« Problem Escalation Process Language | Main | EHR Forum for Hospitals and Medical Centers »

July 31, 2006


Feed You can follow this conversation by subscribing to the comment feed for this post.

Steve Beller, PhD

One challenge is that every healthcare discipline/specialty requires a different clinical model for displaying cumulative diagnostic test results and other patient information. This means that a primary care physician (PCP) needs certain patient data to, for example, help diagnose a patient’s presenting problem or execute a disease management protocol. A physician specialist would probably require some of the same data, but would also likely need other data to support his/her decisions, which the PCP doesn’t have. And a non-physician (e.g., mental healthcare provider, dentist, dietician, chiropractor, etc.) would need much different data. Furthermore, the way the data are combined, analyzed and displayed may vary considerably across the healthcare continuum. In addition, if clinical decision-support tools are offered, the guidelines, protocols and pathways offer must be matched to specific provider types.

So, when I refer to clinical models, I'm talking about the particular data set, analytic, and display models needed for a particular type of practitioner, as well as the particular type of decision support model offered.

Since (a) these diverse data sets are probably stored in different databases, in different formats, at different locations; (b) the displays that present the data in meaningful reports/profiles would differ considerably from one type of practitioner to another; and (c) the decision-support rules/algorithms would vary widely – it means that many different types of clinical models are needed in order to accommodate all providers that may be treating a patient. This calls for a flexible “clinical model exchange” technology that enables collaborative teams across healthcare disciplines and organizational boundaries to develop, share, evaluate, discuss and evolve many different models tailored to the needs of many different end users.

Note that the models can also include different types of PHRs for different patient types, as well as different business intelligence models.

All of the models must be able to obtain data from any required data stores (databases, document files, spreadsheets, flat files, etc.); perform any required computational and/or logical analyses on the data; and integrate, organize and render (format) the data in ways most useful to the end user. They must also be readily modifiable to accommodate different needs of different users, and be easily transmitted, accessed and used.

There is a cost-effective way to do this.



Dear Steve,

I value your comments about my question posted on the Christina's
Considerations' blog on July 31st. (http://thielst.typepad.com)

Helping improve physician and hospital performance and reducing medical
errors and costs will require a number of innovative easy-to-use and
cost-effective software tools. It seems certain that some will need to be
custom-designed and some will be standardized,

From the perspective of the "generic clinician", there are certain software
applications which by combining standardization and data integration will
serve end-user needs better than customization, with or without data

A group of experienced Rhode Island clinicians representing multiple
specialties has developed, patented and tested this type of application and
named it the Diagnostic Information System(SM).

The Diagnostic Information System(SM) is a clinically integrated,
standardized electronic format for efficiently displaying, viewing and
sharing billions of cumulative patient test results across the continuum of

It is a clinical function and content standard that was specifically
designed by clinicians, for clinicians, to help enhance the workflow of all
caregiver types. It accomplishes this by logically standardizing and
clinically integrating the test results reporting format at the user
interface. Displaying the results of the thousands of different diagnostic
tests in this single view dramatically improves the usability of the user
interface and is ideally suited to replace existing infinitely variable,
data-disintegrating formats.

Working prototypes in two Rhode Island hospitals have already demonstrated
its significant value in improving readability, reducing errors and
duplicative testing and media reduction (decreasing screens and printed
pages by an average of 80%).

In order to solve the longstanding problem of inefficient test results
reporting, we are independently trying to apply the value of
standardization, data integration (forming valid relationships between data
sources) and application integration (presenting a unified view of data to a
user through a computer application) to test results for the first time.

Little discussed, but indisputable among clinicians, is the fact that
today's paper-based medical records and existing EHRs in hospitals and
physician offices (all non-interoperable models) neither standardize nor
integrate the format used to display cumulative test results data for
end-users. Since the mainframe time-sharing era, test results have always
been displayed to end-users using variable formats with no effort to unify
tests from multiple data sources in any clinically logical way.

We also believe the same problem exists with the few standards-based,
interoperable RHIOs that are already fully operational and the hundreds
still under development. Determining if this is accurate is why I posed my
question to Christina Thielst on July 16th.

In the emerging standards-based, interoperable Health Information Exchange
(HIE) era in the United States, we believe that the users' need for both
ease-of-use and ease-of-access will mandate that test results (and all other
health care data types) be displayed for viewing and sharing in the same
comprehensive and logically integrated way.

Because of this immense and growing need for more efficient electronic
health care information sharing, we are now seeking to replace the
traditional paper-based process for managing test results statewide by
creating a private-public partnership with Rhode Island's AHRQ-funded HIE.

The best analogy to what we are trying to accomplish for test results is the
ASTM Continuity of Care Record (CCR). This was designed and implemented as a
standard for a complete and detailed patient health care data summary. The
CCR aggregates data from multiple sources, health care records, health care
encounters, and medical legal documents to form a comprehensive overall
picture of a patient's current and relevant historical health status. (ASTM
Standard E2369-06: http://www.centerforhit.org/PreBuilt/chit_ccrhl7.pdf)

We are proposing to create and provide that same type of comprehensive
overall picture of each patient's current and historical test results by
using a clinically indexed preset format. In effect, a standard for
displaying, viewing and sharing test results.

My basic question:

My basic question to Christina and the large network of healthcare
professionals that value her blog is whether the test results viewing and
sharing problem is yet visible outside of Rhode Island, especially among

The initial standards for four basic health IT functions that President Bush
is expected to create by executive order "in a matter of weeks" by signing
an executive order do include reporting test results
Although these initial standards are "pretty basic", they are considered by
Mike Leavitt to be a "cornerstone for a health care system that allows
patients to make much more informed decisions." However, they clearly do not
address the usability of the physician and patient interface for viewing and
sharing test results.

Also, to my knowledge, the patient data aggregation and integration
companies (1) that create customized hospital, physician, patient and
enterprise-wide Web portals do two basic things:
(1) aggregate patient data from diverse databases, applications
and hardware into a customizable, unified Web portal view using interfaces
with leading HIT applications, portal and infrastructure vendors, and
(2) offer providers and patients customizable views of the data
they need to make timely and informed decisions.

I have not yet discussed this with MedSeek or similar companies, but I am
not aware that any of their clients have implemented what we are proposing
for Rhode Island.

From our clinical perspective, we are seeking to avoid substituting a
statewide dysfunctional electronic system for managing test results for an
already maximally dysfunctional paper-based system.

We believe that using the "flexible and user-defined" (customized) model for
displaying test results rather than a "standardized and clinically
integrated" model will continue to prevent this key type of data from being
efficiently displayed, viewed and shared.

The EHR and PEHR vendors and the clinical providers, as represented by the
medical specialty societies, have demanded explicit detailed semantic
interoperability for the CCR data(2). What do you think of using the same
consensus-based SDO approach to standardizing test results reporting?

(1) including Quovadx/Carescience, MedSeek, Carefx, Medicity,
DBMotion, Inc. and NWH.
(2) i.e. demanding that "there be universal compliance with a
defined syntax, semantic content, and explicit structure to all CCRs and
that no local extensions or configurability be allowed."

Best regards,
Bob Coli, MD


Dear Bob,

Your cogent reply indicates that we are approaching the important issue of
standardization from two different paradigms. I'll clarify my point of view
and answer your question accordingly.

We agree that there needs to be a way exchange and integrate data from
disparate sources (EMR/EHR/PHR databases, aggregate research database,
reference literature repositories, etc.). That means there must be some way
to map the data from one source to another so each knows the location,
structure, syntax and context. That allows the data to be transformed as
necessary to accommodate different presentation displays (i.e., different
electronic forms/reports). But that's not what this discussion focuses on.

Instead, it seems to me we agree the mapping and data transformation
processes is required, and are focusing on whether the content (the test
data) should be presented (laid out/formatted/displayed/viewed) in a
standardized way in electronic forms/reports (e.g., a CCR).

You said: > And herein lies
the paradigm shift.

If changing the presentation (display) of the content adversely affected the
data mappings/transformations, allowing users to change the way that content
is presented is a problem. But if the content is *separate* from
presentation, then customized presentation is no problem since the display
parameters do not affect the mappings or transformations required for

There are several ways to separate content from presentation. One is by
using of XML tags. Another is called the CPSplit technology, a process I
invented in the late '90s (see http://nhds.com/te/index.html). These
processes enable the content to be presented in any manner by any user
without adversely affecting the ability to access, integrate and exchange
the data. The electronic forms/reports can then be customized for each
user's needs without problem.

With the CPSplit technology, the underlying data (the test results) are
obtained from multiple sources, transformed, and integrated using automated
(macro-driven) templates (electronic models), such as a simple spreadsheet
program. If you want to perform computational and/or inferential logic
analysis on the data (e.g., to identify individual and combined test results
that are beyond normal limits, do trend analyses for individual patients,
compare patient groups, make recommendations using evidence-based
guidelines, etc.), then the requisite formulas/rules/algorithms/proofs would
be included in the template. This "publication template" then takes the raw
test data and any analytic results and stores them in a small virus-proof
data file ready for transport and presentation. A second template, the
"presentation template," then takes the data file and, using formatting
instructions contained within it, the contents of the data file are
displayed as per the user's preferences. The presentation template "knows"
what each piece of data in the data file is and how to render it, so the
template just "grabs" the data from the data file and displays it as

In this paradigm, the standards are limited to things like data mapping and
transformation for interoperability. There may also be standards for data
analytics (e.g., criteria for judging if a test value is abnormal) and for
defining the specific data sets useful for each specialty. But there is no
need to force users to adhere to inflexible presentation/display standards.
Thousands of test results can still be displayed efficiently, and these
displays can be customized. Furthermore, these test results can be
integrated with any other patient data in the same basic way.

I hope I'm making sense. I look forward to your next reply and would be
happy to speak with you about it.

BTW, Christina, I have no problem should you wish to post any of this on
your blog; assuming Bob's OK with it.



Stephen E. Beller, Ph.D.
National Health Data Systems, Inc.
130 Hastings Ave.
Croton On Hudson, NY 10520
Voice: (914) 271-5434 ext. 2
E-mail: sbeller@nhds.com
Web site: http://www.nhds.com
Wiki: http://wellness.wikispaces.com

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Comments are moderated, and will not appear until the author has approved them.

Your Information

(Name is required. Email address will not be displayed with the comment.)

My Photo

Social Media Tools

Follow Cthielst on Twitter

Twitter Updates

    follow me on Twitter

    Become a Fan

    Your email address:

    Powered by FeedBlitz

    Health Literacy

    • Health Literacy – www.cdc.gov/healthliteracy

    Tip Jar

    Change is good

    Tip Jar

    Learn More


    Influential Networks

    Blog powered by Typepad