profileMolly-Coyepublication23759904 Remote Patient Management Technology-Enabled Innovation and

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/23759904
Remote Patient Management: Technology-Enabled Innovation And Evolving

Business Models For Chronic Disease Care
Article in Health Affairs · January 2009

DOI: 10.1377/hlthaff.28.1.126 · Source: PubMed
CITATIONS READS
90 1,646
3 authors, including:
Molly Coye Steven Demello

University of California, Los Angeles University of California, Berkeley
74 PUBLICATIONS 2,411 CITATIONS 18 PUBLICATIONS 385 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Health Technology Center View project
WirHe - Wireless in Healthcare View project
All content following this page was uploaded by Molly Coye on 18 July 2014.
The user has requested enhancement of the downloaded file.

At the Intersection of Health, Health Care and Policy
Cite this article as:

Molly Joel Coye, Ateret Haselkorn and Steven DeMello
Remote Patient Management: Technology-Enabled Innovation And Evolving Business
Models For Chronic Disease Care
Health Affairs, 28, no.1 (2009):126-135
doi: 10.1377/hlthaff.28.1.126
The online version of this article, along with updated information and services, is
available at:
http://content.healthaffairs.org/content/28/1/126.full.html
For Reprints, Links & Permissions:

http://healthaffairs.org/1340_reprints.php
E-mail Alerts : http://content.healthaffairs.org/subscriptions/etoc.dtl
To Subscribe: http://content.healthaffairs.org/subscriptions/online.shtml
Health Affairs is published monthly by Project HOPE at 7500 Old Georgetown Road, Suite 600,
Bethesda, MD 20814-6133. Copyright © 2009 by Project HOPE - The People-to-People Health
Foundation. As provided by United States copyright law (Title 17, U.S. Code), no part of Health Affairs
may be reproduced, displayed, or transmitted in any form or by any means, electronic or mechanical,
including photocopying or by information storage or retrieval systems, without prior written permission
from the Publisher. All rights reserved.
Not for commercial use or unauthorized distribution
Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

at Indonesia: HEALTH AFFAIRS Sponsored
Ch ro ni c D i se ase Car e
Remote Patient Management:

Technology-Enabled Innovation
And Evolving Business Models
For Chronic Disease Care
Remote patient management technologies are attracting new interest
from organizations at risk for the consequences of poorly managed
chronic disease care.
by Molly Joel Coye, Ateret Haselkorn, and Steven DeMello
ABSTRACT: Remote patient management (RPM) is a transformative technology that im-

proves chronic care management while reducing net spending for chronic disease. Broadly
deployed within the Veterans Health Administration and in many small trials elsewhere,
RPM has been shown to support patient self-management, shift responsibilities to non-
clinical providers, and reduce the use of emergency department and hospital services. Be-
cause transformative technologies offer major opportunities to advance national goals of
improved quality and efficiency in health care, it is important to understand their evolution,
the experiences of early adopters, and the business models that may support their deploy-
ment. [Health Affairs 28, no. 1 (2009): 126–135; 10.1377/hlthaff.28.1.126]
T
h e m o s t p r e s s i n g ta s k o f h e a lt h c a r e is to make care effective and
affordable; this is particularly important in the case of chronic disease. In
this paper we examine current and emerging business models in health care
that use new technologies for the remote management of chronic care, and early
evidence that these technologies may enable new levels of efficiency and patient
self-management.
Enabling Innovation In Chronic Care Management

Health care has been uniquely slow to innovate, especially in core processes, fi-
nancial models, and customer experience.1 In chronic care, new business models
are emerging in response to public and private policies designed to improve out-
comes and reduce spending. Some of these models establish conditions that foster
Molly Coye (mcoye@healthtech.org) is the chief executive officer of HealthTech in San Francisco. Ateret
Haselkorn is a graduate student in the Johns Hopkins University Bloomberg School of Public Health, in Silver
Spring, Maryland. Steven DeMello is director of research and forecasting at HealthTech.
126 Januar y/ Fe br uar y 2009

DOI 10.1377/hlthaff.28.1.126 ©2009 Project HOPE–The People-to-People Health Foundation, Inc.

RPM
the adoption of enabling technologies, including remote patient management

(RPM), and adoption rates are now beginning to accelerate.
! A disruptive technology. RPM is a disruptive technology; its use relies upon a
reorganization of care processes that include physiologic monitoring, protocol-
driven decision support, newly defined roles for clinical and nonclinical providers,
and telecommunications that place patients at a distance in space, and frequently
time, from the providers of their care.2 It also relies on a disruption of the usual busi-
ness model for care of chronic disease, shifting some responsibilities to the patient
and nonclinical providers; reducing use of and revenues for emergency departments
(EDs), hospitals, and skilled nursing facilities; and producing a net reduction in the
total cost of care for chronic disease.
! Drivers of RPM technology. As evidence for these effects has accumulated
over the past decade, RPM has also been defined as a transformative technology,
that is, one that “enable[s] a wide range of disruptive and positive changes in clinical
care and administrative processes, reducing net expenditures and improving the
value of health care.”3 Because such technologies offer important opportunities to ad-
vance national goals of improved quality and efficiency in health care, it is important
to understand the drivers and barriers that will affect their evolution, the experiences
of early adopters, and the business models that might support their deployment.
The evolving business models of interest all include risk associated with the
consequences of poorly managed chronic care, an important driver for the adop-
tion of RPM. Purchaser and payer strategies that shift care out of hospitals and
EDs, reduce variation, and reward compliance with evidence-based medicine con-
stitute additional drivers for the diffusion of RPM technologies.
! Some first-decade results. RPM technologies have evolved steadily over the
past decade. In a comprehensive review, Guy Parè and colleagues found that moni-
toring of patients with chronic diseases produces accurate and reliable data; is ac-
cepted by patients; and positively affects their attitudes, behavior, and satisfaction.
Decreases in ED visits and hospital admissions were consistent across studies of
pulmonary and cardiac disease, but varied in diabetes.4 In four New England hospi-
tals, for example, in-home monitoring and coaching after hospitalization for conges-
tive heart failure (CHF) reduced rehospitalizations for heart failure by 72 percent,
and all cardiac-related hospitalizations by 63 percent.5
VHA Adoption Of RPM Technologies

The Veterans Health Administration (VHA) has evaluated, piloted, reevaluated,
and deployed RPM technologies in a continuing process of learning and improve-
ment that stands in vivid contrast to patterns of adoption in the private sector.
The VHA is an integrated delivery system with a budget fixed by Congress. A
systemwide change in the mid-1990s reorganized care processes across the contin-
uum and provided the opportunity, purpose, and rewards for disruptive advances
such as RPM.6
H E A L T H A F F A I R S ~ Vo l u m e 2 8 , N u m b e r 1 127

! Management of CHF. In 2002 the VHA produced a detailed technology as-

sessment of RPM in the management of CHF. Rita Kobb and colleagues had demon-
strated a 60 percent decrease in hospital admissions, 81 percent decrease in nursing
home admissions, and 66 percent decrease in ED visits among 281 RPM-monitored
veterans with CHF, in comparison to 1,120 veterans who did not use the technology.7
Other VHA studies reported similar conclusions, for other diseases and conditions.
The technology assessment recommended continuing trials to learn more about op-
timal target populations and clinical and financial impact.8
! Use for other conditions. As positive trials continued to be reported from the
field, the VHA expanded the deployment of remote monitoring to a wider range of
conditions. A standardized toolkit for the use of RPM in behavioral health was de-
veloped in 2003, including approaches to inclusion criteria, credentialing, informed
consent, appropriate technologies, networks, and quality review. Although several
RPM technologies have been deployed, the majority of patients use the Health Hero
Network Health Buddy device. In addition to accepting data from glucometers and
other physiological monitoring equipment, the Health Buddy conveys dialogs for ed-
ucation and coaching of patients by condition and preferred language. Care manag-
ers are often social workers or other personnel with limited clinical training.
RPM applications in the VHA now include depression and post-traumatic
stress disorder (PTSD), patients undergoing chemotherapy, and support for pal-
liative end-of-life care.9 More than 2,000 care managers have been trained at three
centers within the VHA, and remote monitoring services serve more than 33,000
veterans with more than thirty-two conditions; in five years, the VHA expects this
total to reach 75,000.10
! Measured improvements. Reports in the literature of VHA studies cite im-
provements in a wide range of metrics. In addition to the expected decreases in ED,
hospital, and nursing home use, use of preventive services and medication adherence
increased, as did patients’ understanding of their condition, confidence in self-
management, communications with physicians and nurses, feeling of connectedness
to the care team, sense of security, and health-related quality-of-life scores. Although
most studies did not include direct cost measures, remote monitoring for end-of-life
care decreased the total combined costs of hospital and ED use over six months for
100 veterans from $151,771 to $25,119.11
! Medicare demonstration. Largely because of the scale and impact of the
VHA’s adoption of RPM technologies, the Centers for Medicare and Medicaid Ser-
vices (CMS) granted Health Hero Network a unique opportunity to organize a
demonstration of RPM deployment within a Medicare fee-for-service (FFS) popu-
lation. That demonstration, now in its second phase, has enrolled more than 600
Medicare beneficiaries with multiple chronic conditions. Its purpose is to investi-
gate whether the savings attained within the VHA can be achieved in FFS settings.

RPM
Drivers Of And Barriers To Adoption Of RPM

! Drivers. Early adopters of RPM technologies, including the VHA, identify six
components of chronic care management that are facilitated by these technologies:
(1) early intervention—to detect deterioration and intervene before unscheduled
and preventable services are needed; (2) integration of care—exchange of data and
communication across multiple comorbidities, multiple providers, and complex dis-
ease states; (3) coaching—motivational interviewing and other techniques to en-
courage patient behavioral change and self-care; (4) increased trust—patients’ sat-
isfaction and feelings of “connectedness” with providers; (5) workforce changes—
shifts to lower-cost and more plentiful health care workers, including medical assis-
tants, community health workers, and social workers; and (6) increased productiv-
ity—decreased home visit travel time and automated documentation.12
The drivers that lead home care agencies, delivery systems, and health plans to
consider adopting RPM for chronic care include familiar pressures to reduce
costs, improve quality and satisfaction, and increase the productivity of primary
care physicians and nurses.
! Barriers. Thomas Bodenheimer, Chad Boult, and Richard Baron have all de-
scribed the principal barriers to innovation in chronic care as the effects of benefit
design and reimbursement mechanisms.13 Each of these general barriers to innova-
tion also affects the adoption of RPM.
Some barriers to the adoption of RPM are particular to these technologies,
however. Most providers and delivery systems have little experience with remote
clinical technologies. They are poorly prepared to evaluate the technologies or to
make decisions about their acquisition or deployment. Given the increasing num-
ber and complexity of RPM solutions and the lack of guidance from the field
about their utility, it is difficult for provider systems or health plans to estimate
the potential effect of a specific technology. Precisely because RPM technologies
are relatively inexpensive, moreover, they do not rise to the dollar thresholds that
delivery systems use to identify candidates for centralized planning and decision
making and therefore are rarely investigated. The financial models and assump-
tions needed to calculate costs and return on investment do not exist. Although
most RPM products on the market today have a functional interface with one or
more electronic medical records (EMRs), installation and maintenance are an ad-
ditional burden on delivery systems’ information technology (IT) staff. Perhaps
most difficult of all, there are few models of implementation by individual physi-
cians, large medical groups, or health care delivery systems to draw upon.
Emerging Business Models

Several types of health care organizations are now fielding RPM-enabled pro-
grams for chronic disease management. In light of the major barriers to adoption
of RPM technologies enumerated above, we consider the potential of each busi-

“Home care agencies may well prove to be the most effective entities
in the adoption of RPM.”
ness model to support the eventual widespread deployment of RPM in chronic
care management.
! The VHA. The VHA represents a unique business model within the U.S. health
sector. In the 1990s, a combination of congressional budget pressure and internal
leadership drove massive cultural and operational changes within the VHA. Hospi-
tal bed capacity was sharply reduced, primary care clinics were expanded, safety
and quality were measured and improved, and significant resources were devoted to
improvements in information systems including an EMR and the coordination of
chronic care management.
In other words, the VHA began with the innovations necessary to realign a
health care delivery system for effective chronic care management—a new frame-
work for creating value—and then began to experiment with RPM. Financial and
other incentives rewarded administrators and clinicians who led the shift to out-
patient and evidence-based care and prepared the ground so that RPM was not as
“disruptive” in the VHA as it has been in other sectors of health care. Most specifi-
cally, when RPM deployment resulted in further decreases in hospital volume,
fewer ED visits and long-term care bed days, as well as a shift in care to outpatient
clinics and into the home, this contributed to important financial and patient care
goals for the VHA and resulted in further deployment of RPM.
! Integrated provider-based health plans. Kaiser Permanente, Group Health
of Puget Sound, and other group- and staff-model provider-based plans resemble
the VHA in some respects. Their business models have also driven increasingly inte-
grated models for chronic care management, with major improvements in outcomes
realized without the use of RPM technologies. Several pilots of RPM within Kaiser
Permanente have also been successful, however, including an early project that re-
sulted in a 33–50 percent decrease in the cost of care delivery and increased patient
satisfaction.14 More recently, the results of the pilots have combined with the pros-
pect of workforce shortages, an aging membership, and rising costs to spark re-
newed interest in the potential utility of RPM technologies.15 Like the VHA, inte-
grated provider-based plans may find that their financial and clinical integration
provides fertile ground for the implementation of RPM.
! Home care agencies. Home care agencies may well prove to be the most ef-
fective entities in the adoption of RPM. The home health agency business model—
largely dependent on fixed case-rate reimbursement—is well positioned to extract
value from RPM deployment. Remote monitoring and management allow home care
nurses and support staff to maintain patient contact, assess needs, provide educa-
tion, and counsel caregivers while reducing staff travel time. Moreover, home health
agency administrators and staff have direct experience with managing patients in

RPM
their homes and communities, and some agencies have the staff and infrastructure
to integrate RPM technologies into patient care. In 2008, 17 percent of agencies sur-
veyed in a large national study were already using some type of telehealth system.16
The principal limitation of this business model lies in the fact that home health
agencies, unlike provider-based plans, ordinarily do not have a means of sharing in
the much larger savings they generate when they reduce ED visits and hospitaliza-
tions. Centura Health at Home, a not-for-profit home health agency in Colorado,
deployed a two-way video and monitoring system in a pilot following CHF pa-
tients after discharge in 2007 and 2008; it found notable productivity increases:
telehealth nurse case managers are able to manage seventeen to twenty patients
per day, rather than seven patients per day for traditional traveling nurse case
managers. These savings make RPM cost-effective for home care agencies, but
many agencies do not have the expertise or initial capital needed to purchase the
technology, reorganize work processes, and train their staffs.
This aspect of the business model for home care is evolving, however. The capi-
tal required for Centura Health at Home to move from an RPM pilot to full deploy-
ment came from its twelve-hospital delivery system parent, Centura Health. The
pilot had produced a 90 percent decrease in ED visits, a 100 percent decrease in
rehospitalizations for CHF, and $3,000–$5,000 in savings per patient. The deliv-
ery system provided a $1 million investment to expand the model, to reduce the
demand for ED services and increase capacity in their hospitals, and also to pre-
pare for the CMS’s potential decision not to reimburse for certain readmissions af-
ter 2009.17
Multihospital delivery systems are an important part of this business model,
because many of them acquired home care agencies as they grew. Catholic Health
East—a system that includes thirty-two acute care hospitals, thirty-six long-term
care facilities, and twenty-five home health and hospice agencies—selected an
RPM vendor in 2008 for deployment through its home health agencies to support
its vision for integrated care across the continuum.18 As this business model is de-
veloped, some home health agencies are exploring opportunities to contract inde-
pendently for the management of patients with chronic diseases.
! Integrated delivery systems. Unlike provider-based plans and home health
agencies, the business model for hospital-based delivery systems has historically
been poorly aligned with chronic care innovations and the RPM technologies that
support them. Chief among the barriers to adoption is the direct effect that RPM has
on hospital revenues by decreasing ED visits and hospital admissions. In a study at
Tufts Medical Center and three other hospitals, reductions of 72 percent in hospital-
izations for heart failure and 63 percent for all cardiac-related hospitalizations rep-
resented a direct loss in revenues.19 For hospital administrators, this creates the
same predicament as the early quality improvement efforts of the 1990s, in which
recognized improvements in care were stymied by the financial incentives of FFS re-
imbursement.

The nature of the business model for hospitals began to shift in 2008, when the
CMS directed hospitals to report rates of readmission within thirty days of dis-
charge for patients admitted for heart failure, pneumonia, and acute myocardial
infarction in 2009, and described options for modifying or denying reimbursement
in 2010.20 To estimate the potential impact of CMS denials, an average 250-bed
hospital with 1,150 CHF Medicare admissions per year can expect approximately
265 thirty-day readmissions a year, at a loss of $1.5 million.21
Some health care delivery systems, such as Catholic Health East and Centura
Health, are now expanding predischarge assessments and planning and following
patients into the community to manage their care and prevent readmissions. An
elegant and effective model for managing care transitions from hospital to home
has been developed by Mary Naylor, but it relies on advanced practice nurses and
other clinical professionals as managers.22 The model is therefore relatively expen-
sive and, if taken to scale, would quickly exhaust the supply of such managers.
Other models that use less expensive personnel and RPM technologies, as in the
VHA, are also being tested in the management of care transitions, and a national
coalition on transitions of care has been formed.23 It is unclear, however, how mod-
els based on a financial imperative to prevent thirty-day readmissions will inte-
grate into the longer-term maintenance of patients with chronic diseases.
! Health plans. After some early experimentation with RPM technologies,
many health plans have relied upon disease management (DM) contractors to iden-
tify opportunities to use RPM in chronic care management. Health plans’ practices
regarding DM itself have fluctuated considerably over the past decade, as plans al-
ternated between bringing DM programs in house and turning to external contrac-
tors.24 DM contractors have in turn faced narrowing margins as their offerings be-
came commoditized, and they have been reluctant to subcontract and share the
margins with RPM vendors. Whether DM is managed in-house by health plans or
externally by contractors, however, fielding RPM technologies in this context is
particularly challenging because most health plans span large regions of widely dis-
persed networks of physicians, and each physician practice must deal with many
other plan- or contractor-operated DM programs as well. PacifiCare Health Plan
was an early and successful adopter of RPM for chronic disease management; its
success led to the acquisition of the RPM vendor by a DM company.25
Health plans serving Medicare populations have the highest-cost chronic care
populations, with the highest concentration of preventable hospitalizations and
therefore the greatest potential savings from innovations in care. This is even more
acute in the case of Special Needs Plans (SNPs), created by Congress in 2003 for
beneficiaries who either are dually eligible (for both Medicare and Medicaid), are
institutionalized, or have severe or disabling chronic conditions. Both the number
of SNPs for people with chronic conditions and the populations enrolled have
grown rapidly over the past five years.26
Although it is too early to know what impact SNPs will have on chronic care

RPM
“RPM technologies are very good candidates for the emerging

concept of coverage with evidence development.”
management, the business model could be expected to migrate toward RPM de-
ployment to capture the efficiencies of improved care management. This potential
can be observed in the case of two of the largest SNPs, both of which emerged from
the federal social health maintenance organization (S/HMO) program established
two decades ago to prevent the placement of frail elders in skilled nursing homes.27
Today SCAN, in Southern California and Arizona, and Elderplan, in Brooklyn, op-
erate well-established and very effective chronic care management programs that
supplement contracted physician networks with community-based services such
as transportation, prescription compliance, homemaker services, and caregiver re-
lief. Using contracted physician networks supplemented by these services, they
have approximated the care coordination redesign in the VHA and provider-based
plans such as Kaiser Permanente. Both SCAN and Elderplan are now embarking
on RPM pilots, with much larger teams of social workers and case managers to
draw upon than exist in most contracted physician networks.28
Observations And Policy Implications

From this review of business models and early adoption patterns, several
broader observations and policy implications can be drawn.
! Observations. First, RPM technologies can facilitate chronic care manage-
ment. They provide essential support for the coordination of care, behavior change
(of providers as well as patients), and evidence-based decision support for patients
and providers. Second, they will be essential in meeting the dual challenges of an ag-
ing workforce and an aging population.29 As Richard Baron and Christine Cassel
have pointed out, solutions to this problem will require the restructuring of primary
care and the use of technology, with much of the care provided by less expensive
team members with less training.30 Third, and most important—especially in the
context of intensifying discussions of health reform—RPM technologies offer a
means of making care more affordable.
! Three policy recommendations. Three policy recommendations would assist
in this regard; each contributes to the ongoing generation of useful information
about the applications and limitations of RPM.
Coverage of new technologies. First, RPM technologies represent an opportunity to
grapple with the coverage issues that arise when a category of technology is con-
tinuously and rapidly evolving. Advances in sensor technology, miniaturization,
information storage, and materials sciences will continue to yield new versions of
products with changes in functions, risks, and benefits. RPM technologies are
therefore very good candidates for the emerging concept of coverage with evi-
dence development (CED), which the CMS uses to extend coverage to new prod-

ucts in exchange for data collection for ongoing monitoring and evaluation.31
Collaborative research program. Second, the incoming federal administration
should establish a collaborative research program on the use of RPM in chronic
disease at the Agency for Healthcare Research and Quality (AHRQ), drawing on
technology vendors, DM companies, and health plans as well as providers and se-
nior living communities. The research program could fund evaluations of RPM de-
ployment and develop guidance for technology developers, with an explicit
charge to decrease the net cost of care for chronic conditions over the next decade.
Coverage and reimbursement approaches. Lastly, coverage and reimbursement poli-
cies remain a barrier to RPM deployment, and we have a great deal to learn about
the most effective means of compensating providers for their use. A number of ap-
proaches could be tested by both public and private payers, independently or as
part of the AHRQ research collaborative: reimbursement for the purchase of RPM
equipment against reimbursement for its ongoing use; FFS payment for perfor-
mance against capitation; and gainsharing arrangements such as the new partner-
ships between home health agencies and integrated delivery systems. Providers,
health plans, purchasers, and policymakers all deserve better information about
RPM technologies, and chronic disease patients will benefit from a concerted ef-
fort to investigate and use these emerging technologies.
NOTES
1. L. Keeley, A Primer: On Building and Innovation Competence (Chicago: Doblin Inc., 2005); and L. Keeley, Innovation
in Healthcare (San Francisco: HealthTech, 2004).
2. The term remote patient management (RPM) is used herein to refer to the management of chronic diseases and
conditions in the home or community and specifically excludes remote health services provided in a hos-
pital setting, such as the tele-ICU.
3. M.J. Coye, “New Healthcare Product Introduction,” in Evidence-Based Medicine and the Changing Nature of
Healthcare, ed. M.B. McClellan et al. (Washington: National Academies Press, 2008), 72–84.
4. G. Parè, M. Jana, and C. Sicotte, “Systematic Review of Home Telemonitoring for Chronic Diseases: The
Evidence Base,” Journal of the American Medical Informatics Association 14, no. 3 (2007): 269–277.
5. A.R. Weintraub et al., “SPAN-CHF II: Specialized Primary and Networked Care in Heart Failure II,” Pro-
gram and Abstracts from the Ninth Annual Scientific Meeting of the Heart Failure Society of America,
Boca Raton, Florida, 18–21 September 2005, http://www.medscape.com/viewarticle/514121 (accessed 19
November 2008).
6. P. Longman, Best Care Anywhere: Why VA Health Care Is Better than Yours (Sausalito, Calif.: PoliPointPress, 2007).
7. R. Kobb et al., “Enhancing Elder Chronic Care through Technology and Care Coordination: Report from a
Pilot,” Telemedicine Journal and e-Health 9, no. 2 (2003): 189–195.
8. Veterans Health Administration, “Technology Assessment Program Short Report: Physiologic Telemoni-
toring in CHF,” Rapidly Produced Brief Assessments of Health Care Technology no. 5 (Washington: VHA,
January 2001).
9. R.S. Schofield et al., “Early Outcomes of a Care Coordination–Enhanced Telehome Care Program for El-
derly Veterans with Chronic Heart Failure,” Telemedicine Journal and e-Health 11, no. 1 (2005): 20–27; J. Maud-
lin et al., “A Road Map for the Last Journey: Home Telehealth for Holistic End-of-Life Care,” American Jour-
nal of Hospice and Palliative Care 23, no. 5 (2006): 399–403; N.R. Chumbler et al., ”Remote Patient-Provider
Communication and Quality of Life: Empirical Test of a Dialogic Model of Cancer Care,” Journal of Telemedi-
cine and Telecare 13, no. 1 (2007): 20–25; and N.R. Chumbler et al., “Healthcare Utilization among Veterans
Undergoing Chemotherapy: The Impact of a Cancer Care Coordination/Home-Telehealth Program,” Jour-
nal of Ambulatory Care Management 30, no. 4 (2004): 308–317.

RPM
10. Adam Darkins, VHA, chief consultant for telemedicine, personal communication, 22 April 2008.
11. Maudlin et al., “A Road Map for the Last Journey.”
12. Regarding early intervention, see Parè et al., “Systematic Review of Home Telemonitoring.” Regarding in-
tegration of care, see Schofield et al., “Early Outcomes.” Regarding increased trust, see D. Ryan et al., “Mo-
bile Phone Technology in the Management of Asthma,” Journal of Telemedicine and Telecare 11, no. 1 Supp.
(2005): 43–46. Regarding workforce changes, see Parè et al., “Systematic Review of Home Telemonitor-
ing.” Regarding increased productivity, see Erin Denholm, CEO, Centura Health at Home, personal com-
munication, 4 March 2008.
13. T. Bodenheimer, “Coordinating Care—A Perilous Journey through the Health Care System,” New England
Journal of Medicine 358, no. 10 (2008): 1064–1071; Boult et al., “Innovative Healthcare for Chronically Ill Older
Persons”; and R.J. Baron and C.K. Cassel, “Twenty-first-Century Primary Care: New Physician Roles Need
New Payment Models,” Journal of the American Medical Association 299, no. 13 (2008): 1595–1597.
14. B. Johnston, L. Wheeler, and J. Deuser, “Kaiser Permanente Medical Center’s Pilot Tele-Home Project,”
Telemedicine Today 5, no. 4 (1997): 16–17, 19.
15. Scott Young, co–executive director, Care Management Institute, Kaiser Permanente, personal communi-
cation, 5 May 2008.
16. Fazzi Associates, Philips National Study on the Future of Technology and Telehealth in Home Care (Washington: Na-
tional Association for Home Care and Hospice, 2008).
17. Denholm, personal communication.
18. Kathleen Popko, executive vice president, Strategy and Ministry Development, Catholic Health East, per-
sonal communication, 22 August 2008.
19. Weintraub et al., “SPAN-CHF II.”
20. U.S. Department of Health and Human Services, “Centers for Medicare and Medicaid Services, Proposed
Rules,” Part II, Federal Register 73, no. 84 (30 April 2008): 23648.
21. B. Harvath, HealthTech Deep Dive: Thirty-Day Readmissions (San Francisco: HealthTech, 2008).
22. M.D. Naylor, “Transitional Care: A Critical Dimension of the Home Healthcare Quality Agenda,” Journal for
Healthcare Quality 28, no. 1 (2006): 48–54.
23. For information, see the National Transitions of Care Coalition home page, http://www.ntocc.org.
24. G.P. Mays, M. Au, and G. Claxton, “Convergence and Dissonance: Evolution in Private-Sector Approaches
to Disease Management and Care Coordination,” Health Affairs 26, no. 6 (2007): 1683–1691.
25. Health Technology Center, The Future of Remote Health Services, Report and Update (San Francisco: HealthTech,
2007); and Sam Ho, executive vice president and chief medical officer, United Health Group, personal
communication, 29 July 2008.
26. J. Verdier, M. Gold, and S. Davis, Do We Know If Medicare Advantage Special Needs Plans Are Special? (Menlo Park,
Calif.: Henry J. Kaiser Family Foundation, 2008).
27. W. Leutz et al., “Financial Performance in the Social Health Maintenance Organization, 1985–88,” Health
Care Financing Review 12, no. 1 (1990): 9–18.
28. Timothy Schwab, medical director, SCAN Health Plan, personal communication, 25 April 2008; and Jay
Gormley, director of strategic planning, Metropolitan Jewish Health System Elderplan, personal commu-
nication, 11 July 2008.
29. Institute of Medicine, Retooling for an Aging America (Washington: National Academies Press, 2008).
30. Baron and Cassel, “Twenty-first-Century Primary Care.”
31. S.R. Tunis and S.D. Pearson, “Coverage Options for Promising Technologies: Medicare’s ‘Coverage with
Evidence Development’,” Health Affairs 25, no. 5 (2006): 1218–1230.

L e t t e r s
Letters
W
e welcome your responses to papers that underperforming system seems a foolish way
appear in Health Affairs. We ask you to resolve the debate.
to keep your comments brief (250–300
John J. Frey
words, including any endnotes) and sharply focused.
University of Wisconsin School of
Health Affairs reserves the right to edit all letters
Medicine and Public Health
for clarity, length, and tone and to publish them in the
Madison, Wisconsin
bound copy or on our Web site. Letters can be submit-
ted by e-mail, letters@healthaffairs.org, or the
Health Affairs Web site, http://www.healthaffairs Primary/Specialty Care: An Author
.org. It is our policy to invite every author to respond Responds
to letters submitted in response to their work. If a re- The hypothesis regarding primary/specialty
sponse letter does not appear, it is because the au- care mix (Web Exclusive, 4 December 2008) is
thor(s) chose not to respond. Richard Cooper’s alone. My preface to his arti-
cle and to the Perspectives by Katherine
Test The Primary/Specialty Care
Baicker and Amitabh Chandra and by Jona-
Hypothesis than Skinner and colleagues was intended
Editor’s Note: This first group of letters are in re- only to introduce the issues in contention. It
sponse to the 4 December 2008 Web Exclusives. They does not take a stand for or against Cooper’s
first appeared online as comments and have been edited reasoning. Cooper argues that more of both
for print format. kinds of physicians per capita are good for
health care quality, and the evidence supports
To test Richard Cooper’s hypothesis (Web that view. But this does not refute the original
Exclusive, 4 December 2008) that it is simply research by Baicker and Chandra, who showed
more doctors, not the mix of specialty/general- that the existence of more specialists reduces
ists, that makes a difference in access, quality, quality if it comes at the expense of having
and cost, why not close down all generalist fewer general practitioners (GPs). These are
training programs (which are well on their two distinct questions: Cooper looks at each
way toward that goal anyway, with the choices type of physician separately, whereas Baicker
made by U.S. medical students) and see what and Chandra hold constant the total number
happens? Managing complex multiple comor- of doctors and consider the effect of substitut-
bidities, managing urgent and unorganized ing one kind for the other. Cooper’s analysis
health complaints, or providing primary and actually agrees with theirs, since his own re-
secondary preventive care to large populations sults show that the presence of more special-
of chronically ill patients would be done by an ists has a much smaller (about one-tenth as
increasing cadre of subspecialty providers. large) effect on quality than the presence of
This would prove interesting for Cooper’s next more GPs has.
analysis. (In his preface to these papers, deputy Two further points are worth mentioning
editor Philip Musgrove appears to be champi- but were not in my preface. First, although
oning this hypothesis.) Cooper’s main finding concerned the relation
The choice for the United States should ei- between numbers of doctors and quality of
ther be developing a robust primary care base care, he also found large differences among
or allowing seemingly random professional states that are independent of that relation
movement of specialists to fill the gaps in care and depend on socioeconomic or cultural fac-
that would develop. Growing the physician tors that merit closer study. Second, both his
supply in the current costly, market-driven, analysis and that of Baicker and Chandra are
594 M a r c h /A p r i l 2 0 0 9
DOI 10.1377/hlthaff.28.2.594 ©2009 Project HOPE–The People-to-People Health Foundation, Inc.

L e t t e r s
conducted at a very high level of aggregation into rates is usually done using some arbitrary
(averages over entire states), whereas much geography based on political boundaries.
more finely grained research on these relations These rates generate “error,” given that we do
has been done; by using each state’s ranking on not restrict use. The creation of the Hospital
quality, both authors discard the cardinal in- Referral Regions (HRRs) in the Dartmouth Atlas
formation contained in the original quality approximated patterns of use and represented
data. Perhaps it is time to get beyond such rel- an estimate of what might be the underlying
atively simple analyses. denominator. The use of state-level indicators
is suggestive at best and ought to be viewed
Philip Musgrove
with skepticism.
Health Affairs
By using analytical methods that loosen the
Bethesda, Maryland
constraints of boundaries, we have tried to
replicate and advance the analysis of physician
Doctors And Quality Of Care supply as it relates to outcomes.1 That analysis
I found the discussion of physicians and suggests that we might need to pay more at-
quality of care in Richard Cooper’s papers and tention to the variability and patterns of the
the surrounding commentaries (Web Exclu- associations to determine if there are alterna-
sive, 4 December 2008) fascinating but not tive geographies. It appears that primary care
conclusive. Given the wide disparity of views physician supply is associated with lower
on this subject, it is very easy for policymakers mortality in some regions but with higher
to pick and choose which theory they would mortality in others. These regions do not over-
like to base their policies on. From a practicing lay political or postal boundaries but tend to
physician’s perspective, I assert that high- constitute subnational regions.
quality care can come only from high-quality The authors of these papers need to con-
doctors, whether generalists or specialists. tinue to develop their analyses to account for
And just like any other service, one only gets these geographic issues. We certainly ac-
what one pays for. It is foolish to debate, as vig- knowledge that the preponderance of readily
orously as the Harvard-Dartmouth group available data tends to be defined by political
does, that one is better than the other. boundaries, often state-level, and that the lack
of data useful for answering interesting ques-
Arvind R. Cavale
tions is a real inhibitor in this line of inquiry.
Feasterville, Pennsylvania
Thomas C. Ricketts and G. Mark
Geographic Analysis: Need For Holmes
University of North Carolina at
Better Data Chapel Hill
The papers by Richard Cooper and others,
including Jonathan Skinner and colleagues
(Web Exclusive, 4 December 2008), that de- NOTE
1. T.C. Ricketts and G.M. Holmes, “Mortality and
bate the relationship between physician sup- Physician Supply: Does Region Hold the Key to
ply, costs, and quality of care tend to ignore the the Paradox?” Health Services Research 42, no. 6,
need to look critically at these relationships Part 1 (2007): 2233–2251.
and explore alternative ways to understand
what is happening. These and related articles Geographic Analysis: An Author
attempt to draw conclusions about the rela- Responds
tionship between supply of physicians and I agree with Tom Ricketts and Mark
rates of mortality and access indicators. Holmes when they note the importance of de-
An overriding problem with any geo- fining the right geographical regions in study-
graphic analysis is that the aggregation of data ing the association between physician supply

L e t t e r s
and health outcomes. However, even perfect is better”) to the wrong question. The issue is
geographic areas won’t solve the problem of the impact of health services on health, not on
untangling causality from simple correlations the quality of care. On average, higher physi-
in cross-sectional aggregated analysis. cian density is inconsistently related to better
So why do cross-sectional analysis at all? health outcomes.1 However, there is consistent
My view is that it provides a reasonable check evidence that what really matters in improv-
on our theories and models. Suppose one’s ing population health is not the number of
view of the world is that more specialists (or physicians but, rather, what those physicians
health care intensity) lead to better health out- do. The availability of an adequate supply of
comes. But if we find in the real world a zero primary care physicians has been consistently
association (or positive in some areas while identified with better health; simply put, per-
negative in others), then it suggests going back son- rather than disease-centered care matters.
to the drawing board and making an attempt Cooper’s work does little to challenge the
to explain why we should observe the correla- existing evidence for the benefits of primary
tion. In other words, these associations can be care on health, which relies on a strong meth-
used to test hypotheses but can never be used odological foundation of multivariate, time se-
to make inferences about causal pathways. ries, and quasi-experimental evidence and
By the same token, I don’t find the associa- based not only on measures of primary care
tion between the proportion of specialists and physician-to-population ratios.2 In fact, evi-
health care quality to say anything about the dence from studies examining health out-
causality of whether specialists are “better” comes of people whose regular source of care is
than primary care physicians, and thus I am a primary care physician, and from studies
sympathetic to Arvind Cavale’s concerns showing that the stronger the achievement of
about the interpretation of the results.1 If after primary care functions, the better the out-
this exchange we can agree that the empirical comes, is even more persuasive than evidence
record indicates a positive association be- regarding workforce numbers. That is why it
tween generalists and process quality, and an is critical to take seriously the importance of
essentially zero association between special- essential primary care functions (including
ists and quality, then we can also agree that person-focused, not disease-focused, care over
any theory of physician workforce should at time; comprehensiveness of services; and coor-
least attempt to explain why this is so. dination of care) and to use populationwide
health outcomes rather than indirect disease-
Jonathan S. Skinner
specific proximate ones as measures of the
Dartmouth College
overall impact of health services resources.
Hanover, New Hampshire
Suboptimal practice does harm, no matter the
number of physicians. Too few true primary
NOTE care physicians and a surfeit of specialists is
1. Some studies using good statistical methods sug- bad for population health, bad for the econ-
gest that specialists are quicker to adopt some
important innovations; for example, see A.M. omy, and even worse for health equity.
Fendrick, R.A. Hirth, and M.E. Chernew, “Differ- Barbara Starfield, Leiyu Shi, and James
ences between Generalist and Specialist Physi-
cians regarding Helicobacter pylori and Peptic Ul- Macinko
cer Disease,” American Journal of Gastroenterology 91, Johns Hopkins Bloomberg School of
no. 8 (1996): 1544–1548. Public Health
Baltimore, Maryland
Physicians And Quality: Answering
The Wrong Question NOTES
Richard Cooper (Web Exclusive, 4 Decem- 1. L. Chen et al., “Human Resources for Health:
ber 2009) has come up with an answer (“more Overcoming the Crisis,” Lancet 364, no. 9449
596 M a r c h /A p r i l 2 0 0 9

L e t t e r s
(2004): 1984–1990. what might occur theoretically is unknowable.

2. B. Starfield, L. Shi, and J. Macinko, “Contribution In actuality, states with more specialists have
of Primary Care to Health Systems and Health,” better quality.
Milbank Quarterly 83, no. 3 (2005): 457–502.
Richard A. Cooper
Physicians And Quality: Cooper School of Medicine, University of
Responds Pennsylvania
I agree with Tom Ricketts and Mark Philadelphia, Pennsylvania
Holmes that units of analysis are critical. ZIP
codes, hospitals, Hospital Referral Regions, Coordination Of Care In Medicare
counties, states, and multistate regions all pro-
In her paper on Medicare Advantage (MA),
vide different information, and each must be
Marsha Gold (Web Exclusive, 24 November
understood in the context of the others, as
2008) misses the mark considerably. Based on
Philip Musgrove suggests. Moreover, as Jona-
our experience, her assertions that little care
than Skinner cautions, the failure to observe
coordination occurs in MA private fee-for-
differences where they actually exist should
service plans (PFFS) are decidedly not true.
not be taken to indicate that they do not exist
With more than 200,000 MA PFFS mem-
but, rather, that the methodology employed
bers, we provide essentially the same robust
may not have been capable of discerning them.
care management services for these members
I also agree with Barbara Starfield and col-
as we do for MA health maintenance organiza-
leagues and with John Frey that primary care
tion (HMO) or preferred provider organiza-
has value, but I don’t need evidence of de-
tion (PPO) members. For example, repeated
creased mortality from cancer, heart disease,
calls are made to all new members each year to
and stroke to prove it, nor would it. These sta-
complete a health risk assessment, so that we
tistics simply reflect the favorable sociodemo-
can quickly offer members with high-risk con-
graphic characteristics of states in the upper
ditions care management services to help im-
Midwest that happen to have more family phy-
prove outcomes. In 2008, 32,000 PFFS mem-
sicians and fewer internists and pediatricians.
bers, or 17 percent of our PFFS membership,
Patients already know the value of primary
received personal evaluations and case man-
care.
agement services from a dedicated team of
As to Arvind Cavale’s question of how con-
nurses, behavioral health specialists, and social
clusive the arguments were, let me summarize
workers helping them address their challenges
two. First, Medicare is anomalous and cannot
and improve their personal care. Our result:
be taken to represent health care spending
program participants experienced 17 percent
overall. Total health care spending correlates
fewer acute hospital days than matched
closely with the number of health care work-
unmanaged Medicare beneficiaries.
ers (a proxy for volume of service), but Medi-
Our coordinated disease management pro-
care does not. More total spending correlates
gram provides nurse engagement and manage-
with better quality, but Medicare does not.
ment through a single point of contact for
Second, Katherine Baicker and Amitabh
members with multiple conditions, to an addi-
Chandra never examined the relationship be-
tional 5 percent of our PFFS members in 2008.
tween quality and the actual numbers of spe-
That year, MedQuery, our program to identify
cialists, or even the actual percentage of spe-
actionable gaps in care, identified 100,000
cialists. Their notion about poor quality came
member-specific opportunities to improve
from a theoretical statistical exchange of fam-
care and shared them with members or physi-
ily physicians for specialists, which has no
cians, or both. The list of private program ben-
real-world equivalent. Their statement that
efits is long, including personal health records,
“states with more specialists have lower qual-
preventive care, and the Aetna Compassionate
ity,” which refers to what is, has no basis, and
Care end-of-life care management program.

L e t t e r s
We take pride in offering our Medicare programs.4 Indeed, a recent Government Ac-
members access to programs that provide countability Office (GAO) report questioned
value and frequently better outcomes. More whether prior notification programs are even
than 75 percent of our MA PFFS members in authorized for private fee-for-service (PFFS).5
our most recent survey said that they were sat- Making the case for care management in
isfied with their plan, and MA, overall, has PFFS plans is particularly challenging when
grown swiftly. The additional services and they were explicitly established as a non-
benefits we provide are a key driver in such managed alternative.6 Today’s PFFS plans do
member satisfaction and membership growth. not use provider networks, cannot put provid-
ers at risk, and are extremely limited in how
Lonny Reisman and Randall S.
they may influence provider practice. Some
Krakauer
PFFS sponsors say that care management is
Aetna
not consistent with PFFS. Reducing hospital-
Hartford, Connecticut
izations and costs in PFFS plans is closely akin
to doing so in Medicare FFS, and the lessons
Coordination Of Care: The Author from the Coordinated Care Demonstration
Responds should apply. PFFS plans also are not required
Aetna is to be commended for using the to provide the kinds of quality data required of
same care management tools across all of its other MA plans, so the quality of care they
Medicare Advantage (MA) products. How- provide cannot even be assessed (this will
ever, that does not mitigate my key point: that change in 2010).
we’ve spent a lot of money expanding choice Marsha Gold
through MA with little evidence of gains in Mathematica Policy Research
quality and efficiency, particularly outside of Washington, D.C.
the most tightly managed plans (Web Exclu-
sive, 4 November 2008).
Although sponsors often believe otherwise, NOTES
1. Congressional Budget Office, “An Analysis of the
the sad fact is that effectively managing care to Literature on Disease Management Programs”
improve quality and promote efficiency has (Washington: CBO, 2004).
been a challenge in both MA and traditional 2. R. Berenson et al., “Cost Containment in Medi-
Medicare. There is limited evidence that third care: A Review of What Works and What
party–administered disease management pro- Doesn’t” (Washington: AARP Public Policy In-
grams have reduced costs, and there are many stitute, December 2008).
flaws in existing studies that suggest that they 3. R. Brown et al., “Fifteen Site Randomized Con-
trol Trial of Coordinated Care in Medicare Fee-
do.1 Targeted approaches at care management
for-Service,” Health Care Financing Review 30, no. 1
focused on high-risk patients have been diffi- (2008): 5–25.
cult to translate to typical settings of commu- 4. R. Brown, “Lessons from CMS Care Coordina-
nity practice.2 Medicare’s coordinated care tion and Disease Management Demonstrations,”
demonstration involving fifteen chronic care Presentation before the Gerontological Society of
programs serving beneficiaries in traditional America Annual Meeting, National Harbor,
fee-for-service (FFS) Medicare found few Maryland, 22 November 2008.
costs savings and only limited and scattered 5. GAO, Medicare Advantage: Characteristics, Financial
Risks, and Disenrollment Rates of Beneficiaries in Private
impacts on quality.3 Better outcomes typically Fee-for-Service Plans, Pub. no. GAO 09-25 (Wash-
involved many in-person contacts, proximity ington: GAO, December 2008).
to patients’ physicians, using the same care co- 6. J. Blum, R. Brown, and M. Frieder, “An Examina-
ordinator for all of a physician’s patients, tion of Medicare Private Fee-for-Service Plans”
timely notification of hospital admission, and (Menlo Park, Calif.: Henry J. Kaiser Family
other features often absent in vendor-supplied Foundation, March 2007).
598 M a r c h /A p r i l 2 0 0 9

L e t t e r s
Some Facts On Rapid Imaging tients. Efforts to discourage self-referral are

Growth the most direct and sensible way to reach this
goal.
Although their paper solely addresses im-
aging growth in a particular health mainte- James H. Thrall
nance organization (HMO), Rebecca Smith- Board of Chancellors, American
Bindman and colleagues (Nov/Dec 08) suggest College of Radiology
that legislation focused on self-referral will not Boston, Massachusetts
sufficiently limit the drivers of rapid imaging
growth nationwide. They state, without sup- NOTES
porting data, that imaging growth in HMO 1. L.G. Aronovitz, “Referrals to Physician-Owned
systems “closely parallel[s]” that in fee-for- Imaging Facilities Warrant HCFA’s Scrutiny,”
service systems. They suggest that self-referral Pub. no. GAO/HEHS-95-2 (Washington: Gov-
is therefore not a primary driver of escalating ernment Accountability Office, 1994), 5; B.J. Hill-
man et al., “Physicians’ Utilization and Charges
overall imaging costs. Neither the scope of for Outpatient Diagnostic Imaging in a Medicare
their paper nor the published results support Population,” Journal of the American Medical Associa-
these claims. tion 268, no. 15 (1992): 2050–2054; and S. Gazelle,
Self-referral, by which providers refer pa- “Utilization of Diagnostic Medical Imaging:
tients to imaging centers or equipment they Comparison of Radiologist Referral versus Same-
Specialty Referral,” Radiology 245, no. 2 (2007):
own, presents a significant conflict of interest 517–522.
and has been identified by private insurers and 2. D.C. Levin and V.M. Rao, “Turf Wars in Radiol-
government agencies as a primary driver of ogy: The Overutilization of Imaging Resulting
spiraling costs. There is no financial incentive from Self-Referral,” Journal of the American College of
for ordering physicians to increase imaging Radiology 1, no. 3 (2004): 169–172.
utilization unless they self-refer. 3. Medicare Payment Advisory Commission, Report
Government Accountability Office (GAO) to the Congress: Medicare Payment Policy (Washing-
ton: MedPAC, March 2005).; and H. Moskowitz
reports and published research document that
et al., “The Effect of Imaging Guidelines on the
imaging skyrockets when providers directly Number and Quality of Outpatient Radio-
profit from ordering scans.1 As much as half of graphic Examinations,” American Journal of Roent-
self-referred imaging may be unnecessary and genology 175, no. 1 (2000): 9–15.
may cost the health care system up to $16 bil- 4. D.A. Schauer, “Medical Radiation Exposure of
lion annually.2 the U.S. Population: Preliminary Results from
Self-referred imaging also presents signifi- NCRP Scientific Committee 6-2 and Other Re-
lated Issues,” slides 20 and 22, Presentation to
cant quality and safety issues for patients. The the International Congress of Radiology, Inter-
Medicare Payment Advisory Commission national Society of Radiology, 5–8 June 2008,
(MedPAC) cited a major insurer study that http://www.ncrponline.org/PDFs/ICR_2008_
found that 78 percent of nonradiologist imag- DAS.pdf (accessed 14 November 2008).
ing facilities had at least one serious defi-
Diagnostic Imaging: The Authors
ciency—many of which could have “tragic”
consequences.3 Also, the National Council on Respond
Radiation Protection and Measurements cited We appreciate the interest of James Thrall
self-referral as a primary driver of a fivefold in- and the American College of Radiology in our
crease in Americans’ exposure to radiation study (Nov/Dec 08) and welcome the oppor-
over the past twenty years.4 tunity to respond. We reported a dramatic rise
Imaging is increasingly replacing more in- in the rates of diagnostic imaging over the past
vasive procedures—enhancing and extending decade at a large nonprofit health maintenance
the lives of patients. Any imaging policy organization (HMO). What we found to be
should curb growth in inappropriate imaging, striking about this rise is that it closely paral-
not imaging that has clearly benefited pa- lels the rise in imaging in fee-for-service set-

L e t t e r s
tings reported in several studies cited in our have started to address this issue, but these
paper (Notes 1 and 3). In fact, a paper by guidelines are typically based on expert opin-
Laurence Baker and colleagues in the same ion instead of outcomes studies. Such studies
Health Affairs issue as our study provides fur- are crucial for identifying when imaging bene-
ther evidence of the similarities between the fits patients and when imaging is useless or
increases in imaging utilization in these two potentially harmful. We hope that Thrall and
settings. Baker and colleagues documented the American College of Radiology will sup-
that the number of computed tomography port investment in the conduct of outcomes
(CT) scans among Medicare beneficiaries ap- studies and in the dissemination of those re-
proximately doubled between 1997 and 2005 sults, so that patients can have access to bene-
(from 260 to 547 CTs per thousand), rates ficial imaging at a cost we can afford.
comparable to Group Health enrollees age
Rebecca Smith-Bindman for the
sixty-five and older during the same years
authors
(from 214 to 476 CTs per thousand). Our re-
University of California, San
sults are presumably not affected by issues of
Francisco
self-referral or the high profitability associated
with imaging in fee-for-service settings, given
that our study was conducted at a nonprofit Comparative Effectiveness
HMO. Research: A Useful Tool
We argue that more research is needed to Brian Firth and colleagues (Nov/Dec 08)
demonstrate whether this increase in imaging use the decision by the National Institute of
is leading to improved health outcomes or effi- Health and Clinical Effectiveness (NICE) on
ciencies. Self-referral is clearly an issue of great drug-eluting stents (DES) in the British Na-
concern, but it is not the main issue raised by tional Health Service (NHS) as an example of
our paper. We need to know more about the why comparative effectiveness research (CER)
benefits and potential harms associated with leads to “erroneous” coverage decisions. How-
the growing use of imaging. Ideally, efforts ever, the authors also identify five reasons why
should be made to use imaging as efficiently as this research is a useful tool in reducing waste
possible with the expectation that more judi- and maximizing health outcomes from invest-
cious and evidence-based utilization would ment in health technology.
lead to cost savings and improved health. With CER allows the following: (1) The synthe-
regard to the evidence cited by Thrall on the sis of good-quality randomized controlled tri-
inappropriateness of imaging related to self- als (RCTs) in establishing clinical effective-
referral, the cited report by Levin and Rao does ness. NICE’s Appraisal Committee considered
not provide evidence that up to 50 percent of twenty-five RCTs and meta-analyses pooling
self-referred imaging may be unnecessary. the results from more than 7,000 patients,
Their work cites considerable variation in im- which allowed comparisons of DES with bare
aging rates, but they provide no data to sup- metal stents (BMS). (2) The assessment of ef-
port an estimation of how much imaging may fectiveness during routine clinical practice by
be inappropriate. We agree that self-referral synthesizing different forms of evidence in-
likely leads to a lowering in the threshold for cluding national epidemiological data. (3) The
imaging, but there are currently few data avail- consideration of value for money; even when
able to assess the appropriateness of this in- improved performance is demonstrated, a
creased use of imaging. decision has to be made as to whether the ad-
In addition to self-referral, we believe that a ditional clinical benefit is worth the additional
large driver of increased imaging is clinical un- cost. NICE’s Appraisals Committee consid-
certainty and lack of evidence-based guide- ered ten full economic evaluations from peer-
lines on when imaging should be used. The reviewed publications, three models submit-
American College of Radiology guidelines
600 M a r c h /A p r i l 2 0 0 9

L e t t e r s
ted by stent manufacturers, and an independ- making what are essentially policy decisions
ently commissioned model. (4) Payers, about patients’ access to particular medical
whether they be the British NHS, Medicare, technologies. We argued that coverage deci-
private insurers, or individual consumers sions should not be overreliant on the cost-
spending out-of-pocket, to take advantage of effectiveness ratio; they should be examined
price competition—a core feature of the free for their wider impact on the health system. In
market. CER empowered the British NHS to the case of DES, a decision not to recommend
use its collective purchasing power to renego- DES would have resulted in the referral of large
tiate regional contracts with DES manufactur- numbers of patients back for coronary artery
ers based on evidence of comparative value for bypass surgery, with consequent increases in
money with BMS. (5) Decisionmakers to make waiting times, lengths of hospital stay, and
their coverage decisions in a fair, contestable, procedural costs. These consequences would
and transparent way, based on a robust analy- have been in contradiction to recent National
sis of the best scientific evidence and on broad Health Service policy priorities and were fed
expert consultation. This is what NICE’s Ap- into NICE’s consultation process by health
praisal Committee did in its evaluation of DES care professionals, patients, and industry alike.
through several rounds of consultation, appeal Economic models do not necessarily capture
hearings, and stakeholder meetings. such consequences; hence our urge for caution
We fully concur with these five principles. against overreliance on the cost-effectiveness
They are all taken into account in the decisions ratios such models generate. Put simply,
made by NICE’s advisory bodies. QALY-based incremental cost-effectiveness
ratios should be “a tool, not a rule” in health
David Barnett, Kalipso Chalkidou, and
policy decisions.
Michael Rawlins
Of course, NICE’s final decision was to re-
National Institute of Health and
tain its previous clinical indications for the use
Clinical Effectiveness
of DES, and patients therefore continue to
London, England
benefit from innovation, choice, and quality of
care, all of which are U.K. policy priorities.
Comparative Effectiveness
Liesl M. Cooper for the authors
Research: The Authors Respond Covidien
David Barnett and colleagues state that we Mansfield, Massachusetts
used the National Institute for Health and
Clinical Effectiveness (NICE) decision on
drug-eluting stents (DES) “as an example of Online Consumer Information In
why comparative effectiveness research (CER) Pennsylvania
leads to erroneous coverage decisions.” As the I was heartened to read Michael Rothberg
abstract to our paper states (Nov/Dec 08), we and colleagues’ paper on “Choosing the Best
used NICE’s DES appraisal as an example of Hospital” (Nov/Dec 08). The authors point to
the challenges faced when using cost- several key issues impeding greater public ac-
effectiveness analysis to make coverage deci- ceptance and understanding of hospital qual-
sions on rapidly evolving medical technolo- ity-reporting Web sites, correctly stating that
gies. Our contention is that cost-effectiveness “information must be accessible, interpretable,
analysis may lead to erroneous conclusions and consistent” and that hospitals must work
when a broader perspective and the impact on with rating agencies and others to develop
health outcomes and costs are considered. consensus on what quality data are most rele-
We cautioned against overreliance on in- vant and actionable for patients.
cremental cost-effectiveness ratios based on Over the past two years, the Pennsylvania
quality-adjusted life-years (QALYs) when Health Care Quality Alliance (PHCQA) has

L e t t e r s
pursued this strategy and focused on develop- narrative provided a reminder of the many set-
ing a uniform and consistent approach to hos- tings in which language barriers can threaten
pital quality measurement. Its Web site, patient safety, the intricate relationship be-
http://www.pahealthcarequality.org, is the tween culture and language, and the chal-
first such site in Pennsylvania that compiles lenges an interpreter is faced with when bridg-
hospital quality data from a variety of sources ing communication gaps over the telephone.
and allows consumers to search and compare Although all of these are crucial issues,
quality data (process and outcome measures) Kelly’s insight into the challenges and neces-
on acute care hospitals in the state. It includes sary qualifications to interpret over the tele-
representation from Pennsylvania’s four Blues phone deserves particular attention. Unfortu-
insurers, its hospital association, its medical nately, we know that many practitioners are
society, and representatives from state and fed- reluctant to use telephone interpreter services
eral governments. Much of its work has been for a variety of reasons, including lack of
spent on developing consensus on what qual- equipment (such as the dual handset), doubts
ity measures are most indicative of overall about interpreters’ qualifications, preferences
quality and meaningful for patients, as well as for in-person interpreters, and even the sense
how those data can best be displayed for inter- that a third-party presence impedes the build-
pretation by consumers. ing of rapport in the exam room. Regardless of
We recently surveyed more than 900 Penn- the underlying rationale, this reluctance is
sylvanians to gain insight into consumers’ worrisome, because telephone interpreters are
awareness and attitudes toward hospital qual- not only the most commonly available lan-
ity-reporting Web sites. The results indicate guage service in U.S. hospitals today, but are
that although most consumers continue to cite often the only qualified interpreters available.
physician recommendation as the most influ- Kelly’s essay helps readers understand why
ential and trustworthy source of information, assessed and trained telephone interpreters
about a third reported visiting a quality- should always be used in favor of less qualified
reporting Web site and using its data to make in-person interpreters, whether they be family
a hospital selection decision. members, friends, or the casual bilingual staff
Although hospital quality reporting is still member. I have circulated this narrative to sev-
in its relative infancy, the PHCQA and similar eral of my colleagues and leadership to share
groups in other states are working hard to es- this point. As a member of that casual bilin-
tablish greater consistency and clarity in re- gual group, I understand that safe and effective
porting data on quality. interpreter services require far more skill than
many of us have, whether interpretation oc-
Erik Muther
curs over the telephone, from an office one
Pennsylvania Health Care Quality
thousand miles away, or at an arm’s reach in-
Alliance
side the same exam room.
Philadelphia, Pennsylvania
Erica Galvez
Telephone Interpreters: Vital Link Joint Commission
Oakbrook Terrace, Illinois
I wanted to offer a personal “thanks” for
Nataly Kelly’s Narrative Matters essay on tele-
phone interpreting (Nov/Dec 08). We have Physicians’ Motives In Imaging
made strides over the past decade in raising As a new subscriber to Health Affairs, I was
awareness of the role that the culture and lan- taken aback at the editor-in chief’s apparent
guage of the patient and provider play during indictment of physicians’ motives in ordering
medical encounters. However, these issues re- computed tomography (CT) scans for their
main to be widely integrated in quality im- patients (Nov/Dec 08).
provement and patient safety efforts. Kelly’s I am aware of the large increase in the num-
602 M a r c h /A p r i l 2 0 0 9

L e t t e r s
ber of imaging studies nationwide. The recent cult to conclude based on the available evi-
decision by Medicare to require accreditation dence that all imaging that is being performed
in the future for certain imaging tests as a re- by doctors in their offices is either cost-effec-
quirement for reimbursement is a major step tive or clinically necessary. As the Government
in the right direction. Any test or procedure Accountability Office (GAO) reported in its 13
ordered for purely financial profit is reprehen- June 2008 analysis of Medicare Part B and im-
sible, and physicians should be prosecuted if aging, analysis over a six-year period showed
proven guilty. However, as I read Susan “certain trends linking spending growth to the
Dentzer’s editorial, I perceived the tone of her provision of imaging services in physician of-
comments about “physicians’ wallets” and fices. The proportion of Medicare spending on
physicians’ “feeling right with their patients imaging services performed in-office rose from
and God” as unfair. Laurence Baker and col- 58 percent to 64 percent. Physicians also ob-
leagues correctly (and rationally) question tained an increasing share of their Medicare
cost versus benefit of CT and computed to- revenue from imaging services. In addition, in-
mography angiography (CTA) in the issue. office imaging spending per beneficiary varied
Similarly, Julie Appleby’s Report from the Field substantially across geographic regions of the
presents both sides of the CTA controversy. country, suggesting that not all utilization was
Physicians and insurers are very often at necessary or appropriate. By 2006, in-office
odds over utilization. We fight the battle over imaging spending per beneficiary varied al-
appropriate tests for our patients every day. As most eight-fold across the states—from $62 in
pointed out, there are other more obvious rea- Vermont to $472 in Florida.” That language
sons for using CT scans and CTA, such as rather pointedly suggests that the problem I
rapid and more precise information (compared referenced in my note—“that many doctors
to a cheaper alternative such as ultrasound), find ample reason to do more lucrative scans
the evolving standard of care, patient demand, and still feel right with their patients and
and, based on personal experience: defensive God”—has been noted by the GAO, as it also
medicine. In my specialty (vascular diseases has been on a number of occasions by the
and surgery), CTA has become a valuable tool Medicare Payment Advisory Commission.
in avoiding invasive arteriography and plan-
Susan Dentzer
ning of minimally invasive endovascular proce-
Editor-In-Chief, Health Affairs
dures. And no, we do not own a CT machine.
Bethesda, Maryland
My object in subscribing to this journal
was to learn more about how scientific meth-
ods are applied in analyzing health policy is- Hip And Knee Implants In Bulgaria
sues. I look forward to enjoying future issues We read with interest the paper by Natalia
of Health Affairs, but perhaps with less pointed Wilson and colleagues (Nov/Dec 08) about
editorial language. ongoing policy considerations for hip and knee
implants. Summarizing recent evidence on
Bhagwan Satiani
hip-spine syndrome (HSS) management, we
Ohio State University College of
contribute recent results from a university
Medicine
hospital in Plovdiv, Bulgaria. HSS is chroni-
Columbus, Ohio
cally progressing and, if untreated appropri-
ately, leads to deterioration of functional fit-
Imaging: The Editor Responds ness and quality of life. In prior research,
As I mentioned in my “From the Editor” surgery (laminectomy) had multiple advan-
note (Nov/Dec 08), “the benefits of many med- tages over nonsurgical therapy in lumbar spi-
ical technologies are real,” and this includes nal stenosis (LSS) at the third month in pro-
the enormous benefits of medical imaging, as spective cohorts. A 7.8-point difference from
Bhagwan Satiani suggests. However, it is diffi- baseline in bodily pain (according to the SF-

L e t t e r s
36) also was evidence of its effectiveness.1 Sup- and After Total Hip Replacement” (in Bulgarian),
porting these findings, we found that total hip Bulg Neurol 8, no. 1 (2008): 24–26.
replacement, alone or after laminectomy, was 3. H. Haro, S. Maekawa, and Y. Hamada, “Prospec-
tive Analysis of Clinical Evaluation and Self-
highly effective in aged patients with both Assessment by Patients after Decompression
lumbar and hip pain. Combined findings from Surgery for Degenerative Lumbar Canal Steno-
our unique observational prospective data in- sis,” Spine Journal 8, no. 2 (2008): 380–384.
dicated very high efficacy levels of THR (in
fifty-eight HSS patients).2 Grim Prognosis For
In particular, twenty-nine patients with Massachusetts Reform
hip osteoarthritis underwent THR; hip pain Jon Gabel and colleagues (Web Exclusive,
disappeared in 68.97 percent of them. Lumbar 28 October 2008) hope that favorable em-
pain remained in eight who also had LSS. Sev- ployer attitudes mean that Massachusetts’
enteen out of another twenty-five patients 2006 health reform will succeed where the
with LSS underwent laminectomy. Lumbar state’s similar 1988 health reform failed. In
pain decreased, but hip pain persisted. Thir- truth, the demise of the 1988 law had more to
teen patients had hip osteoarthritis (bilateral do with economic cycles—the collapse of the
in six of them). Our most important advantage Massachusetts Miracle—than with corporate
was targeted, patient-level decision making by attitudes. Unfortunately, this history, along
a panel of highly-specialized experts (neurolo- with the recent downturn in the economy, im-
gist, neurosurgeon, orthopedic surgeon, roent- plies a grim prognosis for the current Massa-
genologist) acting together to choose the best chusetts reform.
surgical intervention in each case. Assessing The 1988 reform died because health care
their own quality-of-life improvement, many costs continued to soar, while a recession
patients reported relief from pain. Undoubt- shrank tax revenues just as tens of thousands
edly, our individualized surgery showed very lost their jobs and private coverage. (Unem-
high effectiveness as both clinical effects and ployment rose from 3.2 percent in 1988 to 9.1
patient-reported outcomes, especially in LSS percent two years later.) Neither massive ex-
combined with hip osteoarthritis. Interest- pansion of state funding to subsidize coverage
ingly, similar improvements in both patient-re- for the poor nor a costly mandate—the two
ported outcomes and clinical measures were main mechanisms to expand coverage under
reported in comparable prospective cohort of both the 1988 and 2006 laws—was tenable in a
forty-two Japanese patients with LSS.3 cooling economy.
Despite costs of $1.1 billion this year, the
Penka A. Atanassova
2006 law has covered only half of the unin-
Neurology, Medical University
sured and has left many more with inadequate
Plovdiv, Bulgaria
coverage. A recent Boston Globe/Blue Cross Blue
Shield of Massachusetts Foundation survey
Borislav D. Dimitrov found that 9 percent of Massachusetts resi-
Bordani dents avoided or postponed care within the
Alzano Lombardo (BG), Italy past year because of costs; 14 percent had
failed to fill a prescription; and 14 percent had
NOTES run up medical debts.1 The inadequacy of the
1. J.N. Weinstein et al., “Surgical versus Nonsurgi- new coverage is also evident in Robert
cal Therapy for Lumbar Spinal Stenosis,” New Blendon and colleagues’ survey (Web Exclu-
England Journal of Medicine 358, no. 8 (2008): 794–
810. sive, 28 October 2008); those directly affected
2. P. Atanassova et al. “Hip-Spine Syndrome in Pa- by the reform were actually more likely to say
tients with Lumbar Spinal Stenosis” (in Bulgar- that reform had hurt than helped them. This
ian), Neurorehabilitation 1, no. 2 (2007): 24–28; and seemingly paradoxical result probably reflects
P. Atanassova et al., “Hip-Spine Syndrome Before
604 M a r c h /A p r i l 2 0 0 9

L e t t e r s
the fact that most of the newly insured had setts data are not available, the Health Insur-
previously been eligible for completely free ance Association of America (now America’s
care at safety-net hospitals and clinics paid for Health Insurance Plans, or AHIP) reported an
by Massachusetts’ free-care pool. 18 percent national increase in premiums in
Even the partial gains of the 2006 reform 1989.1 Premium increases remained at double
are now in jeopardy, and its collapse may well digits until 1993. From 1994 to 1998, premium
leave patients worse off than ever. The reform increases were at record lows.2 Concurrently,
was partly financed by draining funds from the Massachusetts unemployment rose from 3.5
free-care pool, and hence from safety-net pro- percent in December 1988 to 9.1 percent in
viders. Now, with tax revenues plummeting, April 1992. By December 1993, unemployment
the governor plans to pull another $100 mil- fell to 6.3 percent and to 4.3 percent three
lion from funds owed to the safety-net hospi- years later.3 Hence, conditions for implement-
tals in the Boston area (disclosure: we work as ing the 1988 legislation were highly favorable
primary care doctors at one of them, where when it was repealed in 1996.
these cuts are projected to require hundreds of We bel ieve that Woolhandler and
layoffs and the closure of critical services and Himmelstein don’t sufficiently credit progress
community clinics). Further budget cuts likely achieved in Massachusetts over the past two
lie ahead, and the ranks of the uninsured will years. Compared to national statistics, the fig-
doubtless swell. ures they cite from the Boston Globe/Blue Cross
In the end, Massachusetts’ 2006 reform Blue Shield of Massachusetts Foundation sur-
may be remembered as a short-lived expansion vey are favorable. From a 2007 national house-
of publicly subsidized coverage that served as hold survey of adults ages 19–64, researchers
political cover for the permanent destruction reported that 31 percent of Americans “had a
of institutions that have provided care and ad- medical problem but did not visit the doctor,”
vocacy for New England’s poor for decades. 31 percent “did not fill a prescription,” and 27
percent “had problems paying medical bills.”4
Steffie Woolhandler and David U.
In the face of an economic downturn, pub-
Himmelstein
lic and employer support and a sympathetic
Harvard Medical School and
Obama administration render the future of the
Cambridge Hospital
Massachusetts plan “stable,” not “grim.”
Cambridge, Massachusetts
Jon R. Gabel for the authors
NOTE NORC
1. K. Lazar, “Medical Costs Still Burden Many De- Bethesda, Maryland
spite Insurance,” Boston Globe, 23 October 2008.
Massachusetts Reform: The NOTES

1. J. Gabel et al., “Employer-Sponsored Health In-
Authors Respond surance, 1989,” Health Affairs 9, no. 3 (1990): 161–
We concur with Steffie Woolhandler and 175.
David Himmelstein in their response to our 2. J. Gabel et al., “Job-Based Health Insurance in
paper (Web Exclusive, 28 October 2008) that 2000: Premiums Rise Sharply while Coverage
Grows,” Health Affairs 19, no. 5 (2000): 144–151.
the current economic downturn and rising
3. Data available through query at the Massachu-
cost of health care jeopardizes Massachusetts setts Executive Office of Labor and Workforce
reform. However, we think that their “grim Development, “Labor Force and Unemployment
prognosis” misinterprets the historical record. Data,” http://lmi2.detma.org/Lmi/lmi_lur_a.asp.
After 1988 passage, political opposition 4. C. Schoen et al., “How Many Are Underinsured?
from employers, rising health care costs, and Trends among U.S. Adults, 2003 and 2007,” Health
an economic recession contributed to suspen- Affairs 27, no. 4 (2008): w298–309 (published on-
line 10 June 2008; 10.1377/hlthaff.27.4.w298).
sion of the legislation. Although Massachu-

L e t t e r s
Segmented Approach For China who are better off. A similar analysis of China’s
And India financing gap may involve segmenting along
coverage levels, given the country’s expansive
Responses to our paper on a segmented ap-
basic care health care coverage programs.
proach to health care financing gaps in China
and India (July/Aug 08) underscore the great Ashoke S. Bhattacharjya
contrasts public and private-sector decision- Johnson & Johnson
makers seek to balance: world-class health New Delhi, India
care facilities exist, yet they are sadly out of the
economic reach of the majority (Tom Miller Puneet K. Sapra
and Aparna Mathur, Letters, Jan/Feb 09). Vascular Therapies
Low-cost public health care facilities are avail- New York, New York
able, yet the quality of care provided leaves
little confidence among most consumers. Un-
Errata
defined regulatory systems allow for experi-
mentation, yet they also encourage perverse The January/February 2009 issue of Health
behavior (Chris Conover, Letters, Jan/Feb 09). Affairs contained several errors. All of these ar-
While attempting to balance these contrasts, ticles have been corrected online. The authors
policymakers and consumers are asked to and Health Affairs regret any inconvenience
make difficult choices on how to pay for health these errors might have caused.
! Paez et al., pp. 15–25. A programming
care among a long list of other pressing needs
(Michael Cannon, Letters, Jan/Feb 09). error caused several errors in “Rising Out-of-
In expanding health care financing, we do Pocket Spending for Chronic Conditions: A
not advocate that a one-size-fits-all strategy Ten-Year Trend,” by Kathryn Anne Paez and
(for example, centralized health care) would colleagues (Jan/Feb 2009, pp. 15–25). On page
be effective or sustainable. We prescribe com- 22, the second and third full paragraphs have
bining essential market foundations with a been revised. The new text reads as follows:
segmented approach to health care financing. An out-of-pocket expenditure index (EI) was
Admittedly, public- and private-sector leaders created to measure the overall increase in out-of-
in India and (to a lesser extent) China have pocket spending comparing 2005 to 1996 spend-
ing, holding disease prevalence constant (Ex-
taken steps in this direction. In view of this,
hibit 5).15 The change seen is due solely to in-
the suggestion by Miller and Mathur that we creasing out-of-pocket spending. The EI for the
are endorsing an overregulated or centralized overall population was 1.19, indicating that ex-
approach is rather perplexing. penditures were 19 percent higher in 2005 than
In implementing our approach, health care in 1996, when chronic condition prevalence was
held constant. The younger-old had the greatest
regulators and commercial payers in India and
increase, with an EI of 1.30, followed by young
China should first establish a minimum base of adults, those in midlife, young adults, and the
reference in the form of a mix of regulatory and old-old.
market-based “rules of the game” for health
care financing markets. As Cannon correctly All insurance categories, including Medicaid re-
suggests, this may be the “best thing those gov- cipients, had a sizable increase in out-of-pocket
ernments could do.” The segmented approach spending over the ten-year period. The largest
increase was experienced by those in the “other
to health care financing provides market-
public” insurance category, followed by the un-
appropriate payment mechanisms based on insured and Medicare-only beneficiaries. Al-
consumers’ needs. For example, a segmented though smaller, the EI for Medicaid recipients
approach to address India’s financing gaps may was substantial when controlling for rising
involve government- or community-funded chronic condition prevalence.
health care financing for the rural poor or
needy and private health insurance for those Also on page 22, the last full paragraph has
606 M a r c h /A p r i l 2 0 0 9

L e t t e r s
been revised. The text now reads as follows: ! Grabowski, pp. 136–146. An endnote
was inadvertently left out of “Special Needs
Our study found that out-of-pocket spending
and chronic disease prevalence are increasing
Plans and the Coordination of Benefits and
among not only the old-old but among people in Services for Dual Eligibles,” by David G.
midlife and early old age, without regard to sex, Grabowski (Jan/Feb 09, pp. 136–146). On page
race, ethnicity, or income. The greatest growth 139, endnote 15 should be inserted at the end of
occurred in the number of people reporting mul-
the second sentence below the subheading
tiple chronic diseases; this is also the group with
the most substantial out-of-pocket spending. “For Medicare and Medicaid.” On page 146,
Overall, out-of-pocket spending increased by this note (new note 15) reads as follows: Sauc-
39.4 percent per person over the ten-year period. ier and Burwell, The Impact of Medicare Special
The growth in out-of-pocket spending was not Needs Plans. Every note after this, starting with
evenly distributed across the population. Spend-
the old note 15, is renumbered one higher. In
ing increases were 19 percent higher overall
when holding the rising prevalence of chronic addition, to make room for the added text, the
conditions constant, with the greatest increase first sentence has been deleted from note 14.
among those in early old age, the “other public” ! Etheredge, pp. 148–159. Note 10 in
insured, the uninsured, Medicare beneficiaries, “Medicare’s Future: Cancer Care,” by Lynn M.
the poor, and people who take prescription
Etheredge ( Jan/Feb 09, pp. 148–159), con-
drugs. Medicaid continued to provide financial
protection for people with chronic conditions tained an inadvertent error. The title of the
from high out-of-pocket spending. When pov- book in Note 10 is Ensuring Quality Cancer Care.
erty status was considered, it became evident ! Ham, pp. 190–201 . Note 32 of
that Medicaid is not available to all poor people “Chronic Care in the English National Health
with chronic conditions.
Service: Progress and Challenges,” by Chris
Ham (Jan/Feb 09, pp. 190–201), contained a ty-
On page 23, all of the data in Exhibit 5 have pographical error. The volume and issue num-
been replaced. This exhibit, as well as the cor- ber should be 288, no. 15.
rected text, is available at http://content.health ! Anderson, pp. 202–205. Note 7 in
affairs.org/cgi/content/full/28/1/15. “Missing in Action: International Aid Agencies
! Coye et al., pp. 126–135. Three minor in Poor Countries to Fight Chronic Disease,”
corrections have been made to “Remote Pa- by Gerard F. Anderson (Jan/Feb 09, pp. 202–
tient Management: Technology-Enabled Inno- 205), contained two typographical errors. In
vation and Evolving Business Models for the second citation, the author’s name is
Chronic Disease Care,” by Molly Joel Coye and “Gaziano,” and the volume and issue number
colleagues (Jan/Feb 09, pp. 126–135). On page should be 26, no. 1.
131, the word “potential” has been inserted at ! Hartman et al., pp. 246–261. There
the end of the second full paragraph, as fol- were several minor errors in “National Health
lows: “…and also to prepare for the CMS’s po- Spending in 2007: Slower Drug Spending Con-
tential decision not to reimburse for certain tributes to Lowest Rate of Overall Growth
admissions after 2009.” On page 132, three since 1998,” by Micah Hartman and colleagues
lines from the top, the following revision was (Jan/Feb 09, pp. 245–261). On page 251, the
made: “…patients admitted for heart failure, phrase “the continued shift toward lower-cost
pneumonia, and acute myocardial infarction in mail order channels” has been deleted from the
2009, and described options for modifying or second sentence in the first full paragraph. The
denying reimbursement in 2010.” Finally, at title of Exhibit 6 (page 259) has been changed
the end of the first full paragraph on page 132, to “Expenditure Levels For, And Average An-
the final sentence has been modified as fol- nual Growth In, Health Services And Sup-
lows: “It is unclear, however, how models plies…” Finally, the second source in Note 9,
based on the financial imperative…” The cor- page 260, has been removed. The corrected
rected article is available at http://content article is available online at http://content
.healthaffairs.org/cgi/content/full/28/1/126. .healthaffairs.org/cgi/content/full/28/1/260.

View publication stats

profileMolly-Coyepublication23759904 Remote Patient Management Technology-Enabled Innovation and

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

profileMolly-Coyepublication23759904 Remote Patient Management Technology-Enabled Innovation and

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Remote Patient Management: Technology-Enabled Innovation And Evolving

Article in Health Affairs · January 2009

Molly Coye Steven Demello

SEE PROFILE SEE PROFILE

Health Technology Center View project

WirHe - Wireless in Healthcare View project

The user has requested enhancement of the downloaded file.

Cite this article as:

For Reprints, Links & Permissions:

Not for commercial use or unauthorized distribution

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Remote Patient Management:

ABSTRACT: Remote patient management (RPM) is a transformative technology that im-

Enabling Innovation In Chronic Care Management

126 Januar y/ Fe br uar y 2009

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

the adoption of enabling technologies, including remote patient management

VHA Adoption Of RPM Technologies

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

! Management of CHF. In 2002 the VHA produced a detailed technology as-

128 Januar y/ Fe br uar y 2009

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Drivers Of And Barriers To Adoption Of RPM

Emerging Business Models

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

130 Januar y/ Fe br uar y 2009

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

132 Januar y/ Fe br uar y 2009

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

“RPM technologies are very good candidates for the emerging

Observations And Policy Implications

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

134 Januar y/ Fe br uar y 2009

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

(2004): 1984–1990. what might occur theoretically is unknowable.

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Some Facts On Rapid Imaging tients. Efforts to discourage self-referral are

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Massachusetts Reform: The NOTES

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

Downloaded from content.healthaffairs.org by Health Affairs on May 21, 2011

You might also like