Clinical Change at Intermountain Healthcare
XXXXXXXXXX
R E V : J A N U A R Y 2 2 , XXXXXXXXXX
________________________________________________________________________________________________________________
Professor Richard M. J. Bohmer and Research Associate Erika M. Ferlins prepared this case. HBS cases are developed solely as the basis for class
discussion. Cases are not intended to serve as endorsements, sources of primary data, or illustrations of effective or ineffective management.
Copyright © 2006, 2008 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call XXXXXXXXXX-
7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http:
www.hbsp.harvard.edu. No part of this publication may be
eproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical,
photocopying, recording, or otherwise—without the permission of Harvard Business School.
R I C H A R D M . J . B O H M E R
Clinical Change at Intermountain Healthcare
Intermountain Healthcare (Intermountain) was a nonprofit healthcare system serving the
healthcare needs of Utah and southeastern Idaho residents. Intermountain operated 20 hospitals in
Utah and Idaho, employing over 25,000 people and offering over 2,400 patient beds. Recognized
internationally as a leader in healthcare quality improvement, many of Intermountain’s medical
outcomes were among the best in the nation.1 Furthermore, it provided healthcare at a fraction of the
cost of many competing healthcare organizations. According to a 2006 study by Dartmouth Medical
School, “Medicare spending could be reduced by a third if the nation provided healthcare the way it's
provided [by Intermountain Healthcare].”2
Clinical Programs
In order to improve patient outcomes, Intermountain’s clinical programs integrated care delivery
protocols in 10 areas: Cardiovascular, Neuromusculoskeletal, Surgical Specialties, Women and
Newborns, Intensive (hospital-based) Medicine, Intensive Pediatrics, Intensive Behavioral, Oncology,
Preventive and Health Maintenance, and Primary Care. Clinical programs provided “tools to help
clinicians deliver consistently high-quality clinical care” and defined “appropriate care management
systems and initiatives for medical treatment throughout Intermountain.”3 Each clinical program
was responsible for guiding initiatives, defining new Disease Management Systems, and integrating
them into routine care throughout the IHC system. Each Disease Management System included a
Care Process Model (CPM) consisting of (1) an evidence-based best practice guideline, (2) workflow
tools, to blend the guideline into care delivery as a “shared baseline” (that actively required clincians
to modify the guideline to meet unique individual patient needs, (3) a clinical management
information system that identified and tracked medical, cost, and service outcomes specific to that
particular care process, (4) decision support tools, usually implemented through Intermountain’s
electronic medical record system, and (5) education materials for both health professionals and
patients.
1 Intermountain Healthcare website, http:
intermountainhealthcare.org/xp/public/aboutihc/quality/, accessed May 29,
2006.
2 Ibid., http:
intermountainhealthcare.org/xp/public/aboutihc/news/article38.xml, accessed May 29, 2006.
3 Ibid., http:
www.intermountainhealthcare.org/xp/public/physician/clinicaograms/, accessed April 19, 2006.
Do
N
ot
C
op
y
o
P
os
t
This document is authorized for educator review use only by Sandeep Reddy, Deakin University until July 2017. Copying or posting is an infringement of copyright.
XXXXXXXXXX or XXXXXXXXXX
XXXXXXXXXXClinical Change at Intermountain Healthcare
2
Many physicians at Intermountain were involved in clinical programs. Those physicians who
were particularly passionate or knowledgeable about a particular issue often provided expert advice
to development teams, the engine of clinical programs that designed and implemented CPMs. One
such physician was Dr. Nathan Dean, a nationally recognized pulmonologist. The Primary Care
clinical program administrators asked Dr. Dean to chair the Lower Respiratory Tract Infection team,
which was dealing with community-acquired pneumonia (CAP), a common and potentially serious
illness and a concern for hospitals nationwide. 5.6 million cases of CAP were reported annually
across the nation, with nearly 1.1 million patients requiring hospitalization.
Pneumonia
In 1995, the Lower Respiratory Tract Infection team rolled out its first community-acquired
pneumonia CPM. The team, comprised of physicians, administrative staff, and nurses, first created
the pneumonia CPM based on available medical literature and guidelines published by national
associations such as the American Thoracic Society, which were studying and disseminating
evidence-based recommendations for combating various diseases. Over time, the pneumonia CPM
evolved as changes were made based on feedback from front-line physicians and medical directors as
well as updated antibiotic and clinical care research findings.
In 2000, Intermountain’s protocol system was still paper-based. CPMs were double-sided and
listed clinical tasks and the treatment order sheet on the front, and the decision flow chart for
diagnosis on the back (see Exhibit 1, A and B, respectively). Each protocol form comprised a
guideline and an order sheet, which would accompany those patients who were admitted to the
hospital. As emergency room physicians and attending physicians reached consensus on the optimal
antibiotic choice, this CPM order sheet effectively served as a standard order set so that antibiotic
therapy could begin immediately upon diagnosis, usually before the patient actually reached the
ward. The result—faster antibiotic administration time—reduced the risk of patient mortality and
also reduced physician workload.
Soon the system became electronic, and hospitals across IHC were graded on the extent to which
in-patient prescribing complied with the guideline-recommended antibiotic choice. Central
computer systems could track compliance through pharmacy billing data and produce reports that
the Lower Respiratory development team reviewed on a quarterly basis. The team monitored
compliance at the hospital rather than the physician level for two reasons: first, patient volume by
doctor was too low to track meaningful results, and second, the team felt that monitoring individual
providers would exace
ate an existing feeling by many physicians, especially those working in
managed care situations, that their every move was being watched and criticized. Instead, the
development team compared data across hospitals. According to Dean, “we measured the outcomes
for internal purposes of improving the logic of and compliance with the guideline. We share
compliance and clinical outcome data with clinicians, assuming that each clinician will interpret the
data in relation to their personal level of guideline utilization.”
Measuring guideline compliance for outpatients was harder, primarily because physicians
treating outpatients were reluctant to collect the additional data needed to monitor outpatient
antibiotic choices. However, during a study of outpatient pneumonia management called the
“Instacare”4 study, data were collected retrospectively for one year (through chart review) and
prospectively for one year. In the context of a scientific study, physician compliance with data
4 “Instacares” were Intermountain’s walk-in urgent care clinics. Do
N
ot
C
op
y
o
P
os
t
This document is authorized for educator review use only by Sandeep Reddy, Deakin University until July 2017. Copying or posting is an infringement of copyright.
XXXXXXXXXX or XXXXXXXXXX
Clinical Change at Intermountain Healthcare XXXXXXXXXX
3
collection was 90%. This study found that the level of compliance with the guideline-recommended
antibiotic choice was 72%.5
Although all hospitals were expected to follow the pneumonia guidelines, they were not required
to use the order sheets created by the development team (see Exhibit 2 for an example of orders).
Dean explained, “the process varies between hospitals, as local people choose how they want to
implement the guidelines. Other order sheets exist that have been approved by the development
team. For example, LDS in Salt Lake City likes to use their own sheets.” However, Dean’s personal
e-mail was included with every guideline, regardless of what order sheets were used. Dean relied on
case reports to solicit feedback on failures within the CPM. “In the beginning, when people were less
certain, they e-mailed me often. Now, people are more comfortable, the logic is better, and I receive
fewer e-mails.”
The Lower Respiratory Tract team’s goal was to reassess protocols every two years at a minimum
to keep up with changes in medication, medical technology changes, and research findings.
Antibiotic choice was an important facet not only of the guideline itself, but also in the tracking of
guideline compliance throughout IHC. In 1995, there were a plethora of antibiotic choices available
and across the IHC system physicians were prescribing over 60 different antibiotic regimens for the
treatment of community-acquired pneumonia. Dean felt that “the simpler the process, the better the
compliance. If there is no medical reason to list an additional drug, then what’s the advantage to
having it?” Even internationally accepted CAP guidelines listed 10 to 12 antibiotic possibilities.
However, Dean aimed to have one standard antibiotic combination across the entire Intermountain
system. Because outcome (i.e., death from pneumonia) had been linked to the accuracy of antibiotic
choice and the speed with which the first dose was delivered, Dean wanted a process that was as
simple as possible so compliance would be as high as possible.
Dean’s problem was that the medical literature provided limited guidance on the optimal
antibiotic choice. This was because of the way in which randomized controlled studies were
designed and executed. The sickest patients (i.e., those at the highest risk of dying—the ones most
like the population Dean and his colleagues treated daily) tended not to be included in scientific
trials. The reduction in observed death rate of those patients who were included in a randomized
trial was often very small, typically not reaching statistical significance