|
E L E C T R O N I C P R E S C R I B I N G
E X P E R T C O N S E N S U S
W E B E X C L U S I V E
25 May 2004
Recommendations For Comparing
Electronic Prescribing Systems:
Results Of An Expert Consensus Process
Guidance to help early adopters
and policymakers
select the systems most likely to benefit patients.
By Douglas S. Bell, Richard
S. Marken, Robin C. Meili,
C. Jason Wang, Mayde Rosen, Robert H. Brook,
and the RAND Electronic Prescribing Expert Advisory Panel
ABSTRACT:
Commercially available electronic prescribing systems may differ in
their effects on patients’ health outcomes and on patients’ ability
to manage costs. An expert panel convened to recommend specific features that
would enable electronic prescribing systems to advance these goals. The panel
authored sixty recommendations and rated each using a modified Delphi process.
Ratings identified fifty-two recommendations as clearly positive for patient
safety and health outcomes and forty-three recommendations as achievable in
the average clinician’s office within three years. Overall, these recommendations
offer a synthesis of evidence and expert opinion that can help guide the development
of electronic prescribing policy.
Electronic prescribing systems are computerized systems that clinicians use
to prescribe medications. Both clinicians and policymakers expect these systems
to improve health care quality.1 Although the evidence
to support these expectations is limited, the implementation of electronic prescribing
in a few academic hospitals has been associated with decreased medication errors,
improved formulary adherence, and shorter lengths-of-stay.2
Few providers use electronic prescribing, but new incentives for its adoption
are emerging.3 The Medicare Prescription Drug, Improvement,
and Modernization Act (MMA) of 2003 will fund grants for implementation of electronic
prescribing, and it specifically permits third parties to fund such systems
on behalf of providers. In addition, the Leapfrog Group has made electronic
prescribing a quality criterion for hospitals.4
Provider organizations wishing to implement electronic prescribing can choose
from a variety of systems, but implementation can be difficult, and providers
have no basis for selecting those that can improve health care quality.5
Furthermore, funding from third-party organizations could shape these systems
in ways that distort medication choices toward the funders’ interests.
Recommendations are therefore needed to ensure that electronic prescribing systems
meet patients’ needs for better health outcomes and a greater ability to
manage their own costs.
Recommendations for ensuring the effectiveness of electronic prescribing would
ideally be based on scientific evidence, but the available evidence is insufficient
to guide comparisons among alternative systems.6
When evidence is incomplete, expert opinion can offer valid guidance through
the use of the modified Delphi panel process.7 In
this process, a panel of experts initially rates a set of recommendations, they
then meet to discuss their ratings, and finally each panelist has a chance to
revise his or her ratings in light of the discussion.8
In this paper we present the results of a modified Delphi panel process that
produced recommendations for judging an electronic prescribing system’s
ability to meet the goals of improving patient safety and patients’ ability
to manage costs without damaging patients’ privacy or clinicians’
acceptance of electronic prescribing. The recommendations are intended for clinicians
and provider organizations to use in selecting among electronic prescribing
alternatives, for manufacturers to use in designing future electronic prescribing
systems, and for policymakers to use in promoting the adoption of electronic
prescribing systems that advance health care quality. The recommendations focus
primarily on systems for outpatient office environments. However, inpatient
and outpatient prescribing have many basic processes in common; thus, wherever
possible, we constructed the recommendations to make them applicable to both
environments.
Study Methods
We recruited a panel of recognized experts from professional domains related
to electronic prescribing: medicine, nursing, pharmacy, managed care, pharmacy
benefit management, consumer advocacy, medical informatics, health care oversight,
health care quality and safety, and health economics. The panel chair was recruited
first; he then worked with project staff to select the remaining panelists.
(Panel members are listed at the end of the paper.) The project team provided
panelists with a manuscript that included a conceptual framework for electronic
prescribing evaluation and a systematic review of the relevant evidence.9
At an initial meeting in December 2001, the panelists used a nominal group process
to create forty-one initial draft recommendations for electronic prescribing.
Using an affinity-grouping procedure, project team members merged similar recommendations.10
The team then rephrased the draft recommendations to read as standards that
could be objectively assessed in comparing electronic prescribing systems and
wrote a rationale statement for each based on the evidence, conceptual framework,
and panel discussion. From June through December 2002, the recommendations and
rationale statements were substantially revised and extended by the panel through
two rounds of preliminary ratings, teleconference discussions, and written comments.
This process resulted in a set of sixty final recommendations.
Panelists rated the expected effects of each final recommendation along four
dimensions: patient safety and health outcomes, helping patients manage costs,
maintaining patient privacy, and promoting clinician acceptance. The fifteen-point
rating scale ranged from –7 to +7, where –7 represented the largest
negative impact on the dimension and +7 represented the largest positive impact.
Panelists were asked to use ratings of +3 or higher to indicate “clearly
positive” effects and ratings of –3 or lower to indicate “clearly
negative” effects.11 Clearly positive and
negative effects were those deemed “major” enough to warrant consideration
by anyone choosing among electronic prescribing alternatives.
Panelists were also asked to judge the time frame within which each recommendation
could be implemented in the average clinician’s office, using two criteria:
whether the recommendation is technically feasible, and whether the implied
workflow changes could be achieved by the average clinician. To assist panelists
in assessing technical feasibility, we provided them with preliminary results
of a field study that assessed the implementation of each recommendation by
six commercially available outpatient electronic prescribing systems.
In March 2003 the panel held a teleconference to discuss its initial ratings
of the final recommendations. Panelists then submitted final revisions of their
ratings. For each recommendation we determined the median rating on each dimension.
For each median, we also examined the spread of the panelists’ ratings
for “disagreement,” which was defined as having more than one panelist
giving a “clearly positive” rating and more than one panelist giving
a “clearly negative” rating. To identify relationships among the recommendations,
four authors (Douglas Bell, Richard Marken, Robin Meili, and Jason Wang) independently
reviewed each recommendation and listed dependencies or underlying mechanisms.
Team members met to discuss and achieve consensus on these interpretations.
Study Results
Final recommendations.
All sixty final recommendations received a median rating in the clearly positive
range (+3 or greater) for at least one of the health care dimensions. For patient
safety and health outcomes, fifty-two had clearly positive median ratings. For
helping patients manage costs, eighteen had clearly positive medians. None of
the median ratings were in the clearly negative range (–3 or less) for
any of the final recommendations. Forty-three recommendations were rated as
being achievable in the average physician’s office within three years.
The complete final recommendations and median ratings are available in an online
data supplement.12
The final recommendations were organized into ten categories, seven of which
correspond to the steps that prescribers follow in managing a patient’s
medications, and three of which relate to broader aspects of electronic prescribing
(for example, security and confidentiality).13
Here we summarize the highest-rated recommendations for patient safety and health
outcomes that were also considered achievable within three years.
Patient identification. Electronic prescribing usually begins with selecting
the patient’s name from a menu. To mitigate the risk that slips in these
menu selections would send prescriptions to the wrong patient, systems should
display the selected patient’s identifying information throughout the process
of creating a prescription (recommendation 1). To facilitate integration of
prescribing with other patient information, patient-identifying data should
be imported from electronic medical record or practice management systems (recommendation
2), but because these linkages may intermittently fail, electronic prescribing
systems should also provide for manual entry of these data (recommendation 3).
Exhibit
1 shows the median ratings for these recommendations.
Access to patients’ historical data. Before selecting a medication,
prescribers often need to review elements of the patient’s history. The
patient’s current medication list was rated as a critically important element
for electronic prescribing systems to provide (recommendation 7). The current
medication list should integrate data from all clinicians caring for the patient,
but to mitigate privacy threats, this list should be available only to prescribers
with care responsibility for the patient. To improve the accuracy and completeness
of the list, systems should extract the patient’s prescription data from
external pharmacy and hospital systems (recommendation 5) and support the manual
entry of prescription information that is not available from external systems
(recommendation 9). Exhibit
2 shows the highest-rated, immediately achievable
recommendations in this category.
Medication selection. After deciding that a medication is indicated, prescribers
must specify a medication and dose. Some prescribing systems assist prescribers
at this step by listing the medications that are indicated for particular diagnoses.14
As shown in Exhibit
3, panel members considered this feature important for patient safety and
health outcomes (recommendation 13), but their ratings also indicate that systems
should not force the assignment of a diagnosis (recommendation 14). Some electronic
prescribing systems attempt to influence prescribers by altering the order in
which medications are presented or by displaying special symbols (such as an
asterisk) next to favored or disfavored options. The panel recognized that this
potentially beneficial feature could also be used to create commercial advantages
for third parties. To curb these potential conflicts of interest, the panel
strongly recommended that the display of medication options should not be influenced
by promotional considerations (recommendation 16).15
Furthermore, the meaning of any symbols or special typefaces used to differentiate
medication choices should be made clear (recommendation 17), and the rationale
for these and any other medication recommendations should be immediately available
(recommendation 18).
Alerts and other messages to prescribers. After a medication is selected,
many electronic prescribing systems attempt to alert prescribers to potential
problems. Exhibit
4 shows that the panel strongly recommended safety alerts for contraindications
or significant precautions (recommendation 27). However, systems should implement
methods for prioritizing alerts and allow prescribers to suppress less clinically
important alerts (recommendations 31 and 32), and prescribers should be able
to clearly distinguish safety-related messages from other types of messages,
such as formulary alerts (recommendation 30).
Patient education. Before finalizing a new prescription, clinicians should
educate patients about the new medication and negotiate patients’ adherence
to treatment. The panel recommended that electronic prescribing systems provide
information tailored for patients on how and why the medication should be used
(recommendation 39) and provide a printout of a complete current medication
list in nontechnical language (recommendation 40).
Data transmission and storage. Once completed, prescription orders must
be transmitted to a pharmacy. As shown in Exhibit
5, the panel recommended that prescribing systems enable these transmissions
following established standards from Health Level 7 (HL7) or the National Council
for Prescription Drug Programs (recommendations 41 and 42). Data transmissions
should use National Provider Identifiers when these become available (recommendation
44).16 Prescribers should also be notified when
transmissions fail (recommendation 46).
Monitoring and renewals. Monitoring is an important but underappreciated
step in medication management.17 The panel expected
important benefits from alerts for patient adherence problems (recommendation
48) and from assistance with ordering and tracking laboratory monitoring tests
(recommendations 49–51), but none of these recommendations was rated as
achievable in three years.
Transparency and accountability. Many developers and implementers of
electronic prescribing are receiving support from third-party organizations
that have incentives to influence the prescribing process. To help deter this
sponsorship from introducing biases in prescribing, the panel made several recommendations
intended to open the results of third-party support to scrutiny. Some of these
recommendations modify individual steps in the prescribing process (recommendations
18, 29, 34, and 45).18 In addition, disclosing
the details underlying clinical decision-support rules would further expose
potential biases (recommendation 54), and full disclosure of third-party support
should help focus scrutiny where biases are most likely (recommendation 53).
Prescriber-level feedback. Most systems collect data that could be used
to give feedback about prescribing behavior. Systems could promote practice-based
learning by giving clinicians access to their own prescribing patterns (recommendation
55). Profiling the use of alert overrides may help to identify unintended problems
with alerts (recommendation 56), but some panelists expressed the belief that
clinicians would find this capability unacceptable, given the potential for
its use in a punitive manner. The rating of this recommendation on the dimension
of “promoting clinician acceptance” was one of only two ratings that
met the formal criteria for disagreement among panelists.19
Security and confidentiality. Patients need assurances that prescribing
data will not be used to harm them—for example, through disclosure to a
prospective employer. However, the panel concluded that adherence to the privacy
and security rules arising from the Health Insurance Portability and Accountability
Act (HIPAA) of 1998 would provide the necessary protections (recommendation
57).20 In addition, audit trails (recommendation
58) and role-based access privileges (recommendation 59) are strongly recommended
for privacy (Exhibit
6).
Core electronic prescribing capabilities.
In examining the relationships among recommendations, we noted that some recommendations
exist to “refine” other recommendations, meaning that they specify
particular details of another recommendation’s implementation. These “refining”
recommendations would be applicable only if the core capability were implemented.
Other recommendations serve to “support” a core capability by contributing
toward its implementation. Here we summarize four core capabilities to which
many of the recommendations relate.
Diagnosis-based medication menus. Several recommendations (15–19)
serve to “refine” the selection of medications by diagnosis (recommendation
13) because they call for specific modifications of diagnosis-based menus. Two
other recommendations serve to “support” diagnosis-based menus—external
interfaces could make historical diagnosis data available (recommendation 5),
and permitting prescribing without entering a diagnosis would help prevent the
accumulation of inaccurate diagnoses (recommendation 14).
Safety alerts. Nine recommendations (29–37) refine the core safety-alerting
capability (recommendation 27); the highest-rated ones call for systems to clearly
distinguish safety messages from other messages and to prioritize safety alerts
according to their clinical importance. Safety alerts would also be supported
by having a current medication list (recommendation 7) to provide data for detecting
drug interactions, and by the ability to correct or flag erroneous patient data
(recommendation 38) to reduce the triggering of nuisance alerts.
Current medication list. Four recommendations (5, 8, 9, and 12) serve
to support the achievement of an accurate and comprehensive current medication
list (recommendation 7). For example, when prescribers issue an order to discontinue
a medication (recommendation 12), the pharmacy system would record that fact
and take the drug out of the patient’s current medication list.
Systems integration. Three separate recommendations (2, 5, and 47) call
specifically for the integration of electronic prescribing systems with practice
management systems, electronic medical records, and other external systems.
Four recommendations (6, 42–44) refine systems-integration features, and
two recommendations (2 and 4) support integration. Systems integration, in turn,
is a prerequisite for accurate safety alerts, current medication lists, patient
monitoring, and other recommended capabilities.
Discussion
Electronic prescribing is expected to improve many aspects of health care, but
these high expectations will be achieved only if provider organizations select
systems that have the appropriate features. We report a set of sixty recommendations
for electronic prescribing that resulted from a modified Delphi process integrating
scientific evidence, expert opinion, and observations from a field study of
electronic prescribing systems. Each recommendation describes a feature expected
to have a positive effect on at least one important dimension of health care
performance. The recommendations should aid purchasers of electronic prescribing
in comparing alternative systems, and they may also help electronic prescribing
vendors to prioritize their development efforts.
The results of this study should contribute to current policy efforts. MMA contains
provisions intended to accelerate the adoption of electronic prescribing, both
through grants and by authorizing third-party funding for electronic prescribing
systems. Qualifying systems will need to meet federal standards, which are to
be developed with the goals of promoting safety, quality of care, and cost savings.
Our results provide information that federal policymakers can use in developing
standards that are likely to achieve these goals.
Some of the recommendations with the greatest expected benefits were among those
considered not immediately achievable. These included the highest-rated recommendations
(21 and 22) for helping patients manage their costs and recommendations (28
and 48) that would reduce the underuse of highly effective medications, a major
quality problem for elderly Medicare beneficiaries.21
The main barrier to achieving these recommendations is the lack of integration
among information systems used by physicians, pharmacies, laboratories, hospitals,
payers, and others responsible for patient care in a given community. In the
current environment, these separate entities have little incentive to share
data. Policy initiatives are needed to encourage the development of business,
political, and technical infrastructure that would enable communitywide integration
of health information systems. The Agency for Healthcare Research and Quality
(AHRQ) is beginning to fund the development of such infrastructure, but these
efforts will likely need to be sustained and expanded.22
In the meantime, providers considering the implementation of electronic prescribing
should not lose sight of the benefits that can be gained from more achievable
electronic prescribing capabilities.
Although federal grants under MMA are not scheduled to begin until 2007, third-party
initiatives to promote electronic prescribing are already under way.23
Our panel’s recommendations for transparency and accountability can address
the immediate need for preventing biases as third-party funding in this area
grows. In general, the panel believed that substantial biases could be deterred
through full disclosure of any third-party sponsorship coupled with full disclosure
of decision-support rules. However, as the complexity and power of electronic
prescribing grow, subtle distortions could become more difficult for providers
and consumer groups to detect. Further research should address this possibility,
and if the potential for bias appears significant, an independent group with
expertise in evidence-based medicine may be needed to monitor the development
of decision support.
More research is needed into most of the effects that the panel postulated,
but the most fruitful questions for research may be those on which panelists
disagreed. The panel’s most significant disagreement centered on whether
fallibility of clinicians or of electronic prescribing systems should pose a
greater concern. Although the patient safety literature provides ample evidence
of clinician fallibility, emerging evidence suggests that prescribing system
design flaws can also introduce unanticipated hazards.24
Research is needed into the ways that electronic prescribing systems may create
new hazards. Until more evidence is available, providers who install an electronic
prescribing system should be vigilant for new types of adverse events.
More research is also needed into physician adoption of electronic prescribing.
The most rigorous surveys available show that adoption remains low overall.25
Although the panel included several practicing clinicians and clinician-managers,
it did not represent the broad range of clinicians in community practice. Thus,
its ratings for promoting clinician acceptance should be followed up with additional
research on adoption among representative groups of clinicians. As new evidence
emerges, ongoing efforts will be needed to keep these recommendations updated.
New electronic prescribing tools or approaches may have emerged in the year
since the recommendations were finalized, but we are not aware of any new evidence
that would warrant immediate changes in the recommendations.
The recommendations we
have presented here offer a starting point to guide early adopters of electronic
prescribing toward the systems that are most likely to benefit patients. However,
for the majority of patients to experience these benefits, substantial additional
investments will be needed to foster adoption and to gather adequately detailed
evidence about the effectiveness of electronic prescribing. Electronic prescribing
will achieve its full potential if these investments are made with continuing
attention to the features that would most improve patients’ safety and
health.
The following people were members of the RAND Electronic Prescribing Expert
Advisory Panel. The professional domains that each panelist represented are
shown in parentheses. Donald M. Berwick, Institute for Healthcare Improvement
(panel chair; health care quality and safety); Phyllis Borzi, George Washington
University (consumer advocacy); Lonnie R. Bristow (medicine); Christine K. Cassel,
American Board of Internal Medicine (medicine); Schumarry H. Chao, MedImpact
(pharmacy benefit management); Paul B. Ginsburg, Center for Studying Health
System Change (health economics); Peter I. Juhn, CareTouch Inc. (managed care,
health care quality and safety); Helene Levens Lipton, University of California,
San Francisco (pharmacy); Clement J. McDonald, Regenstrief Institute For Health
Care (medical informatics); Mary O’Neil Mundinger, Columbia University
(nursing); and Margaret E. O’Kane, National Committee for Quality Assurance
(health care oversight). This study was funded through a contract from Pfizer
Inc. that obligated RAND to publish its findings regardless of the study’s
outcome. Pfizer’s aim for this study was to make health care quality and
patients’ health outcomes the focus of competition among electronic prescribing
systems. Pfizer did not select any of the panelists. Pfizer representatives
were permitted to observe panel meetings without comment. Pfizer had the right
to comment on manuscripts for publication, but RAND was not obliged to respond
to Pfizer’s comments. The study was approved by institutional review boards
at RAND and the University of California, Los Angeles. The authors are grateful
for editorial assistance from Sydne Newberry and Kristen Mukae.
NOTES
1. G.D. Schiff and T.D. Rucker, “Computerized Prescribing:
Building the Electronic Infrastructure for Better Medication Usage,” Journal
of the American Medical Association 279, no. 13 (1998): 1024–1029.
2. R. Kaushal and D.W. Bates, “Computerized Physician Order
Entry (CPOE) with Clinical Decision Support Systems (CDSSs),” in Making
Health Care Safer: A Critical Analysis of Patient Safety Practices, ed.
K.G. Shojania et al., Pub. no. AHRQ-01-E058 (Rockville, Md.: Agency for Healthcare
Research and Quality, July 2001); D.W. Bates et al., “Effect of Computerized
Physician Order Entry and a Team Intervention on Prevention of Serious Medication
Errors,” Journal of the American Medical Association 280, no. 15
(1998): 1311–1316; R.S. Evans et al., “A Computer-Assisted Management
Program for Antibiotics and Other Antiinfective Agents,” New England Journal
of Medicine 338, no. 4 (1998): 232–238; D.W. Bates et al., “The Impact
of Computerized Physician Order Entry on Medication Error Prevention,”
Journal of the American Medical Informatics Association 6, no. 4 (1999):
313–321; J.M. Teich et al., “Effects of Computerized Physician Order
Entry on Prescribing Practices,” Archives of Internal Medicine 160,
no. 18 (2000): 2741–2747; and W.M. Tierney et al., “Physician Inpatient
Order Writing on Microcomputer Workstations: Effects on Resource Utilization,”
Journal of the American Medical Association 269, no. 3 (1993): 379–383.
3. J.S. Ash et al., “Computerized Physician Order Entry
in U.S. Hospitals: Results of a 2002 Survey,” Journal of the American
Medical Informatics Association 11, no. 2 (2004): 95–99; and D.J. Brailer
and E.L. Terasawa, Use and Adoption of Computer-Based Patient Records
(Oakland: California HealthCare Foundation, October 2003).
4. Leapfrog Group, “Patient Safety: Setting Standards:
An Rx for Rx,” www.leapfroggroup.org/safety1.htm#CPOE
(3 May 2004).
5. D.F. Doolan and D.W. Bates, “Computerized Physician
Order Entry Systems in Hospitals: Mandates and Incentives,” Health Affairs
21, no. 4 (2002): 180–188, highlights the difficulty of implementation.
6. Kaushal and Bates, “Computerized Physician Order Entry”;
and D.S. Bell et al., “A Conceptual Framework for Evaluating Outpatient
Electronic Prescribing Systems Based on Their Functional Capabilities,”
Journal of the American Medical Informatics Association 11, no. 1 (2004):
60–70.
7. P.G. Shekelle et al., “The Reproducibility of a Method
to Identify the Overuse and Underuse of Medical Procedures,” New England
Journal of Medicine 338, no. 26 (1998): 1888–1895; R.L. Kravitz et
al., “Validity of Criteria Used for Detecting Underuse of Coronary Revascularization,”
Journal of the American Medical Association 274, no. 8 (1995): 632–638;
and J.V. Selby et al., “Variation among Hospitals in Coronary-Angiography
Practices and Outcomes after Myocardial Infarction in a Large Health Maintenance
Organization,” New England Journal of Medicine 335, no. 25 (1996):
1888–1896.
8. R.H. Brook et al., “A Method for the Detailed Assessment
of the Appropriateness of Medical Technologies,” International Journal
of Technology Assessment in Health Care 2, no. 1 (1986): 53–63.
9. Bell et al., “A Conceptual Framework.”
10. W. Nicholas et al., “Group Process Methods,”
in Putting Practice Guidelines to Work in the Department of Defense Medical
System, Pub. no. MR-1267-A (Santa Monica, Calif.: RAND, 2001).
11. Previous Delphi panels have used a nine-point scale, with
the upper third of the scale (7–9) being used for clearly appropriate indications,
and the lower third of the scale (1–3) being used for clearly inappropriate
indications, as described in Brook et al., “A Method for the Detailed Assessment.”
The scale used in our study spans fifteen points, but the upper and lower thirds
of the scale have analogous interpretations.
12. The online supplement is available at content.healthaffairs.org/cgi/content/full/hlthaff.w4.305v1/DC2.
13. Bell et al., “A Conceptual Framework.”
14. Commercial drug-indication databases exist to support this
feature, such as First DataBank, “Indications Module,”
www.firstdatabank.com/integrated_content/drug_information/clincial_content_us/index.asp?module_id=13
(2 February 2004); and Medi-Span, “Drug Indications Database,”
www.medi-span.com/products/product_drugindications.asp
(2 February 2004). In the Netherlands a national database of drug indications
is maintained for this purpose. See Nederlands Huisartsen Genootschap (Netherlands
College of General Practitioners), EVS (Electronic Prescribing System), 2004,
nhg.artsennet.nl/content/resources/AMGATE_6059_104_TICH_L1310214546/
(17 May 2004).
15. “Promotional considerations” were defined as
payments or in-kind support given to providers, prescribing system vendors,
or any other third party, in exchange for their promoting a medication or group
of medications. This definition is intended to encompass circumstances that
could put the provider or vendor’s financial interests in competition with
the patient’s best interests. Note that medication formularies and other
information related to the patient’s health care coverage would not fall
within this definition.
16. U.S. Department of Health and Human Services, Office of
the Secretary, “45 CFR Part 162. Standard Unique Health Identifier for
Health Care Providers,” Federal Register 69, no. 15 (2004): 3434–3469.
17. T.K. Gandhi et al., “Adverse Drug Events in Ambulatory
Care,” New England Journal of Medicine 348, no. 16 (2003): 1556–1564
.
18. To make vendors accountable without giving prescribers
unwanted information, recommendations 18 and 29 require only that disclosures
be available in the system upon the user’s request. The sponsorship referred
to in recommendations 34 and 45 could be disclosed by simply including the sponsor’s
name in the message. Thus, the panel considered all of these recommendation
to be achievable within three years.
19. See Note 12.
20. U.S. Department of Health and Human Services, “Administrative
Simplification in the Health Care Industry,” aspe.hhs.gov/admnsimp/index.shtml
(5 February 2004).
21. T. Higashi et al., “The Quality of Pharmacologic Care
for Vulnerable Older Patients,” Annals of Internal Medicine 140,
no. 9 (2004): 714–720. See Note 12.
22. AHRQ, “Transforming Healthcare Quality through Information
Technology (THQIT)—Implementation Grants,” grants.nih.gov/grants/guide/rfa-files/RFA-HS-04-011.html
(10 April 2004).
23. R.L. Rundle, “WellPoint to Pay $30 Million for Physicians’
Computers,” Wall Street Journal, 15 January 2004.
24. J.S. Ash, M. Berg, and E. Coiera, “Some Unintended
Consequences of Information Technology in Health Care: The Nature of Patient
Care Information System Related Errors,” Journal of the American Medical
Informatics Association 11, no. 2 (2004): 104–112; and Bates et al.,
“The Impact of Computerized Physician Order Entry.”
25. Brailer and Terasawa, Use and Adoption of Computer-Based
Patient Records; and Ash et al., “Computerized Physician Order Entry.”
Douglas Bell (dbell{at}rand.org)
is an associate natural scientist at RAND in Santa Monica, California; an assistant
professor in the Department of Medicine, David Geffen School of Medicine, University
of California, Los Angeles; and a Robert Wood Johnson Foundation Generalist
Physician Faculty Scholar. Richard Marken is a senior behavioral scientist at
RAND, where Robin Meili is a senior management systems analyst. Jason Wang is
a doctoral fellow in the Pardee RAND Graduate School. Mayde Rosen is an associate
natural scientist at RAND. Robert Brook, vice president, directs the RAND Health
Program. Names and affiliations of the Expert Panel members are listed in the
paper's acknowledgments.
Read related papers by:
David
Brailer, Jonathan
Javitt, and W.
Ed Hammond.
DOI: 10.1377/hlthaff.W4.305
©2004 Project HOPE–The People-to-People Health Foundation, Inc.
|
