To The Moon In A Hot Air Balloon: Why Is Clinical IT Difficult?

Scot M. Silverstein, MD Dec. 10, 2007

Suss Consulting, Inc.
Drexel University, College of Information Science & Technology

1
– Introduction to Medical Informatics
– EMR vision, knowledge gaps, realities
– Case examples
– Recommendations
– Questions

2
 On Medical Informatics
• Medical Informatics is a cross-disciplinary field
that studies information-seeking activities and
tools, analytic processes, and workflows in
biomedical research and clinical care delivery. It
focuses upon the innovative use of computers in
clinical medicine, molecular biology,
neuroscience, and other areas of biomedical
research.

3
 NIH training programs in Medical Informatics
http://www.nlm.nih.gov/ep/GrantTrainInstitute.html

1 - University of California Irvine (Irvine, CA), 2 -University of California Los Angeles (Los Angeles, CA), 3- Stanford University (Stanford, CA), 4-
University of Colorado Denver/HSC Aurora (Denver, CO), 5- Yale University (New Haven, CT), 6- Regenstrief/Indiana University (Indianapolis, IN),
7- Harvard University (Medical School) (Boston, MA), 8- Johns Hopkins University (Baltimore, MD), 9- University of Minnesota Twin Cities
(Minneapolis, MN), 10 - University of Missouri-Columbia (Columbia, MO), 11- Columbia University Health Sciences (New York, NY), 12- Oregon
Health & Science University (Portland, OR), 13- University of Pittsburgh at Pittsburgh (Pittsburgh, PA), 14- Medical University of South Carolina
(Charleston, SC), 15- Vanderbilt University (Nashville, TN), 16 - Rice University (Houston, TX), 17- University of Utah (Salt Lake City, UT), 18-
University of Virginia Charlottesville (Charlottesville, VA), 19 - University of Washington (Seattle, WA), 20 - University of Wisconsin Madison
(Madison, WI). Training is provided by other universities via internal funds as well.

4
The Informatics Subspecialties

5
 Why is Medical Informatics important?
• Healthcare IT and its environment are core competence
• Recognition that organizations are simultaneously social (people, values,
norms, cultures) and technical (tools, equipment, technology). These
elements are deeply interdependent and interrelated. Good design and
implementation is not just a technology issue but also one of jointly
optimizing the combined sociotechnical systems.
• Medical Informatics training recognizes these issues and trains cross-
disciplinary specialists accordingly. Curriculum example:
– Fundamental of computer science
Medical language and terminology systems
Modeling of medical observations and data
Medical coding systems
Medical knowledge structures
Information organization and flows in medical practice
Quantitative models for medical decision making
Clinical decision support
Medical image processing
User interfaces and ergonomics in healthcare
Health information systems architecture
Security and confidentiality
Ethical and legal issues in electronic medical records
Organizational and sociological issues in clinical IT projects
Metrics and methods for evaluating healthcare information systems
Cost and investment issues in healthcare IT

6
 The Vision …
• "We will make wider use of electronic records and
other health information technology to help control
costs and reduce dangerous medical errors."
-- President Bush, January 31, 2006

• "..to link all health records through an interoperable

system that protects privacy as it connects patients,
providers and payers, resulting in fewer medical
mistakes, less hassle, lower costs and better health."
-- HHS Secretary Mike Leavitt

7
 Revolutionizing Healthcare
• In April 2004, President George W. Bush revealed his vision for the future of healthcare in the
United States. The President's plan involves a health care system that puts the needs of the
patient first, is more efficient, and is cost-effective. The President's plan is based on the
following tenets:

– Medical information will follow consumers so that they are at the center of their own care
– Consumers will be able to choose physicians and hospitals based on clinical performance results
made available to them
– Clinicians will have a patient's complete medical history, computerized ordering systems, and
electronic reminders
– Quality initiatives will measure performance and drive quality-based competition in the industry
– Public health and bioterrorism surveillance will be seamlessly integrated into care
– Clinical research will be accelerated and post-marketing surveillance will be expanded.

• Together, these tenets will revolutionize healthcare, making it more consumer-centric, and
will improve both the quality and the efficiency of healthcare in the United States.
Source: http://www.hhs.gov/healthit/vision.html

8
 Revolutionizing Healthcare?
• MS-HUG meeting, late 1990’s, CEO’s of
Cerner, HBOC and other vendors calling for
room to “revolutionize healthcare”
– How many people here think they will revolutionize
healthcare?
– How many people here have true clinical experience, or
have read a medical text such as Harrison’s or The Merck
Manual?

9
 Achilles’ Heel: False assumptions and underestimations
• Balloons go up in the sky.
• The moon is up in the sky.
• With enough process and resources, we can travel to the moon in a balloon.
• What we don’t know can’t hurt us, and it’s bad for business anyway.

10
 Implementing EMR not as easy as it seems
• Not a technology problem any longer.
• Medical informaticists study the issues that impede EMR
adoption in clinical settings.
• What if clinical computing and business computing are
different, distinct subspecialties of computing, much as
neurosurgery and psychiatry are highly distinct medical
subspecialties?
• What few talk about regarding clinical information technology:
social and organizational causes of healthcare IT difficulties
– “Sociotechnologic issues in clinical computing: Common examples of
healthcare IT difficulties” at http://
www.ischool.drexel.edu/faculty/ssilverstein/medinfo.htm

11
 To an information scientist,
Google is revealing

• Relevant hits on “Healthcare IT Failure” or similar

concepts: almost none

• Hits on “Medical Malpractice” 294,000

• Why?

12
 Access Patterns to a Website on Healthcare IT Failure
Scot M. Silverstein MD, Yunan Chen, Christine Wania
College of Information Science & Technology, Drexel University

Background Visitor types

2842
3000

Organizational and human factors (sociotechnical) issues 2500

associated with healthcare IT have led to project difficulties and 2000

failures. Detailed case accounts might improve knowledge sharing

1500
between healthcare organizations on lessons learned and best
implementation practices. Based on an online search for web- 1000
564
based resources, there appears to be few relevant sites containing 500
332
239

either high-level or case-level information regarding healthcare IT 149 93 97

11

difficulty and failure. Case-level accounts describing issues at a 0

ISP University / Other Healthcare Unknown Government Healthcare Other
Educational Organization Organization Organization Related Individual
fine level of granularity, such as detailed accounts of interactions Institution Industry
Note: The terms in this diagram represent our categorization of the concepts used by Web
among clinicians, IT personnel and healthcare executives, are
searchers to find our website. For example: “healthcare IT” might have been expressed as Visitor Countries distribution
potentially useful learning resources. Knowledge sharing via the
“medical computing” or “clinical information technology.”
Web on best practices for management of sociotechnical
impediments to healthcare IT is thus suboptimal.

Case Study Results

Research Questions We studied access patterns to our website on healthcare IT difficulties. The
distribution of countries, website visitor types and query types are shown in the
following charts:
• To illustrate the level of interest and knowledge sharing about
Healthcare IT difficulties, we explored two research questions: Visitor Input Types
 What case-level information about healthcare IT project
difficulty and failure is available via the Web? 2500 2252

 Who is seeking this information?

2000

1500
Methods
1000
Conclusion
500 206
We used our website, entitled “Sociotechnologic Issues in Clinical 166 132

Computing: Common Examples of Healthcare IT Failure”, online It appears there is an ongoing interest in information about
0
since 1999, as a representative information resource. Search Engine Healthcare IT Site Other Non Healthcare IT healthcare IT difficulty among searchers of varied
Site
demographics, as evidenced by searches on these issues and
We first conducted a detailed Web search on various search resultant “hits” on our website, one of the few websites that
phrases representing the concept “healthcare IT failure” using
Visitor Search Engine Referrer Input specifically addresses these issues. At present such material is
three major search engines (Google, Yahoo, and MSN). The uncommon. We believe Medical Informatics specialists can
results showed that our website was nearly unique. We then 800
774
contribute significantly to filling this information gap, and that
tracked views of our website from Sep 27th, 2005 to June 30th, 700
doing so would be very helpful to the healthcare IT community.
2006 with a free public tracking service, eXTReMe Tracking 600
(http://extremetracking.com). 423
500
389
Our website “Sociotechnologic Issues in Clinical Computing:
400
We examined the access logs to determine viewer IP and referrer Common Examples of Healthcare IT Failure” is at URL:
(primarily search engine query, or link from another site) to 300 227
193

identify viewer demographics including country and organization 200 116

http://www.ischool.drexel.edu/faculty/ssilverstein/medinfo.htm
type, where available. No personally-identifiable information was 100
It was created by our first author via an informal collaboration of
sought or obtained. 0 AMIA clinical information systems working group members (CIS-
13
Healthcare OR Healthcare IT Failure Healthcare IT Unrelated Project
IT OR Failure AND IT Failure Management WG) during 1998-2001.
 Why is good clinical IT difficult?
• Lindberg: “Computer Failures and Successes”, Southern Medical Bulletin 1969;57:18-21

– Computer experts per se have virtually no idea of the real problems of medical or even hospital
practice, and furthermore have consistently underestimated the complexity of the problems…in no
cases can [building appropriate clinical information systems] be done, simply because they have not
been defined with the physician as the continuing major contributor and user of the information.

• Nemeth & Cook: “Hiding in Plain Sight”, Journal of Biomedical Informatics 38 [2005], 262–
263

– Just beneath the apparently smooth-running operations [of healthcare] is a complex, poorly
bounded, conflicted, highly variable, uncertain, and high-tempo work domain. The technical
work that clinicians perform resolves these complex and conflicting elements into a productive work
domain. Occasional visitors to this setting see the smooth surface that clinicians have created and
remain unaware of the conflicts that lie beneath it. The technical work that clinicians perform is
hiding in plain sight. Those who know how to do research in this domain can see through the
smooth surface and understand its complex and challenging reality. Occasional visitors cannot
fathom this demanding work, much less create IT systems to support it.

14
 The Gaps …
• IT Industry View
– Computers are the solution.
– If it’s information, we do it, and with a few
domain experts helping us, we can automate
anything.
– maybe – assuming no unpredictables intervene.
Clinical computing is a subspecies of MIS.
– We take specs, build the system employing SLC
and other methodologies, deliver product, do a few
tweaks, and we’re done.
– End users are to be managed, and occasionally
listened to.
– The world operates like an assembly line.
– With enough process, “skill sets” and resources,
we can go to the moon in a hot air balloon.
– We don’t need space scientists interfering in our
endeavors. What do they know about IT and about
project management?
– Our way is the only way. Agile methodologies?
Iterative and incremental development? Doctors
leading computing projects? No way.
• Medical Informatics View
– Computers are inanimate objects. A computer
with appropriate software developed incorporating
appropriate domain expertise is a tool. A tool in
the right hands and right contexts is a solution –
– Clinical computing is a subspecies of “scientific”
computing.
– Throwing IT at a health care system to remedy its
problems will not be effective unless the
organizational reasons for those failures also are
addressed.
– These reasons are hidden in the "messy details" of
clinical work: complexity; uncertainty; conflicting
goals; gaps in supplies, procedures, and
coordination; brittleness of tools and
organizational routines.
– These cannot be “modeled” and automated in the
traditional MIS manner.
– “Changing the culture” of healthcare involves
changing the culture of clinicians AND changing
the culture of healthcare IT professionals.
15
 The Gaps …
• IT Industry View • Medical Informatics View

16
From Kling, Rosenbaum &
Sawyer: Understanding and
Communicating Social
Informatics, Information
Today (1st Edition, 2005)

17
 Hubris? – Observed by me …
• Medical data is the same as anyone else’s data.
• Don’t worry about it, you can’t understand it, we’ll make it better.
• I don't think a degree gets you anything. There's nothing like the school of
Hard Knocks [for healthcare CIO’s].
• I don’t think clinical experience yields healthcare IT workers who have
broad enough perspective. Physicians in particular make poor choices for
CIOs. They don't think of the business issues at hand because they're
consumed with patient care issues.
• Only satellites in Earth orbit need protection from x-rays.
• Doctors don’t do things with computers.
• Medical Informatics is too academic.
• We only support Compaq computers in our hospital.
• We have a “turnkey system only” rule. No customizations.
• Doctors just don’t want to do things the way the application requires.

18
 The Realities
• "Many information systems simply don't reflect the health care
professional's hectic work environment with its all too frequent
interruptions from phone calls, pages, colleagues and patients. Instead
these are designed for people who work in calm and solitary environments.
This design disconnect is the source of both types of silent errors …Some
patient care information systems require data entry that is so elaborate that
time spent recording patient data is significantly greater than it was with its
paper predecessors," the authors wrote. "What is worse, on several
occasions during our studies, overly structured data entry led to a loss of
cognitive focus by the clinician.“ – OHSU research on CPOE

Source: http://www.eurekalert.org/pub_releases/2003-11/ohs-mhd112403.php

19
20
21
• In 2005 University of Pennsylvania sociologist Ross Koppel,
PhD and clinical colleagues, at some risk to their professional
reputations, published findings that a leading Computerized
Practitioner Order Entry (CPOE) system often facilitated
medication error risks, with many reported to occur frequently.
• “Role of Computerized Physician Order Entry Systems in
Facilitating Medication Errors” (JAMA 2005;293:1197-1203)
– link to abstract: http://jama.ama-assn.org/cgi/content/abstract/293/10/1197

22
• In a 2007 research study involving the University at Buffalo and other
institutions, "Design of Patient Tracking Tools May Have Unintended
Consequences", researchers found that “Virtual Patient Status
Whiteboard” systems to replace traditional, dry-erase patient status boards
in hospital Emergency Departments (ED’s), but designed by technologists
without adequate domain expertise, often do not work as expected. The
computer-based system interferes with staff communications.

• In some cases, providers noted that computer systems hid some of the information; if
only three comments could be viewed per screen, they had to click to get to another
screen, requiring them to search for information that might demand immediate
attention. The study also found that there were fewer visual cues with the
computational system. Some providers noted that they used to be able to get a sense
of the status of the emergency department just by walking through the room and
visually checking the manual whiteboard. “Without that public display, providers
have to sit down at the computer and check it, which can add time or reduce
awareness,” said a principal investigator.

Reference: http://www.eurekalert.org/pub_releases/2007-11/uab-dop112607.php

23
• Research article in the Archives of Internal Medicine -
“Electronic Health Record Use and the Quality of Ambulatory
Care in the United States” (Arch Intern Med. 2007;167:1400-
1405, link to abstract here):

– The authors examined electronic health records (EHR) use throughout

the U.S. and the association of EHR use with 17 basic quality
indicators. They concluded that “as implemented, EHR’s were not
associated with better quality ambulatory care.”

– On several indicators, physicians using paper did better.

24
 Case Example
• Serious clinical computing problems in the
worst of places: an ICU

– http://www.ischool.drexel.edu/faculty/ssilverstein/failurecases/?loc
=cases&sloc=clinical%20computing%20problems%20in%20ICU

25
 Recommendations
• Temper the irrational exuberance that now prevails.
• Recognize sociotechnical barriers and work with them, not
against them.
• Address need for specialized approaches to clinical computing
(vs. business computing).
• Improve Medical Informatics professionals’ representation in
health IT management
• Incorporate medicine’s rigor into the medical IT leadership
and staff – end “pre-Flexner report healthcare IT”
– Familiarity with nuances of healthcare and healthcare IT
– Specialized educational experience
– Standards and certification for healthcare IT workers

26
 Medical Informatics can take us to exotic places!

Red Sand Dunes area, ~

50 km w. of Riyadh

27
Supplementary slides

28
 The iSchools
Focus is on how people seek, use or interact with information using technology,
not simply on technologic devices and computer programs.

• University of California, BerkeleySchool of Information Management and Systems University of MichiganThe School of Information

• University of North CarolinaSchool of Information and Library Science

University of California, IrvineThe Don Bren School of Information and Computer Sciences

• University of California, Los AngelesGraduate School of Education and Information Studies

The Pennsylvania State UniversitySchool of Information Sciences and Technology

• Drexel UniversityCollege of Information Science and Technology

University of PittsburghSchool of Information Sciences
• Florida State UniversityCollege of Information
Rutgers, the State University of New JerseySchool of Communication, Information, and L
• Georgia Institute of TechnologyCollege of Computing

• University of Illinois Urbana-ChampaignThe Graduate School of Library and Information Science

• Indiana UniversitySchool of Informatics
• Indiana UniversitySchool of Library and Information Science
• University of MarylandCollege of Information Studies
Syracuse UniversitySchool of Information Studies
University of Texas, AustinSchool of Information
University of TorontoFaculty of Information Studies
University of WashingtonInformation School

29
 Medical Informatics: What it is not
• Medical informatics is a formal field of study and a scientific discipline. Computers in
medicine is not the equivalent of medical informatics.

• Position descriptions with requirements for neither clinical nor medical informatics training or
experience, are often seen. From the Philadelphia Inquirer:

– Medical informatics analyst. [Company name] seeks a Medical Informatics Analyst to support resource
management and medical affairs in their data analysis needs. Through downloading of MCS database, PC-
based analysis of clinical and clinically-relevant financial data will be performed. Qualifications: BA/BS in
computer science or related discipline, 3-5 years experience in PC-based data analysis of health care
information, knowledge of SAS or similar analysis software, knowledge of mainframe DB2 database
architectures, working knowledge of CPT-4 and ICD-9-CM coding."

– From an Internet biomedical employment service, Medzilla.com: [Company name] seeks a Director of
Clinical Informatics. Overall coordination of design specifications, implementation and support for all
internet/browser based systems. Assurance of continued, reliable and consistent resources and
applications to all corporate personnel and external users who may rely upon these systems.
Documentation and control of said software systems including package systems and license control if
necessary. Provide ongoing maintenance oversight and management support for said systems. Organize
and train personnel, both internal and external, who will be using said products. Client contact and
development and assist Sales and Marketing as necessary in client presentations. Minimum of a BA. MBA
preferred.

30