Healthcare Information Management Systems Cases

Strategies And Solutions 5th Edition Joan M Kiel

download

https://ebookbell.com/product/healthcare-information-management-
systems-cases-strategies-and-solutions-5th-edition-joan-m-
kiel-48673620

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

Healthcare Information Management Systems Cases Strategies And

Solutions 3rd Edition Neal L Patterson Mba Auth

https://ebookbell.com/product/healthcare-information-management-
systems-cases-strategies-and-solutions-3rd-edition-neal-l-patterson-
mba-auth-4500970

Healthcare Information Management Systems Cases Strategies And

Solutions 4th Edition Charlotte A Weaver

https://ebookbell.com/product/healthcare-information-management-
systems-cases-strategies-and-solutions-4th-edition-charlotte-a-
weaver-5353850

The Cphims Review Guide 4th Edition Preparing For Success In

Healthcare Information And Management System 4th Edition Healthcare
Information Management Systems Society Himss

https://ebookbell.com/product/the-cphims-review-guide-4th-edition-
preparing-for-success-in-healthcare-information-and-management-
system-4th-edition-healthcare-information-management-systems-society-
himss-36065484

The Cphims Review Guidepreparing For Success In Healthcare 4th Edition

Healthcare Information Management Systems Society Himss

https://ebookbell.com/product/the-cphims-review-guidepreparing-for-
success-in-healthcare-4th-edition-healthcare-information-management-
systems-society-himss-36149150
Himss Dictionary Of Health Information Technology Terms Acronyms And
Organizations 5th Edition Healthcare Information And Management
Systems Society Himss

https://ebookbell.com/product/himss-dictionary-of-health-information-
technology-terms-acronyms-and-organizations-5th-edition-healthcare-
information-and-management-systems-society-himss-9968778

Improving Outcomes With Clinical Decision Support An Implementers

Guide Second Edition Paperback Healthcare Information And Management
Systems Society

https://ebookbell.com/product/improving-outcomes-with-clinical-
decision-support-an-implementers-guide-second-edition-paperback-
healthcare-information-and-management-systems-society-10022398

Cphims Review Guide Preparing For Success In Healthcare Information

And Management Systems 3rd Edition Healthcare Information And
Management Systems Society Himss

https://ebookbell.com/product/cphims-review-guide-preparing-for-
success-in-healthcare-information-and-management-systems-3rd-edition-
healthcare-information-and-management-systems-society-himss-10022684

The Cahims Review Guide Preparing For Success In Healthcare

Information And Management Systems 2nd Edition Himss

https://ebookbell.com/product/the-cahims-review-guide-preparing-for-
success-in-healthcare-information-and-management-systems-2nd-edition-
himss-43852272

The Cphims Review Guide Third Edition Preparing For Success In

Healthcare Information And Management Systems 3rd Edition Himss

https://ebookbell.com/product/the-cphims-review-guide-third-edition-
preparing-for-success-in-healthcare-information-and-management-
systems-3rd-edition-himss-5699240
Health Informatics

Joan M. Kiel
George R. Kim
Marion J. Ball Editors

Healthcare
Information
Management
Systems
Cases, Strategies, and Solutions
Fifth Edition
Health Informatics
This series is directed to healthcare professionals leading the transformation
of healthcare by using information and knowledge. For over 20 years, Health
Informatics has offered a broad range of titles: some address specific
professions such as nursing, medicine, and health administration; others
cover special areas of practice such as trauma and radiology; still other books
in the series focus on interdisciplinary issues, such as the computer based
patient record, electronic health records, and networked healthcare systems.
Editors and authors, eminent experts in their fields, offer their accounts of
innovations in health informatics. Increasingly, these accounts go beyond
hardware and software to address the role of information in influencing the
transformation of healthcare delivery systems around the world. The series
also increasingly focuses on the users of the information and systems: the
organizational, behavioral, and societal changes that accompany the diffusion
of information technology in health services environments.
Developments in healthcare delivery are constant; in recent years,
bioinformatics has emerged as a new field in health informatics to support
emerging and ongoing developments in molecular biology. At the same time,
further evolution of the field of health informatics is reflected in the
introduction of concepts at the macro or health systems delivery level with
major national initiatives related to electronic health records (EHR), data
standards, and public health informatics.
These changes will continue to shape health services in the twenty-first
century. By making full and creative use of the technology to tame data and to
transform information, Health Informatics will foster the development and
use of new knowledge in healthcare.
Joan M. Kiel • George R. Kim
Marion J. Ball
Editors

Healthcare Information
Management Systems
Cases, Strategies, and Solutions

Fifth Edition
Editors
Joan M. Kiel George R. Kim
Duquesne University The Johns Hopkins University
Pittsburgh, PA, USA Baltimore, MD, USA

Marion J. Ball
The University of Texas at Arlington
Arlington, TX, USA

ISSN 1431-1917     ISSN 2197-3741 (electronic)

Health Informatics
ISBN 978-3-031-07911-5    ISBN 978-3-031-07912-2 (eBook)
https://doi.org/10.1007/978-3-031-07912-2

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Switzerland AG 1991, 1995, 2004, 2016, 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher,
whether the whole or part of the material is concerned, specifically the rights of translation,
reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any
other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, expressed or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
The editorial work of this volume is dedicated to the
memory of:

Thomas D. Kiel, M.D.

by Joan M. Kiel, Ph.D.

Meg Koppel, Ph.D. (in honor of Ross Koppel, Ph.D.)

My parents
Andrew R. Scholnick

by George R. Kim, M.D., F.A.A.P.

 y beloved husband John Charles Ball who, throughout my

M
career, had been:
– a guiding light and supporter of my professional activities,
– a dedicated father to our beloved children: Charles Jokl Ball
and Elizabeth Ball Concordia, as well as
– a devoted grandfather to our five grandchildren: Alexis
Marion Concordia, Alexander John Ball, Michael John
Concordia, Ryan Jokl Ball, and Erica Adelaide Concordia.
I was blessed to have John at my side for fifty-eight years.
Without him, I would not be where I am today.

In loving memory,
Marion Jokl Ball, Ed.D.
 e join our co-editor, Marion Jokl Ball, in honoring and
W
remembering Dr. John Charles Ball, a scholar and a friend.

Joan and George

Foreword

Information is life. Or not. Quite literally in healthcare, the correct (or incor-
rect) information at the right time in the right context creates a bright line that
separates life from death for many patients. The technological systems which
accept, store, retrieve, and present this critical information continue to
evolve—but the history of their development is littered with more failures
than successes. While we should learn from our failures, too often we simply
press forward and communally repeat the same failures devolving into a
vicious cycle of innovation for innovation’s sake. Technology can help us
solve problems in healthcare; however, it is when we start by clearly under-
standing those problems that the most fertile environment for virtuous cycles
of information, process, and technology innovation occurs. When we apply
the creative capability of our collective efforts toward understanding the jobs
to be done, the problems to be solved—that is when we shine the brightest in
our success.
I write this foreword as we begin to emerge from the global pandemic
wrought by COVID-19. In the last year, we have seen telemedicine and
remote care advance by leaps and bounds. Concurrently, we are seeing more
successful cyber-attacks shut down healthcare around the world. The pan-
demic has created not just an environment for change, but an imperative to
change. Barriers—political, economic, logistical, even technological—have
given way to the irresistible force of immediacy of need, a metaphorical
“burning platform” (à la Clayton Christensen).
In this past year we have lost far too many souls to the tragedy of this
global pandemic. My reasons for serving in healthcare have always been
rooted in the losses of family and friends to death and disease as well as the
injury and illness suffered daily by those I love. When we choose to serve in
healthcare, we choose to be healers. Whether our path is through research,
technology, clinical care, or any of the other innumerable branches of health
and care—we make a difference, we contribute, we matter.
Throughout the pandemic, healthcare researchers and practitioners have
rallied in an incredible effort to provide care and compassion to the billions
impacted. We have heard stories of unrelenting resolve alongside stories of
tragedy and loss. As I experience connections with students in my classes, we
consider the future of healthcare through a “visioning” experiment. In this
experiment, each student closes their eyes and is led to consider what health-
care at their organization will look like in another 10 years. Across all the
students I serve—including nurses, doctors, technologists, informaticists, and

vii
viii Foreword

many more—a recurring theme of optimism can be heard. When they con-
sider the potential futures within their own organizational healthcare context,
all see a more convenient, efficient, and personal healthcare system. While it
may seem obvious, that system will not come about on its own—it depends
on the efforts and faith of those who serve in it. May each of us, whether serv-
ing on the front lines of care, the backstage of support, or the academic com-
munity of preparation, find our personal vision of a bright future which makes
that future inevitable.
This book describes numerous successes as well as some useful failures in
our work toward more virtuous healthcare information technology. Through
the narration of the community of contributors in this edition, we experience
the breadth of actualities and possibilities in our world of healthcare IT.
There are many books about healthcare, technology, and even healthcare
information technology. However, none of those other books rival the one
you hold in your hands (literally or figuratively) at better representing the
current collective knowledge and experience of healthcare information sys-
tems experts. Within the chapters of this text, you will encounter nurses, phy-
sicians, informaticists, technologists, researchers, analysts, educators,
scientists, and many others with titles and positions that represent the broad
world of healthcare information professionals. The scope of their experience
spans the globe and represents the best and brightest our professions have to
offer. Many I am privileged to know as friends, others I know by reputation,
and I welcome you to partake in their collective wisdom through this text.
This volume offers the opportunity for students, professionals, and leaders
who are responding to a calling to care to walk through a door of understand-
ing into a room of knowledge. This knowledge represents the exponential
impact that healthcare information systems bring when designed, imple-
mented, and used well for the power of healing. Information is the lifeblood
of healthcare and technology is the catalyst for information efficacy.
Transformation, leadership, informatics, ethics, technology, analytics,
vision, architectures, innovations, modeling, interoperability, remote care,
regulations, economics, safety, patient empowerment, public health, disaster
management, and virtual care are some of the essential topics addressed by
the experts assembled within this volume. While the topics venture far and
wide, the consistent theme across every chapter is the inevitability and the
imperative for advances in healthcare information technology.
We have entered the fourth revolution. Boundaries are shifting, blurring,
and disappearing between the organic and inorganic, the digital and the phys-
ical. The possibilities for good have never been greater, but like prior indus-
trial revolutions the potential for great harm also exists. Our choices in how
we apply advanced digital capabilities to health and care will be a signal to
the rest of society about the likelihood of our collective direction: good or
harm, healing or hurt, division or unification.
We are always faced with the options of entrenchment in our own current
environment as well as the risk-laden option to venture out into the newness
of innovation and change. While our minds tell us change is required, our
hearts sometimes fear and seek solace in the familiar. As leaders in health and
care, our obligation is to press our hearts and minds into an integrated whole
Foreword ix

and face the needs of our patients, communities, organizations, and even our-
selves as we step out of the familiar into the bright future set before us.

The College of Healthcare Information Timothy Stettheimer

Management Executives (CHIME)
Ann Arbor, MI, USA
Woods College of Advancing Studies
Boston College
Birmingham, AL, USA
Acknowledgments

With the rapid growth of information technology, a global pandemic, and

discoveries in healthcare, it was clearly time for the fifth edition of Healthcare
Information Management Systems: Cases, Strategies, and Solutions. This
book showcases the theory to practice approach of information technology
transforming healthcare delivery and the people driving those transforma-
tions. In a time of so much change and challenge, this fifth edition could not
have been created without a myriad of people whom we acknowledge.
We thank the Springer staff, Grant Weston, Raagai Priya Chandrasekaran,
and Rakesh Kumar Jotheeswaran, who were invaluable in the organization
and production of the final product. We also thank the industry and domain
leaders who contributed their expertise and experience in the foreword and
chapters of this book. Their perspectives and personal stories bring to light
the incredible excitement of information technology in healthcare. The result
is a compendium of ideas and realities that are enhancing healthcare and its
delivery.
Joan M. Kiel
George R. Kim
Marion J. Ball

xi
Contents

Part I The Current State

1 Estimating the United States’ Cost of Healthcare

Information Technology������������������������������������������������������������������   3
Ross Koppel
2 
Innovating Payment Models for High-Value Healthcare ������������ 39
Christopher P. Tompkins and Stephen Bandeian
3 Leadership and Change������������������������������������������������������������������ 53
Patricia Hinton Walker, Bonnie Blueford, and John M. Walker
4 Promoting Informatics Workforce Development
Through Global Initiatives�������������������������������������������������������������� 65
Man Qing Liang, Trisha Pongco, and Toria Shaw Morawski
5 Preparing Clinicians and Patients for the
Future of Virtual Medicine and Telehealth������������������������������������ 81
Bridget C. Calhoun
6 Privacy and Security������������������������������������������������������������������������ 93
Darren Lacey
7 Interoperability: Current Considerations ������������������������������������ 113
Hans J. Buitendijk
8 Health Information Exchange�������������������������������������������������������� 133
David Horrocks, Lindsey Ferris, and Hadi Kharrazi

Part II Innovations and Trends

9 
Telemedicine: Its Past, Present and Future ���������������������������������� 149
Richard S. Bakalar
10 
The Telehealth Challenge During COVID-19 Emergency
Preparedness and Response������������������������������������������������������������ 161
Anne M. Hewitt and Joan M. Kiel
11 Information Technology and Operational Issues for
Emergency Preparedness and Response���������������������������������������� 167
Stephen L. Wagner

xiii
xiv Contents

12 Data
 Use in Public Health �������������������������������������������������������������� 181
Musa A. Kana, Ahmad Khanijahani, Ismail A. Raji, Abdu
Adamu, and Faina Linkov
13 Patient
 Safety and Health Information Technology���������������������� 201
Yushi Yang, Samantha Pitts, Allen Chen, Nicole Mollenkopf,
Taylor Woodroof, and Bridgette Thomas
14 Digital
 Health in Chronic Care and Self-Management���������������� 209
Malinda Peeples and Bhagyashree (Disha) Maity
15 Algorithmic
 Fairness and AI Justice in
Addressing Health Equity �������������������������������������������������������������� 223
Yoonyoung Park, Moninder Singh, Eileen Koski,
Daby M. Sow, Elisabeth Lee Scheufele, and Tiffani J. Bright
16 Managing
 Clinical Data in Neurocritical Care ���������������������������� 235
Peter H. Dziedzic and Jose I. Suarez
17 Data-Driven
 Disease Progression Modeling���������������������������������� 247
Kenney Ng, Mohamed Ghalwash, Prithwish Chakraborty,
Daby M. Sow, Akira Koseki, Hiroki Yanagisawa, and
Michiharu Kudo
18 Virtual
 Health in Patient Care and Clinical Research ���������������� 277
Tianna M. Umann, Molly McCarthy, Clifford Goldsmith,
Paul Slater, and Christopher Regan
19 Digital
 Health Solutions Transforming Long-Term
Care and Rehabilitation������������������������������������������������������������������ 301
Mohamed-Amine Choukou, XinXin (Katie) Zhu,
Shwetambara Malwade, Eshita Dhar, and Shabbir Syed Abdul
20 Learning
 Interprofessionally from a Real-Life
Simulation in a Smart Home���������������������������������������������������������� 317
Gabriela Mustata Wilson and Ruth E. Metzger
21 Predicting
 Preventive Care Service Usage in a
Direct Primary Care Setting Using Machine Learning���������������� 325
Sugato Bagchi, Ching-Hua Chen, George R. Kim,
Judy George, Thomas A. Gagliardi, Marion J. Ball,
Sasha E. Ballen, and Jane L. Snowdon

Part III Horizons

22 Healthcare
 Delivery in the Digital Age������������������������������������������ 341
M. Chris Gibbons, Yahya Shaihk, and Frances
Ayalasomayajula
23 I nformatics and Clinical Workforce
Competencies and Education���������������������������������������������������������� 355
William Hersh
Contents xv

24 
Emerging Need for a New Vision of Multi-Interprofessional
Training in Health Informatics������������������������������������������������������ 363
Gabriela Mustata Wilson, Patricia Hinton Walker,
and Marion J. Ball
25 Understanding Disparities in Healthcare:
Implications for Health Systems and AI Applications������������������ 375
Eileen Koski, Elisabeth Lee Scheufele, Hema Karunakaram,
Morgan A. Foreman, Winnie Felix, and Irene Dankwa-Mullan
26 Addressing Health Equity: Sources, Impact and
Mitigation of Biased Data���������������������������������������������������������������� 389
Eileen Koski, Fernando Suarez Saiz, Yoonyoung Park,
Brett R. South, Elisabeth Lee Scheufele,
and Irene Dankwa-Mullan
27 A Future Health Care Analytic System:
Part 1—What the Destination Looks Like������������������������������������ 401
Stephen Bandeian, Christopher P. Tompkins,
and Ashwini Davison
28 A Future Health Care Analytic System (Part 2):
What is Needed and ‘Getting It Done’������������������������������������������ 419
Stephen Bandeian, Christopher P. Tompkins,
and Ashwini Davison
29 
HIT, Informatics and Ethics ���������������������������������������������������������� 435
David L. Meyers
30 
Nurse Informaticists and the Coming
Transformation of the U.S. Healthcare System ���������������������������� 453
Mark Hagland
31 
The Future of Health Systems: Health Intelligence���������������������� 461
John S. Silva, Marion J. Ball, Mark Polyak,
and Gabriela Mustata Wilson
32 Health IT for the Future – It Isn’t (Just) About
the Technology���������������������������������������������������������������������������������� 471
Stephanie L. Reel and Steven F. Mandell

Index�������������������������������������������������������������������������������������������������������� 479
Contributors

Shabbir Syed Abdul Artificial Intelligence and Digital Health, Taipei

Medical University, Taipei, Taiwan
Abdu Adamu, MD, MSc, PhD Cochrane South Africa, South African
Medical Research Council, Cape Town, South Africa
Frances Ayalasomayajula, MPH, PMP Reach Thought Leadership, San
Diego, CA, USA
HP Inc., San Diego, CA, USA
Sugato Bagchi, PhD IBM T. J. Watson Research Center, Yorktown Heights,
NY, USA
Richard S. Bakalar, MD, FATA ViTel Net, McLean, VA, USA
Sasha E. Ballen, MS R-Health Inc, Elkins Park, PA, USA
Marion J. Ball, EdD Multi-Interprofessional Center for Health Informatics
(MICHI), The University of Texas at Arlington, Arlington, TX, USA
IBM T. J. Watson Research Center, Yorktown Heights, NY, USA
Johns Hopkins School of Medicine, Baltimore, MD, USA
University of Texas, Arlington, TX, USA
Stephen Bandeian, MD, JD Elevance Health, Bethesda, MD, USA
Bonnie Blueford, MA The Blueford Gorup, LLC, Coaching/Leadership
Development/Communications, Annapolis, MD, USA
Tiffani J. Bright, PhD, FACMI IBM Watson Health, Sandy Springs, GA,
USA
Hans J. Buitendijk, MSc, FHL7 Chester Springs, PA, USA
Bridget C. Calhoun, DrPH, PA-C John G. Rangos Sr. School of Health
Sciences, Duquesne University, Pittsburgh, PA, USA
Prithwish Chakraborty IBM Research, International Business Machines
Corporation, Yorktown Heights, NY, USA
Allen Chen, MD, PhD, MHS The Johns Hopkins University School of
Medicine, Baltimore, MD, USA

xvii
xviii Contributors

Ching-Hua Chen, RN, PhD IBM T. J. Watson Research Center, Yorktown

Heights, NY, USA
Mohamed-Amine Choukou Department of Occupational Therapy, College
of Rehabilitation Sciences, University of Manitoba, Winnipeg, MB, Canada
Irene Dankwa-Mullan, MD, MPH Merative (formerly IBM Watson
Health), Bethesda, MD, USA
Ashwini Davison Johns Hopkins University, Baltimore, MD, USA
Eshita Dhar Graduate Institute of Biomedical Informatics, Taipei Medical
University, Taipei, Taiwan
Peter H. Dziedzic, MS Division of Neurosciences Critical Care, Precision
Medicine Center of Excellence for Neurocritical Care, Center of mHealth
and Innovations, Department of Neurology, The Johns Hopkins University
School of Medicine, Baltimore, MD, USA
Winnie Felix, MS, MPH Merative (formerly IBM Watson Health),
Bethesda, MD, USA
Lindsey Ferris, DrPH, MPH Chesapeake Region Information System for
Our Patients (CRISP), Columbia, MD, USA
Morgan A. Foreman, BS IBM T. J. Watson Research Center, Yorktown
Heights, NY, USA
Thomas A. Gagliardi, BS IBM Watson Health, Cambridge, MA, USA
Judy George, PhD IBM Watson Health, Cambridge, MA, USA
Mohamed Ghalwash IBM Research, International Business Machines
Corporation, Yorktown Heights, NY, USA
M. Chris Gibbons, MD, MPH The Johns Hopkins School of Medicine,
Baltimore, MD, USA
The Greystone Group, Inc., Greenbelt, MD, USA
Reach Thought Leadership, San Diego, CA, USA
Clifford Goldsmith Microsoft Corporation, Health and Life Sciences,
Redmond, WA, USA
Mark Hagland Healthcare Innovation, Chicago, IL, USA
William Hersh Oregon Health & Science University, Portland, OR, USA
Anne M. Hewitt, PhD, MA Seton Hall University, Nutley, NJ, USA
David Horrocks, MBA, MPH New York eHealth Collaborative, New York,
NY, USA
Musa A. Kana, MD, MPH, PhD National Institute of Environmental
Health Sciences, Epidemiology Branch, Durham, NC, USA
Hema Karunakaram, MPH IBM Watson Health, Chicago, IL, USA
Contributors xix

Ahmad Khanijahani, PhD, CPH, CHDA John G. Rangos, Sr. School of

Health Sciences, Duquesne University, Pittsburgh, PA, USA
Hadi Kharrazi, MD, PhD, FAMIA Johns Hopkins School of Public Health,
Baltimore, MD, USA
Joan M. Kiel, PhD, CHPS Duquesne University, Pittsburgh, PA, USA
George R. Kim, MD, FAAP Johns Hopkins School of Medicine, Baltimore,
MD, USA
Ross Koppel, PhD, FACMI Department of Sociology, School of Arts and
Sciences, University of Pennsylvania, Philadelphia, PA, USA
Department of Biomedical Informatics, Perelman School of Medicine,
University of Pennsylvania, Philadelphia, PA, USA
Department of Biomedical Informatics, Jacobs School of Medicine,
University at Buffalo, Buffalo, NY, USA
Akira Koseki IBM Research, International Business Machines Corporation,
Tokyo, Japan
Eileen Koski, MPhil, FAMIA IBM T. J. Watson Research Center, Yorktown
Heights, NY, USA
Michiharu Kudo IBM Research, International Business Machines
Corporation, Tokyo, Japan
Darren Lacey, JD Johns Hopkins University and Johns Hopkins Medicine,
Baltimore, MD, USA
Man Qing Liang, PharmD, MSc Student TIGER Initiative, HIMSS,
Montreal, Canada
Faina Linkov, MPH, PhD John G. Rangos, Sr. School of Health Sciences,
Duquesne University, Pittsburgh, PA, USA
Bhagyashree (Disha) Maity, MS, BSN, RN Clinical Informatics Lead, San
Francisco, CA, USA
Shwetambara Malwade Taipei Medical University, ICHIT, Taipei, Taiwan
Steven F. Mandell Division of General Internal Medicine, Biomedical
Informatics and Data Science, The Johns Hopkins University, Baltimore,
MD, USA
Healthcare Information Systems, Johns Hopkins Hospital, Baltimore, MD,
USA
Molly McCarthy, MBA, RN-BC Microsoft Corporation, US Health and
Life Sciences, Redmond, WA, USA
Ruth E. Metzger, PhD, MBA, BA, BSN Indiana State Department of
Health, Division of Long-Term Care, Evansville, IN, USA
David L. Meyers, MD, MBE, FACEP Adjunct Faculty, Johns Hopkins
University, Baltimore, MD, USA
xx Contributors

Emergency Medicine, Sinai Hospital of Baltimore, Baltimore, MD, USA

Nicole Mollenkopf, PharmD, MBA The Johns Hopkins University School
of Medicine, Baltimore, MD, USA
Toria Shaw Morawski, MSW Professional Development, HIMSS,
Schaumburg, IL, USA
Kenney Ng IBM Research, International Business Machines Corporation,
Cambridge, MA, USA
Yoonyoung Park, ScD IBM T. J. Watson Research Center, Cambridge, MA,
USA
Malinda Peeples, MS, RN, CDCES, FADCES Clinical Services and
Research, Welldoc, Inc, Columbia, MD, USA
Department of Medicine, Division of General Internal Medicine, Johns
Hopkins Medicine, Baltimore, MD, USA
Samantha Pitts, MD, MPH The Johns Hopkins University School of
Medicine, Baltimore, MD, USA
Mark Polyak Ipsos, Washington, DC, USA
Trisha Pongco, CAHIMS Clinical Informatics, HIMSS, Brooklyn, NY,
USA
Ismail A. Raji, MD, MSc, MPH Department of Community Medicine,
Usmanu Danfodiyo University Teaching Hospital, Sokoto, Nigeria
Stephanie L. Reel Johns Hopkins Medicine, Johns Hopkins Health System,
Johns Hopkins University, Baltimore, MD, USA
Division of General Internal Medicine, Biomedical Informatics and Data
Science, The Johns Hopkins University, Baltimore, MD, USA
Christopher Regan Microsoft Corporation, Office of the CTO, Health and
Life Sciences, Redmond, WA, USA
Fernando Suarez Saiz, MD IBM Watson Health, Toronto, ON, Canada
Elisabeth Lee Scheufele, MD, MS IBM Watson Health, Boston, MA, USA
Yahya Shaihk, MD, MPH The MITRE Corporation, McLean, VA, USA
John S. Silva, MD, FACMI Silva Consulting Services, Phoenix, MD, USA
Moninder Singh, PhD IBM T. J. Watson Research Center, Yorktown
Heights, NY, USA
Paul Slater BillionMinds, Woodinville, WA, USA
Jane L. Snowdon, PhD IBM Watson Health, Cambridge, MA, USA
Brett R. South, MS, PhD, FAMIA IBM Watson Health, Salt Lake City, UT,
USA
Contributors xxi

Daby M. Sow IBM Research, International Business Machines Corporation,

Yorktown Heights, NY, USA
IBM T. J. Watson Research Center, Yorktown Heights, NY, USA
Timothy Stettheimer, PhD, CHCIO, FACHE The College of Healthcare
Information Management Executives (CHIME), Ann Arbor, MI, USA
Woods College of Advancing Studies, Boston College, Birmingham, AL,
USA
Jose I. Suarez, MD, FNCS, FANA Division of Neurosciences Critical Care,
Precision Medicine Center of Excellence for Neurocritical Care, Departments
of Anesthesiology and Critical Care Medicine, Neurology, and Neurosurgery,
The Johns Hopkins University School of Medicine, Baltimore, MD, USA
Bridgette Thomas, PharmD Johns Hopkins Home Care Group, Baltimore,
MD, USA
Christopher P. Tompkins, PhD Brandeis University, Waltham, MA, USA
Tianna M. Umann Microsoft Corporation, Office of the CTO, Technology
Strategy, New York, NY, USA
Stephen L. Wagner, PhD, FACHE, LFACMPE Seton Hall University-
Interprofessional Health Sciences Campus, Nutley, NJ, USA
John M. Walker, MA, IT Professional, Canton, GA, USA
Patricia Hinton Walker, PhD, RN, FAAN, MCC CoachingSteppingStones.
com, Canton, GA, USA
Canton, GA, USA
Gabriela Mustata Wilson, PhD, MSc, FHIMSS, SNAI Multi-
Interprofessional Center for Health Informatics (MICHI), The University of
Texas at Arlington, Arlington, TX, USA
Taylor Woodroof, PharmD, MHIIM Johns Hopkins Home Care Group,
Baltimore, MD, USA
Hiroki Yanagisawa IBM Research, International Business Machines
Corporation, Tokyo, Japan
Yushi Yang, PhD Armstrong Institute for Patient Safety and Quality, The
Johns Hopkins Health System, Baltimore, MD, USA
XinXin (Katie) Zhu Center for Biomedical Data Science, School of
Medicine, Yale University, New Haven, CT, USA
 Part I
The Current State

Two decades of federal incentives have moved certified electronic health

record technology (CEHRT) into all sectors of US health care and have estab-
lished the centrality of electronic data in the daily work of clinical medicine.
The mutual assimilation of information technology (HIT) and US healthcare
infrastructures has led to significant changes to how healthcare gets done:

• The availability of up-to-date electronic clinical and administrative data

for individuals and populations is providing data infrastructures for main-
taining institutional awareness of health and healthcare metrics for timely
response, improvement, and prevention.
• The ability to measure outcomes from aggregate data is providing new
opportunities to link health quality and remuneration, moving from fee-
for-service to value-based payment models based on provider and institu-
tional performance.
• With these affordances have arisen new needs and responsibilities to
assure the confidentiality, integrity and availability of healthcare data,
infrastructures, and functionalities. This in turn has led to the need for new
expertise, resources, and expenditures to protect the content, privacy, and
security.

Topics covered in this Section include:

• A consideration and estimation of ongoing annual costs of electronic

health record (EHR) and other healthcare information technologies (HIT)
in the United States by Ross Koppel
• An overview of the evolution of models for healthcare payment/remunera-
tion with respect to value and quality by Chris Tompkins and Steve
Bandeian
• A discussion on the ongoing and changing needs of healthcare leadership
in the era of electronic health records by Patricia Hinton Walker and
colleagues
• Considerations for developing the healthcare informatics workforce on a
global scale by Man Qing Liang and colleagues
2 The Current State

• An exploration of the new needs of clinicians and patients in the era of

virtual and remote healthcare by Bridget Calhoun
• Expositions on developments in

–– Healthcare privacy and security by Darren Lacey

–– Interoperability by Hans Buitendijk
–– Healthcare Information Exchange by David Horrocks, Lindsey Ferris
and Hadi Kharrazi
 Estimating the United States’ Cost
of Healthcare Information 1
Technology

Ross Koppel

Abstract Keywords

Current US healthcare involves extensive use Electronic Health Records: classification

of shared electronic health records (EHRs) Economics · Statistics and numerical data
and other data from health information tech- Trends · Clinical pharmacy information
nologies for clinical care, data collection, bill- systems · Classification · Economics
ing and regulatory reporting. Business and Statistics and numerical data · Telemedicine
regulatory processes also are dependent on Economics · Organization and administration
EHRs and a cornucopia of other medical ser- Nurses · Physicians · Community health
vices, devices and platforms (wearables, home Economics · Medical care
health monitors, medical imaging software,
etc.). The healthcare world has created an
enormous medical information infrastructure Learning Objectives
that itself has ongoing operating, maintenance On completing this chapter, the reader shall be
and updating costs. This chapter is a first able to:
attempt to estimate the scope and magnitude
of those software costs, and an invitation for • Describe/Discuss the impact and annual costs
others to join in the discussion. of electronic health records and other clinical
software in US healthcare
• Enumerate and describe considerations
Supplementary Information The online version con-
tains supplementary material available at [https://doi.
involved in estimating HIT costs for different
org/10.1007/978-­3-­031-­07912-­2_1]. clinical functions and specialties
• Appreciate/articulate the magnitude of ongo-
R. Koppel (*) ing costs to US healthcare
Department of Sociology, School of Arts and
Sciences, University of Pennsylvania,
Philadelphia, PA, USA
Department of Biomedical Informatics, Perelman
School of Medicine, University of Pennsylvania,
Philadelphia, PA, USA
Department of Biomedical Informatics, Jacobs
School of Medicine, University at Buffalo,
Buffalo, NY, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2022 3

J. M. Kiel et al. (eds.), Healthcare Information Management Systems, Health Informatics,
https://doi.org/10.1007/978-3-031-07912-2_1
4 R. Koppel

1.1 Introduction 2009) Act to spur EHR adoption and use.

However, these funds were only seed money to
In recent years, researchers, clinicians, journal- incentivize hospitals and physicians to purchase,
ists and even supporters of healthcare technology implement and use EHRs. This effort was stun-
have renewed questions about the effectiveness ningly successful, resulting in extraordinary
and burdens of electronic health records (EHRs) increases in HIT sales and implementation, with
and other healthcare information applications. hospital EHR adoption growing from about 9%
Although EHRs have been cited as a major rea- in 2009 to 97% in 2021, and with an equally
son for clinician rage, burnout and early retire- remarkable uptake of EHRs and related technolo-
ment, we do not address these collateral costs/ gies by doctors’ offices.1
damages of healthcare IT (HIT) herein, but the
topic is well documented [1–11].
1.1.3 “Carrot and Stick”

1.1.1 The Myriad Benefits of HIT The “carrot” of federal incentives and subsidies
for EHR adoption was reinforced by a “stick” of
A full analysis of the cost of HIT must reflect the regulations for non-participation in the form of
many dramatic efficiencies and advantages HIT Medicare and Medicaid reimbursement deduc-
affords. These benefits go far beyond clinical ser- tions for non-adopters. As Medicare and Medicaid
vices, business operations and scheduling—from cover about 42% and 16% of all US patients
speeding information to any and all clinicians, to respectively, such deductions would be de facto
supporting clinical decisions, improving legibil- bankruptcy for most medical providers and hos-
ity, error-checking and tracking. In addition, HIT pitals. And in fact, inability to face EHR imple-
has become an essential and regulatory require- mentation and use (in other words, to accept
ment for billing and reimbursement by govern- financially infeasible Medicare/Medicaid reduc-
ment (e.g., Medicare, Medicaid and CMS) and tions), caused many medical practices, usually
insurance companies. older and smaller ones, to close.
Our aim here, however, is more modest than
enumerating HIT’s many benefits. Our focus here
is on the direct financial costs of US healthcare’s 1.1.4 The “Real” Costs
software; specifically, on the ongoing costs of
buying, implementing and maintaining the soft- Software, and its implementation, while very
ware on which US clinical care has become expensive, is a small part of total US healthcare
dependent: e.g., EHRs, dental IT, pharmacy IT, costs. As we document below, it represents about
visiting nurses IT, laboratory IT, telemedicine, 7–8% of the $4 trillion the US spends on
and so on. In examining the published literature, healthcare.
we find that few have examined the aggregate Significant parts of healthcare system IT
cost of healthcare software for the range of medi- costs are supported by taxpayers through
cal services described herein. Medicare and Medicaid, the National Library
of Medicine (NLM), the Office of the National

1.1.2 A Frequent Misunderstanding 1

If instead of asking about the ROI to the government’s
$36 billion in seed money, one might ask about vendors’
Many articles on the cost of healthcare IT (HIT) return on convincing the government to enact the HITECH
[4] focus on the $36 billion in incentives, part of legislation. That ROI is almost beyond calculation.
Certainly, the billions of dollars returned for every dollar
the American Recovery and Reinvestment Act spent convincing the government to require hospitals and
(ARRA) and of the HITECH (Health Information clinicians to purchase the technology was the best invest-
Technology for Economic and Clinical Health of ment ever made.
1 Estimating the United States’ Cost of Healthcare Information Technology 5

Coordinator of Health Care IT (ONC), in addi- (users, patients) and tiers (e.g., stepwise
tion to other HHS and NSF programs. We esti- increases for numbers of users: 1–10,
mate the actual amount for HIT software and 11–100, etc.).
implementation to be almost just over 300 bil- (b) Add-on services are additional costs for
lion dollars (almost 9 times the original $36 bil- specific functionalities, e.g., scheduling
lion in seed money). These estimates reflect and patient reminders, SMS texting to
costs of maintenance, customization, modifica- patients, plus set up and training fees--
tions and connection of EHRs to existing/leg- some at no extra cost, others at thousands
acy healthcare technology (e.g., patient or tens of thousands of dollars.
monitoring devices and displays). They also 3. However, we do not include in these cost esti-
reflect a very small proportion of the costs of mates the frequently discussed negative costs
training, consultants, additional IT staff and of EHRs, which reflect time-intensive data
associated access utilities and devices. entry tasks [12] often associated with lost pro-
ductivity, inefficiencies, burnout and addi-
tional training [12].
1.1.5 Our Scope and Plan Nor do we include:
(a) Losses associated with legal efforts
In this work, we first enumerate the costs of related to HIT-linked problems, a focus of
healthcare IT (HIT) that we include in this exam- work by the Health Law Group of the
ination and estimation. We focus only on the cost American Bar Association [13].
of the software and its implementation. We (b) Losses due to ransomware and data
exclude all hardware costs. We also list the users breaches (fines, recovery, reputational
of these software products. We lay out the general loss, and remediation).
method by which we obtained estimates and cal-
culated costs. This is followed by a domain-by-­
domain accounting of our findings (with source 1.3 Who Buys and Uses HIT?
citations and rationales) with notations on the
special considerations for each domain in annual 1.3.1 Inclusions
cost estimations. What follows is a tally of these
costs as a “first” estimate of annual US health IT Government Systems
costs. Our brief conclusion is an invitation for Ambulance services/EMS (if not part of a fire
ongoing dialogue and discourse on the impact department system)
and ongoing costs of healthcare information Dept of Defense health care software
technology in US healthcare. Indian Health system
Prison health system software
State and County Health Systems IT
1.2 Our Task Dept of Veterans Affairs health system (VHA)

In this examination of HIT costs, we endeavor to: Home Care, Long Term Care, and Elderly
1. Enumerate and document the clinical services Care
that purchase and employ HIT software. Adult Day Care
2. Outline and estimate costs of the healthcare Home Health Care
software and its implementation in US medi- Hospices
cal facilities and offices, including setup and Remote Patient Monitoring Systems—if separate
maintenance. software and costs from hospital or medical
(a) System pricing varies, with some ven- practice systems
dors charging periodic (yearly, monthly) Skilled Nursing Facilities
fees, and others charging by person Visiting Nurses
6 R. Koppel

Clinical and Other Services (Usually Not in mats, additional information). Another is the need
Hospitals) to accommodate systems for unusual local condi-
Acupuncture tions (such as climate). An example of the latter is
Chiropractic a southern US hospital system with one mountain-
Clinical trial software based facility with a need for not previously loaded
Dental Care protocols for frostbite. Such variances create addi-
Dialysis centers (free standing) tional costs for healthcare systems.
Medical Offices/PCPs
Occupational Therapy (if not included in general
hospital software) 1.3.2 Exclusions
Optometry
Pharmacies (chain and independent) We exclude many types of HIT, some because of
Pharmacy Benefits Mangers’ software uncertainty in the accuracy of estimating soft-
Physical Therapy ware or implementation costs, and some because
Podiatry of concerns about double counting. Undoubtedly,
Telehealth Systems/Virtual Medical Visits-- their absence will result in noteworthy under-
Outpatient counting, but we would rather err on the side of
under- rather than over-estimation. These exclu-
Hospitals and Usually Linked Services sions are:
Hospitals
ICU telemetry and monitoring within the hospital • Medical billing software, including revenue
system. cycle software
Medical Imaging (X-rays, MRIs, etc.) • Data warehouse software
Medical Laboratory Information Systems • Blood bank IT software
PACS (Picture Archiving and Communications • Psychologists’ software and all behavioral
Systems) software if not included health software, including all mental health
in the total facility costs facilities (hospitals and clinics)
Medical social services • Software for most social workers
• Pharmaceutical manufacturers’ or distribu-
Also included is cybersecurity software insur- tors’ software
ance premiums (excluding hardware) • Artificial Intelligence (AI) clinical software,
Our list of users, although extensive and probably e.g., used to predict sepsis, treatment proto-
unique, obscures the complexity of deriving cost cols, discharge, bed use, etc. These exclusions
estimates. For example, EHR costs may be shared included the vast new AI firms and programs
across hospital chains, groups and facilities. Each linked to medical facilities and services
implementation, even in the same medical system, • AI for robotic devices, such as robotic surgery
however, may demand separate efforts because of or brain-to-speech functions, etc.
different legacy devices and software, staff experi- • Data sharing, e.g., Health Information
ence with previous software, patient populations, Exchanges (HIEs), direct exchange, query-
state regulations, and availability of ancillary ser- based exchange, etc.
vices (e.g., nearby labs or pharmacies), etc. • Connectivity costs for telehealth IT
As facilities (such as private medical practices) • Extra time (pajama time) of clinicians enter-
merge with larger or more prosperous entities (such ing data after workhours
as hospital groups) [14] , there may be shifting and/ • Implementation costs for software that is
or modification of existing EHR systems. One shared with other systems
example of this is the need of national pharmacy • Pharmaco-surveillance software
information systems to accommodate varying state • Software for the increasingly popular mail
regulations on medication labeling (font size, for- order pharmacies
1 Estimating the United States’ Cost of Healthcare Information Technology 7

• Public or other health clinics—including AIDs for 302 price quotes from HIT vendors. We estab-
centers, etc. lished a separate website, email, and phone num-
• All rehabilitation facilities (separate EHRs, ber to avoid overwhelming our personal and
not part of regular hospital systems) university email and other systems.
• Urgent care centers’ software In addition to “direct” efforts, we reviewed:
• Nurse Practitioners (NPs) using telemedicine
software who work in hospitals and NPs who • The peer-reviewed literature that addresses
work in medical offices but are not in primary software costs and EHRs, with articles
care. addressing implementation costs (e.g., train-
• Costs to the VHA for annual software licens- ing, consultants, retrofitting devices, IT
ing and maintenance. (Excluded because of costs). These included JAMIA, ACI, JAMA,
the long implementation time) NEJM, etc.
• The value of thousands of VHA clinician • The literature on healthcare finance, cyberse-
hours assisting the VHA implementation curity costs, and information systems
effort at the VHA facilities. Neither do we • Regulatory agency websites and publications
include the $4.9 billion the VHA is spending for software and implementation costs, includ-
to maintain VistA during the Cerner roll-out ing DHHS’ Office of the National Coordinator
• Physicians who left the profession because of for Health Information Technology (ONC)
the requirements to use and bill via EHRs and DHHS Information Technology Agency
• The lost productivity and the implementation Summary for software expenses [15]
costs for many of the services enumerated • Relevant professional and trade publications,
below. (We only include such costs when we e.g., American Medical Association, American
have hard numbers for those functions for each Dental Association, Visiting Nurse
specific type of software implementation Associations of America, HIMSS, CHIME
• User fees for physicians at SNFs and AMDIS
• Budgets in city, county, state, and federal
Adding these to our totals might well double reports on healthcare software costs. For
our total estimated costs. But we do not do that example, the Indian Health Service and the
for the reasons articulated above. Veterans Administration are obliged to publish
budgets for congress. Many publicly funded
hospitals and services are also obliged to sub-
1.4 Cost Estimates: Method mit budget documents. Also, most public agen-
cies must disclose their budgets in state and
1.4.1 Information Sources county databases and reports [16, 17]
• Scholarly literature and reports on these
We have enumerated many clinical services that issues, e.g., from AMIA, IMIA, European and
use healthcare software (“Who Buys and Uses other Informatics Associations
HIT? Inclusions”) and sought to contact all • Budget analyses for software costs by indus-
known vendors in all relevant service fields (e.g., try, from data marketing firms [18], published
software for dental practices, visiting nurse asso- comparisons of software spending by indus-
ciations, hospices, chiropractors, EHRs for medi- try. We note also that healthcare software costs
cal enterprises, medical imaging centers, are expected to increase significantly as “soft-
ambulatory clinics, nursing homes, etc.). We col- ware as a service” (SaaS) becomes more pop-
lected information via direct phone calls, website ular vs on-premises software
request forms, contract forms, and email requests • Associations of medical CFOs, CTOs, CIOs
to vendors and others. We often were obliged to and others
view demonstration videos and talks before we • Professional organizations focused on health-
could seek price information. We also signed up care institutions’ HIT, e.g., CHIME and
8 R. Koppel

AMDIS—often via personal contacts and of forthrightness/knowledge of interviewees, bias

reports or documents (intentional or unintentional) in presenting costs
• Hospitals and other reports on costs of soft- or functionality, etc. In addition, there may be
ware--both reports to CMS and other errors in assumptions or calculations. In all cases,
agencies however, we have sought to make those calcula-
• Webinars from healthcare staff and vendors- tions and assumptions transparent. We invite
-often sharing their improvements to efficien- those with more recent or detailed data to access
cies, cost-effectiveness, staff satisfaction, etc. the following website: universityhitsoftware@
[19] socialresearchcorp.com and share additional
• Additional clarifications in response to our sources or more precise data. We also urge others
questions, e.g., in one case the software con- to continue this initial effort and to advance the
tract disclosures (revealed via FOIA from understanding of HIT’s costs.
known disputes involving publicly funded
hospitals and from publicly available pre-trial
disclosures) 1.5 Software Cost Estimates:
• Many requests to listservs of many of the Findings
above associations and organizations.
1.5.1 Introductory Notes

1.4.2 Responses from Vendors These findings exclude all hardware costs. In all
and Limitations cases, we distinguish estimates for one-time costs
(implementation/customizations) from yearly
Many vendors responded to direct and indirect costs (e.g. monthly or annual licenses). For the
requests for information. Almost one-third (32%) one-time costs we amortize the expense over at
of vendors for many service areas were forth- least 5 years to provide an annual cost estimate
coming about prices, especially after we probed for each of the thirty-two categories listed below.
repeatedly for additional costs, such as fees for
set-ups, data migration, patient reminder func-
tions, user training (in-person and remote, the lat- 1.5.2 Government Systems
ter because of COVID-19), additional billing
functions (if not included), revenue and person- 1.5.2.1 Emergency Medical Services
nel management, linkages to CMS or insurance (not part of Fire Departments)
companies (if not included), etc. Often, we re-­ There are 21,283 separate EMS units in the
contacted them for clarification on price charts or US. However, 40% are linked to fire departments
fee schedules. as part of many communities’ protective services.
These estimates represent the most encom- We therefore remove them from inclusion in cal-
passing effort of which we are aware. We realize, culating HIT software costs, with the resulting
of course, that they are undoubtedly subject to number equaling 12,770. These remaining EMS
error. Many estimates are based on available data, departments are funded by over 34 different
augmented by interviews with providers, admin- agencies and sources—including state funds,
istrators, vendors, government reports (state and medical systems, local agencies, and a very wide
federal), industry research, scholarly reports, and range of federal funds, including federal emer-
financial documents (e.g., reports from hospitals gency preparedness funds and the US Department
and government agencies, etc.). While we always of Homeland Security [20].
sought to triangulate data, estimates and sources, Based on review of EMS EHR vendor data
these are subject to limitations: censored data and interviews with both vendors and local offi-
(e.g., truncated on one side or the other), under- cials [21–23], the implementation costs for these
and over-statements of costs or efficiencies, lack EMS units (i.e., the 60% of EMS units not linked
1 Estimating the United States’ Cost of Healthcare Information Technology 9

to fire departments) is approximately $5,342 per $2,500,000 and $831,000 respectively [28]. We
system, which for 12,770 systems equals $68.216 do not include costs of implementation, licens-
million as a one-time cost. Amortized over 5 ing, maintenance, patches, or repairs.
years, this equals $13.653 million. The total of these is $37,831,000.
EMS HIT software license fees are based on Based on budget reports, we add four percent
the number of ambulances in each unit. There are for management, operations and quality and
52,000 ambulances in the US (not counting the oversight budgets (4% of $9,898,000)
military, tribal systems and scores of other agen- [28] = $395,920.
cies). With a mean license fee of $57/month/per Estimated combined Indian Health Service
ambulance = $654/yr per ambulance. This totals EHR/HIT costs: $38,226,920
to $34,008,000 million for the US. Combining
the amortized implementation costs and the 1.5.2.4 Prison and Jail EHR Software
license fees equals $47,661,000. Persons in prisons (state and federal institutions) are
medically vulnerable populations, with rates of
1.5.2.2 Department of Defense: HIV, drug dependence, hepatitis, and mental illness
Cerner-Leidos-Accenture EHR far exceeding those of the general population. Sixty-
Implementation three percent receive medical care and 53% are tak-
Originally, the cost of the US Department of ing prescription medications. For persons in jails
Defense project to replace its existing EHR and (local short-term holding facilities), 45% receive
medical information system (the Composite medical care and 39% take prescription medica-
Health Care System/Armed Forces Health tions [29–31].,, In addition, there are required intake,
Longitudinal Technology Application (CHCS/ yearly, and exit medical inspections.
AHLTA) [24] ) with a commercial solution, was Federal law dictates that all care provided in
$4 billion, which subsequently increased to $5.2 prisons and jails must be medico-legally docu-
billion by 2018 with the proviso then that the cost mented and (sometimes billed) via software that
would increase [25]. This cost was recently is similar to a regular EHR (some with less func-
increased to account for issues with interopera- tionality than those of hospitals). Licensing costs
bility with the VHA system and with the expan- are usually by prisoner (or bed), with fees ranging
sion and implementation of the scheduled 23 from $15 per prisoner per month to $90 per pris-
“waves,” each targeting specific regions over the oner per month [32]. It should be remembered
next few years [26, 27]., that the number of comorbidities and the severity
We estimate the cost at $5.6 billion, and we do of their illnesses often involved hospitalization
not add costs for the thousands of military and (within the prison). We use a mean of $30/mo per
civilian personnel assisting in this effort. prisoner (Annual = $360). Moreover, because of
Amortized over 7 years because of delays caused the transient population in jails (vs. prisons) we
by COVID and other factors, the estimated cost is reduce the population by 40%, from 2.3 million to
$5.6 billion/7 or $800,000,000. 1.38 million. We do not include any cost for
The annual maintenance contract is estimated implementation, training, lost productivity or set
at $850 million. Thus, the combined implementa- up fees, although these are considerable.
tion and maintenance costs are estimated at: Multiplying only the reduced population numbers
$1,650,000,000 by the annual software license fee = $496,800,000

1.5.2.3 US Indian Health Service (IHS) 1.5.2.5 State and County Health
In a detailed public report to Congress, the 2021 Departments
enacted budget for the Indian Health Service’s Collectively, state healthcare agencies spend over
(IHS) Electronic Health Record System, its $1.5 trillion on direct expenses. Separately, county
Electronic Dental Health Record and its healthcare agencies spend over $1.7 trillion on
TeleBehavioral Health Center are $34,500,000, direct expenses. Federal budget reports and the
10 R. Koppel

healthcare literature reveals that healthcare (not over 5 years, but rather over 7 years. Thus, our
healthcare IT) represents 9.4% of those totals estimate for the cost of the VHA Cerner imple-
[33]. Estimated IT expenses for state and county mentation and software cost at $26.5 bil-
health departments is 2.9%, of those costs [34]. lion/7 = $3,785,700,000 [46–48].
Thus, the collective state health depart-
ments’ software costs [35] are estimated at
$98,500,000. 1.5.3 Home Care, Long Term Care,
The collective healthcare software costs [36] and Elderly Care
for 3006 counties in the US, based on data from
the National Association of Counties, totals 1.5.3.1 Adult Day Care
$1,856,000,000. According to the US Centers for Disease Control
Together, these equal $2,398,500,000. and Prevention (CDC) [49] there are 4,600 adult
daycare centers that each serve between 2–530
1.5.2.6 The Veterans Health adults (a mean of 66 adult patients per center).
Administration (VHA) Cerner The total number of FTE staff for the centers
EHR Implementation (users) is 19,900 [50].
Originally, the Veterans Health Administration Interviews with vendors and their literature
(VHA) project to convert the VHA hospitals’ reveal that the mean cost for software licenses in
EHR from the Veterans Health Information adult day care (ADC) is $1,596.00 per staff user
Systems and Technology Architecture (VistA) to per year. For 19,900 FTE US staff users, this is a
a Cerner implementation was listed as a $16 bil- total cost of $31.76 million per year [51].
lion effort [37], but quickly increased to $20 bil- Most vendors in this sector appear to include
lion [38]. Recent reports, and the several Inspector set up and implementation costs as part of the
General and congressional reports now put the package (i.e., no additional cost for this func-
figure at $26.5 billion [39–44]. tion). Estimates for lost productivity (mean = 3
Many interviews with VHA clinicians cite weeks) and implementation (mean = 1 week) add
this $26.5 billion as a significant underesti- $3,500.00 per center (not per user), which, for
mate because many thousands of clinician 4,600 centers, yields a one-time total of $16.1
person-­hours are being assigned to work with million, which amortized over 5 years is $3.222
the implementation effort at the many target million.
facilities, diverting direct patient care to rates The estimated US cost for ADC software is
lower than before the implementation process $31.76 million plus $3.22 million, which totals to
began. However, we do not have an accurate $34,980,000.
metric for this loss of service and extra
expense, hence we do not include it in our esti- 1.5.3.2 Home Health Care (HHC)
mate. Neither do we include the $4.9 billion According to the US Bureau of Labor Statistics
the VHA is spending to maintain VistA during (BLS), there are 3.5 million US home healthcare
the Cerner roll-out [45]. workers (HHW) whose work is organized, sched-
Licensing/maintenance/repair costs, ordinar- uled, billed and paid for via 12,200 agencies and
ily assessed on the software portion of the effort, services [52, 53].
would be only a tiny fraction of the implementa- Costs per HHW: Based on CDC data, inter-
tion cost, and because of the long implementation views with providers and vendors, and on indus-
time, we assess no cost to those usual fees. try reports, we estimate software costs to average
Because the Cerner implementation is occur- ~$40 per month per HHW [54]. Thus, an initial
ring in many VHA hospitals and clinics, and estimate of home health care software costs in the
because of the repeated delays and ordered US is: $40 per month per HHW × 12 months per
“stand-downs,” we have not amortized the cost year = 480 per HHW.
1 Estimating the United States’ Cost of Healthcare Information Technology 11

Therefore, the initial estimate of $480 per tals, but similar to SNFs, hospices are generally
HHW per year × 3.5 million HHW in the US = (65.2%) for-profit institutions [58, 60].
$1.680 billion before adjustments (below) For Medicare and Medicaid funding and to
To put this into perspective, total US annual bill insurance carriers, hospices need digital sys-
spending for home health care, in 2021 dollars, is tems similar to EHRs, often, however, with fewer
$17.985 billion [55] features. The large vendors are Epic (the largest),
However, we then reduce the cost to reflect the Homecare, Homebase, Brightree, Wellsky, and
reality that the number of users should be reduced Netsmart. As the US population ages, the “silver
by 25% [56] (estimated) due to the high turnover tsunami” is causing a growth in the number of
rates of HHWs--with the expectation that soft- hospice patients and in the number of hospice
ware services do not need to be paid for those not beds, with increases in the cost of hospitaliza-
working for a month. Thus the $1.680 bil- tions and acceptance of end-of-life care.
lion × 0.75 (reduction) = $1.260 billion Hospice HIT Cost: Using similar metrics for
other long-term care facilities (see SNFs, below),
• Addition of one-time costs (per agency) for albeit with significantly reduced fees and denom-
set up (average $1,117.5), training (average inators, we reduce implementation costs, produc-
$140.00 which is included in some packages), tivity loss, et cetera by 20% of the SNF’s costs
implementation (average $310) and lost pro- discussed below ($24,000, compared to the SNFs
ductivity during implementation and learning at $30,000); and we use a per bed cost of $10 per
(based on a mean of 3 weeks, we estimate at a bed per month ($120 per bed per year) with no
lower end of range of costs: $12,000 per additions for laboratory linkages, pharmacy con-
agency). nections, etc.
For 12,200 US home health agencies, this To determine the number of facilities, we use
one-time cost: the mean of the two hospice estimates (4,515 and
($1,117.50 + $140.00 + $310.00 + 6,800/2 = 5,672) and we take only 70% of that
$12,000.00) per agency = $13,567.50, number (= 3,970) because we wish to avoid the
amortized over 5 years is $2,713.50 per possibility of double counting any home care
agency, which for 12,200 agencies yields services.
$33.105 million estimated US home health For implementation costs (training, set up and
care initial costs [57]. lost productivity), we use both a reduced number
• Adding these adjustments ($33.105 million + of facilities (70% = 3970) and a reduced cost of
$1.260 billion) yields an adjusted annual cost $24,000 per facility. This equals a total of
of for home healthcare software of $95,280,000, which amortized over 5 years is
$1,293,105,000. $19,560,000 per year. The bed fee is $120 per
bed per year times 1,162,500 = $139,500,000.
The combined total is $159,060,000 [61]
1.5.3.3 Hospices
There are between 4,515 and 6,800 hospices in 1.5.3.4 Remote Patient Monitoring
the USA. The numbers differ by sources and by (RPM)
definitions of in-home vs institutional [58, 59] Remote Patient Monitoring (RPM) is experienc-
with CDS data indicating the workforce is 48% ing exponential growth. Increasing availability
RNs, 8% LPNs, 31.8% nursing aids, and 11.4% and reliability of sensors and devices, real-time
social workers. physiologic and biochemical monitors, tracking
Hospices are generally much smaller than tools (for weight, vital signs, glucose levels,
hospitals (the mean average daily census is 63, activity), wireless and cloud technologies etc. are
but the median is 31, and most (62%) have fewer improving the scope and quality of care. The con-
than 50 patients. Nevertheless, they serve vergence of technology with drivers such as an
1,162,500 patients each year [59]. Unlike hospi- increasing older population that is better served
12 R. Koppel

at home than in hospitals [62], the desire to about 5% of all US healthcare expenses [73].
reduce healthcare costs and improve outcomes SNFs employ 1,534,120 workers of whom
through awareness of preventable morbidity (in 43,420 are in management and 6,410 are top
chronic disease) and reduction of hospitalizations executives. The vast majority (630,550) are nurs-
(admissions and lengths of stay), the need to ing assistants or in related fields, and only about
extend clinical person-power (to care for more 153,000 are RNs [74].
patients with fewer physicians and nurses), and EHR software for SNFs incorporate many
more recently, the COVID-19 pandemic [63], hospital EHRs functions as well as administrative
have allowed insurance reimbursement for the and scheduling tasks such as patient scheduling
use of RPM—a point made repeatedly in the soft- and tracking, nursing care and patient flow and
ware vendors’ advertisements. Over 88% of hos- regulatory reporting [75, 76] SNFs are served by
pitals are investing in RPM and the ability to many software vendors.
implement and document remote care. As with most products of this type, ongoing
RPM software packages vary depending on costs are based on either the number of users or
the technology (sensors, recording, storage and number of beds, with a few combining charges
retrieval of clinical data), and with linkage to for both. Several vendors’ charges are based on
EHR software for documentation and billing. the number of beds and have caps on the number
Pricing packages are based per patient and range of users.
from $30 to $98 per patient per month, depend-
ing on the amount and level of communication • Estimated SNF HIT Costs per Bed For ven-
with the patient, the functionality for analysis, dors that charge by the bed, the costs for basic
interpretations, and servicing for devices services average $10/bed/month ($120 per
[64–68]. bed per year) [77, 78].
Based on review of the literature [69, 70] and • Estimated SNF HIT Costs per User For ven-
interviews with providers, we use a mean toward dors that charge by the user, the costs range
the low end of the range: $45 per patient per from $280 per user per month to $1,000 per
month ($540 per patient per year). The literature user per month, with a mean of $430 per user
from the Society of Critical Care Medicine reveal per month. This figure may be high because
[71] that there are 25.8 million US patients who some vendors offer discounts based on the
are monitored remotely. We do not include imple- number of users above some minimum. Thus,
mentation, set up, or delivery costs and focus on we use a far lower number: $80 per “basic”
the software licensing costs. When the 25.8 mil- user per month ($960 per user per year). Some
lion is multiplied by $45 per patient per month plans charge physician users more than three
($540 per patient per year) it yields an estimated times that cost (i.e. $3,000 per physician per
RPM software cost of $13.932 billion per year. year). We do not include that in our calcula-
This number, however, would overestimate the tions [79].
cost because not all patients are remotely moni-
tored for a full year. We therefore reduce the Not all SNF have full-functionality EHRs.
number of patients by 25% [72] to equal a cost of Based on CMS and industry reports, we estimate
$10,449,000,000. that only 80% of licensed patient beds are linked
to full EHRs. Therefore, we reduce the number of
1.5.3.5 Skilled Nursing Facilities (SNFs) beds from 1.7 million to 1.36 million.
There are 15,600 skilled nursing facilities in the The following calculations are based on the
US, with a total of 1.7 million licensed beds, remaining 80% of the facilities. Based on inter-
housing over 1.35 million people. More than views with vendors, SNF industry representatives
70% of facilities are for-profit institutions; with and publications, we estimate that that 60% of
4% run by hospitals. Skilled nursing facilities the SNFs are charged for their software by bed,
(SNFs) and continuing care facilities reflect and 40% are charged by user.
1 Estimating the United States’ Cost of Healthcare Information Technology 13

Base Assumptions and Estimates This is charged for each facility; not directly
• Number of Beds: As from the previous sec- based on per bed or user numbers. Similarly,
tion, the total number of licensed US SNF training and lost productivity can “cost” three
beds is estimated to be 1.36 million, adjusted weeks’ worth of activity.
from 1.7 million. From that, 60% (816,000
beds) are charged software costs based on a In addition, integration with laboratories can
per bed per year charge. Thus, 60% of the esti- be an additional $300 per facility per month
mate annual US SNF HIT Costs (by bed) is ($3,600 per facility per year) for each facility
816,000 beds × $120 per bed per year which with that functionality. However, we do not
yields $97,920,000 per year. include these costs.
• Number of Users (non-physician): According
to the BLS, there are 630,550 nursing assis- • Summary of One-Time Costs: Based on
tants or related professionals and 153,000 interviews with vendors, vendor contracts
RNs. (Total = 783,550). Using a similar esti- and interviews with hospice leaders, we use
mate of 80% of those users yields an estimate a mean implementation cost of $30,000 per
of 628,840 non-physician users for all facili- facility, which includes lost productivity, set
ties. To estimate costs by HIT that charge by up fees, and training time. For these one-
users, we take 40% (from the above section’s time implementation cost calculations (train-
description), which yields 425,736 users. ing, lost productivity, set up, etc.) we use the
However, because of the high employee turn- number of facilities, not the beds or the
over rate at SNFs and because some assistants users.
may not use the HIT, we estimate a reduction Because we assume only 80% of facilities have
of that number by another third to yield EHRs, we apply this only to 80% of facilities we
283,824. Thus, 40% of the estimated annual conservatively estimate who use EHRs (80% ×
US SNF HIT Costs (by Users) is 283,824 15,600 facilities = 12,480 facilities). Thus 12,480
users × $960 per user per year to equal facilities × $30,000 mean implementation cost
$272,471,040 per year. per facility = $374.4 million, which amortized
over 5 years is $74.88 million.
As noted, we do not include any fees for the • Additional software for direct links to labora-
physicians, who pay about 3 times the usual user tories: Based on vendor and SNF interviews
fees. and using vendor contract data, we estimate
that only 40% of the reduced number of SNFs
• Analytics software package: Another ongoing pay for the links to laboratories. Thus, the cost
cost for those SNFs that use it, is the analytics of direct links to laboratories from SNFs is:
package sold as an “add on” to the EHRs. The 40% of the remaining SNFS (40% of 2480 =
cost, according to interviews with vendors, is 4992 SNFs) X the cost of $10,000 = $49.920
$21,600 per year. Based on interviews with million, which amortized over 5 years =
vendors and SNF leaders, we estimate that $9,984,000.
only 20% of SNFs pay for that service (Note
this is 20% of the 80% of SNFs with EHRs). The table below summarizes the findings
Thus, the estimated cost of analytics software (Table 1.1):
(as used) is the number of SNFs that purchase
it (20% of 12,480 or 2,396 SNFs) which is Summary
multiplied by the cost of $21,600 per SNF per The total cost of software for SNFs is estimated
year to equal $51.753 million per year. to be $497,024,640.
• One-Time Costs: Independent of the per bed For comparison, the US spends about 5% of
or per user fee structure, implementation costs its health care budget (5% of $4 trillion = $200
range from less than $4,000 to over $48,000. billion) on nursing homes and continuing care
14 R. Koppel

Table 1.1 Cost of SNF Software (minus 20% of SNFs cost of $5,080 per agency/implementation. This
b/c they may not have EHRs)
yields an estimated total implementation cost of
Beds = 1.7 million reduced to 1.36 $30,480,000, which amortized over 5 years to be:
million
$6,096,000 per year.
Users: 783,550 reduced to 628,840
We know 60% of vendors charge by
We estimated the total number of users as
beds, and 40% charge by users. 95,000 nurses, 200,000 aides, and 12,000
40% of users = 628,840 × .40 = $272,471,040 administrators, billing and insurance personnel
251,536 × $960,471,040 which yields 307,000 “users.” Licensing costs/
60% of beds = 1.36 × .6 = 816,000 × $97,920,000 fees ($480 per user per year) yields an esti-
$120/bed = 97,920,000
mated $147,360,000 per year, which added to
Analytics package: after removing $51,753,600
80% of facilities = 2,396 × cost of amortized implementation costs of $6,096,000
$21,600 per year yields a total of: $153,456,000 per
One-time: implementation for 80% of $74,880,000 year.
SNFs (N = 12,480) × $30,000 =
$374,400,000. This is amortized over
5 years to equal
Total $497,024,640 1.5.4 Clinical and Other Services
(Usually Not in Hospitals)

facilities. Thus, as a percentage of SNF’s bud- 1.5.4.1 Acupuncturists

gets, HIT costs are remarkably modest. Acupuncture is covered by Medicare/Medicaid
and private insurance and is approved by the
1.5.3.6 Visiting Nurses VHA for pain management. There are 24,954
In the US, over 500 visiting nurse associations, practitioners working in the US, providing more
representing over 12,000 agencies, employ than 10 million treatments yearly [83], creating a
approximately 95,000 nurses, and many multi- revenue stream of $650 million per year. As with
ples of that number for other healthcare workers, all medical services, software is required for bill-
usually with more restricted licenses, e.g., nurs- ing, record keeping and insurance.
ing assistants, LPNs) [80]. From vendor literature and interviews [84–
Many software providers offer programs to 86], the mean cost of software is $122.00 per
these agencies. Their charges incorporate both month ($1,464 per year) per user, which for
the number of users (nurses, aides, back-office 24,954 US practitioners is $36,532,636.
personnel) and the number of patients. The latter Not included in this calculation because of
numbers are substantial, with visiting nurses partial or incomplete data coverage, are fees of
serving over 4 million people annually [81]. $20-$45 month per practice for data migration,
Interviews with vendors and visiting nurse custom medical forms, training, etc.)
agency officials plus review of vendor contract
details reveal that the mean EHR/HIT licensure 1.5.4.2 Chiropractic
costs per user (both clinicians and administra- There are 70,000 independent US chiropractors
tors) averages to $40 per user per month ($480 and 40,000 chiropractic assistants (CAs) [87],
per user per year), with a cap on the number of who pay user fees averaging $1,747.00 per user
patients [82]. per year (but which can be as high as $3,380 plus
Implementation costs average about $5,080 additional fees other services (e.g., training, set
per agency with fewer than 40 nurses/nurse assis- up). Using the lowest estimates, this results in a
tants/back-office personnel. Larger agencies’ US cost of $192.17 million for yearly chiroprac-
software installations cost more. tic software fees.
To avoid double-counting other home health Vendor estimates of implementation and lost
services, we cut the number of visiting nurse productivity costs are estimated at $7,600.00 per
agencies from 12,000 to 6,000, using the mean practice, which amortized over 5 years is
1 Estimating the United States’ Cost of Healthcare Information Technology 15

$1,520.00 per practice [87]. We apply this only to software for trial and pharmacovigilance work,
fully licensed chiropractors (not CAs), yielding a which is estimated at 360,000 users [93, 94].
total of $106.40 million per year. Added to the Licensing Costs: These are quoted at a mean of
yearly software fees yields and annual software $2300 per user/per year. Larger firms (with
cost of $298,570,000 bigger trials and more employees) may pay
less per user than smaller firms with fewer
1.5.4.3 Clinical Trials Software employees. Using these averages, total licens-
At the time of writing this, there were 387,063 ing costs are estimated to be $828 million per
registered clinical trials (CT) known to the US year [95].
government, of which 144,853 were operated by Implementation Costs (and lost productivity dur-
US researchers/firms (of which 125,323 were ing training): These are estimated as a cost per
entirely in the US) [88]. Software for these enti- site (location where patients are recruited for
ties and research groups (such as independent data collection, which is neither the number of
contract research organizations (CROs) must sup- companies performing a research project nor
port institutional board review, ­ recruitment/ the number of subjects). In the US, the num-
enrollment of human subjects, informed consent, ber of research sites is estimated to be 2300 at
payment tracking (to subjects and all involved any given time. Implementation Costs are esti-
entities), coordination of teams and research sites mated at $40,000 per site, which amortized
and their documentation. The software must over 5 years yields $5,000 per site per year.
assure collection and secure storage of all of the Thus, Implementation Costs are conserva-
participant information, compliance, outcomes, tively estimated to be $11.5 million per year.
adverse events, et cetera according to regulatory Combining Licensing and amortized
requirements for protecting human subjects. Implementation Costs yields an estimate of
Increasingly, CT software must be remotely and $839,500,000 per year.
securely accessible by a number of devices:
smartphones, tablets, notepads, laptops. It must 1.5.4.4 Dental Care
also manage compliance needs, protocol changes, There are 201,117 licensed dentists in the
reminders, and updates. USA. Most are working at about 193,000 dental
Information assurance (confidentiality, integ- office establishments [96]. Almost all employ
rity, availability) of data and subject privacy must some form of software for medical records, bill-
be managed and updated constantly. There are at ing, scheduling, insurance claims management,
least 60 firms that provide software to this indus- etc.
try. Pricing is based on the number (of trials) and Based on the software license agreements and
trial size (number of human subjects) or the num- interviews with vendors and providers, we
ber of users (research staff). Costs for pharmaco- include dental hygienists (N = 150,000) and den-
vigilance software is usually based on the number tal assistants (N = 354,600), who must pay
of cases (human subjects) rather than the number monthly user fees [97, 98]. Thus, the total num-
of staff (users) [89–91]. ber of “users” is 555,867.
The collective CRO workforce is estimated at From that number we subtract 10% of dentists
450,000 persons. This large labor force is, in part, (not assistants or hygienists) who are involved in
due to the very high employee turnover rate research, administration or who work in other
(ranging from 22% to 27%), to the vast number settings. Thus, the final N of users is 536,567.
of trials conducted at any one time and the wide From the same ADA sources we determine the
range of skills required for the tasks [92]. Job annual license fee averages to $4,464. When we
titles include: case managers/coordinators, bio- multiply that fee times the number of users, the
statisticians, payment specialists, physicians, total is $2.396 billion
medical device technicians, and pharmacists, Implementation, lost productivity etc.: Based
among others. We focus only on those using paid on the vendor data and market research reports,
16 R. Koppel

we calculate set up fees, lost productivity, train- • The number of clinicians (MDs/DOs/NPs)
ing and implementation costs plus commonly with independent offices that are not entirely
assessed additional fees for connections to part of a larger hospital system
patients’ insurance companies, labs, dental • Disaggregating clinicians who work within
device manufacturers, connections to patient’s hospitals or other large organizations with
EHRs and patient reminders, etc. This brings EHRs, but who also have independent prac-
the total to $37,640, which we amortize over 5 tices (to avoid double counting) while also
years to equal $7528 per office/practice counting clinicians with both hospital and
(n = 193,000). This cost is not per user, but per “private” practices who pay for their EHRs
office. • The cost of implementation, set up, et cetera
Summary: 193,000 Times $7528 = $1.453 bil- of such systems.
lion. When added to the license fees ($2.396 bil-
lion) the total = $3,849,000,000 Considerations: In 2020, almost 40% of phy-
Note that $3.849 billion is only 2.85% of sicians worked directly or partially for a hospital
dental expenses in the USA in 2021. The 2.85% or for a practice owned by a hospital or health
is considerably lower than the expected ratio system (increased from 29% in 2012 to 34.7% in
spent on software in most healthcare settings 2018). Those working solely for a hospital
[99]. increased from 5.6% in 2012 to 9.3% in 2021.
An increasing number of clinicians who work
1.5.4.5 Dialysis within (hospital) systems and within other health-
In the US, 554,038 patients received dialysis. care contexts (e.g., pharmaceutical firms, public
Cost estimates for the basic software fee is $2.30 health organizations, research institutions, etc.)
per treatment. However, dialysis software must that do not require individual EHR licensure. The
be connected to EHRs (for the dialysis unit as number of physicians with separate practices var-
well as for clinics and primary physicians), clini- ies widely by specialty and by business consider-
cal services (pharmacy and laboratory informa- ations [103, 104]. “Surgical specialties had the
tion systems), scheduling and reminder systems, highest share of owners (64.5%) followed by
billing (Medicare/Medicaid, private insurance) obstetrics/gynecology (53.8%) and internal med-
and revenue management. This total cost for icine subspecialties (51.7%)” [105].
(routine) dialysis software operation alone is esti- Physicians: Based on the several studies of phy-
mated at $3.85 per treatment/patient (excluding sicians with EHRs (both independent of larger
costs of set up/training or of other aspects of dial- organizations or with additional EHRs for pri-
ysis care) [100–102]. Thus, an estimate for US vate practice), we estimate that 38% of MDs/
dialysis software cost for one year (for a standard ODs pay for separate EHRs [105]. In terms of
3 times a week schedule) is: $ 3.85 per treatment the numbers of relevant clinicians, this means
per week per patient × 3 treatments per week × that rather than 525,000 active physicians with
52 weeks per year =11.5. their own EHRS, we estimate that the number
For the US patient population requiring is 199,500 physicians.
dialysis: $6,006 per patient per year x 554,038 Nurse Practitioners: We further estimate that of
US patients provides an estimate of the 190,000 NPs, only 18%, or 34,200, should
$3,327,552,228 be included as incurring expenses for EHR
licenses. Moreover, we do not include any
implementation costs for the NPs because we
1.5.4.6 Medical Practices’ EHRs (MD, assume those expenses are absorbed by the
DO, some NPs) separate offices in which they work.
Calculating the cost of general medical practice The AMA survey data show 49.1% of patient
EHRs requires estimation of: care physicians worked in physician-owned prac-
1 Estimating the United States’ Cost of Healthcare Information Technology 17

tices, down from 54% of physicians in the 2018 Licensing and total Costs: This ranges from $99
AMA survey [106]. The cost of implementation per practitioner/user per month to over $600
of physician practice EHRs is in part determined per practitioner/user per month. Based on
by the number of services covered. These include: review of the contracts, interviews with the
billing and links to insurers, scheduling, patient vendors and OT leaders, we use a mean of
reminders, patient portals, links to laboratories, $245 per practitioner/user per month ($2,940
referral services, revenue management, adminis- per practitioner/user per year) for the 31% of
trative oversight, and analytic functions. Based OT practitioners). With no implementation
on industry data, interviews with vendors, and costs, the total estimated OT EHR/HIT licens-
interviews with clinicians, we determine that ing cost is (31% of 126,610 × $2,940 or
implementation costs are very conservatively $115,392,060 per year.
estimated at $215,000 per physician, which we
amortize over 5 years to equal $43,000. 1.5.4.8 Optometry
EHR license fees vary considerably by amount There are 41,000 optometrists in the US [110].
of functions, and if there is “free” access by Software vendors charge by user, and there are,
nurses and other team members, etc. We c­ alculate on average, usually two or more users per facility,
that a cost of $5,200 per month per clinician, or or a total of 82,000 users (assuming only two per
$62,400 per year is a median price. practice). Mean software cost is $2,988 per year
To summarize: Implementation costs for the per user, which yields $245 million per year in
199,500 physicians are calculated at the amortized optometry software license costs.
cost of $43,000 each. License fees are calculated One-time costs for set up and implementation
at $62,400 for the 199,500 physicians and 34,200 costs average $500 and $400 per practice respec-
NPs (total = 233,700). Thus: Implementation costs tively. Lost productivity and implementation
of $43,000amortized for 199,500 physicians = times are said by the vendors to be three weeks,
$8,578,500,000; and license fees for 233,700 cli- but we do not count the entire time because
nicians, totals to $14,528,880,000. The combined optometrists continue to operate during the
total is $23,107,380,00. implementation, but at a lower rate. We thus use
a very low cost for total estimated onetime costs
of only $3,700 per user, counting only the optom-
1.5.4.7 Occupational Therapy etrist (i.e., not staff). Thus, $3,700 X 41,000 =
In the US, of 126,610 occupational therapists $151.7 million, which amortized over 5 years =
(OTs), approximately 69% work in hospitals, $30.34 million.
academia, mental health facilities, and long-term This yields $245.016 million for the license
health facilities whom we assume do not pay plus $30.34 million of amortized one-time costs.
licensing fees for separate software. The 31% Thus, total US optometry software costs per year
who work in other settings are obliged to use is $275,360,000
EHR/HIT for OT (with licensing fees) that pro-
vides additional clinical/administrative func- 1.5.4.9 Pharmacies--Chain
tions: e.g. telehealth, mobile access, scheduling, and Independent
and appointment reminders [107–109]. There are 88,000 pharmacies in the US. Of these,
On average, OTs see 5–8 patients per day. approximately a quarter, 23,000, are “indepen-
Software pricing is predominantly based on the dent” (of chains). There is a critical distinction in
number of practitioners/users, although a few the pricing of pharmacy software for chain vs.
companies charge by number of patients. We independent pharmacies.
found there are few if any charge for implementa- Chain Pharmacies: Chain pharmacies use one
tion, setup or training. There is lost productivity of several “back-office” software products (e.g.,
when learning the system, but we do not include SAP, Salesforce) that coordinate individual
that cost. Point-of-Sale (PoS) software at the physical loca-
18 R. Koppel

tions in the chain. PoS software, in combination remaining (smaller) chains have a total of 18,000
with the back office software, connect chain loca- stores (Total = 65,000)
tions to other systems, including pharmacy ben- As noted above, the mean cost of “back-­
efits managers, insurance companies, state office” software licenses differs by chain size.
agencies and local healthcare providers. Note Based on the large software sellers, the largest
that both the back-office software and the PoS groups pay about $10 million for their software—
software must be customized to meet local regu- both the pharmacy chains and the large super-
latory requirements for labeling of prescriptions market and big box stores. (Note this is not the
and pharmacy data management program cost per pharmacy, but rather the cost for the cor-
(PDMP) reporting for controlled substances. porate headquarters IT group that covers all of
Chain pharmacy software requirements: the outlets. Note also this is not, for example, for
a CVS located in a Target store, which shares real
• Yearly maintenance fees for corporate license estate.) Medium size chains pay a mean of $5.4
(usually about 20% of the cost) million for their enterprise software. The smaller
• Customization, adjustments and repairs chains have lower costs, at $1.7 million.
(includes repairs after installation by Each chain site requires a separate PoS license
consultants, e.g., Deloitte, Accenture). This
­ that is integrated with the corporate “back-office”
includes updating the formulary, regulatory software, customized to local needs (links to
changes and price adjustments (both price and local providers, hospitals, regional health sys-
copays). tems, nursing homes, billing, inventory, in addi-
• Key here are the links to the pharmacy benefit tion to maintenance, repairs, integration with new
managers (PBMs) that both serve as middle- systems, etc.). Excluded from this are: individual
men for payers--insurance companies, site licenses ($47,000 per site), additional builds
employers—and for the retail sales/copays for and web-based enterprises (i.e. Amazon).
individuals. The table below summarizes the types of
• Additional builds (added or modified chains, costs and market sizes Table 1.2 [111].
software) Independent Pharmacies The 23,000 indepen-
• Local point-of-sale software; not part of the dent pharmacies must have software that per-
back-office software noted above forms PoS functions, just as for the chain
• Local state by state for PoS software, e.g., operations—and links to: PBMs, PDMPs, bill-
state labeling rules, PDMP links. ing; claims processing, inventory, ordering, cus-
• Connections to local and regional hospitals tomization for state labeling rules; connections to
and clinicians local and regional hospitals, clinicians, formular-
• Connections to local/regional insurance com- ies, and to local/regional insurance companies.
panies, etc. Software services to independent pharmacies
must also address repairs and modifications, as
The “back-office” software for the largest required [112]. There is a very large market of
chains costs as much as 10 million dollars, and software vendors to serve independent pharma-
far less for smaller chains (costs outlined below). cies, and many offer combined services. Thus,
Chain store data: The largest 24 big pharmacy independent pharmacies can purchase combined
chains have a collective 39,914 stores, with a packages for as little as $44,600 that accomplish
range from 9,900 stores to 88 stores); another what they need [113].
large group are supermarket pharmacies and “big For independent pharmacies, we multiply
box” pharmacies, e.g., Costco, Giant, Publix, $44,600 times the 23,000 sites = $1.026 billion.
Sam’s Club, Kroger. Many of these are large Independent pharmacies do not obtain the cost
enterprises, selling millions of prescriptions via savings of the larger chains, which benefit signifi-
thousands of outlets. There are also medium size cantly by using one contract for “back-office”
chains that have a total of 7,086 stores; and the services.
1 Estimating the United States’ Cost of Healthcare Information Technology 19

Table 1.2 For chain pharmacies

Backend SW fee for each
Chain size N sites chain (not store) Total
Major chain N = 24 39,914 $10 million $240
million
Major sellers that are part of supermarket or big box store, 4,453 $10 million 1.330
e.g., Kroger, Giant) N = 133 billion
Subtotal big stores 44,367
Medium N = 92 2,643 $5.4 million $496.8
million
Small N = 943 18,000 $1.7 million $1.603
billion
Total for chain stores 65,000 $3.67
billion
Individual PoS licenses for each location is $47,000. For the 65,000 chain sites $3.055
billion
Total for back end and PoS for chain pharmacies = $6.725
billion

Total Software Cost for Independent and of approved/covered drugs. They are paid by
Chain Pharmacies: Thus, the combined phar- insurance companies and similar entities, e.g.,
macy software costs for independent pharmacies CMS.
($1.026 billion) and chain pharmacies, from As might be expected, the industry is both
above, ($6.725 billion) in the US is $7,863,000,000 capital intensive (with ratios of 1:1 for labor
We do not include implementation or hard- and capital expenses) and software intensive.
ware costs, or even the mail-order software Based on review of the PBM software vendors,
license costs (for the increasingly popular mail interviews with vendors and PBM providers,
order pharmacies). plus US BLS data, we estimate the number of
employees in PBMs totals 114,800. However,
1.5.4.10 Pharmacy Benefit Manager the proportion of that work with paid software
(PBM) Software is only about 50% of those employees, which
In the past few years, pharmacy benefit man- equals 57,400. Multiplying the 57,400 by the
ager firms (PBMs) have consolidated and mean software licensing cost of $3,400 per
become vertically integrated into insurance employee/user per year, produces a total of
companies; often also vertically integrating $195,160,000.
with healthcare ­providers. Although there are We do not include any costs for implementa-
66 listed PBMs [114], most of the market is tion, training, lost productivity, etc.
concentrated in 11 firms, with much of the work
controlled by 6 firms that are vertically inte- 1.5.4.11 Physical Therapy (PT)
grated with insurance carriers and providers. Of There are an estimated 258,200 physical thera-
these, CVS, Express Scripts and Optum control pists (PTs) working in the US, aided by an addi-
32%, 24% and 21% of the market respectively. tional 149,300 assistants and aids (Total 407,500)
The market size is listed as $458 billion [118, 119], with many working at hospitals,
[115–117]. home health agencies, and residential care facili-
PBMs manage insurance coverage of pre- ties. After interviewing PTs and vendors of ser-
scribed medications by: “advising” (limiting) vices, we estimate half of PTs at such facilities
patients and prescribers of “approved” medica- use clinical and other software provided by those
tions and indications through formularies, institutions as part of the general EHR or separate
assuring prior approvals for coverage of pre- PT-specific modules. For our calculations, we do
scribed drug regimens, and ensuring availability not include these PTs in our cost equations.
20 R. Koppel

Thus, we focus on the PTs [120–123] who 1.5.4.12 Podiatry

require software to document, bill (submit insur- The number of practicing podiatrists in the US is
ance claims), provide exercise libraries, patient listed as between 15,000 [125] and 18,000 [126]
portals and other functions: We use the lower number for estimates—15,000.
The average cost for podiatric software is $268.00
• Users in PT offices/clinics (33% of 407,500 = [127] per user per month. This results in:
134,475);
• Self-employed PTs (8% of 407,500 = 32,600); • $268.00 per user per month x 12 months ×
and 15,000 users = $48.24 million per year
• PTs who work in facilities but must have sepa- • One-time costs (lost productivity, implemen-
rate clinical software licenses (50% of the tation/setup) are estimated to be $4,850 per
remaining 240,425 = 165,034). The total is user, which amortized over 5 years is $970.00
332,068 users. per user per year for 15,000 users = $14.55
million per year
Software License Costs: Software license
costs are based on the number of users (clini- These sum to an estimated annual US podiat-
cians). Based on a review of the vendors’ mar- ric software cost of $62,790,000
keting literature, advertisements, and interview We do not include non-podiatric staff, which
responses, we find the mean fees are $85 per are usually counted as “users,” and would signifi-
user per month. Annualized, this equals $1,020 cantly increase the estimates. Nor do we include
per user per year. For the 332,068 PTs, assis- $0.10 cost for each SMS appointment reminder
tants and aides who pay separate licensing fees that the HIT vendors charge their podiatrist users.
(at $1,020 per user per year), the total is
$338,709,360. 1.5.4.13 Telehealth
Implementation Costs: We calculate PT-­ Telehealth expanded exponentially during the
specific EHR/HIT (i.e., for PT centers and not for COVID 19 pandemic. A recent AMA study found
other facilities that employ PTs (e.g., hospitals, it was used by over two-thirds of physicians, with
nursing homes etc.) at an average of $600 per some areas—psychiatrists—employing it almost
facility per year. We conservatively estimate lost universally [128].
productivity (a mean of 3 weeks) on the assump- A Web search discovered over 170 vendors, with
tion that clinicians can trade off time/patients names running from Zipnosis to Anytime Pediatrics
during the installation and deployment. The US to DoctorConnect, to well-known vendors such as
Dept of Labor’s Occupational Outlook Handbook Nextgen, Athenahealth, or EpicCare EMR [129]
reports that PTs earn an average of $91,010/year; In addition, we benefited from extensive
and PT assistants earn $49,970 per year [124]. PubMed searches. Many research reports, plus
For each PT center, we estimate the loss of one government and consultant services, provided
week for only one PT to calculate a loss of useful breakdowns of costs for licenses, imple-
$3,640, not including costs for training. Each PT mentation, per user costs, integration with EHRs,
center is assumed to spend $2,100 on HIPAA compliance, etc.[130, 131]
Implementation and $3,640 on lost productivity.
For the 38,000 facilities the total is $5,749 X Excluded: We do not include costs of hardware,
38,000 = $218,120,000, which we amortize over used in telemedicine, e.g., devices that are sent
5 years to equal $43,624,000 to patients to enhance cell phones data for clini-
In summary: For PTs, the estimate of EHR-­ cians, including: digital telescopes, examination
type licensing costs is $338,709,360; and amor- cameras, or ENT scopes; or the myriad devices
tized implementation costs ($43,624,000) is for what is called “telehome” such as home-
$382,333,360 installed digital blood pressure cuffs, scales, etc.
1 Estimating the United States’ Cost of Healthcare Information Technology 21

Connectivity: Estimates of costs to set up “con- To this we add the nurse practitioners adjusted
nectivity to portable devices or integrating as follows: Total N = 325,000, of which 290,000
additional APIs” were between $5000 and have active licenses [133]. We then subtract all
$10,000,” but we excluded them with the those in hospital settings and all those not in pri-
assumption that such costs will be shared with mary care, which reduces the number to only
a practice offices’ other IT costs. one-quarter of their number—to only 72,500
Consultants: Training is estimated at two-to-­ NPs—undoubtedly a low estimate.
three weeks (of lost productivity) with train- Cost and numbers: The two, reduced numbers,
ing program costs ranging from $200 to (174,825 MDs and DOs) and NPs (72,500) totals
$2,000 per location, depending on the number to 247,325 clinicians, which we multiply by the
of users, additional installations, and technical amortized implementation costs of $12,740
training for IT staff. We use the lowest esti- (from above) to = $3,150,920,500
mates at $200 for the training, and lost pro-
ductivity at $5,000—assuming learning times Yearly software licenses: The reports reveal that
will be distributed in cost efficient ways. The the software license fees range from $420 a
amortized cost of $5,200 is incorporated month to “several thousand dollars” a month.
below, in the implementation cost analysis. We use the near-lowest end of the range ($550/
Implementation: Implementation (one-time cost) month). Thus, our estimate for a one year the
can be surprisingly expensive. The higher com- license is $6,600. Allocating this to the num-
plexity or functionality of a telehealth app, the bers of clinicians from the above paragraph =
higher the cost. Telemedicine software cost can ($6,600 X 247,325) determined above equals:
exceed $370,000 if one chooses a highly com- $1,632,345,000
plex and feature-rich solution. On the other We add legal fees from estimates of the American
hand, solutions are available from $50,000 Telemedicine Association’s website for state-­
[132]. We use a lower range estimate of total cost by-­state estimates. Originally, we were not
for software for telehealth at $62,600 (before going to include legal fees because pandemic
amortizing). Thus the implementation total for related legislation allowed telemedicine clini-
telehealth totals: training ($200), lost productiv- cians to practice without additional restric-
ity ($5,000), software set up and integration into tions. However, after June of 2021, many states
the EHR ($62,600), which totals to $63,700. are again imposing restrictions that are being
Then we also amortize all costs over 5 years, challenged in court. Because, however, these
which reduces the onetime cost to $12,740. issues are not universal across the US, we use
Relevant population: We limit the costs to only only 30% of the legal fee reported rates of
the MDs/DOs and NPs who use telehealth and $75,000 ($75,000 X 30% = $22,000). Thus,
who are not part of a larger system’s setups we only add on $22,500 per state, and only
(e.g., not part of a hospital or larger system’s multiply that times 25 states, which equals $
telehealth system). To determine the applica- 1.875 million to reflect the entire nation’s tele-
ble users, we calculate that only one-third of health legal costs.
the MDs/DOs in active private practice use Summary: Total of all amortized implementation
telehealth. Thus, while there are 900,000 phy- fees = $3,150,920,500; total of all license fees
sicians with licenses, only 525,000 are actively = $1,632,345,000; total legal fees for states =
practicing physicians. Moreover, a portion of $1,875,000. Thus, the total for telehealth and
these 525,000 are with larger systems, or related services not part of hospitals or other
working in the insurance or pharmaceutical entities is $4,785,140,500
industry, public health, etc. We therefore use
only one-third of the active physicians to deter- Note that the 4.785 billion is less than 1/4th of
mine a population of 174,825 MDs/DOs [132]. the $20 billion estimates of other studies.
22 R. Koppel

1.5.5 Hospitals and Usually Linked Hospitals now own more than one-quarter of
Services physician practices; and corporate entities now
own more than one-fifth (22.1% ref) of physi-
1.5.5.1 Hospital EHR Costs cian practices [137, 138] This last datum is
Calculating the cost of EHRs for hospitals is especially important in estimating the number
challenging. The systematic cost underestimate and percent of physicians who do not pay for
biases noted earlier [134] apply here with great separate practice EHRs because either they are
force; and we outline them below. On the other employees of hospitals, or because their prac-
hand, there are literally hundreds of cost esti- tices use hospital software.
mates, many from reliable sources.
In contrast, there are many sources of reliable,
Data Sources for Hospital Software Costs vetted data-supported information on estimates
Biases and challenges in obtaining estimates and ratios with which triangulation is possible. A
include: few simple examples include:

• Underreporting of software costs in trade jour- • Gartner and other reports on the mean cost of
nals, dependent on EHR vendors and consul- HIT per healthcare employee [139], in which
tants for advertising and other expenses, with hardware costs can be disaggregated from
the added observation that journals may be software costs [140].
published by trade associations • Detailed data on IT operating expenses in
• Attractively low cost and staffing estimates relation to hospital operating expenses.
quoted by vendors, including their suggested • Data on implementation costs, including the
EHR implementation time requirements [14] software, implementation, training, retrofits,
• Vendor statements about the need for fewer IT builds, extra IT staff, etc. (Note that imple-
personnel post-implementation—statements mentation is generally 3–5 times the cost of
that emerge as marketing efforts, and that are the software) (see Box 1.1).
documented as unrealistic [14]. • Vendor industry sales and investment figures
• Not including and not counting the use of • Software sales figures and predictions from
existing staff in the implementation process, vendors, market researchers, and others
e.g., not including the cost of staff to design [141–144]
order sets, CDS alerts, review problems,
address issues, staff help desks, train, help
optimize the system (an ongoing expense) and
the training and use of “superusers” BOX 1.1 Examples of Data Sources on Cost
• Self-interest of CMIOs and CIOs to not reflect of Software, Implementation, Training,
the full costs of implementing their recom- Retrofits, Builds, Extra IT Staff, etc.
mended EHR choice
• Federal regulators who saw their mission as • The percent of hospitals’ budgets
encouraging the sale and use of EHRs. Note devoted to software, including percent
we do not suggest these motivations were ill-­ of operating budgets, percent of person-
intended [1, 3, 135, 136] nel working on clinical software
• Increasing purchase and incorporation of phy- • Percent of hospital IT budgets devoted
sicians’ practices by hospitals and often by to capital (25%) vs operating expenses
larger entrepreneurial enterprises—with sig- (75%)
nificant effects on how EHRs will be merged, • Spending by I.T. functional areas, i.e.,
serviced, licensed, and invoiced. Hospitals, data centers 20%; end user computing
especially, are predicted to become even more 10%
of the “hub” for medical care in an area.
1 Estimating the United States’ Cost of Healthcare Information Technology 23

First Method: HIT Operating Cost per hospital

• Cost percentages: service desk 5%; net- employee according to PWC reports is $6,850
work 13%; application development 9%; per employee [153]. For 7.6 million US hospi-
application support 32%; IT. manage- tal employees [154] at a mean HIT cost per
ment, finance, and administration 11% hospital employee ($6,850) the total in 2021
• Estimates of all hospital spending on dollars is $52.060 billion.
software and other studies by Gartner This is a conservative estimate as the industry is
and others [145] estimated to have spent $120 billion on HIT in
• Operating IT budgets for The Cleveland 2021 [155]
Clinic’s main campus; NYU Langone Second Method: The American Hospital
Tisch Hospital, Vanderbilt Univ Medical Association provides an annual US hospital
Center, MGH and 21 more institutions HIT cost of per bed of $56,614.28, which for
[146] 919,599 US hospital beds yields a figure
• Budget details for all hospitals in vari- within .01 percent of the cost based on per
ous states [147] hospital employee or $52.0624 billion [156]
• Several guides to IT budget processes, Third Method: HIT Operating Cost based on total
e.g., IT costs from “Gartner IT Key healthcare organizations/hospital spending.
Metrics Data” [148] Hospitals represent about a third of all health-
• Publications from the ONC and AHRQ care spending, that is, 1/3 of $4 trillion or
on cost of software, e.g., “How much is $1.320 trillion [157]. If the estimate of 4% of
this going to cost me?” [149] total budget is spent on US hospital HIT
• *Budget reports on hospitals, e.g., Operating Cost, the estimate is $52.8 billion.
reports on the cost of hospital IT service Fourth Method: HIT Operating Cost based on
desks [150] recently published and available data for 15
• *Other government estimates, e.g., hospitals’ HIT Operating Budgets, number of
“Medical Practice Efficiencies & Cost beds, and total patient revenue (see Box 1.S1
Savings” [151] in supplemental notes), the mean IT operating
• Electronic Health Records – “Health IT cost is $132,030.80 per bed. For the total num-
Playbook,” which defines line-item ber of US hospital beds (AHA: 919,599), the
costs for EHR software, implementa- estimated cost is $121.416 billion. (This is
tion, training, and support — for both considerably more than twice the other three
on-site licensing models and cloud-­ estimates, but we average the four. However,
based platforms [152] as a reality check of this method, we note that
IT operations form 3.85% (range 2.79% -
4.77%) of total patient revenue = 3.85%,
Cost Estimates which are consistent across many industry
Cost estimates for hospital EHRs can be divided measures.
into three groups:
Averaging these 4 estimates of hospital HIT
1. HIT Operating Costs Operating Costs listed above, we find the mean costs
2. HIT Implementation Costs for hospitals with of US HIT Operating Costs to be $56.085 billion.
new or recently replaced EHRs As a second “reality check”, we multiply our
3. Licensing Costs (including maintenance, mean US HIT hospital operating costs from
updates and service) above ($56.085 billion) as a percent of all hospi-
tal costs of $1,475.5 billion that the US spends on
HIT Operating Costs hospital care [158]. The percent is 3.8%, which is
We used four methods to estimate hospital IT very close to the national average for most
operating costs: industries.
24 R. Koppel

HIT Purchase and Implementation Costs (One • Lehigh-Valley-Health system (Epic): $45,122
Time Costs Amortized Over 5 Years) per bed X 4 = $180,448
To calculate US hospital HIT implementation
costs for software, we first use data from four Based on the calculations presented above in
medical centers with listed and recent purchase Box 1.1, we calculate that the mean for hospital
costs (including per bed costs) for their EHRs. EHR implementations is $170,548 per bed.
Again, this reflects software and implementation Which, when added to the mean amortized pur-
costs only, not total system costs. The 4 systems chase cost of $53,917.5 totals to $244,465.5 per
are: bed. We do not, however, multiply that n­ umber
times the 919,519 beds because that would gen-
• University of Pennsylvania (Epic): $62,906 erate an exaggerated figure that fails to reflect the
per bed. reality that only a portion of hospitals are recent
• University of Arizona (Epic): $41,641 per bed buyers and installers of EHRs. Instead, we take
• Cape Cod Health Center (CCHC) (Epic): only the 27.6% of all hospital beds that reflect
$66,001 per bed more recent EHR purchases [159]. The 27.6% of
• Lehigh-Valley-Health system (Epic): $45,122 919,599 beds equals 253,809 beds. Multiplying
per bed that bed number by the $244,465.5 of cost equals:
$62,047,544,090. We thus exclude almost three-
The data for each system are presented in quarters of hospitals’ costs in this part of the
Supplemental Box 1.S1. calculation.
We take the average per bed of purchased HIT
software cost at the four institutions. The mean Licensing Costs (Including Maintenance,
cost after amortization (i.e., divided by 5) is Updates and Service)
$53,917.5 per bed. This cost reflects some of the Annual maintenance, service and licensing fees
costs of training, retrofitting to legacy software for hospitals with EHRs range from 18-23% of
and customizations (order sets, etc.). but does not the purchase price of the software. However,
reflect the years typically required for full we must account for the fact that many US hos-
implementation. pitals have older systems with lower annual
To estimate the implementation cost we multi- HIT licensing costs. Using full recent costs
ply the purchase price by the usual ratio of imple- would exaggerate the licensing fees. Thus, we
mentation cost to purchase price, which varies segment hospitals by dates of EHR purchase
from three-fold to five-fold. We use a four-fold [159] and apply adjusted licensing costs as
ratio. However, we do not use the actual purchase follows:
price because implementation takes several years.
It also does not reflect the cost of additional IT • “New” EHRs – 27.6% of all hospital beds
personnel required with each EHR implementa- (919,599) x full licensing cost of $53,917.5
tion, which can be considerable [14]. Thus, we (from above) X 253,809 beds =
base our multiple on the amortized costs, i.e., $13,684,746,758.
1/5th of the purchase price. Thus, for the four sys- • “Older” EHRs 72.4% of all hospital beds
tems, the amortized implementation costs are: x 65% of full licensing cost = ($35,046
per bed) for 665,790 beds =
• University of Pennsylvania (Epic): $62,906 $23,333,276,340
per bed X 4 = $251,624
• University of Arizona (Epic): $41,641 per bed Adding the $13,684,746,75 and the
X 4 $166,564 $23,333,276,340 totals to $37,018,023,098
• Cape Cod Health Center (CCHC) (Epic): In summary, US Annual Hospital HIT
$66,001 per bed X 4 = $264,004 Costs:
1 Estimating the United States’ Cost of Healthcare Information Technology 25

 • HIT Operating Costs $56,085,000,000 The total for the initial software and monitoring
 • HIT Purchase and $62,047,544,090 stations, amortized, is $27,945,884,000 per
Implementation Costs: year.
 • Licensing Costs (including $37,018,023,098
Licensing Fees: The annual cost per bed (license,
maintenance, updates)
 • Total: $155,150,567,188 support, repairs, patches etc.) is between
$50,000 and $100,000/year. We use the esti-
Again, this does not reflect the many years mate of $75,000. Thus, for 98,401 beds, licen-
required to implement an EHR (noted as typi- sure costs are $7,380,075,000 per year.
cally four years), three quarters of all hospitals’ Combining the amortized costs and the annual
older purchases of EHRs, and all of the inhouse fees = $35,325,959,000 [164] per year.
work by clinicians and others.
1.5.5.3 Medical Imaging
1.5.5.2 ICU Monitoring: Tele-­ US expenditure on medical imaging is estimated
Monitoring for ICU Beds to be 10% of all healthcare expenses, equal to
Of the 68% of US hospitals with ICU units, $400 billion [165]. Imaging has expanded to
36% of those have telemonitoring systems. include not only radiology, tomography, sonog-
Data on ICU beds indicate 68,558 adult ICU raphy, nuclear medicine, et cetera; and medical
beds (medical-­surgical 46,795, cardiac 14,445, imaging is now incorporated into many “new
and other ICU 7,318), 5137 pediatric ICU beds, areas of image-based work in: ophthalmology,
and 22,901 neonatal ICU beds. Additionally, surgical specialties (inclusive of laparoscopic
there are 25,157 step-down beds, and 1,183 surgery), invasive cardiology, pulmonology,
burn-care beds: for a total of 237,684 ICU-type neurosurgery, urology, speech pathology, der-
beds [160] matology and burn/wound medicine, etc.
More recent studies [161] indicate a signifi- Imaging is a ubiquitous service and is also
cant increase in ICU beds (in part due to COVID offered and performed now in 10.7% of urgent
19) to conservatively increase the number by an care visits [166].
additional 15%, which equals another 35,653 We focus only on the cost of the software por-
beds, or a total of 273,337 beds, of which we tion of imaging, its implementation, and user
assume only 36% have telemonitoring (same per- licensing costs. We used cost information based
cent as the estimate above, which is almost cer- on industry data [167], federal agencies (CMS)
tainly an underestimate). Thus, we take 36% of [168], disaggregated hospital reports to federal
those beds to equal the resulting figure is 98,401 agencies [169], interviews with hospital leaders
ICU-T tele-monitored (remotely monitored) (CFOs, CIOs), medical imaging vendors and
beds. healthcare providers, plus review of medical
One-Time Costs: Based on several industry and imaging contracts (via FOIA from others’ legal
critical care medicine reports [162, 163], ini- cases).
tial software costs for these beds range from Implementation Costs: Implementation cost is
$2.5 million to $1 million each. We use $1.3 based on the facility and the types of equip-
million per bed. Multiplied by the number of ment used. In the US, there are approximately
ICU-T beds (98,401) = $127,921,300,000, 19,985 facilities: ~5,200 hospitals, 7,000 free
which amortized over 5 years yields standing medical clinics and 7,885 urgent care
$25,584,260,000 per year centers that provide imaging services. To
Central monitoring units cost $3 million each, focus on software implementation costs, we
and we calculate that one monitoring unit is must first disaggregate the cost of software
needed for every 25 beds. Therefore, with 1 from that of hardware and the intrinsic struc-
station for every 25 beds, or 1/25th of 98,401, tural costs of certain types of imaging modali-
is 3,936.02 X $3 million = $11,808,120,000, ties, such as facilities for magnetic resonance
which amortized over 5 years is $2,361,624,000 imaging (MRIs) centers.
26 R. Koppel

Based on the sources previously described, we high-capacity, low-cost, cloud-based storage

also used: data on industry growth rates [170], and high-bandwidth connections, most soft-
reports to CMS from providers, and industry ware and licensing costs are borne by institu-
reports to estimate medical imaging software tions for integrating PACS with EHRs and other
implementation costs at $60 billion, which, data capture/rendering devices throughout
amortized over 5 years, is $12 billion [171] facilities (and to individual offices) and manag-
Licensing Costs: The US Bureau of Labor ing user access.
Statistics indicates there are currently 36,134 Software costs include IT support, and ongo-
practicing US radiologists and 250,700 radiol- ing services (e.g., patches and updates) [174].
ogy and MRI technologists, for a total of Software licenses vary by size and function
286,834 “users” [172]. but the mean is $60,000 per hospital per year,
Unlike implementation costs, which are based on which we multiply by the number of US hospitals
the physical location and the equipment, fees (6,090, less 208 federal hospitals = 5,882). This
are often associated with the number of “users.” equals an annual US PACS software cost of
Note that “fees” here is an expansive term that $352,920,000
often includes maintenance, upgrades, many
repairs, and patches. These combined fees and 1.5.5.5 Medical Laboratory
services average $100,000 per “user,” with Management Systems (LIMS)
variations by type of equipment, e.g., a sono- Medical laboratories, either tethered to other
gram license fee and an MRI license fee differ medical institutions or separate enterprises,
widely. When multiplied by the number of require software to operate and bill. Required
“users” in those many settings, the total US functionality includes: sample management, con-
medical imaging licensing costs is 286,834 X nection to instruments and EHRs, results report-
$100,000 = $28.683 billion. ing and tracking, quality assurance/control and
quality control, workflow automation, regulatory
Adding Implementation and licensing costs, management/compliance, and invoicing, among
the total estimate for ongoing US medical imag- others. In 2021, the US listed 29,227 “Diagnostic
ing HIT software costs is $12 billion + $28.683 & Medical Laboratories Businesses” [175]. In
billion = $40,683,000,000 addition, laboratories must share and report
While seemingly large, software costs for results to local and state governments (health
imaging are only 1.0% of the US healthcare bill information exchange) [176].
in contrast to the 10% of all healthcare costs Laboratory information management software
that is the estimate of medical imaging ($400 (LIMS) configure and store data in 4 methods:
billion out of total US healthcare bill of $4
trillion. • Client servers (in house)
• Web-based
• Standalone
1.5.5.4 Medical Image Management-- • Thin-client servers (information is housed
Picture Archiving & elsewhere).
Communications Systems
(PACS) Two major vendors of LIMS that serve 30% of
PACS are high-capacity, high-speed hardware the industry are: Abbott Laboratories and Thermo
systems for handling medical imaging, and Fisher Scientific. The rest (70%) are served by
over 96% of hospitals have them either in house over 400 other software providers.
or as cloud services used [173] for multiple Reports on revenue and cost data are available
purposes, and no longer limited to imaging/ from industry publications and software compa-
radiology. Our focus is only on the software nies [177–183]. Based on these reports, and on
costs, not the hardware. With the advent of interviews with medical providers who use their
1 Estimating the United States’ Cost of Healthcare Information Technology 27

services, we find that US LIMS Costs for the such collateral damage to the reputation of
29,227 services are a mean of $15,742.29 each, healthcare institutions due to data breaches is
which yields an annual total of $460,100,000 per often considerable.
year. We also note that:

1.5.5.6 Medical Social Workers • Healthcare software and data are especially
In the US, there are 176,110 healthcare/medical vulnerable because of the many stakeholders
social workers, most of whom work in hospitals, and users that are involved, i.e., clinicians,
family service clinics, home healthcare, skilled clinical services laboratories, pharmacies,
nursing facilities, etc. It is estimated that only administrators, business associates for billing,
13,700 (or 7.7%) work in outpatient care centers insurance, etc. [192, 193]
[184], on which we base estimates for profes- • EHR and medical insurance claims data are
sional software costs. especially valuable to cyberthieves because
Licensing Costs: Based on quotes, literature and they contain protected health information
digital displays from vendors [185, 186], we (PHI), patient financial data (credit cards,
estimate the mean fees to be $68 per user per SSNs, insurance accounts), and personal and
month ($816 per user per year). Thus, for 13,700 private health data, and information that can be
medical social workers the total licensing costs used to fraudulently bill insurance companies
are estimated to be: $11,179,200 per year. and CMS.
Implementation Costs: These are minimal and
we estimate them as only $400 per user, which As a result, cybersecurity insurance premi-
amortized over 5 years is $80 per user per ums have increased significantly—both in
year. For 13,700 users, this yields annual esti- actual premium amounts and areas covered
mated total Implementation Costs to be: (e.g., more software under the insurance
$1,096,000 per year. umbrella and more expansive coverage (data
Literature and vendor quotes indicate that a few recovery, data breach, business interruption,
weeks is allocated for training and lost productiv- cyber extortion and 3rd party liability), result-
ity, but we do not include any cost for this. ing in general increases in cybersecurity bud-
Summing the annual licensing costs gets by about 53% [194].
($11,179,200) and amortized Implementation Premium Increases: Interviews with medi-
Costs ($1,096,000) for Medical Social Worker/ cal facility CIOs and review of insurance carri-
EHRs yields an estimate of $12,275,200. ers’ data [195] reveal cybersecurity premiums,
that have traditionally been about 4% of the
total institutional software costs, have dramati-
1.5.6 Cybersecurity Risk Insurance cally increased, with premiums in Q1-4 2020
Premiums and Q1-2 in 2021 now estimated at 6.5 to 7.5%.
We use 7% in our estimate [196]. Thus, we
The rise in ongoing ransomware attacks, data take the total software and implementation
breaches, and hacking has accelerated cyberse- costs of HIT (= $298,126,041,732 (see
curity (cybersec) to be a prime concern for Table 1.2, directly below) and add 7%, which
CIMOs, CIOs, and CISOs [187]. In this esti- is $20,868,822,921.
mate, we include only the HIT cybersecurity
software premiums. We do not include increases
in cybersecurity staff, processes or equipment- 1.6 The Final Tally and Estimate
-limiting our scope to only the software insur-
ance premium costs [188–191]. We do not, for Combining the total costs for the software and
example, include the “cost” of reputational implementations for each of the services outlined
damages, rebuilding databases, etc. However, above provides a grand total. We show this for the
28 R. Koppel

Table 1.3 Cost Estimate for HIT Software and Implementation

All data for 1 year. Percent of total with and without cybersecurity Insurance premiums
Cost Percent w/o ins Percent w/ ins
Hospitals $155,150,567,188 48.64% 52.04%
Medical Imaging software $40,683,000,000 12.75% 13.65%
Hospital ICU telemonitoring $35,325,959,000 11.07% 11.85%
Medical practices’ EHRs (PCPs MDs ODs) $23,107,380,000 7.24% 7.75%
Remote Pt Monitoring SW only $10,449,000,000 3.28% 3.50%
Pharmacies $7,863,000,000 2.46% 2.64%
Telehealth $4,785,140,500 1.50% 1.61%
Dental care $3,849,000,000 1.21% 1.29%
VHA/Cerner $3,785,700,000 1.19% 1.27%
Dialysis SW (separate facilities) $3,327,552,228 1.04% 1.12%
County and State Health Dept $2,398,500,000 0.75% 0.80%
DoD Cerner, Leidos EHR $1,650,000,000 0.52% 0.55%
Home Healthcare $1,293,105,000 0.41% 0.43%
Clinical Trial software $839,500,000 0.26% 0.28%
SNFs $497,024,640 0.16% 0.17%
Prison Health Services $496,800,000 0.16% 0.17%
Medical Laboratory Information Systems SW $460,100,000 0.14% 0.15%
Physical Therapy $382,333,360 0.12% 0.13%
PACs (med images) SW $352,920,000 0.11% 0.12%
Chiropractic $298,570,000 0.09% 0.10%
Optometrists $275,356,000 0.09% 0.09%
Pharmacy Benefits Managers $195,160,000 0.06% 0.07%
Hospice $159,060,000 0.05% 0.05%
Visiting nurses $153,456,000 0.05% 0.05%
Occupational Therapy $115,392,060 0.04% 0.04%
Podiatry SW $62,790,000 0.02% 0.02%
Emergency Medical Services (not part of fire depts) $47,661,000.00 0.01% 0.02%
Indian Health Service $38,226,920 0.01% 0.01%
Acupuncturists $36,532,636 0.01% 0.0%
Adult Day Care $34,980,000 0.01% 0.01%
Med Social Workers $12,275,200 0.00% 0.00%
Total Excluding Cybersecurity Premiums $298,126,041,732 93.46% 100%
Cybersecurity Premiums $20,868,822,921 6.54%
Grand total $318,994,864,653

sum of all HIT costs before the cybersecurity As a percent of total US healthcare spending
insurance premiums ($298,126,041,732), and of $4 trillion, the cost ($298,126,041,732)
then with cybersecurity insurance premiums before adding the insurance premiums) reveal
added. Table 1.3 illustrates the cost in descending that software and the many implementation
cost order; first column without cybersec premi- costs reflect 7.45% of the nation’s healthcare
ums and the second column with the premiums costs. When we add cybersecurity premiums,
added. (Supplemental Table 1.S1 shows the esti- the combined totals are 7.97% of total US
mates in alphabetic order.) healthcare costs.
As can be seen, after inclusion of cybersecu- The above table presents the data arrayed in
rity premiums, the combined cost of software and descending cost order and with percentages of
the cybersecurity premiums is $318,994,864,653 total for both with and without cybersecurity
per year. premium costs.
1 Estimating the United States’ Cost of Healthcare Information Technology 29

1.7 Conclusion has been regulatory: federal incentives for

CEHRT adoption and meaningful use
A common refrain in the medical and informatics linked to essential Medicare/Medicaid
literature is that healthcare is more about infor- payments and penalties for non-adoption.
mation than anything else. As such, it is not sur- A third driver is the current focus on
prising that US healthcare has become so value-based payments and pay-for-­
dependent on software to collect, share, access, performance, linked to aggregate elec-
process, and analyze that information. It is there- tronic clinical quality metrics (eCQM).
fore understandable that software has a signifi- Note, however, that many dispute the util-
cant and ongoing (annual) cost. Herein, we have ity of, and metrics used to account for,
attempted to estimate the cost of this critical ele- “value-based care.” Currently, also, the
ment. This cost has increased over time and is use of medical practice EHRs that are
under 7.5% of the total cost of US healthcare— linked to other health care entities (hospi-
both a staggering figure and yet quite modest tals, health information exchange, public
compared to its vital and encompassing role. health agencies) is viewed by some of the
We have based these estimates on available public as a sign of quality and as a mar-
data, augmented by interviews with providers, keting driver.
administrators, vendors and reports from state
and federal, industry and scholarly sources. Acknowledgements This effort was helped by several
These numbers are undoubtedly subject to error colleagues and friends. Catherine Mavrich did extraordi-
nary and successful work obtaining software costs by call-
and should be augmented by more complete ing hundreds of vendors, reading their materials, and
information from vendors, providers, consul- sitting through too many demonstrations (before they
tants, IT staff, and others. Also, we have consis- would provide more data and sales personnel with whom
tently underestimated the costs—usually because to talk). She also constructed and populated the original
data structure for collecting and organizing the cost data.
we decided not to include costs without hard The editor of this volume, Dr. George Kim (Johns
numbers. We also recognize this is a first effort at Hopkins), spent endless hours improving the text and the
a comprehensive estimate, and we await others to format of this work. He also served by questioning most
add to or to amend these figures. Be that as it of the numbers or requesting additional documentation.
He was invaluable in suggesting additional sources and
may, the numbers are clearly consequential, and people from whom I learned much about HIT cost struc-
further study is needed as the role of US HIT con- tures. George was, above all, a kind and encouraging
tinues to evolve and expand. editor.
We encourage others to build on and refine Many healthcare providers were invaluable sources of
information. In this category, Darren Lacey, the Chief
this work. Information Security Officer (CISO) at Johns Hopkins
was invaluable in helping me obtain cybersecurity insur-
ance costs and the increasingly shifting scope of cyberse-
Question and Answer curity insurance coverage.
Dr. Joel L. Telles, a wonderful friend and colleague,
1. What are primary drivers for the increased helped mightily with some of the calculations and spread-
adoption and widespread use of certified elec- sheets. Joel had previously been a key player in analyzing
tronic health record technologies (CEHRTs) data for hospital systems and hospital associations.
in US healthcare? My colleague and friend, Dr. Stephen Soumerai of
Harvard, was always urging me to include additional and
(a) One driver is the call to improve the qual- important costs. Because this document is focused only
ity of healthcare (safety, effectiveness, on the software and implementation costs—and does not
patient-centeredness, equity, timeliness, engage in the debates about HIT’s negative aspects--
efficiency) through continuous perfor- Steve’s advice was confined to the list of items that I
excluded from the calculations. But they are serious issues
mance measurement through HIT, inno- that should be examined in later analyses.
vation in care, such as patient-centered I’m not going to name the many HIT vendor sales-
medical homes (PCMH). Another driver people who spent hours with me providing numbers and
30 R. Koppel

cost parameters. But I’d like to thank them very much. and recommendations. Appl Clin Informatics.
Their information was invaluable, even if sometimes they 2020;11(5):742–54.
assumed I was a potential customer. Also, I’m not going to 9. Koppel R, Lehmann CU. Implications of an emerg-
name the leaders and others of the many associations ing EHR monoculture for hospitals and healthcare
(e.g., pharmacy associations, hospice associations, physi- systems: an analysis of Epic’s market power. J Am
cal therapy associations, to name only a very few) because Medical Inform. 2015;22:465–71.
they were sometimes concerned about what was permitted 10. Koppel R. Keynote chapter: is HIT evidenced-­
to discuss about these complex money matters. But they based? International medical informatics yearbook.
were very helpful and even led me to others who were also Stuttgart: Schattauer Publisher; 2013.
knowledgeable. 11. Nelson H Can low EHR use time, clinician demand
My son, Jonah Koppel, was a wiz with formatting predict physician departure? EHR Intelligence
issues and with helping with the over 200 references. 10/14/21
Other friends and colleagues from Penn, from The 12. Lee BY. How doctors may be spending more time
University at Buffalo and from AMIA have been very with electronic health records than patients. Forbes.
encouraging about this effort, and often offered sources 2020; Jan
and insights. I thank them all. 13. American Bar Association. Health law section.
Any errors, of course, are my responsibility. Available from: https://www.americanbar.org/
groups/health_law/ (accessed 11/24/21)
14. Huang C, Koppel R, McGreevey JD III, Craven C,
Schreiber R. Transitions from one electronic health
record to another: challenges, pitfalls, and recom-
References mendations. Appl Clin Informatics. 2019;
15. US Department of Health and Human Services.
1. Koppel R. Great promises of healthcare information Information Technology Agency Summary https://
technology deliver less. In: Weaver CA, et al., edi- itdashboard.gov/drupal/summary/009 (accessed
tors. Healthcare information management systems: 11/23/21)
cases, strategies, and solutions, health informatics. 16. National Association of State Budget Offices.
Switzerland: Springer International Publishing; 2020 State expenditure report. Available from:
2016. https://higherlogicdownload.s3.amazonaws.com/
2. Koppel R, Gordon S. First, do less harm: confronting NASBO/9d2d2db1-­c943-­4f1b-­b750-­0fca152d64c2/
the inconvenient problems of patient safety. Ithaca, UploadedImages/SER%20Archive/2020_State_
NY: Cornell University Press; 2012. Expenditure_Report_S.pdf (accessed Nov 27, 2021)
3. Moy AJ, Schwartz JM, Chen R, Sadri S, Lucas E, 17. Trust for America’s Health. Investing in America’s
Cato KD, et al. Measurement of clinical documenta- Health: a state-by-state look at public health funding
tion burden among physicians and nurses using elec- and key health facts. Available from: https://www.
tronic health records: a scoping review. J Am Med tfah.org/report-­details/investing-­in-­americas-­health-­
Inform Assoc. 2021;28(6):998–1008. a-­state-­by-­state-­look-­at-­public-­health-­funding-­and-­
4. Fry E, Schulte F. Death by a thousand clicks: where key-­health-­facts/ (accessed Nov 24, 2021)
electronic health records went wrong. Fortune 18. Weins K, IT Spending by Industry. Flexera [inter-
Magazine. 2019. net]. Mar 5, 2020. Available from: 2020 https://www.
5. Dymek C, Kim B, Melton GB, Payne TH, Singh H, flexera.com/blog/technology-­value-­optimization/it-­
Hsiao C. Building the evidence-base to reduce elec- spending-­by-­industry/ (accessed Nov 27, 2021)
tronic health record–related clinician burden. J Am 19. Kaelber D The journey to enterprise imag-
Med Inform Assoc. 2021;28(5):1057–61. https://doi. ing: Best practices and lessons learned from
org/10.1093/jamia/ocaa238. (accessed 12/1/2021) MetroHealth's CMIO Becker’s Hospital
6. Karsh BT, Weinger MB, Abbott PA, Wears Review 2021 June 23 Available from: https://
RL. Health information technology: fallacies go.beckershospitalreview.com/the-­j ourney-­t o-­
and sober realities. J Am Med Inform Assoc. enterprise-­i maging-­b est-­p ractices-­a nd-­l essons-­
2010;17(6):617–23. from-­metrohealths-­cmio?utm_campaign=Hyland_
7. Amato MG, Salazar A, Hickman TT, Quist AJ, Volk Webinar_6.23.2021&utm_source=website&utm_
LA, Wright A, McEvoy D, Galanter WL, Koppel R, content=featured (accessed Nov 27, 2021)
Loudin B, Adelman J, McGreevey JD 3rd, Smith 20. US Department of Transportation. Traffic Safety
DH, Bates DW, Schiff GD. Computerized prescriber Facts EMS Research Note. June 2014 Available
order entry-related patient safety reports: analysis of From: https://www.ems.gov/pdf/National_EMS_
2522 medication errors. J Am Med Inform Assoc. Assessment_Demographics_2011.pdf (accessed
2017;24(2):316–22. Nov 17, 2021)
8. Huang C, Koppel R, McGreevey JD III, Craven 21. Capterra. EMS Software. Available from: https://
C, Schreiber R. Transitions from one electronic www.capterra.com/ems-­software/?pricing_options=
health record to another: challenges, pitfalls, %5B2%5D&sortOrder=sponsored
1 Estimating the United States’ Cost of Healthcare Information Technology 31

22. ESO Resource Library. www.eso. 33. McCullough JM, Leider JP, Riley WJ. Local fis-
com. Available from: https://www.eso. cal allocation for public health departments. Am
com/resources/#category=Industry%20 J Prevent Med. 2015;49(6):921–9. Available
I n s i g h t s & r e s o u r c e _ t y p e = P r o d u c t % 2 0 O n e -­ from: https://www.ajpmonline.org/article/S0749-­
Pa g e r & r e s o u r c e _ p r o d u c t = E H R % 2 0 ( e P C R ) 3797(15)00205-­6/fulltext. https://doi.org/10.1016/j.
(accessed Nov 17, 2021) amepre.2015.04.021.
23. ESO Resource Library. www.eso.com. Available 34. Hyde JK, Shortell SM. The structure and orga-
from: https://www.eso.com/ems/ https://www. nization of local and state public health agencies
eso.com/resources/#category=Industry%20 in the U.S.: a systematic review. Am J Prev Med.
I n s i g h t s & r e s o u r c e _ t y p e = P r o d u c t % 2 0 O n e -­ 2012;42(Suppl. 5):S29–41. https://doi.org/10.1016/j.
Pa g e r & r e s o u r c e _ p r o d u c t = E H R % 2 0 ( e P C R ) amepre.2012.01.021.
(accessed October 13, 2021) 35. GovPIlot Health Inspection Software https://www.
24. Defense Health Agency, October, 2020 Fact sheet, govpilot.com/health-­department-­software https://
composite health care system order entry and results www.govpilot.com/health-­d epartment-­s oftware
retrieval system. Available from: https://www.health. (accessed Dec 1, 2021)
mil/Reference-­C enter/Fact-­S heets/2020/10/26/ 36. National Association of Counties, Advocacy
Composite-­Health-­Care-­System (Accessed 2020 Center https://www.naco.org/resources/
Sept 21) featured/2020-­advocacy
25. Serbu J. DoD: Price tag for new electronic health 37. Mitchell H. Former VA IT exec: 6 insights on VA's
record might grow by $1.2 billion. Federal News $16B Cerner EHR [internet] Becker's Health IT. July
Network. July 25, 2018 Available from: https:// 27th, 2021 Available from: https://www.beckershos-
federalnewsnetwork.com/defense-­n ews/2018/07/ pitalreview.com/ehrs/viewpoint-­6-­reasons-­the-­va-­
dod-­price-­tag-­for-­new-­electronic-­health-­record-­might-­ s-­16b-­cerner-­ehr-­rollout-­is-­failing.html (Accessed
grow-­by-­1-­2-­billion/ (Accessed 2021 August 21) 2021 Nov 11)
26. Jercich K. Department of Defense expands Cerner 38. Haxel C. Inspector general finds fault in rollout of
EHR implementation to 10,000 more provid- Cerner's $10 billion contract for VA health records.
ers Healthcare IT News. April 28,2021 Available KCUR [internet]. July 12, 2021. Available from:
from: https://www.healthcareitnews.com/ https://www.kcur.org/news/2021-­07-­12/inspector-­
news/department-­d efense-­e xpands-­c erner-­e hr-­ general-­finds-­fault-­in-­rollout-­of-­cerners-­10-­billion-­
implementation-­10000-­more-­providers (Accessed contract-­for-­va-­health-­records (Accessed 2021 Nov
2021 Jun 22) 11)
27. Martin N. Cerner to Continue VA EHR System 39. Cohen JK. VA underreported EHR project costs by
Deployment Under $134M Task Order GovCon $2.5B, OIG finds. Modern Healthcare. July 09, 2021
Wire September 10,2021 Available from: https:// Available from: https://www.modernhealthcare.
www.govconwire.com/2021/09/cerner-­awarded-­ com/information-­technology/va-­underreported-­ehr-­
113m-­t o-­c ontinue-­v a-­e hr-­s ystem-­d eployment project-­costs-­25b-­oig-­finds Accessed 2021 Oct 3)
(Accessed 2021 Dec 4) 40. Landi H Tech VA underestimated infrastructure
28. Indian Health Service. Operating Plan for FY 2021 costs that could raise EHR price tag by up to $5B:
Available from: https://www.ihs.gov/sites/bud- OIG 2021 July 10. Available from: https://www.
getformulation/themes/responsive2017/display_ fiercehealthcare.com/tech/va-­u nderestimated-­
objects/documents/FY2021_OperatingPlan.pdf physical-­i t-­i nfrastructure-­c osts-­c ould-­r aise-­e hr-­
(Accessed 2021 Nov 12) price-­tag-­by-­up-­to-­5billion-­oig (Accessed 2021 Nov
29. Allen SA, Wakeman S, Cohen R, Rich JD. Int J 12)
Prison Health. 2010;6(3):100–6. 41. US Government. VA EHR modernization fact sheet.
30. Department of Nursing Nursing USC Blog. What to Available from: https://www.ehrm.va.gov/resources/
expect as a correctional care nurse and how to avoid factsheet (Accessed 2021 Nov 12)
burnout in challenging settings. Available from: 42. Meagher E There’s still a chance for VA to save
https://nursing.usc.edu/blog/correctional-­n urse-­ its electronic health record from looming disaster
career/ (Accessed 2021 Nov 30) Federal Healthcare News 2021 April 5 Available
31. Careers in Public Health. Net correctional medicine from: [https://federalnewsnetwork.com/commen-
physician. Available from: https://www.careersin- tary/2021/04/theres-­still-­a-­chance-­for-­va-­to-­save-­
publichealth.net/careers/correctional-­m edicine-­ its-­electronic-­health-­record-­from-­looming-­disaster/
physician/ (Accessed 11/27/21) (Accessed 2021 Nov 1)
32. Fusion. Fusion EHR for correctional facili- 43. Phillips D Hidden Costs and Flawed Training Plague
ties. Available from: https://fusionehr.com/? the V.A.'s Huge Software Upgrade. A $16 billion
g c l i d = C j w K C A i A s 9 2 M B h A X E i w A X Ti 2 5 effort to modernize health records at the Department
7 AG b RW q 3 b j q U 3 5 c _ 9 2 O a I F 6 O 5 u Q Z c v d _ of Veterans Affairs ran into major problems in its first
w37pMqlCVgQv_5U8fdAjhoC8vsQAvD_BwE installation, two watchdog reports say. NYTimes
(Accessed 11/27/21) July 9, 2021 (Accessed 2021 Aug 12)
Random documents with unrelated
content Scribd suggests to you:
 “I hope Cousin Elizabeth will lend you to us again,” Cousin Honoria
said, and Cousin Augusta added that it was wonderful how a young
person brightened up a house.
 CHAPTER XX

CONCERNING THE SARGENT

To go into a thing half-heartedly was not Blue Bonnet’s fashion;

before she was half-way to Woodford she was deep in plans for her
paper. It should not be hard, just to tell the story of The Alamo, as
her father used to tell it to her. She must find out about that
Woodford man, but there were any amount of old record books at
the Woodford Library; Alec had shown them to her one afternoon,—
she had thought them very dull-looking.

No one else would have thought of this subject; and she would
say nothing about it to anyone—not even at home—until her paper
was finished. Then Grandmother should be allowed to see it before it
was handed in.

It was mighty good of her and Aunt Lucinda not to have bothered
her about it; perhaps—Blue Bonnet straightened herself at the
thought—they had not considered it worth while,—had been sure
that in spite of her protestations she would come around in the end.

“They came near being disappointed,” she said to herself; “if

Cousin Tracy hadn’t given me such a good subject, I shouldn’t be
going to try.”

Alec was waiting when the train drew into the Woodford station; “I
thought Bruce and the cart would make better time than Peter and
the phaeton,” he explained. “You don’t want to start the week being
late to school, I suppose? So they did get you off in time?”
 “They didn’t have to ‘get’ me; I met all their efforts more than
half-way. I’ve had a beautiful time—and I hope Woodford’s missed
me a little bit?”

“Some of it has. Mind you don’t go and do it again.”

“I may not get the opportunity.”

Alec was not the only one glad to see her; as for Solomon, he was
all over her, before she was well out of the cart. There was only time
to kiss Grandmother and Aunt Lucinda, before snatching up her
school-books.

“Well!” Kitty demanded, waiting for her at the parsonage gate with
Sarah; “I hope you’re glad to get back.”

“Even if I were not, I hope I am too polite to say so,” Blue Bonnet
laughed, falling into step. Going to and coming from school was fun;
it was the staying there that was apt to prove irksome.

She did not go directly home from school that afternoon; instead,
she turned off in the direction of the Library, standing well back from
the street in its own square of green. It had been easy to put Sarah
and Amanda off; the rest of the club were busy “making up” these
afternoons. It seemed to Blue Bonnet, that, on the whole, it was
Miss Fellows who was paying the penalty for the fourteen’s act of
insubordination.

Once at the Library, Blue Bonnet hurried to the little room at one
side, devoted to the books concerning local history. There was no
one else there, though the reading-room was filling fast with pupils
on Sargent thoughts intent. Standing before the rows of musty-
looking old volumes, Blue Bonnet gave an impatient thought to the
originator of so much trouble. It was positively wicked to waste such
a glorious Spring afternoon indoors. Perhaps, if she hurried there
would still be time for a ride.
 Blue Bonnet found that it was not going to be as easy to keep her
secret as she had thought, neither at home nor at school. Some of
the fourteen had already been granted the longed-for permission,
and on the big board up at the front of the assembly-room, the list
of papers turned in—including titles and names of competitors—was
lengthening daily.

“I think,” Blue Bonnet confided one afternoon to Chula, as they

started briskly off down the drive, “that I’ll begin to write mine on
Saturday morning; I’ve got all the dates and details about ready.”

At the sound of quick steps behind her, she looked around. “Two is
company, you know,” Boyd said, riding up beside her; “I hope you
are in a mood for company—present company, at that.”

“Then you don’t call a horse and dog company?”

“Do you?”

“Certainly, and very good company.” Blue Bonnet leaned forward

to pat Victor; they had become good friends since that ride together
last October. “You’ve been riding Victor too hard—again,” she added,
with sudden severity.

“Victor has been spoiled ridiculously. He and I have been having a

bit of an argument.”

Blue Bonnet’s eyes flashed; “He is not spoiled; but he is used to

his owner.”

“He will get used to me—after a while; he’s been learning a thing
or two lately.”

By way of answer, Blue Bonnet wheeled Chula around towards

home. She knew now why she had not liked Boyd Trent; underneath
that smiling, easy politeness were selfishness and cruelty.
 Boyd turned too; she was a queer girl, but she was interesting,—
which was more than could be said for some of her friends,—and
she rode well. “Are you always so extremely sociable?” he asked.

Blue Bonnet flushed; Aunt Lucinda would say that she had been
showing her dislike too plainly. “I was thinking of—something,” she
said; “I suppose you are looking forward to summer?” After all, he
was even more of a newcomer in Woodford than she was, and he
hadn’t half as many friends; even if one were horrid, one might have
feelings like other people.

“Well, rather!” Boyd laughed; “I’ve seen livelier spots.”

“Don’t you like it at the academy?”

“Slow like all the rest of the place.” He pulled out a note-book; “I’ll
show you some snap-shots of my school at home.”

Blue Bonnet brought Chula nearer; the snap-shots though small

were clear, and the bits of school-life they gave interested her. She
decided that she would like a camera; she would like some
Woodford views to take back to the ranch.

“Did you take these?” she asked.

“Yes,” Boyd answered. “I’ll overhaul the camera, and we’ll go

picture-hunting some Saturday morning.” He was returning the views
to his note-book, and, as he spoke, some papers fell from it to the
ground.

“One would think you were taking notes for a book—” Blue Bonnet
began, then she stopped. They were notes, and they were all in
Alec’s handwriting.

Boyd had slipped down from his horse, and was gathering the
slips of paper up hurriedly; he looked confused, Blue Bonnet
thought.
 The little incident came back to her the next morning, as Kitty
drew her to a standstill before the bulletin board in the assembly-
room. “Three more names,” Kitty commented; “they’re coming in
fast. Why, there’s Boyd Trent’s. I didn’t know he meant to try; it not
being the regulation thing, apparently, for outsiders to do.”

Blue Bonnet let the little dig pass; she was bending to read the
title of Boyd’s paper—“The After Stories of Some Sargent Winners.”
Suddenly, Blue Bonnet saw again the little pile of papers lying in the
dusty road, and Boyd’s face as he bent to pick them up.

“What’s the matter?” Kitty asked; “Are you beginning to repent?

It’s not too late even yet! Billy’s still on the tenterhooks,—I think Mr.
Hunt might temper judgment with mercy a little more quickly,—and
if there’s time for Billy Slade to get up a paper, there’s time enough
for you. Nothing happening, you’ll be reading Katherine Clark’s name
there before many days.”

“Come on!” Blue Bonnet said. “No, I’m not beginning to repent;
I’ve always understood that it was a very uncomfortable process to
go through with.” Her thoughts were in a whirl. Had Boyd really
taken Alec’s—She couldn’t think that.

She thought about it all during opening exercises; also, all through
the Latin recitation afterwards, with the result that she failed twice
on questions that she knew quite as well as the girl next her who
answered them so glibly.

“So like the dear old days!” Kitty murmured provokingly; and Blue
Bonnet decided to put the matter out of her thoughts until after
school. Just what she intended to do then, was not clear to her; she
could hardly go to Boyd and accuse him of—that.

She wouldn’t ride that afternoon; Boyd would probably have Victor
—she wished General Trent knew how seldom Alec had the use of
his own horse nowadays; she and Alec would go for a walk, and—
 “Elizabeth!” Miss Fellows said, “I am afraid that you are not
attending to the matter in hand.”

“But I’m going to, really and truly!” Blue Bonnet promised, with an
earnestness not all for Miss Fellows. “Mind you do,” she told herself,
“or there won’t be any time for walking this afternoon.”

“No, I can’t go home with you!” she assured Kitty after school. “I
can’t go home with any of you girls! Yes, there is something on,
Little Miss Why; but I am not going to tell you what it is.”

Kitty looked impatient; “You’re the greatest girl for wrapping

yourself up in mysteries!”

“I’m not!” Blue Bonnet answered; “but little girls mustn’t ask
impertinent questions; good-bye, I’ll see you to-morrow morning.”

“Or before—perhaps,” Kitty retorted. “As I take the notion.”

Blue Bonnet found Alec reading on the side piazza; he was looking
troubled about something, she told herself. “If you don’t mind, I
would like to follow our brook this afternoon,” she said.

“And I am to follow you?”

“It would be more sociable if we kept together.”

They went out across the back meadow, the dogs leaping and
barking on ahead, just as they had that August afternoon. A good
deal had happened in the eight months since, Blue Bonnet thought;
it did not seem as if any other eight months could ever bring so
many new experiences; she felt considerably more than eight
months older.

“What are you looking so sober over?” Alec asked.

“A great many things.”

 They had reached the brook, and turning they followed it back
along the way it had come until the woods were reached; here they
went more slowly. The April woods were too lovely to be hurried
through, Blue Bonnet thought, with the light falling soft and
shimmering through the young green of the trees, and the Spring
beauties making a delicate border for the brook, which laughed and
splashed over the stones, as if it knew that at last the long winter
were gone for good.

“Let’s go up to our old picnic place,” Blue Bonnet suggested, and

they came at last to the open space where they had lunched that
afternoon, with, it would seem, the very same squirrel eying them
askance from the upper bough of a tall tree.

“Isn’t it nice here!” Blue Bonnet leaned back against the moss-
covered trunk of an old tree. “Why couldn’t we come out here for
school! It would be much more sensible!”

“From your point of view!”

Blue Bonnet passed a hand lovingly over the pink and white
beauties which seemed to be smiling up at her. “And isn’t it good
that at last all the fourteen can try for the Sargent? Billy got his
discharge papers this noon.”

“I thought Mr. Hunt would prove amenable.”

“How soon do you send your paper in?” Blue Bonnet was picking a
knot of the flowers for her blouse and did not look up; she hoped
her question sounded sufficiently casual.

“I—oh, I’ve decided to follow your example.”

“You mean you’ve given up trying?”

“Sounds that way, doesn’t it?” Alec was looking straight ahead of
him; there was a little pucker between his brows.
 Blue Bonnet seemed for the moment to be giving her attention to
her flowers. It was just as she had expected; by some means,
evidently not fair ones, Boyd must have secured Alec’s notes and
used them. Of course she had not liked him—he was selfish and
cruel and mean! And she would have to pretend not to know, unless
Alec made some sign, which he would not—she wasn’t good at
pretending.

“‘BUT I THOUGHT,’ SHE SAID, ‘THAT IT WAS A GIRL’S

PRIVILEGE TO CHANGE HER MIND?’”
 “But I thought,” she said, “that it was a girl’s privilege to change
her mind?”

“Mayn’t we borrow one of your privileges occasionally? You borrow

some of ours. Besides, I won a prize last year—suppose I should do
it again, wouldn’t too much glory be bad for a fellow?”

“Aunt Lucinda won it three times running when she was a girl.”

“Yes, but she was—Miss Lucinda! Come to think of it, my lady, you
are not precisely in a position to lecture me for not trying.”

“But I—” Blue Bonnet caught herself up; “I don’t want to lecture
anyone—to-day,” she ended, and leaning back again she looked
thoughtfully up at the soft stretch of blue showing between the tree
tops.

She wished Alec would up and fight Boyd on his own ground! But
then, Boyd had stolen his ammunition. Good subjects for the
Sargent were not lying around waiting to be picked up; no wonder,
when one remembered all the papers that had been written since
the originating of the competition.

Blue Bonnet caught her breath; suppose—

But he would not take her subject. Very well, he would have to be
managed. She could not help feeling a very real sense of regret. She
had meant to begin writing her paper to-morrow morning; she had
become honestly interested in the doing of it, and she was looking
forward to Grandmother’s and Aunt Lucinda’s surprise and pleasure
when she told them. As for the girls—

Fortunately, she had said nothing about it. There would not be
time to hunt up another subject; besides, she didn’t want any other,
she knew how Alec felt about that; still, she was offering him a really
new idea. It was the manner of offering it that was troubling her
now.
 “We aren’t very talkative, are we?” she said.

“We don’t seem to be,” Alec agreed.

“Shall I tell you about Cousin Tracy’s medals? He has a fine

collection;” and presently she had him interested in the short
accounts Mr. Winthrop had given her, introducing—much as he had
done—the subject of the Alamo, and the fact that the father of one
of its heroes had been a Woodford man.

“I never knew that,” Alec said.

“I’m glad, somehow,—so long as I belong to both places,—that

Woodford can claim a share in the Alamo.” And Blue Bonnet went on
to tell the story as her father used to tell it to her; seeing, and
making Alec see the tragic drama enacted there in that little church
near San Antonio during those memorable three weeks; the
struggle, the heroic courage, the no less heroic endurance of the
men, who, like the Old Guard, could die, but would not surrender.

“I don’t wonder your Texans took ‘Remember the Alamo’ for their
war-cry afterwards!” Alec said. There was an eager light in the boy’s
gray eyes; he had not come of a race of soldiers for nothing.

He was not much more talkative going home than he had been
coming, but from a different reason, Blue Bonnet felt sure; and she
lingered a moment on the porch, watching him cross the lawn after
saying good night. “Will he, or won’t he, Solomon?” she asked.

As she came up the drive the next afternoon, after her ride with
the club, Alec came to meet her. “See here,” he said, stroking the
head Chula stretched towards him, “I’ve been thinking—”

“Did it come hard?” Blue Bonnet laughed.

“I’ll settle that score later! We’ll stick to business now, if you
please. My New England thrift makes me hate to see good material
going to waste.”

“He will do it!” Blue Bonnet told herself. “Then why not prevent
it?” she asked.

“Don’t you feel an inner call to turn that Alamo business into a
Sargent?”

Blue Bonnet stroked Chula’s mane thoughtfully; “No,” she

answered, “I don’t think I do;” and to herself, she added, that she
didn’t—now.

“I’ve a notion that if you don’t do something of the sort your

Woodford relatives will be a bit disappointed.”

“They might be more disappointed if I did.”

“Then you are quite sure?”

“Perfectly.”

“In that case—it’s such splendid material, I really don’t see how
you have strength to let it alone—I believe I’ll change my mind a
second time.”

“You may; only don’t get into the habit—and change it again,”
Blue Bonnet warned.

“I won’t,” Alec promised; “I’m going straight to work. I’m no end

obliged to you for telling that story; it’s the best subject ever.”

Spring came early that year, and no one rejoiced more in its
coming than Blue Bonnet. Now that the winter was over, she began
to realize how long it had seemed; and, as the days went by, Miss
Fellows began to realize with equal vividness something of what Miss
Rankin had gone through with last fall.

There was no wilful breaking of rules, Blue Bonnet had not

forgotten her promise, but there was much inward rebellion and
outward struggle, resulting in more or less inattention during school
hours. Blue Bonnet’s eyes would wander again and again to the
window, her thoughts drifting even further afield. The remembrance
of what the ranch must be like now grew daily more insistent.

The long rides and walks after school, the hunts for wild flowers,
the tennis which, with the coming of Spring, the Woodford young
people had promptly instituted, helped a good deal.

By the fifteenth of May, all of the papers for the Sargent had to be
in.

“And to-morrow is the fifteenth!” Blue Bonnet rejoiced one

afternoon. “Now, perhaps, the old thing can drop!”

“Ah, but the waiting will begin now,” Ruth said.

“Can’t you wait in silence?”

“You’re a very disrespectful girl!” Debby said severely.

Blue Bonnet smiled agreeingly; “I have learned a lot of things

since I came East, haven’t I?”

The “We are Seven’s” were sitting under the trees in Kitty’s front
yard, resting after a long walk. “I’m going to have a birthday next
Saturday week,” Amanda announced.

“Is there to be a celebration?” Kitty inquired.

Amanda nodded importantly.

 “Of course there is, little Miss Why!” Debby said. “There’s some
use in having a birthday in Woodford. If you were wise, Blue Bonnet,
you’d arrange to have yours while you were here—there would be
something doing then.”

“In August I’ll be on the ranch—and there’ll be something doing

there. There’s some good in having a birthday on the Blue Bonnet
Ranch.”

“Aunt Huldah”—Amanda looked still more important—“says I may

bring a party out there for supper and—”

Kitty came nearer; “‘Codlin’s your friend!’ And look here,” she
turned to the others, “we’ll appoint a body-guard right now to see
that Blue Bonnet doesn’t pay any visits to the Poor Farm between
now and a week from Saturday.”

“I’ve never been there but that once!” Blue Bonnet protested.

“That’s not saying you wouldn’t go again if the fancy seized you,”
Kitty rejoined.

“I wish you would listen,” Amanda objected; “I thought I’d ask you
girls—”

“If you didn’t some of us would be asking the reason why,” Debby
interposed.

“And the boys who were at the ‘skating-rink party’ that day. I
couldn’t take any larger party than that.”

“Making it Gentlemen’s Day?” Blue Bonnet asked.

“Uncle Dave’s just finished building a new barn,” Amanda went on.

Kitty clapped her hands—“And we’re to dance in it after supper!

Oh, what fun!”
 “It’ll be moonlight coming home, I looked it up in the almanac.”
Amanda leaned back with a sigh of satisfaction.

“Amanda Parker, you’re the sensiblest girl!” Kitty declared. “Now I

don’t believe Blue Bonnet or I would ever have thought of providing
a full moon too. Sarah might’ve.”

Blue Bonnet carried her good news home. “And I may go this
time?” she said. “I won’t ask anybody to tea for that night. I’d just
love to see a real farm. I suppose it’s what Uncle Joe would call a
‘juvenile ranch.’ Twelve days is going to be an awful long while to
wait.”

“A what, my dear?” Aunt Lucinda suggested.

“Very—spelled like—awful,” Blue Bonnet laughed.

“The days are going pretty fast the past weeks,” Grandmother
said, thinking sadly that already May was half gone and that June
would soon be here; even now, Mr. Ashe was writing of coming East
for Blue Bonnet. The summer seemed to stretch ahead, unusually
long and quiet; and who knew what the fall would bring forth? Blue
Bonnet had not said as much lately about coming back; and once Mr.
Ashe had her safely on the ranch, would he be willing to part with
her again?

Grandmother roused herself; at least, Blue Bonnet had not gone

yet. Looking up, she found Blue Bonnet watching her rather soberly;
and presently, when supper was over, the latter ran hastily upstairs
to her own room.

“I’ve the best plan ever, Solomon!” she confided to him, as he

danced on before her. Five minutes later, she was down again. “I’m
going to the office to mail a letter,” she announced from the sitting-
room doorway; “I won’t be gone long.”
 Those twelve days were not so long in passing. That all of the
invitations should have been promptly accepted was only to be
expected.

“It’s about the only thorough-going jollification we’ll have time for
between now and closing of school,” Debby told Blue Bonnet; “the
exams will be beginning soon.”

“And we’ll have all last winter’s agony to go through with again?”

“That depends upon how easily you agonize.”

“I’m not quite so scared as I was then,” Blue Bonnet said; “I

wonder if one would ever get where an exam didn’t really bother
one at all?”

“I’m not wasting my time over any such nonsense,” Kitty declared;
“I’m wondering why the wagon doesn’t come.”

The party were waiting on the Parker front steps for the big hay
wagon from the farm; the girls, in their fresh summer dresses,
making a bright spot of color against the green background of the
vine covering the piazza.

“Here it comes!” one of the boys said.

Billy had provided himself with a horn, a battered old affair which
had seen much service but was still capable of more, as Billy
proceeded to prove, waking the echoes of the quiet old street.

“Billy!” Mrs. Parker implored, coming out, “you’re not going to take
that thing?”

“I am surprised at you!” Billy eyed her reproachfully. “Don’t I

always take it?”
 “We won’t let him blow it too often,” Alec promised; “if he tries to,
we’ll drop him and it overboard.”

“Isn’t living in a village ever and ever so much more fun than
living on a ranch?” Kitty demanded of Blue Bonnet as the wagon
started.

“Tell her ‘no,’” Alec said.

“Tell her comparisons are odious,” another of the boys suggested.

“Tell me to come and see,” Billy urged.

And suddenly Blue Bonnet found herself wishing that it were

possible to take all the “We are Seven’s” and some of their friends
back to Texas with her. Would they find the life there as strange and
as confusing as she had found it here? At least, there would be no
school; just long happy care-free days to be spent out-of-doors. She
would like Uncle Joe Terry to know Kitty—she could see the twinkle
in his shrewd kindly eyes as he looked down into the freckled,
piquant little face; she would like him to know Sarah, too, and all the
girls, and Alec. And she would like them all to know Uncle Joe. So
long as there were no fences making choice of side imperative, even
Amanda was good fun; besides, she was a club member.

But of course, it was not to be thought of.

“If I were the ‘rankin’ officer,’” Kitty announced, “I should be

calling you to attention just about now, Blue Bonnet Ashe. You are
the unhearingest girl that ever was!”

“But you’re not, you know,” Blue Bonnet answered; “and I was
thinking of something.”

“You mostly are—when you shouldn’t be; and mostly aren’t when
you should be,” Kitty observed.
 “The ‘rankin’ officer’ is a part of the past, so far as we are
concerned,” Debby said comfortably.

“And so will the ‘jolly good’ be soon,” Billy said.

“And will you tell me,” Kitty looked from one to another, as if the
question were a momentous one, “what we are going to do next
term with a teacher named Kent!”

“You haven’t got her yet,” one of the boys reminded her. “‘There’s
many a slip ’twixt the cup and the lip.’”

“‘Spell it with a we, my lord, spell it with a we,’” Alec quoted.

“And have her Vent it all on us?” Ruth laughed.

“Somebody kindly head Sarah off! She’s getting ready to

remonstrate!” Kitty added.

“I see the new barn!” Susy called; “I guess you’re glad we’re
nearly there.” She looked up at Mrs. Parker, in the seat of honor
beside the driver.

“I’ve chaperoned you young people before,” Mrs. Parker answered,

—a remark, which, as Alec said, could be construed in more than
one way.

“Choose your partners,” Billy called; “it’ll save time afterwards.”

They were within sight of the low, stone farmhouse by now; from
the front porch, Amanda’s Aunt Huldah was waving a welcome to
them.

Boyd gave Billy a sudden shove into the road, slipping into his
place beside Blue Bonnet. “May I have the first dance?” he asked.

“It’s promised,” she answered; Alec had seen to that the night
before.
 “Well, I like that!” Billy stood staring after the wagon. “A nice way
to treat a fellow.”

“He thought you needed exercise, Billy,” Kitty called.

“Then, the second?” Boyd asked; she had seemed to avoid him
whenever possible lately,—he half wanted to find out why; and
outside of that, she was the best dancer there.

The wagon was stopping, but Blue Bonnet did not appear to have
noticed; she was looking off down the road they had come by, a
doubtful expression in her blue eyes; then she turned, meeting
Boyd’s glance fully, “I’ll give you the next to the last.”

“The next to the last!” She was a queer girl.

“Come on, Blue Bonnet!” Amanda called; “I want to introduce you

to Aunt Huldah—you and Boyd too.”

“I’m coming!” Blue Bonnet did not seem to see the helping hand
Boyd held out.

As she went up the steps with the other girls, he stood a moment
looking after her. He was not so sure now that he did want to find
out why she had—she had some nonsense in her mind. It couldn’t
be about—

With a little shake of the shoulders, Boyd followed the rest.

 CHAPTER XXI

THE END OF THE TERM

Boyd was in two minds about claiming that dance—it wouldn’t do

the little Texan any harm to be called down; but when the time
came, he presented himself before Blue Bonnet, outwardly as
smiling as usual.

“Would you mind if we sat it out?” she asked.

Boyd looked his surprise; she had not been sitting out any of the
other dances, and again that uneasy feeling came over him. “As you
like, of course,” he answered, leading the way to the old bench
under a big apple tree just outside.

“I wanted to tell you,” Blue Bonnet began at once,—“I’ve thought

it all over, and it doesn’t seem fair not to tell you—that I know about
—”

Boyd’s quick glance of astonishment, even though she felt it to be

half assumed, made it hard to go on.

“About your Sargent paper,” she added determinedly.

“Is that to be wondered at? It is down on the board with the rest.”

“I think you know what I mean. You know that those notes you
dropped the other day belonged to Alec.”

“Upon my word, that is—”

 “And that the subject you used was really the one he was using.”

“Aren’t you taking a good deal for granted?” Boyd broke in; she
should not have it all her own way.

“You know what I say is so,” Blue Bonnet insisted. “Those were
Alec’s notes, the subject was his, and all at once he gave up sending
in a paper. It’s very plain.”

“It has not occurred to you that Alec might have given me those
notes?”

“Then, in that case, you would not have looked so—ashamed,

while you were picking them up.”

Boyd sprang to his feet, his face crimson. “I don’t wonder they
sent you East to be taught—manners!”

It was Blue Bonnet’s turn to crimson, but she held back the retort
trembling at the edge of her tongue; she had come out there to tell
Boyd Trent what she knew, and she had told him. It was
inconceivable that a Trent—the General’s grandson, and Alec’s cousin
—should have done this thing.

“I only wish you were a boy!” Boyd said.

“I’d like well to be—for a few moments,” Blue Bonnet answered,

turning away.

Boyd did not follow her; instead he wandered off to the lower end
of the yard, out of sight of the lantern-lighted barn, but not out of
hearing of the fiddle played by Amanda’s Uncle Dave. Leaning
against the old stone wall, the boy stared miserably out over the
broad moonlit meadow.

The worst of it was that he did not know what Blue Bonnet would
do now. As things were, it would be just his luck for that paper to
take a prize. It ought to, considering how carefully Alec had
prepared those notes; there had been very little left for him to do,
beyond putting them together. He wouldn’t have bothered about
writing a paper at all—what did he care for Woodford customs?—
except that his grandfather had seemed to expect it, and he wanted
to keep on the right side of his grandfather—for various reasons.
Alec shouldn’t have left the notes lying around, he knew he had
been hunting for a subject; and anyhow, they were only notes—
taken from books; he wouldn’t have thought of taking a real paper.
There would have been plenty of time for Alec to get up another
one; it was the sort of thing he liked doing. If only Blue Bonnet had
not—Alec could have been depended on not to tell; he had not
referred to the matter since—Boyd moved impatiently; that brief
interview between his cousin and himself was one of the things he
preferred to forget.

It was all a horrid mess whatever way you looked at it; he would
be mighty glad when school closed; next fall he should be going
back to his own school; he never wanted to see Woodford again.

In the meantime, he supposed that Amanda girl was wondering

where her partner for this last dance was? She would have to
wonder, that was all.

They were finishing the dance as he went back to the barn.

Amanda received his murmured apology about a sudden headache
in indignant silence; she didn’t believe he had a headache.

More than once, during the ride home, Boyd felt Kitty’s inquisitive
eyes upon him. “Why aren’t you singing with the rest of us?” she
demanded at last.

“I’d rather listen.”

“You didn’t look as if you were doing even that,” Kitty remarked.
 Alec glanced at his cousin; something had happened during that
sitting out.

“Don’t let’s wait to talk,” Susy urged; “we’ll be home before we

know it now. Mrs. Parker, mayn’t we go around the long way? It’s
such a beautiful night.”

But Mrs. Parker vetoed this request; the short way ’round was fully
long enough in her opinion.

Two or three days later, Blue Bonnet came in after school waving a
letter. “I met the carrier! It’s from Uncle Cliff! He expects to get here
by the twelfth. He will be here in two weeks! And then in ten days
school will be out!” Blue Bonnet waltzed Solomon about the room
excitedly.

There was a litter of sewing about the sitting-room; Blue Bonnet

was to take her summer things back with her, and Grandmother
insisted on having a share in the making of them. Being fitted by
Grandmother was much pleasanter than being fitted by Mrs. Morrow,
Blue Bonnet thought; she didn’t fill her mouth full of pins, and then
sigh if one so much as stirred.

Not that there were no fittings to be gone through with at the old-
fashioned house at the further end of the village; Mrs. Morrow was
making the new white dress for “Closing Day” right now, and Blue
Bonnet was due in her little trying-on room right now, too.

“To think that it’s only two weeks!” Blue Bonnet looked about the
sitting-room a little soberly; would she be homesick for it after she
got back to the ranch? The great living-room there was not much
like this, certainly.
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com