Using virtual reasoning to redefine health care

is the CEO of Enhanced Medical Decisions, which is the company behind DoubleCheckMD.com.

The Internet is redefining the health care industry. Major transformations can be expected because Internet-based technology will deliver certain health care services more effectively and at lower costs. In the near future, much of the information that is currently imparted to consumers by clinicians will be delivered through and by web-based technology. If the web-based tools that deliver this information mature to the point of becoming reimbursable, beyond their current usefulness as value add-ons, the health care industry could experience a dramatic shift.


Early stage Health 1.0 information has predominantly been available
in the form of an “e-pamphlet” with a one-size-fits-all approach.
Subsequent generations of Health 2.0 tools are interactive and deliver
personalized, and therefore more valuable information that is geared
specifically to the user’s input. With the advent of advanced tools,
consumers are no longer limited to being passive recipients of
pamphlet-style information; in these milieus they interact with each
other or with sophisticated software that analyzes cases on an
individualized basis. Advanced tools are being used by Internet
entities as bait both to drive traffic and to capture detailed user
profiling data.

Several categories of interactive Health 2.0 tools are available,
including advanced search engines that deliver more accurate results;
social media sites in which individuals hone their medical knowledge
through interactions with each other; and finally, expert systems –
sophisticated software programs that analyze a consumer’s profile and,
based on the analysis, pinpoint the most relevant educational
information necessary to support the consumer’s health care decisions.
Expert system tools basically simulate human reasoning. 

Virtual Reasoning – a New Model

Speculation about where Internet health care might be taking us can
be approached, in part, as a projection of the next generation of the
most advanced web-based health care tools.  If what consumers and
payers want from health care is an acceptable cost-quality trade off,
assurance that care is appropriate for the situation, that the
diagnosis is correct, and that errors of commission and omission are
kept to a minimum, web-based tools need to move beyond simple
information retrieval to the level of analytic services – expert
systems that are virtually capable of reasoning, rather than only
presenting facts.

If Health 2.0 products bring you cholesterol guidelines, then
virtual reasoning tools, in counter-point, should calculate your LDL
(bad cholesterol) goal based on your level of risk for heart disease
and assess whether or not you are taking the right medicine.  If Health
2.0 delivers personalized information, virtual reasoning systems offer
the equivalent of a virtual second opinion.    

As virtual reasoning tools reach a greater level of maturity, they
will represent cost-effective alternatives to certain health care
services traditionally performed face-to-face by health care
professionals. At some point, they will cease to function merely as
value add-ons and become reimbursable, revenue-realizing businesses in
their own right, augmenting some of the educational and analytic
services now in the purview of health care providers. The logical
end-point of expert system and decision-support adoption will be a new
reimbursable segment of the service industry in which lower cost
services are rendered through technology solutions.  When physicians
are reimbursed to use expert system technology, they will have
sufficient incentive to adopt electronic medical records (EMR) as well.

For tools to replace certain traditional face-to-face services they
have to be integrated into the clinician’s normal workflow and
electronic medical records systems.  EMR companies will first
incorporate expert system tools for much the same reasons as the PHR
initiatives do – for the value-added benefits they offer to the
customer base. In future models, though, fully integrated tools will
sit “under the hood” of an EMR, continually combing the medical record
data for errors and oversights as new data is entered. They will
compare treatment to evidence-based recommendations, follow response to
treatment over time, generate outcomes data, and generally function as
an automated quality assurance system.   

Physician Adoption

Physicians will ultimately be reimbursed for the time they spend
administering care through the medium of information technology. Early
pilots will likely be initiated by payers (large, self-insured
employers) who believe that the potential for improved, more
appropriate care will result in significant cost-savings. These
technology adjuncts will free health care professional from the mundane
functions of data gathering, recording, and administering; they will
enable physicians to focus on the more rewarding cognitive aspects of
medical practice. With more free time, physicians can move squarely
into a consultancy role in which they help their patients assimilate
and analyze increasingly complex choices.   

Significant activity around tool adoption is already underway:
Microsoft, Google and other Internet companies have been acquiring
advanced search and expert system tools; programs in which web-based
service delivery is reimbursed are being piloted; and Google Health has
taken the first steps to make a large number of tools available on
their platform. How quickly the process eventually unfolds depends on
a number of factors that mostly revolve around any upcoming changes in
health care financing and in the political climate. No matter how
health care financing is structured, however, the need for tools that
enhance health care quality and efficiency still applies.

2 replies »

  1. bev MD’s response highlights some of the systemic intellectual resistance (on the provider side)to what is clearly an enlightened perspective offered by Marlene Beggelman’s article. The underlying assumption speaks to a belief that today’s provider role will always be central to the healthcare equation. A brief glance at Levitt’s classic “Marketing Myopia” should place every provider on notice.
    If one takes musculoskeletal medicine (the 900 lb gorilla that includes back pain) as an example, most ‘care’ breaks down to information therapy and pattern recognition. The provider solicits historical information from the patient(which can be transformed into machine readable structured data),adds key physical findings (which do not necessarily require high levels of education or training to detect, and can also take the form of structured data), then maps this information against the providers database (the sum total of education, experience and testing tools)to make decisions about the ‘care’ that follows.
    This is perhaps a simplistic model, but I would submit that there is not one part of this ‘care’ paradigm that cannot and will not be supplanted (in majority part) by virtual reasoning tools in the future. The providers will continue to function as “domain experts” but many of the routine decisions will be made on a very cost effective level by machine “supported” by human oversight. This you will notice is decidely different from the more palatable decision support role artificial intelligence takes in todays dialogue.
    While we may debate how acceptable this may be to the medical consumer (or provider for that matter), I submit that costs will ultimately drive these changes. We know and accept the fact that machines contain larger databases and are much more efficient in many decision processes. What we may find harder to accept is the notion that physicians or other providers may not necessarily be in a position to be reimbursed for ‘care’ administered through the medium of information technology.

  2. Artificial intelligence systems in medicine have been around for awhile; although they have their benefits as reminders and prompters of physician diagnosis and treatment, it will be awhile before they can be standalone systems. See “Happy Hospitalist”‘s post of 6.30.08 for an example. As we used to say, the patients don’t read the books (e.g. how a disease presents itself does not always follow textbook precepts). Also, who will pay when they get it wrong? Can computers carry malpractice insurance?