FOCUS: CPOE AND PATIENT SAFETY

CPOE Systems:
Success Factors
and Implementation
Issues
A B S T R A C T
The medication error dilemma has come to the forefront of most hospitals’ improvement
agendas. The most often cited solution to the problem has been computerized provider order
entry (CPOE) systems. These systems have significant potential to improve errors associated
with illegibility as well as inappropriate drug use and dosing. On the other hand, CPOE
system implementation is fraught with barriers that impede acceptance and use of these
systems. Knowing what strategies have proven successful and what upfront analysis is
required can help increase the chances of success and ultimately improve the quality of
patient care.

Patricia P. Sengstack RN, MS, and Brian Gugerty DNS, RN

ealthcare workers on the front lines of patient Medication errors top the list of mistakes that increase mor-

H care delivery were probably not as shocked

as the general public when the Institute of
Medicine (IOM) report, “To Err is Human”
(Kohn, Corrigan, and Donaldson, 1999), came out. It made
bidity and mortality in hospitalized patients. The literature
related to medication errors in terms of patient outcomes and
financial ramifications is startling. One solution to this partic-
ular problem is for hospitals to implement a CPOE system.
public the fact that healthcare workers make mistakes, and Some authors and researchers document very positive out-
to a significant degree. This eye-opening document has comes, while others are still wary about its ability to solve
resulted in unequalled public insistence for change. the entire medication error problem.

K E Y W O R D S
Computerized provider order entry (CPOE) Medication errors
System implementation Physician involvement

36 Journal of Healthcare Information Management — Vol. 18, No. 1

 FOCUS: CPOE AND PATIENT SAFETY

Installation of CPOE systems in hospitals nationally has (ISMP), a nonprofit organization that works closely with the
been described as slow. Several reasons have been noted healthcare industry to provide education regarding adverse
for the delayed penetration of this technology. High cost drug events, supports the IOM report and encourages the
and complexity of the task top the list of these reasons. use of a national adverse drug event (ADE) reporting system
Strategies that have resulted in successful implementations (ISMP, 2000). The Joint Commission for the Accreditation of
as well as lessons learned have been published in increas- Healthcare Organizations (JCAHO) added new patient safety
ing numbers. This paper reviews the literature associated standards beginning July 2001, which include enhanced error
with medication errors, and then analyzes the barriers to reporting, designation of qualified individuals to lead the
CPOE implementation. Strategies for successful implementa- safety improvement effort, and the development of risk iden-
tions proposed by various authors will then be reviewed tification and reduction programs (JCAHO, 2001a, 2001b).
along with a discussion of the complexity of these substan- The Medicare Payment Advisory Commission’s report to
tial projects. Congress recognizes the medical error problem and has
made recommendations for congressional action (Medicare
Medical Mistakes Payment Advisory Committee, 1999). The Leapfrog Group,
The landmark document, published by the IOM, “To Err is made up of a group of influential healthcare purchasers, has
Human” (Kohn and others, 1999), was the impetus for action also been active in education and support of patient safety
that brought patient safety and error reduction to the fore- issues (The Leapfrog Group, n.d.).
front of many healthcare improvement agendas. It
points out that literature on medical errors has Medication Errors
not been plentiful, but has increased over Unfortunately, medication errors
the past several years. The report con- are not uncommon in hospitals. There is
tains data from several studies that have “…the literature much literature documenting their inci-
been cited numerous times in subse- dence. The largest numbers of serious
quent healthcare literature (Armstrong consistently demonstrates that errors in hospitals involve medica-
and Chrischilles, 2000; Business tions (Thomas and others, 1999). In
Roundtable Press Release, 2000; medication errors are costly and a nine-year study conducted at a
Chaiken, 2001; Clark, 2001; First teaching hospital, Lesar, Briceland,
Consulting Group, 2000; The Institute occurring at an alarming and Stein (1997) reported that the
for Safe Medication Practice (ISMP), most common types of medication
2000; Jech, 2001; McConnell, 2001; errors were errors that involved pre-
Millennium Health Imperative, 2001; The
rate.” scribing medications to which the patient
National Academy of Sciences, 2001; was allergic and prescribing inappropriate
Simpson, 2000). dosage forms. Lesar, Lomaestro, and Pohl (1997)
Two of these often cited studies are the Harvard reported that hospitalized patients experienced an
Medical Practice Study and a similar study of adverse events overall rate of medication errors of 3.99 errors per 1,000
in two hospitals in Colorado and Utah. The Harvard Medical medication orders. Classen, Pestonik, Evans, and Burke
Practice Study found that adverse events occurred in 3.7 per- (1997) reported that adverse drug events increased hospital
cent of the hospitalizations, with 58 percent of those attribut- length of stay by 1.74 days, while in another study by
able to preventable adverse events with 13.6 percent of these Bates, Spell and Cullen (1997), it was estimated that pre-
adverse events leading to death. In the Colorado and Utah ventable adverse drug events increased length of stay by
study, adverse events took place in 2.9 percent of hospital- 4.6 days.
izations with 53 percent attributable to preventable errors When evaluating the financial ramifications of medication
and 8.8 percent leading to death (Kohn and others, 1999). errors, Schneider, Gift, Lee Rothermich and Sill (1995) esti-
When these numbers were extrapolated to the more than mated costs of medication-related problems at a university
33.6 million admissions in U.S. hospitals in 1997, it implies hospital as being $1.5 million in a year. Another study at
that at least 44,000 and perhaps as many as 98,000 people two large teaching hospitals discovered that 2 percent of
die in hospitals annually as a result of medical errors. admissions experienced a preventable ADE, resulting in an
The IOM document alone is more than likely responsible average increase in hospital costs of $4,700 per admission
for the majority of the increased efforts surrounding medical or about $2.8 million annually for a 700-bed hospital (Bates
error reduction in the United States. Government agencies and others, 1997). Drug related morbidity and mortality was
and task forces have sprung up to answer the call for estimated to cost $76.6 billion in the ambulatory setting in
improvement. The Institute for Safe Medication Practices the United States (Johnson and Bootman, 1995). Although

Journal of Healthcare Information Management — Vol. 18, No. 1 37

 FOCUS: CPOE AND PATIENT SAFETY

the numbers vary, the literature consistently demonstrates Order-entry systems can also be a powerful tool for
that medication errors are costly and occurring at an alarm- reducing unnecessary variation in care by encouraging rec-
ing rate. ommended practices through the use of online guidelines
or pathways (Teich and others, 1996). In another study by
The CPOE Solution Teich and others (2000), it was reported that the tools
Solutions to this problem have been offered by many included in these systems reduced errors associated with
authors, but the resounding theme is the use of technology inappropriate medication selection, inappropriate dosing for
and the application of computerized provider order entry several drugs, and inappropriate frequency of medication
(CPOE) systems (Armstrong, 2000; Armstrong and administration. Bates and others (1998) concluded that, in
Chrischilles, 2000; Bates and others, 1998; Chaiken, 2001; their institution during a six-month study period, CPOE
Dwyer, 1997; First Consulting Group, 2000; General Atlantic decreased the rate of non-intercepted serious medication
Partners, 2001; Hume, 1999; Jech, 2001; The Leapfrog errors by more than half, although this decrease was larger
Group; McConnell, 2001; Millennium Health Imperative, for potential ADEs than for errors that actually resulted in
2001; The National Academies Press Release, 2001; an ADE.
Simpson, 2000; Teich and others, 2000). The potential benefits of these computerized order-entry
In a report written for the American Hospital Association, systems range from legible orders, a non-trivial issue, and
Armstrong (2000) defines CPOE as “a system for direct completeness of orders, to alerts of possible contraindica-
entry of one or more types of medical orders by a tions based on unique patient information.
physician into a system that transmits those Common order entry errors can be reduced,
orders electronically to the appropriate such as (a) selecting the wrong drug for a
department” (p. 6). Many CPOE systems solution; (b) selecting the wrong dose,
include alerts, drug information, access
...order entry systems route, interval or duration; (c) over-
to evidence-based clinical guidelines looking drug allergies; (d) overlook-
and some degree of decision-support
tout the ability to create a ing drug-drug interactions; (e) over-
functionality. Since clinicians other legible, complete order and apply looking drug-laboratory value inter-
than physicians have the authority to actions; and (f) overlooking drug-
write orders, all types of practitioners logic-based rules to patient disease interactions. To sum it all up,
should be considered when referring order-entry systems tout the ability to
to CPOE (Chaiken, 2001; Simpson, information to prevent create a legible, complete order and
2000). apply logic-based rules to patient infor-
Electronic ordering of medications as a errors.” mation to prevent errors. Studies that
solution to the error problem seems to make evaluate the effectiveness of CPOE systems
sense given that medication errors are reported are generally positive and show results that
to be the largest cause of adverse hospital events CPOE systems do result in a reduction of medication
(Thomas, 1999, Simpson, 2000). A study by Bates, Spell, errors. At the same time, most do include the fact that this
and Cullen (1997) found that ordering is the largest source strategy alone will not be the complete panacea.
of medication-related error and that computerized ordering
reduced these errors. Teich and others (2000), in a study of Will CPOE Be Enough?
hospitalized patients, found that 56 percent of preventable In December 2000, the United States Pharmacopoeia
adverse drug events were primarily related to errors in pre- (USP) released the “Summary of 1999 Information
scribing. These studies conclude that the implementation of Submitted to MedMARx™, A National Database for Hospital
a CPOE system has the potential to significantly reduce Medication Error Reporting.” This report indicated that the
medication errors. three most frequently reported types of medication errors
One of the early users of CPOE has estimated that these were (a) omission errors (failure to administer an ordered
systems could potentially save an organization between $5 medication dose); (b) improper dose/quantity error (any
million and $10 million annually in medication costs by medication dose, strength, or quantity that differs from that
guiding physicians to effective lower doses or alternative prescribed); and (c) unauthorized drug errors (the medica-
medications and reducing adverse events through decision tion dispensed and /or administered was not authorized by
support tools (Teich and others, 2000). Other research has the prescriber, which includes wrong drug errors) (USP,
shown reduced use of resources, reduced length of stay, 2000, ANA News 2000). This differs from the earlier
and an overall reduction of costs with CPOE (Tierney and studies that concluded the ordering/transcribing phase was
Miller, 1993). most problematic.

38 Journal of Healthcare Information Management — Vol. 18, No. 1

 FOCUS: CPOE AND PATIENT SAFETY

The five phases of the medication management process touts CPOE as being a powerful and effective tool for med-
have been identified as ordering, dispensing, administering, ication error reduction. In the same article, there is mention
documenting, and monitoring, each with the potential for that CPOE is not the complete solution. In their research,
error commission or omission (ANA News, 2000). When they noted that physicians would often reject guidelines for
CPOE systems are analyzed based on these phases, appropriate drug use if they did not agree. The
it becomes clear that these systems will be Agency for Healthcare Research and Quality
primarily effective in the ordering phase,
but not as beneficial in other phases.
“W hile (2001) offers a combined approach to error
reduction including root cause analysis,
While it is obvious that a CPOE sys- it is obvious that a CPOE CPOE with decision support, automat-
tem would assist in reducing errors ed medication dispensing systems,
related to illegibility and inappro- system would assist in reducing errors patient wrist-band bar coding tech-
priate drug use and dosing, it
related to illegibility and inappropriate nology, aviation style preoperative
does not address the aspects of checklists, and integrating human
medication use that involve the drug use and dosing, it does not address factors theory into the design of
dispensing, administering, docu- medical devices
menting, and evaluating the the aspects of medication use that involve and alarms.
effectiveness of the drug. The respected author and
the dispensing, administering,
A feature article on the researcher Lucian Leape has been a
MEDerrors.com (2000) Web site is documenting, and evaluating the driving force in medical error reduc-
quick to point out that CPOE systems tion. He recognizes that CPOE alone
lacking drug information and access to effectiveness of the drug.” will not be adequate. He states that if an
patient data and rules have the potential to organization is serious about preventing
be more error prone. They argue for the imple- errors, the problem must be attacked on three
mentation of safety measures beyond CPOE. While sup- levels simultaneously: (a) culture change that allows for
porting the use of CPOE systems, First Consulting Group risk-free reporting of errors by employees, (b) the use of
(2000) suggests that hospitals could attempt several alterna- technology and CPOE systems, and (c) smaller incremental
tives, instead of the costly and complex CPOE solution. They changes. These smaller changes can take the form of having
suggest process changes and related interventions that have a pharmacist on each nursing unit, removing potassium chlo-
been shown to reduce errors such as (a) improving policies ride from the unit shelves, and initiating pre-printed orders
and procedures for medication history taking; (b) creating for high-risk situations, such as chemotherapy (Hume, 1999).
preprinted standing orders for common conditions; (c) estab- As shown in table 1, various improvement strategies will
lishing protocols for high-risk medications (anticoagulants, assist in the reduction of medication errors at different phas-
insulin, etc.); (d) including pharmacists in unit rounds, espe- es of the medication management process.
cially in ICUs; and (e) standardizing infusion devices across The overall opinion from the literature is that CPOE sys-
care units. While the Institute for Safe Medical Practice (n.d.) tems do significantly reduce medication errors, but not all
also encourages the use of CPOE systems, it notes there are of them.
other activities that must occur concurrently for more global
error reduction. Overcoming Barriers to Implementing CPOE
Research mentioned earlier by Teich and others (2000) The statistics in the literature are somewhat disappoint-

Journal of Healthcare Information Management — Vol. 18, No. 1 39

 FOCUS: CPOE AND PATIENT SAFETY

ing regarding the prevalence of use of CPOE in the United infrastructure, hardware, software, and considerable process
States. A survey of 1,000 hospitals showed that 66 percent redesign and change management (Chaiken, 2001). First
do not have any form of computer order entry available. Of Consulting Group (2000, p. 4) states, “Successful CPOE is
the 33 percent that do have such systems, more than half not a technology implementation but a redesign of a com-
reported that less than 10 percent of orders are entered into plex clinical process integrating the technology at key
the system and less than 10 percent of physicians use the points to enhance and optimize ordering decisions.” It is
system (Jech, 2001). apparent from the literature that there is not one foolproof
The literature, discussing possible rationale for low use, recipe for a successful CPOE implementation; however,
identifies several factors. For many years, hospital execu- there are many authors who offer their advice based on
tives and most vendors of hospital information systems experience. The following sections will discuss a number of
believed that physicians would not use CPOE. Another rea- factors that have reportedly led to successful implementa-
son for slow adoption is that demand has been low, and tions of CPOE systems.
has resulted in minimal motivation for development on the
part of the vendor community. Those systems that were Continuous Administrative Support
available were perceived as less than adequate. This Support for the project at the executive level is a neces-
appears to be changing as demand and interest increase sary prerequisite, even prior to considering this huge
(McConnell, 2001; First Consulting Group, 2000). undertaking. Not only is there a need for support at the ini-
Ferren (2001) reports that physicians at a tial stages, but it is required on an ongoing basis
Pennsylvania hospital, where CPOE was (Ferren, 2001). Management has a responsi-
recently implemented, were reluctant to
embrace the technology for a number
“A bility to develop a clear vision of CPOE
and address such issues as (a) why
of reasons. The reason most com- survey of 1,000 hospitals do we wish to pursue this, (b)
monly stated was that of fear of what do we expect to get out of
using the system. Some physi-
showed that 66 percent do not have it, and (c) how will we measure
cians claimed that their hand- any form of computer order entry success.
writing was perfect and illegi- Ahmad and others (2002)
bility was not an issue. Others available. Of the 33 percent that do have reported several strategies for
held the belief that handwritten success as a result of their
orders were faster to execute
such systems, more than half reported that nine-month pilot study at Ohio
and that computerized order less than 10 percent of orders are entered State University. Hospital
entry would be more cumber- administrators realized the
some. into the system and less than 10 importance of having physicians
The costs surrounding an imple- involved. They recruited and fund-
mentation of this type of system are
percent of physicians use ed a physician team to lead this pro-
extremely high and would be a signifi- the system.” ject. These physicians were required to
cant capital investment. Brigham and sign a contract outlining their responsibili-
Women’s Hospital, one of the pioneers of ties. With strong backing from physician leader-
CPOE, has estimated the cost of developing and imple- ship and administration, numerous efforts to standardize
menting such a system in 1992 at $1.9 million, with ongo- practices and policies across the organization became an
ing maintenance costs of $500,000 per year. Purchasing and integral part of the CPOE project. For example, only elec-
implementing large vendor-developed systems today can tronic ordering was allowed. Administration and physicians
vary greatly in cost, but could be in the tens of millions of backed this policy.
dollars (Kaushal and Bates, n.d.). Government support is Once implemented, ongoing CPOE support is reported
improving and taking the form of grant money. U.S. Sens. to be crucial for success. The administrative team should
Bob Graham and Olympia Snowe introduced legislation embrace the philosophy of continued support and provide
that would establish a 10-year, $1-billion hospital and adequate funds to ensure that help is always available.
skilled nursing facility grant program to offset the prohibi- When Ohio State University Health System implemented its
tively high cost of developing and implementing CPOE CPOE system, administration created permanent support
systems (Lovern, 2001). positions that remained on-site from 6 a.m. to 11 p.m.,
Probably the largest reason for the slow implementation seven days a week. These support personnel wore red
of CPOE systems is the complexity of the entire task. tunics and were dubbed the “red coats” (Ahmad and
Implementing CPOE is admittedly costly because it involves others, 2002).

40 Journal of Healthcare Information Management — Vol. 18, No. 1

 FOCUS: CPOE AND PATIENT SAFETY

During implementation of CPOE at Abington Memorial various levels and from various departments. Departments
Hospital in Pennsylvania, a group of super-users were impacted by CPOE should play a role in the implementa-
trained and placed on each unit. They wore identifying but- tion of the system. Because CPOE has such far-reaching
tons that read, “TDS Super User – May I Help You?” They effects, the task of determining who should be on the team
were trained to approach physicians and ask if they could can be somewhat difficult. Most teams are made up of a
help them with the system. They distributed a User Tips core group of administrators, physicians, nurses, and infor-
Booklet and ensured that the phone number for the help mation services staff, with ancillary department involvement
desk was prominently displayed (Ferren, 2001). as needed. The team could potentially be too large to be
Several authors note that a CPOE system rollout is a sig- effective or too small to garner adequate support if all
nificant milestone, but only the beginning of intensive sup- departments are not included. Team composition is very
port. Providing this assistance on a continuous basis will important and unique to each institution. The task of team
have a strong bearing on the system’s success or failure. selection should not be taken lightly.
Planting a seed and neglecting to water or fertilize it will
ultimately lead to a weak and unhealthy plant. The analogy Analysis and Considerations Prior to Implementation
can be made to implementing CPOE and not providing The preceding sections have emphasized the importance
strong ongoing administrative support at all levels and from of human resources required to undertake this important
all departments involved. project. For an implementation to be successful, these dedi-
cated resources will be involved in an extensive
Physician Empowerment amount of upfront work that includes system
and Involvement analysis, integration between ancillary sys-
Physician involvement, in combination tems, work process analysis and
with administrative support, tops the
list for the most cited reason for suc-
“P hysician redesign, and review of organizational
culture. Upfront analysis is as impera-
cess. In fact, it was noted in one arti- involvement, in combination tive as effective project leadership.
cle that a failed CPOE implementa- This somewhat abstract, seeming-
tion was the result of not having
with administrative support, ly interminable task is probably the
adequate physician representation on tops the list for the most-cited most difficult of all phases of the
the implementation team. It is recom- project. Yet without thorough consid-
mended that a physician champion be reason for success..” eration and detail given to this task,
committed to this project’s efforts who CPOE implementation could be fraught
is a trusted, clinically active, well-respect- with obstacles and frustrations. It is, in
ed physician (First Consulting Group, 2000). the author’s opinion, one of the least under-
Successful implementations in the literature stood and poorly documented components of
describe a physician team that is intimately involved the CPOE implementation puzzle. Several articles have
with all steps of the system’s implementation. At one hospi- stated that CPOE is not solely a technological solution to
tal, a Physician Advisory Group was charged with the over- the medication error problem. They emphasize that it
sight of all clinical aspects of CPOE implementation. This involves much more than computer technology and, if
group comprised the department chairs from general medi- there is not adequate support to work through the organi-
cine, internal medicine, pulmonary medicine, urology, and zational culture and dynamics surrounding the implementa-
surgery (Ferren, 2001). Representation from such a wide tion of CPOE, it may result in failure (Kaushal and Bates,
range of departments led to a powerful support base for n.d.).
CPOE. An important task for the implementation team will be to
Physician buy-in, involvement, and championship are assess and analyze current processes and determine how
common themes in the literature that are reported to result these will be affected by CPOE. Eisenburg and Barbell
in successful CPOE implementations. The message to hospi- (n.d.) have broken down the physician order entry work-
tal administrators contemplating acquisition of a CPOE sys- flow into eight steps, (1) access the system, (2) select a
tem is clear – plan on making a resource commitment patient, (3) review patient data, (4) enter data, (5) sign/con-
toward physician involvement. firm order, (6) order is processed, (7) receive results/take
action, and (8) outcomes and accountability are measured.
An Effective Implementation Team Each of these steps will be affected by the implementa-
In addition to administrative and physician support, the tion of CPOE and could benefit from evaluation. For exam-
implementation team should represent the organization at ple, a hospital’s current process for the first step of access-

Journal of Healthcare Information Management — Vol. 18, No. 1 41

 FOCUS: CPOE AND PATIENT SAFETY

ing the system may be to find the chart and open it. With sure to improve efficiency? The development of the mission
CPOE, access includes system availability (uptime, down- and vision statements for this project will stem from the
time), network connections and interfaces, speed of sign-on motivation to proceed and begin to build a relationship
and screen flips, and availability of an adequate number of among implementation team members.
devices. Each of these considerations must be discussed Another consideration is questioning whether there
prior to CPOE implementation to determine levels of accep- exists a strong foundation prior to implementation that
tance. Inadequate review of each of these process steps includes adequate resources in terms of people and
prior to implementation can create implementation delays finances. This foundation relies upon an administration that
as issues are worked through. is trusted and is an advocate for change and embraces
The complexity of CPOE implementation has spawned a continuous learning. Because of the high costs associated
group that calls itself the Physician Order Entry Team with CPOE, organizations need to have a long-term finan-
(POET). POET is based in the Division of Medical cial plan and understand the total costs of CPOE.
Informatics and Outcomes Research at Oregon Health and Organizations must ask themselves, “Can we afford a tem-
Science University. It consists of representatives from a vari- porary loss of productivity?” as is often seen during the
ety of healthcare-related disciplines. Membership of the period of time just after implementation.
team varies over time, but all members use research meth- Hospitals are advised by the Menucha group to give
ods that consist of cross-site comparisons using multiple thought to their organization-wide change strategy as they
qualitative and quantitative data-gathering techniques to consider CPOE. Because it is a given that work processes
assess implementation issues regarding CPOE. POET is will change, a plan to analyze and work through these
funded by grants from the National Library of Medicine, the changes is necessary. Response times, ordering times, and
U.S. Department of Energy, and other resources (Physician communication times all need to be reviewed in terms of
Order Entry Team, 2002). what effect CPOE will bring. How will retrieval of infor-
In May 2001, at a retreat center called Menucha, 13 mation occur and how will this affect integration with
CPOE experts met with seven members of POET for the other systems? A plan for organizational readiness for
purpose of developing recommendations for CPOE imple- CPOE that key stakeholders mutually agree upon is
mentation. The Menucha Consensus Conference succeeded strongly advised.
in its goals of identifying and agreeing on a list of consider- The value that CPOE will provide to clinicians is anoth-
ations and questions organizations may choose to work er consideration that should be discussed prior to imple-
through to increase their potential for an implementation of mentation. Analyzing this concept will assist organizations
CPOE that is judged to be successful. A list of 10 high-level in promoting and marketing the system. Will the proposed
considerations was generated to benefit organizations think- system include decision support, order sets, and alerts? Will
ing about implementing CPOE (The 2001 Menucha efficiency and patient care be improved? In other words,
Conference List, 2002). Discussion and analysis of some of what’s in it for the individual physicians and practitioners
the highlights of these considerations follow this list: using the system, and how will it benefit their practice?
1. Motivation for Implementing POE One of the primary technological considerations that
2. Foundations Needed Prior to Implementing POE should be addressed is the security plan. It is recommend-
3. Costs ed that security of access and confidentiality issues be
4. Integration/Workflow/Health Care Processes carefully planned, especially with impending enforcement
5. Value to Users/Decision Support Systems of the Health Insurance Portability and Accountability Act
6. Vision/Leadership/People (HIPAA). Advantages and disadvantages of a single sign-on
7. Technical Considerations need consideration.
8. Management of Project or In addition, organizations should assess the ability to and
Program/Strategies/Processes from Concept the desire to customize the CPOE system. It is advisable to
to Implementation determine if a system administrator will be able to perform
9. Training/Support/Help at the Elbow customizations or whether the vendor will need to perform
10. Learning/Evaluation/Improvement this task, and for what fee. Other technical questions to
Motivating factors for implementing CPOE at the organi- consider include: (a) How will data integrity be ensured?
zation level need to be identified and agreed upon. A gen- (b) How will the system interact with current systems? (c)
eral consensus among stakeholders of why CPOE is being Will remote access be needed and is the network infrastruc-
pursued should be the impetus for moving forward with ture stable? (d) Is there an easy way to exit the system for
CPOE implementation. Is there pressure from administra- frustrated users? (e) Is there a consistent and friendly graph-
tion, clinicians, and/or regulatory agencies, or is there pres- ic user interface?

42 Journal of Healthcare Information Management — Vol. 18, No. 1

 FOCUS: CPOE AND PATIENT SAFETY

Project management strategies to consider include over- cians were more clinically oriented, administrators were
all impact of changing workflow processes. Managing more fiscally focused. Information technology personnel
people issues will come into play and should be tended to focus on the complexities of system integration
addressed. The project manager will need to ensure that and the frustrations with vendors for not producing a per-
plans are detailed enough – but not too detailed. It is rec- fect product. The point the author makes is that it is imper-
ommended that there is not exaggerated attention to ative that those implementing CPOE understand all views
details that would jeopardize the overall implementation and plan implementation strategies with this in mind (Ash,
goal. Keep it simple; strive for excellence, not perfection. Gorman, Lavelle, and Lyman, 2000). Given the large num-
Are there clear and measurable goals with a clear commu- ber of factors to consider prior to implementation of CPOE,
nication plan? it is not a wonder that it takes organizations years to imple-
Considerations for training and support include the con- ment such systems.
cept of “help at the elbow.” This means ongoing, readily
available help. What will be the training methodology Conclusion
employed (one-on-one, group sessions, train the trainer, Medication errors in hospitals are significant in volume,
etc.) ? A formalized training program is the normal process contribute to the morbidity and mortality of patients, and
that occurs from two to four weeks prior to “go live” of the increase resource utilization and costs. The CPOE solution,
system. However, getting busy physicians to attend training while costly and complex, is one way healthcare organiza-
can be difficult. One organization dealt with this tions can reduce a significant amount of medica-
by having the chairperson of its Patient Safety tion errors. To be on the realistic side, how-
Committee send a letter to all division
chiefs. Letters were also sent to all tar-
“S uccess ever, organizations should be aware that
CPOE systems will not solve the entire
geted physicians. The trainers followed stories and lessons learned medication error problem. This
up with non-responders with three becomes evident as one examines the
personal phone calls. If these follow- from experienced institutions medication management process in
up calls did not result in attendance conjunction with the functionality of
should be shared and ideas
at a training session, the chairperson CPOE systems.
of the Patient Safety Committee was incorporated into vendor products Several reasons are cited for orga-
notified. This very respected and per- nizational delay in implementing
suasive leader then made a personal and in organizational CPOE systems. Complexity of the task
call to the physician. This strategy was and high costs are common factors,
implementation plans.”
reported as being very effective (Ferren, while other reasons include lack of physi-
2001). In addition to initial training efforts, cian acceptance and inadequate develop-
the plan for ongoing education and competency ment of the product by vendors. The work of
assessment should be addressed. process re-engineering adds to the hesitancy to move
The final consideration involves the methodology for forward with CPOE implementation. Despite these barriers,
system evaluation and testing. Consider carefully planning a there have been successful implementations documented at
process for problem identification and problem resolution many facilities. These organizations attribute their success
involving the users. How will you continuously improve the to factors such as ongoing administrative support, physician
system? Understand that you are never done. This is truly involvement, an effective implementation team, and sub-
an iterative process. stantial planning.
In a qualitative study by Ash and others (2002), a Ongoing research conducted in hospital settings with
Multiple Perspective Model was used for organizing the differing characteristics is recommended. Success stories
descriptions of CPOE by physicians, administrators, and and lessons learned from experienced institutions should
information technology staff. This was done to aid in the be shared and ideas incorporated into vendor products
understanding of all points of view of CPOE. Discussion and in organizational implementation plans. As more orga-
comments regarding CPOE were organized into the cate- nizations move forward with CPOE systems, there will be a
gories of technical, organizational, and personal systems. broader knowledge base from which to draw. Continued
Notes from 120 person-hours of observation and audiotapes publication of the CPOE implementation experience will
from 22 hours of formal interview and focus groups were assist organizations contemplating what seems to be an
transcribed and resulted in more than 400 pages of tran- enormous undertaking, but one that is well worth the
scripts. The results demonstrated the unique differences of effort in terms of making improvements to the care and
perspective of CPOE among the three groups. While physi- safety of our patients.

Journal of Healthcare Information Management — Vol. 18, No. 1 43

 FOCUS: CPOE AND PATIENT SAFETY

About the Authors Dr. Brian Gugerty, DNS, RN, is an assistant professor of
Patricia P. Sengstack, RN, MS, is currently working as a nursing informatics at the University of Maryland School of
clinical coordinator/educator for the Information Services Nursing. He co-developed the Clinical Information System
department at Adventist Healthcare in suburban Maryland Questionnaire (CISQ) family of measurement instruments to
with responsibilities of project management, implementa- assess staff involvement in and attitudes toward the imple-
tion, systems analysis, and informatics education. mentation of clinical information systems.

References
Agency for Healthcare Research and Quality (2001). Making health care Hume, M. (1999). Changing hospital culture and systems reduces drug
safer: A critical analysis of patient safety practices. Retrieved Aug. errors and adverse events. The Quality Letter for Healthcare
22, 2001, from http://www.ahrq.gov/clinic/ptsafety/. Leaders, 11, 3, 1-9.
Ahmad, A., Teater, P., Bentley, T. D., Kuen, L., Kumar, R. R., Thomas A, The Institute for Safe Medication Practices (2000). Discussion paper on
and Mekhjian, H. (2002). Key attributes of a successful physician adverse event and error reporting in healthcare. Retrieved Aug. 8,
order entry system implementation in a multi-hospital environment. 2001, from http://www.ismp.org/Pages/IOM.html.
Journal of the American Informatics Association, 9, 16-24. The Institute for Safe Medication Practices. (n.d.). Prevent adverse drug
American Nurses Association / ANA News (2000). Report analyzes events before they occur! ISMP error reduction activities. Retrieved
common medication errors. Retrieved Aug. 30, 2001, from Aug. 8, 2001, from http://www.ismp.org/Pages/ismp_erract.html.
http://nursingworld.org/news/nws12_00.htm. Jech, A. O. (2001). The next step in preventing errors. RN, 64, 4, 46-49.
Armstrong, C. W. (2000). American Hospital Association guide to Johnson, J. A., and Bootman, L. J. (1995). Drug-related morbidity and
computerized physician order-entry systems. Chicago, IL: American mortality: A cost-of-illness model. Archives of Internal Medicine,
Hospital Association. 155, 18, 1949-1956.
Armstrong, E. P. and Chrischilles, E. A. (2000). Electronic prescribing and The Joint Commission on Accreditation of Healthcare Organizations
monitoring are needed to improve drug use. Archives of Internal (2001a). Medication errors related to potentially dangerous
Medicine, 160, 2713-2714. abbreviations. Sentinel Event Alert. Issue 23. Retrieved Oct. 18,
Ash, J. S., Gorman, P. N., Lavelle, M. and Lyman, J. (2000). Multiple 2001, from http://172.16.17.109/sea/sea23.html.
perspectives on physician order entry. Proceedings of the AMIA The Joint Commission for the Accreditation of Healthcare Organizations
Annual Symposium, USA, 27-31. (2001b). Revisions to Joint Commission standards in support of
Bates, D. W., Leape, L. L., Cullen, D. J., Laird, N., Petersen, L. A., Teich, patient safety and medical/health care error reduction: Effective
J. M. and others. (1998). Effect of Computerized Provider Order July 1, 2001. Retrieved July 3, 2001, from http://www.jcaho.org/
Entry and a team intervention on prevention of serious medication standard/fr_ptsafety.html.
errors. Journal of the American Medical Association, 280, 15, 1311- Kaushal, R. and Bates, D. (n.d.). Computerized physician order entry
1316. (CPOE) with clinical decision support systems (CDSSs). Retrieved
Bates, D. W., Spell, N., Cullen, D. J. (1997). The costs of adverse drug Sept. 5, 2002, from http://www.ahcpr.gov/clinic/ptsafety/chap6.htm.
events in hospitalized patients. Journal of the American Medical Kohn, L. T., Corrigan, J. M. and Donaldson, M. S. (eds.). (1999). To err is
Association, 277, 307-311. human: Building a safer health system. Washington, DC: National
The Business Roundtable Press Release. (2000). BRT – sponsored Academy Press.
initiative focuses on patient safety. Retrieved Jan. 29, 2001, from The Leapfrog Group. (n.d.). Patient safety; Leapfrog initiatives to drive
http://www.brtable.org/press.cfm/375. great leaps in patient safety. Retrieved Oct. 1, 2001, from
Chaiken, B. P. (2001). Enhancing patient safety with clinically intelligent http://www.leapfroggroup.org/safety1.htm.
physician order entry. Nursing Economics, 19, 3, 119-120. Lesar, T. S., Briceland, L. and Stein, D. S. (1997). Factors related to errors
Clark, A. P. (2001). What will it take to reduce errors in healthcare in medication prescribing. Journal of the American Medical
settings? Clinical Nurse Specialist, 15, 4, 182-183. Association, 277, 4, 312 – 317.
Classen, D., Pestonik, S. L., Evans, R. S., Burke J. P. (1997). Adverse drug Lesar, T. S., Lomaestro, B. M. and Pohl, H. (1997). Medication prescribing
events in hospitalized patients: Excess length of stay, extra costs, errors in a teaching hospital: A 9-year experience. Archives in
and attributable mortality. Journal of the American Medical Internal Medicine, 157, 14, 1569-1576.
Association, 277, 301-306. Lovern, E. (2001). A dose of safety: Senate bill would help facilities buy
Dwyer, K. (1997). The role of human factors research in reducing technology to reduce medication errors. Modern Healthcare.
medical errors: A conversation with Dr. Lucian Leape. Forum, 17, 5, Retrieved Aug. 10, 2001, from http://www.modernhealthcare.com/
6. Retrieved March 3, 2000, from http://www.rmf.org/b7395.html. topten.php3?refid=7117&db=mh99up&published=2001050.
Eisenberg, F. and Barbell, A. (n.d.). Clinical informatics: Computerized McConnell, T. (2001). Safer, cheaper, smarter: Computerized physician
physician order entry: Eight steps to optimize physician workflow. order entry promises to streamline and improve healthcare delivery.
Journal of Healthcare Information Management. Retrieved Sept. 5, Health Management Technology. Retrieved Oct. 1, 2001, from
2002, from http://www.himss.org/ASP/ContentRedriector.asp? http://www.healthmgttech.com/cgibin/arttop.asp?Page=h0301
ContentID=14163. safer.htm.
Ferren, A. (2001). Gaining MD buy-in: Physician order entry. Journal of Mederrors.com. (2000). Reality check: Computerized physician order
Healthcare Information Management, 16, 2, 66-70. entry is not enough. Retrieved from http://www.mederrors.com/
First Consulting Group (2000). A primer on physician order entry feature/feature4.html.
[Brochure]. Oakland, CA: First Consulting Group. Medicare Payment Advisory Commission. (1999). Addressing health care
General Atlantic Partners (2001). Two independent surveys rate Eclipsys’ errors under Medicare. In Report to Congress: Selected Medicare
Sunrise Solutions Physicians Order Entry, clinical alert. Retrieved issues.
October 1, 2001, from http://www.gapartners.com/ Millennium Health Imperative. (2001). Prescribed pathways for an
newseclipsys31july2001.asp. industry on the critical list. Kansas City, MO: Millennium Health
Imperative.

44 Journal of Healthcare Information Management — Vol. 18, No. 1

 FOCUS: CPOE AND PATIENT SAFETY

The National Academy of Sciences (2001). U.S. Healthcare delivery The 2001 Menucha Conference List. (2002). Considerations concerning
system needs major overhaul to improve quality and safety. computerized physician order entry implementation. Retrieved
Retrieved Aug. 10, 2001, from http:www4.nationalacademies.org/ Sept. 5, 2002, from http://www.ohsu.edu/biccinformatics/
news.nsf/isbn/0309072802?OpenDocument. research/cpoe/consider_allv1_1.html#1.
Physician Order Entry Team (2002). Studying success factors for POE Tierney, W. M., and Miller, M. E. (1993). Physician inpatient order writing
implementation. Retrieved from http://www.ohsu.edu/bicc- on microcomputer workstations: Effects on resource utilization.
informatics/research/cpoe/poet.htm. Journal of the American Medical Association, 269, 379-383
Schneider, P. J., Gift, M. G., Lee, Y., Rothermich, E. A. and Sill, B. E. Thomas, E. J., Studdert, D. M., Newhouse, J. P., Zbar, B. I., Howard, K.
(1995). Cost of medication-related problems at a university hospital. M., Williams, E. J., and others. (1999). Costs of medical injuries in
American Journal of Health-System Pharmacy, 52, 21, 2415-2418. Utah and Colorado. Inquiry, 36, 255-264.
Simpson, R. L. (2000). Nurses – yes, nurses – improve physician order United States Pharmacopoeia (2000). USP releases first report on
entry. Nursing Management, 31, 9, 20-22. medication errors: Trends in severity, types and causes of errors
Teich J. M., Glaser, J. P., Beckley, R. F., Aranow, M., Bates, D. W., identified. News Release. Retrieved Oct. 23, 2001 from:
Kuperman, G. J. and others. (1996). The Brigham integrated http://www.usp.org/aboutusp/releases/2000/pr_2000-39.htm.
computing system: Advanced clinical systems in an academic
hospital environment. International Journal of Medical Informatics,
54, 3,197-208.
Teich, J. M., Merchis, P. R., Schmiz, B. D., Kuperman, M. .D., Spurr, C. D.
and Bates, M. D. (2000). Effects of computerized physician order
entry on prescribing practices. Archives of Internal Medicine, 160,
2741-2747

Journal of Healthcare Information Management — Vol. 18, No. 1 45