, became one of
SMS‟ better IDs (aced my 2nd and 3rd hospitals), and was
eventually promoted to be Education Manager, in charge of
teaching all new IDs the ropes. I told this story to every trainee!
• So is it better to get a rookie who‟s very bright and hard-working,
or a stogy old veteran who just repeats the same formula over &
over? I‟d look for both: a veteran who is smart & willing to learn!
• And never be any IC‟s first implementation! Send them back…
14. The Takeaway?
• So what can one take away from this story of SMS‟ early days –
should a CIO stick with large proven giants like today‟s leaders:
– McKesson, Cerner, Siemens, and other “Top 10” HIS vendors,
– or take a risk with daring new “cloud-based” products from
early start-ups like CSS HealthTech, or RazorInisghts?
• Like so many HIS issues, the answer has both pros & cons. Pros:
– Giants forget their own past when they too were start-ups
themselves, viz: Huff, Barrington & Owen in Walt‟s kitchen
struggling to write an order entry system on a Four Phase…
– Small start-ups generally give the best service as any of their
early clients can get the CEO on the phone & they‟ll listen!
• And on the other side of the coin, there are cons, like:
– Who can remember hot new start-ups Bulldog IT, IntraNexus
and American Health Net, who rocked just a few years ago?
– An adage from the 60s had it that “No One Ever Got Fired For
Buying IBM” – dare take an unknown name to your Board?
The answer is different for every hospital and every HIS-tory epoch…