The future of health care innovation is best understood if we take a moment to look at how we got where we are today. What have been the obstacles? How is the industry overcoming these obstacles? Who are the leaders in terms of change, and what kind of changes can the industry expect in the future?

The use of computers in health care evolved through a process driven by need.
Over the last 30 years, health care has seen few real visionaries who have developed
systems to aid in the clinical process. Only in the past five years has the emphasis
shifted from applying computer technology to hospital administration, to applying
technology to the clinical process.

The use of computers in health care began to emerge in the early 1970s (through
the use of shared systems) after Medicare reimbursement legislation was enacted,
but it wasn’t until the late 1970s, when minicomputers began to become available,
that computers began to be widely used in health care.

Initial market penetration into health care came from two directions, hospital
systems and physician practice management systems. Not surprisingly, the initial
use of computers focused on computerizing the administrative process. Computerization
primarily involved hospital billing, financial applications, and physician billing.
To a lesser degree, computers also began to automate the administrative functions
of clinical departments such as laboratory, radiology, and pharmacy.


Health care automation was quite limited for the first 10 years because of
the insufficient file capacities of most systems. Most vendors focused on a
limited family of applications, and almost every system was designed for a single
hospital or physician practice. As a result, the industry saw the emergence
of a myriad of limited product vendors with little integration between systems.
It wasn’t until the last half of the 1980s and early ’90s that the industry
began to see integrated systems become commonplace. At the same time, the health
care industry began to see two other important factors emerge: market dominance
by a number of large multi-application vendors and the beginning of market application

Until the mid-1990s, very little work was done on integrating computers into
the clinical process. The industry saw the automation of order entry, result
reporting, and some nursing applications, but innovative automation of the clinical
process was still lacking for four primary reasons. First, the industry was
just beginning to see the extended use of clinical data repositories with large
databases available for advanced applications. Second, until the mid-1990s most
vendors perceived no need to develop advanced applications while there was still
market opportunity to sell existing systems. Third, there was a lack of supportive
data to show that advanced clinical applications were worth the effort and cost.
At that time, computer applications had not yet shown real results. Fourth,
unrealizable growth expectations by stock analysts resulted in very few vendor
funds allocated to the development of new systems. Consequently, today there
are 10,000 to 15,000 applications that were installed more than 10 years ago
that should have been candidates for replacement long ago. In fact, a number
of 20-year-old systems are still being sold today.

One of the biggest obstacles to the advanced use of computers in health care
is the reluctance of vendors to document both the economic and patient health
improvements that can be realized through computerization. In fact, health care
is the only industry where products have not been sold based on improving results,
gaining a competitive advantage, or producing savings.

Health Care Innovation Today

After 20-plus years of automation, every major health care administrative function
has been automated. Actually, the market requirements of traditional systems
are now quite well defined.

In some respects, the health care industry can still be called a cottage industry,
with several hundred vendors having annual revenues of less than $5 million.
At the same time, only a few large multi-product vendors have emerged to dominate
the industry. These include Cerner Corporation, Eclipsys, Epic, IDX, McKesson/HBOC,
Medication Information Technology, Misys, Quadramed, Siemens Health Care, and
WebMD, each of which has annual revenues of at least $100 million.

There have been some changes in health care systems over the past two decades,
but health care systems up until now have merely provided a solid base for future
care innovation.

Emerging opportunities have largely come from three technology advancements
in the past 10 years: the availability of large-scale databases; the development
of true relational database software; and the emergence of the Internet into
the commercial marketplace. Because of expanded file capacities, images and
voice can now be stored as part of a patient medical record. Vendors have quickly
adapted these technologies into their existing systems development. In reality,
today’s systems are now capable of keeping a complete medical history of a patient.

Market saturation of traditional systems and the availability of several new
hospital systems, like Cerner and Epic, have forced major health care system
vendors to either develop new replacement products, face the possibility of
losing clients, or realize that growth will be stagnant at best. As a result,
all of the major system vendors have either developed new products or are in
the midst of developing such systems. Moreover, hospital and private practice
management vendors and applications have begun to overlap. For the first time,
a computerized record of a patient’s complete medical history can be achieved,
including hospital and physician office data. The net result of computer automation
in health care is that the industry is at a threshold of opportunity centered
around clinical care innovation.

Innovating Clinical Care — The Future

The easy tasks of clinical care automation have been done; now the fun starts.
The health care industry has for some time automated orders, results, nursing
care plans, documentation, and to a limited extent physician order entry. As
a result, large databases of clinical findings have been established, but the
health care industry is only now beginning to use this data in clinical care
innovation. So, what areas of clinical care innovation can realistically be
expected in the next five to 10 years?

It appears that the greatest effort in innovating clinical care will occur
in three areas: medical error, computerized physician order entry (CPOE), and
disease management.

Medical Error

It has been estimated that medical error costs an average hospital more than
$2 million per year for every 100 beds. Yet, little was done to address this
problem until the release of the 1991 government report on patient safety. Why
was so little attention given to decreasing medical error prior to the government
report? The answer is simple but alarming. Few hospitals want to admit they
have a medical error problem. Instead, they ignore the problem and say they
are doing as well as the next hospital. Furthermore, to admit to medical error
would increase the possibility of increased legal liability claims, and no hospital
wants to be known for providing unsafe care.

It has been documented that 11 types of medical error occur in the hospital:
emergency room, surgical, medical, medication, adverse drug reactions, nosocomial
infections, patient falls, laboratory, maternity, device complications, and
dicubiti. What is interesting about medical error is that very little has been
done to quantify it. In addition, medical error is sometimes difficult to diagnose,
and it’s even more difficult to develop an avoidance response.

Industry interest in medical error took a giant leap forward with the 1999
government report on the magnitude of medical error. It seemed that every vendor
immediately took notice and had some form of computerized solution. Almost every
solution dealt with medication errors. However, little is being done to provide
computerized solutions to other types of medical error. Yet, today’s databases
contain a huge amount of clinical data that should help identify the magnitude
of the medical error problem. One of the problems is that it has been difficult
to measure the magnitude of each type of medical error. For example, surgical
delays are classified as medical error because of the causes of delay, such
as not administering the correct pre-operation procedure. One hospital studied
surgical delays and found that they were costing the hospital a significant
amount each year. As a result, an inter-disciplinary team and the IT department
were asked to work together to develop protocols and procedures that, when implemented,
led to a savings of more than $4 million a year. Had the magnitude of medical
error loss been identified earlier, it would have received attention much earlier.

Another area where computerization can play a big part in eliminating medical
error is in the development of medical alerts and protocols that make health
care professionals aware of the potential for a medical error. For example,
an admission system might flash an alert for proper restraints when a 70-plus-year-old
female with osteoporosis is being admitted.

The magnitude of medical error is huge and thus provides an opportunity for
big savings; however, even bigger is the impact it will have on patients.

Computerized Physician Order Entry

One of the earliest examples of Computerized Physician Order Entry (CPOE) occurred
30 years ago with El Camino Hospital’s (Mountain View, CA) installation of the
Technicon Medical Information System; however, until the last few years the
industry has had to push and pull physicians to perform computerized physician
order entry. Then things changed in two ways. The 1999 Leapfrog declaration
by a number of large corporations on CPOE really got the industry’s attention.
The Leapfrog declaration required physicians to begin to use computers for order
entry in the next few years or hospitals would face the prospect of being “blackballed”
by these firms. At the same time, a number of practice management vendors began
to develop electronic medical record (EMR) systems, which automated the physician’s
clipboard and provided a medical history of patient treatment within the physician’s
office. However, CPOE required changes to many state regulations that previously
mandated that caregivers see a handwritten copy of a doctor’s order before acting
on the instructions.

Health care vendors are great copycats. As soon as the Leapfrog mandate was
published, hospital and practice management system vendors developed what they
thought was the perfect CPOE solution. It would only be fair to add that the
coming of age of handheld entry devices has also made it easier to perform CPOE.

Whereas most CPOE focus has been on the automation of physician orders with
in-hospital systems, the automation of the physician clipboard offers at least
as much potential for improving patient care. Currently, there are more than
20 practice management firms developing EMR systems. The EMR market is currently
less than 10 percent penetrated, which means that there are more than 200,000
physicians without an EMR system. It is somewhat surprising that it has taken
so long for EMR to draw industry attention. Certainly, doctors prefer not to
spend their own money on systems, particularly when the benefits of an EMR are
questionable. At least one vendor study has shown that its EMR system could
save physicians enough time to reap an annual $70,000 increase in business.
EMR also provides much better treatment documentation, which reduces legal liability.

The automation of the physician’s clipboard is a major step toward a complete
patient health history. It will be interesting to watch who emerges as the provider
of a complete patient record (a combination of both in-hospital and physician
office data), major providers of hospital systems (IDX, Cerner McKesson/HBOC,
Meditech, QuadraMed, Siemens) or the leading providers of practice management
systems (EPIC, Medical Manager, Misys).

Disease Management

Disease management covers a wide swath of opportunities. All of these opportunities
center around ways of providing the best treatment for specific diseases and
illnesses. Thus far, two types of disease management systems are leading the

One effort is the development of “best treatment plans” or protocols for the
treatment of specific types of diseases, such as diabetes. Much of this development
would not have been possible without large databases of different types of treatment.
It is still hard to say that the best treatment protocol has been found for
any disease, but what is being developed are “recommended” treatments, while
further treatment improvements continue to evolve.

Non-vendors such as insurance and pharmaceutical companies are spearheading
a second effort. These companies are interested in gathering large databases
of patient diseases and treatments from hospitals and physicians in an effort
to determine the most efficacious and cost-effective treatments for a given
disease or illness.

The ultimate intent of much of the disease management work is the development
of computerized system protocols, treatment disciplines and treatment plans,
and the monitoring of successful treatment against the best available treatment

A discussion of disease management would be remiss without mentioning the increasing
shift of disease management from the hospital and physician to the patient.
The increasing use of computers by most people has led to efforts to have patients
perform disease management from home. Such efforts range from an initial self-diagnosis
of a health problem to the monitoring of patients who are on specific types
of therapy. The remote entry by patients of blood pressure, temperature, pulse,
respiration, and conformance with treatment plans is already being accomplished
through some home health systems.

It is estimated that there may be as many as 100 vendors currently working
on the development of an interactive patient disease management system. A number
of these vendors are small companies with expertise in a single medical specialty.

Innovation in clinical care would be sharply reduced were it not for the way
the health care industry has flocked to the Internet. Probably no other technology
since personal computers has gained such attention. Five years ago, the primary
use of the Internet was for messages and promotional purposes. Today, the Internet
has become an integral part of many vendors’ systems. Most important, from the
perspective of clinical care, is the access to a myriad of disease and medical
databases through the Internet. Such access provides medical data to even the
most remote parts of the world. In fact, best-treatment plans that were developed
in advanced countries can be made available to physicians in even the least-advanced
counties over the Internet.

High-speed communication becomes important in sending large amounts of data
and images. A major Internet restriction is speed of communication; however,
this limitation will be quickly overcome with broadband communication.

The biggest problem in clinical care is data security. Today’s legal society
is increasingly focused on violations of access to patient data. While HIPAA
compliance provides guidelines for securing data, it also makes it easy to identify
innocent and inconsequential violations and for lawyers to sue.

A second potential problem facing clinical care innovation is the Food and
Drug Administration. The FDA has already placed very restrictive controls on
blood bank systems. The imposition of similar measures on clinical care systems
could have a very restrictive impact on the development of such systems and
would certainly delay their implementation.


The health care industry is entering a new era as it moves from the use of
computers to handle administrative tasks into becoming more directly involved
in the care process. In the next 10 years, the computer may well become the
physician’s best friend.

1 Kohn, L.T., Corrigan, Jim, Donaldson, M.S., Eds. To Err is Human: Building
a Safer Health System. Washington: National Academy Press. 1999.