workcomp.kaiserpapers.info
Article
placed here for historical purposes.
Title:
Thomas Fleishman., By: James, Philip
N., Information Systems Management, 10580530, Fall91, Vol. 8, Issue 4
THOMAS FLEISMAN
Section: INTERVIEW On The Role of IS in Health Care
KAISER PERMANENTE
Medical Care Program is an
association of health care professionals and patients that is organized
as a health plan, a hospital corporation, and an association of
separate and independent groups of physicians who are contracted to
provide services to the health plan and hospital corporation. Kaiser
Foundation Health Plan Inc is a California nonprofit
corporation that enrolls members and arranges for their medical,
hospital, and related services. Kaiser Foundation
Hospitals is a separate nonprofit corporation that owns and operates
hospitals and outpatient facilities. Permanente
Medical Groups is an association of separate and independent groups of
physicians who are contracted to provide medical care for Health Plan
members. A small corporate organization drawn from these entities is
located in Oakland CA.
There are 12 autonomous Regions, which are subsidiaries of these
corporations and groups and are similarly organized. The Northern California and Southern California Regions are the
oldest and largest. The Regions span the nation from Hawaii to New England, but they
currently do not cover all states. The Southern California Region, the
second largest, encompasses 10 counties from Kern to San Diego and retains
2,800 physicians to provide health care for about 2.3 million members.
Kaiser Permanente
was a pioneer in the provision of
prepaid health care. It functioned as a health maintenance organization
before the term HMO was coined. Once a minor part of the health care
scene in California,
it has benefited from changes in the social contract governing health
care in the US and is now a significant force in the delivery of
quality health care and in pioneering new cooperative arrangements
between those in need of health care and those who provide it. On a
national basis, the annual revenues for the organization exceed $7
billion and it currently has 7 million members.
Thomas Fleishman, director of the Information Services Department of Kaiser
Permanente's Southern California
region, presides over an organization of 500 employees with an annual
operating budget of about $50 million. Fleishman's career began when he
joined a unit of Columbia University as a computer programmer trainee,
where he worked on software supporting sonar submarine detection
systems. He was recruited to System Development Corp in Santa Monica
and then, after a brief stint in Northrop Corp's internal audit
department, joined TRW Space Systems Division, where he spent nearly 10
years before joining Kaiser Permanente
in May 1979.
After about 18 months as manager of computer operations and support
services, he was appointed assistant director of the division, then
director in January 1982. He has one of the longest tenures among chief
information executives in southern California and is
currently a member of the Southern California Chapter of the
Society for Information Management. He was recently elected to the
Board of Directors of the Pasadena
Chamber of Commerce and was asked to join IBM's Information Systems
Customer Advisory Council. Fleishman discusses the changes in the
health care industry during the past decade and the effects they are
having on information systems management.
The health care system seems to be in a tremendously dynamic
state these days, more so than most industries in this turbulent time.
Why is that?
I think there are two primary reasons. First, when inflation
began to recede during the Reagan years, people noticed that the
increase in the cost of health care wasn't receding proportionately. In
fact, it was continuing to grow. There has been continuous debate in
many circles about the rising costs of health care. The most recent
debates have been in the corporate world, where health coverage is
changing dramatically for both employees and retirees as a part of the
accelerating effort to reduce business costs.
Second, for a variety of reasons, people are now less in awe of
physicians and are likely to want to participate in decisions affecting
their care. More patients are challenging physician authority, which
has undermined the deity status physicians once enjoyed. One result of
this is an increase in the number of malpractice cases, and this change
has driven many physicians out of health care. It has also dramatically
increased the number of procedures performed to reduce liability rather
than for actual medical need.
These factors are producing changes in the way health care is
viewed--by the public, by Congress, by regulatory agencies, by
Insurance companies, and by the patients themselves. And, there's so
little real information about many of these fundamental issues that
everything we learn tends to cause a change in the way we do business.
The regulatory agencies have probably had the most significant impact,
and that's related to the rising costs of Medicare. Although some of
these costs have been brought about by new medical technologies, which
are more and more sophisticated, effective, and expensive, there are
other factors as well. But I think one of the most significant change
agents in the way we do business has been the establishment of the DRGs
as the means for reimbursing health care costs.
What is a DRG?
DRG stands for Diagnostic Related Group. A set of government panels
determined that all illnesses could be classified into approximately
500 DRGs. Within a DRG, there's an established protocol or procedure
for treating an illness or a set of illnesses. Of course, circumstances
differ, but on the average all patients within a DRG can be considered
as if they were treated alike.
Clearly, everyone expects that each patient will be treated
according to his or her needs. But reimbursement is made in accordance
with the specified DRG. This means that the cost to the health care
delivery system--the physician, the hospital, the medications- should
match the DRG reimbursement level. If the treatment cost is higher, the
delivery system swallows the excess; if it's lower, the delivery system
enjoys the benefits. To have a lower delivery cost, the institution
must be relatively efficient in comparison with other delivery
organizations.
That approach makes some sense to me. Why is it a problem?
The real problem is that there's very little consensus in the
medical community about what treatments produce what outcomes.
Physicians legitimately and reasonably differ depending on their own
backgrounds and experiences. The heart of the problem is that there is
no one body of information that makes it clear, in most cases, what the
relative effectiveness of differing treatment styles is on the patient.
We see a very strong sentiment growing for the collection and
analysis of information of this kind. As the results of such analysis
accumulate, the cost-effective options will become clearer, and we may
begin to get a handle on an approach to bringing health care costs
down. That's going to have profound effects on the way information is
managed.
With all of these changes, what has been the role of information
technology during the past decade?
Well, first of all, I'd say that the health care industry has been
somewhat behind other industries in its use of sophisticated
information technology. It was not too long ago at this facility, for
example, that patient admission statistics were almost literally
computed on the backs of envelopes. The development of hospital
information systems to handle business transactions has matured pretty
well during the decade, and there has been a lot of fine work in the
areas of pharmacy and laboratory management systems. This has helped
the health care industry at large.
A significant problem for Kaiser Permanente
is that
the market for these systems is generally the single community hospital
with fewer than 300 beds. Most of the 6,000 or so hospitals in the US
fall into that category. When you get into large multifacility
organizations like ours, the volume of transactions is huge. The
readily available systems simply cannot handle those volumes.
Furthermore, in that category, the needs are usually unique enough that
applications must be custom developed.
What are the basics of your shop? Do you have a substantial
telecommunications network?
Well, we have three IBM 3090600's (two J models and one S model)
plus substantial installations of Digital Equipment Corp, Tandem, and
Data General equipment. We also have a fully digital, T1,
state-of-the-art multipath backbone network with six major nodes. A lot
of it is fiber optics wherever it's available, and we're working to
complete that. We have approximately 15,000 devices connected to the
network, including computers, terminals, controllers, and printers.
Did you build this network yourself? Is it proprietary, or do you use a
telecommunications vendor?
We primarily use Pacific Bell because most of the network is within
its territory, but we also have a substantial presence in GTE
territory. For long distance, we use both AT&T and MCI. The
10-county Southern California Region
includes several LATAs (local access and transport areas), and our
network covers both inter-LATA and intra-LATA links.
How many transactions a day does this network handle, and where do they
come from? What drives the principal load?
I'd say in a typical day, the IBM part of the network alone handles
about 2.2 million transactions. Most of these are from the medical
centers, and they're all complex transactions not just single items
like processing a check. Almost 1 million per day come from our
scheduling system as appointment booking transactions. Another 400,000
or so come from our inpatient management/admission/discharge system.
And another 200,000 to 250,000 are from our materiel management system.
What does your whole application portfolio look like?
At this point, we have approximately 120 applications. These fall
into four major groups. Accounting and finance (the controller
functions) obviously compose one of these groups. Service and support
for the health plan composes another, and is primarily in the area of
membership service, marketing, and membership accounting. The two
medical areas form the third and fourth groups. The hospital
information systems deal with the business and financial support of the
acute care facilities. It's larger than most systems, so although it's
based on a vendor system, most of it has been modified in-house.
Finally, we have the outpatient and clinical support systems, which are
perhaps the most innovative and interesting. Among other things, we
have the largest private medical lab in the country.
One of our systems is a very sophisticated appointment scheduling
system. A physician's time is obviously one of our most valuable
resources, and we try to manage that as effectively and efficiently as
we can. This system has many of the characteristics of an airline
reservation system except that we're scheduling physician time slots
instead of airline seats.
Do you have what you'd call a patient-centered information system
that includes automated medical records, bedside terminals, nursing
support, lab, pharmacy, and radiology support?
The automated medical record is the holy grail of health care.
Very few organizations have made a substantial commitment to it. One
group that has is the Harvard Community Health Plan in the
Boston-Cambridge area; this has no connection with Harvard University,
though many in the Harvard family are members.
Why did Harvard feel that the automated medical record commitment was
important?
I don't really know. I'd guess perhaps they feel that in that
community it's justified. There's a large concentration of
distinguished medical schools there, as you know, and the system was
originally built as a research project.
Why don't you do it?
It's primarily a cost issue, but the technology is also lacking at
this time. The Harvard group spends about $0.64 per member per month
for that system alone, whereas we provide our full range of support for
about $1.40 per member per month. At $0.64 per member per month, the
automated medical record represents an annual cost of approximately
$17.5 million, in addition to our existing budget. At this time such
cost is prohibitive for us.
Bedside terminals are a different issue. I have no doubt that
some day there will be automation arrayed around the patient. But I
don't think we currently know what's required in such a workstation.
There's been so much talk about bedside terminals in the industry and
so much written in the literature--but so little actual
implementation--that I get the feeling that this still is a solution
looking for a problem. We've had demonstrations of several systems, but
we don't see anything that we feel would be a significant help to us at
this time. However, we are actively working with several vendors to
eventually acquire a Clinical Work Station that is functionally robust
and meets our needs in terms of cost/benefit trade-offs.
What are you doing that you'd consider innovative?
Well, I mentioned our appointment scheduling system earlier. We
think that's pretty innovative; it's the only one like it that I'm
aware of. It works much like an airline reservation system and provides
a great deal of management information about patient and procedure
characteristics and physician work loads that was never available
before. In that sense, it's our SABRE system.
Another application that we think helps a lot is our automated
prescription refill system. A patient can call a number, enter his or
her medical record number and prescription number, and the system
checks to see if a refill is authorized. If it is, it tells the patient
to pick up the prescription in 24 hours and transmits the order to the
pharmacy. If it's not, it tells the patient that the physician will be
called and asks the member to call back. When the refill authorization
has been entered, it works the same way. If a refill is not authorized
by the physician, the system notifies the patient to call his or her
doctor to schedule an appointment.
In the old system that this replaced, phone messages accumulated
on an answering machine. First, there was a lot of work matching the
message to the patient and the prescription. Then, if the refill wasn't
authorized, the pharmacist had to call the patient back to keep him or
her informed. The new system is much more efficient and frees up the
pharmacy staff for more substantive work. We also have innovative
inventory control and laboratory management systems that use bar codes,
and we're doing electronic data interchange with a couple of our major
vendors.
How do you measure your success here? How do you justify your budget to
senior management?
We do it largely by the applause level. During the 1960s and 1970s,
our data processing achievements were relatively modest. We were in a
kind of theoretical stargazer mode during those years. Since 1980,
we've implemented a substantial number of the systems that are now in
place. I've also been able to move the shop into a situation in which
our applications come in reasonably close to the schedule and budget we
establish. This has done a lot to build management's confidence in us,
as has the fact that we have managed our tremendous growth within
reasonable costs and with relatively favorable results.
In the 1970s, we realized, as many others did, that a data management
orientation was the wave of the future, but Kaiser
made the same mistake that many other shops made. We thought that meant
a single global data base accessed by all the applications. Kaiser Permanente
spent many years and dollars trying to make that approach work. The
output, in terms of improved applications useful to the various parts
of the business, was few and far between. So we stepped back and took a
long look at our approach.
We asked Nolan, Norton to come in and review our situation. They
found a fair number of areas that needed attention, and they suggested
a different philosophical approach. I was a beneficiary of their study.
As part of the process to rebuild the IS management structure, I was
asked to come on board to manage operations and support services, and I
did so in May 1979.1 became assistant director in October 1980 and
director in January of 1982.
We acknowledge the data management orientation, which was the
objective during the 1970s; indeed, we see it as the means for
integrating our systems. But our approach benefits from all that was
learned about data management during that period. We'll be developing a
collection of data bases and applications that will provide a steady
stream of results of value but in accordance with an architecture that
describes our ultimate objectives and strategies.
That sounds like a strategic business plan. How did you develop that?
Well, of course a lot of the planning occurs as a result of my
reporting relationship. I'm one of eight people who reports directly to
the CEO of the Southern California
Region, and I'm accepted as part of executive management. This gives me
a clear perspective on the business needs and direction of the
organization, and I can translate those into supportive information
technology implementation projects. This has been reasonably informal
until recently because we've been operating with a relatively informal
business plan. We've just recently formalized the process, and it has
produced the first truly strategic plan. That should help us a great
deal for the future.
During the 1980s, we were guided by a strategic information
systems plan which was prepared about the time of the Nolan, Norton
study. About a year-and-a-half ago, we asked Deloitte & Touche
to
look at what we were doing and how we were doing it, and to give us
some feedback. They did a fine job. They recommended that we needed an
updated strategic information systems plan, arid we developed one last
year.
Do you have a process in place for maintaining that plan as the
business environment changes?
I think we're in pretty fine shape on that score. As I've
mentioned, my position as a member of senior management of the Region
is the backbone of the effort, and we have an efficient Information
Systems Steering Committee. I also spend a large part of my time
working directly with my customers in the medical centers of the
Region, and I encourage my managers to do the same. That keeps us
pretty well abreast of the needs as they evolve and allows us to keep
the plan current, at least informally.
You're familiar with the survey issues that come out of studies
by various organizations from time to time. How do you relate to those
issues?
We track pretty well with the issues, although of course the
priorities vary. We're trying to reengineer the firm using information
technology and we're moving toward a workable datacentered philosophy
and strategic planning. We're also trying to improve software
development quality and productivity, including CASE. We've created an
information architecture, and we're addressing new technology
integration in our applications. We're not considering outsourcing, and
we're not international. Telecommunications improvement is a continuing
issue, but upgrading old systems, fortunately, is not; most of our
systems are relatively new. We continue to emphasize career development
and training, and this has paid off in a staff turnover rate that is
substantially below that of the industry. And, as in the past, we
continuously emphasize the information systems department's role as a
service organization that supports the Region's business direction and
vision. That is the primary mission; technology is secondary and serves
as the tool we implement to achieve the mission of the department.
PHOTO (BLACK & WHITE): PHILIP N. JAMES
~~~~~~~~
By Philip N. James
PHILIP N. JAMES is a member of the information systems faculty at
California State University, Long Beach.
KAISER PAPERS
workcomp.kaiserpapers.info
|
|