Search tips
Search criteria 


Logo of jamiaAlertsAuthor InstructionsSubmitAboutJAMIA - The Journal of the American Medical Informatics Association
J Am Med Inform Assoc. 2011 Nov-Dec; 18(6): 774–782.
Published online 2011 July 29. doi:  10.1136/amiajnl-2011-000255
PMCID: PMC3198002

Factors contributing to an increase in duplicate medication order errors after CPOE implementation



To evaluate the incidence of duplicate medication orders before and after computerized provider order entry (CPOE) with clinical decision support (CDS) implementation and identify contributing factors.


CPOE with duplicate medication order alerts was implemented in a 400-bed Northeastern US community tertiary care teaching hospital. In a pre-implementation post-implementation design, trained nurses used chart review, computer-generated reports of medication orders, provider alerts, and staff reports to identify medication errors in two intensive care units (ICUs).


Medication error data were adjudicated by a physician and a human factors engineer for error stage and type. A qualitative analysis of duplicate medication ordering errors was performed to identify contributing factors.


Data were collected for 4147 patient-days pre-implementation and 4013 patient-days post-implementation. Duplicate medication ordering errors increased after CPOE implementation (pre: 48 errors, 2.6% total; post: 167 errors, 8.1% total; p<0.0001). Most post-implementation duplicate orders were either for the identical order or the same medication. Contributing factors included: (1) provider ordering practices and computer availability, for example, two orders placed within minutes by different providers on rounds; (2) communication and hand-offs, for example, duplicate orders around shift change; (3) CDS and medication database design, for example confusing alert content, high false-positive alert rate, and CDS algorithms missing true duplicates; (4) CPOE data display, for example, difficulty reviewing existing orders; and (5) local CDS design, for example, medications in order sets defaulted as ordered.


Duplicate medication order errors increased with CPOE and CDS implementation. Many work system factors, including the CPOE, CDS, and medication database design, contributed to their occurrence.

Keywords: Human factors, EHR, informatics


Computerized provider order entry (CPOE) with clinical decision support (CDS) is a critical component of the electronic health record's (EHR's) order-management system. To be safe and effective, this order-management system must support the patient's care team as it cooperates in performing the subprocesses of: (1) care-plan development and communication (by physicians, nurses, and pharmacists); (2) order planning (by physicians, nurses, and pharmacists); (3) order entry (by nurses and physicians); (4) order review, modification, and fulfillment (by pharmacists, nurses, and physicians); and (5) order review and administration (by nurses and physicians). Computerized provider order entry has been shown to decrease medication ordering errors,1 although increased errors have also been noted.2 CDS has been described as ‘making the right thing to do the easy thing to do.’3 It is specifically designed to support the user in safe medication order planning and entry and the real-time recognition of potential medication errors. Basic decision support for medication ordering includes warnings regarding potential duplicate orders, potential drug–drug interactions, and patient allergy information.4 Although CDS can help users decrease medication ordering errors, CDS warnings are commonly overridden by users. In the 10–60% of the cases in which the warnings represent true positives, ignoring the alert could compromise care quality.5

Despite overall reductions in medication errors, multiple studies have noted continued or increased duplicate order entry after CPOE implementation.4 6–9 Most of these studies have not examined why the errors persist or increase after implementation, particularly when CDS is implemented to reduce a specific error type. Studies that evaluate provider response to CDS alerts often focus on provider attitudes or perceptions about alert usability or usefulness without considering the design of the alert, for example, the decision support algorithms behind the alert and how the alert is presented to the user. Research into factors that may lead to inappropriate warning overrides is needed to improve system and technology design.10

We performed a prospective pre-observational post-observational study to evaluate the impact of the implementation of CPOE with CDS on the frequency of duplicate medication ordering. We then performed an in-depth analysis of work system factors contributing to the increased overall rates of duplicate medication orders post-implementation.


The study was performed in two intensive care units (ICUs) of a 400-bed rural, community tertiary care teaching hospital in the Northeast US. The adult ICU (AICU) specializes in critical care, trauma, and post-surgical care. It expanded from 18 to 24 beds during the 3 months of data collection prior to CPOE implementation. The 18-bed cardiac ICU (CICU) specializes in cardiothoracic surgery, cardiovascular critical care, liver transplantation, and adult critical care overflow.

The medication ordering processes in the two ICUs are similar. An ICU pharmacist processes all medication orders from 06:30 to 15:00 h daily and joins physician team rounds when possible. At other times the medication orders are reviewed by a pharmacist in the central pharmacy. Before CPOE implementation, the pharmacist transcribed the order into the pharmacy computer system and reviewed orders for appropriateness, dosing, allergies, duplications, and drug–drug interactions. If this review indicated a change in a medication order, the pharmacist consulted with the ordering physician or made changes to the order by protocol (eg, dosing adjustment based on renal function). The pharmacy computer system supported this review with alerts for duplicate therapy, allergies, and drug–drug interactions. Pre-CPOE, all medication orders were entered on paper order sheets or pre-printed order sets. Post-CPOE, all orders are entered electronically by providers except for chemotherapy orders, which are entered by trained pharmacists and are rare in the ICU. Clinical decision support alerts are sent to providers at the time of signing the orders. All CDS alerts can be overridden by the provider. Pharmacists use the EHR pharmacy computer system which is integrated with CPOE. Pharmacists no longer transcribe orders. Clinical decision support alerts regarding duplicate therapy, drug–drug interactions, and allergies are presented to the pharmacist at the time of electronic order verification. Pharmacists make changes to medication orders by modifying an existing order or discontinuing an order and entering a new order using the same protocols as previously.

The EHR product studied was the EpicCare Inpatient Clinical System version Spring 2006 (Epic Systems, Verona, Wisconsin, USA). The CPOE with a CDS system, clinical documentation (nurse and physician), pharmacy system, and electronic medication administration record were implemented together throughout the hospital in October 2007. All hospital providers, except the neonatal ICU, had been using the integrated outpatient EHR prior to October 2007. Hospital nurses had been using the EHR nursing documentation component for 6 months. All clinicians had electronic access to test results, radiology images, and electronic messaging since 2002. Clinical decision support for medication ordering included the standard medication package for drug–drug interactions, drug allergy, and duplicate order alerts. Medi-Span provides the content for the medication database in the EHR and determines both the sensitivity and specificity of drug–drug interaction and duplicate medication alerting. Specifically, Medi-Span classifies a medication product as the combination of medication name, route of administration, and manufacturer. Medication ingredients are not included in this algorithm. The EHR's duplicate alerts are designed to identify duplications of the same medication product or the same therapeutic medication class, for example, two antibiotics. Same medication (medication product) duplicate checking considers the same medication orders to be duplicates if the routes of administration are treated as being the same. For example, the oral and nasogastric routes are treated as oral administration and would trigger a duplicate alert. However, because the oral and intravenous forms of a medication are treated as two different medication products, duplicate checking would not trigger an alert if both routes were ordered. To identify duplicate medications, the EHR compares the new orders to each other as well as to currently active medication orders for the patient.


This study is part of a larger study, ‘Computerized Provider Order Entry (CPOE) Implementation in ICUs’ (, which is designed to evaluate the impact of computerized order management (ie, computerized order entry, pharmacy management, and medication administration) on worker, organizational, and patient outcomes. We conducted a prospective, pre-intervention post-intervention observational trial. The data collection pre-intervention spanned 30 weeks, from October 29, 2006 to March 3, 2007 in the AICU and from January 1 to March 22, 2007 in the CICU. Post-intervention data collection occurred in the AICU from March 3 to June 13, 2008 and in the CICU from April 7 to June 23, 2008 (25 weeks in total). Post-intervention data collection was completed before the new class of residents started on July 1.

Data were collected on all patients admitted to the ICUs during the study period until the pre-determined sample size of 300 patients in the AICU and 325 patients in the CICU was achieved. Patients were excluded from the study if they were under 18 years of age, prisoners, or if they spent less than 4 h in the ICU.

Medication error data collection

Medication error and adverse drug event (ADE) data were collected by trained nurses following a protocol adapted from Bates et al.11 Events were identified by chart review, pharmacist's records, staff report, and review of staff-generated hospital event reports. In the post-intervention data collection period, electronic chart review was supplemented with computer-generated reports of medication orders, alerts to providers, and pharmacist modifications of orders. After a medication error was found, standard data were collected. Events were adjudicated by a physician and a human factors engineer to determine whether a medication error had occurred, the stage of medication management (eg, ordering, administration), and the type of error. A duplicate medication order was defined as two or more orders for the same medication or for medications in the same therapeutic class, one of which was clinically redundant. Duplicate medication orders were considered potential errors if the provider overrode a true positive alert or if an intervention was performed by pharmacy or nursing staff to correct duplicate orders. If providers received a duplicate alert and changed the order, these orders were not considered duplicate medication errors as the alert was the purpose of the CDS and the order was never finalized in the computer system. Duplicate ordering errors were further categorized into three categories: (1) an identical order, (2) the same medication (but different dose, form, frequency, or route), and (3) a different medication of the same therapeutic class.

Survey questionnaire

A survey questionnaire was administered to attending and resident physicians, physician assistants, and nurse practitioners in the ICUs 3 months and 1 year post-CPOE implementation to evaluate their perceptions of the order entry system.12 The design and procedure of the questionnaire are described in detail elsewhere.13 Two questions were adapted from a previously validated questionnaire assessing provider perceptions of the usefulness of the duplicate medication CDS alerts: (1) How useful are duplicate medication order warnings in identifying a problem with a medication order? and (2) How useful are duplicate medication order warnings in correcting a problem with a medication order?14 A 7-point Likert scale was used with 1 representing ‘not useful at all’ to 7, ‘extremely useful.’

Data analysis

Descriptive statistics evaluated the frequency of medication errors by stage and type. χ2 Analyses compared the frequency of duplicate orders pre- and post-CPOE implementation and user perceptions of duplicate alerts at 3 months and 1 year post-implementation. SPSS version 16 was used for quantitative analyses. A content analysis was performed (TBW) of all duplicate orders to identify themes related to duplicate order errors and work system factors that contributed to the duplicate orders. Available data for each duplicate order included an event description from the nurse data collector, the ICU it occurred in, the medication orders placed, and the time of the orders. The Systems Engineering Initiative for Patient Safety (SEIPS) model was used to identify and categorize contributing factors from all elements of the work system: tasks, tools and technology, organization, environment, and people.15 16 The results of the content analysis were reviewed and verified by research team members (JMW, PC, RSC) and triangulated with findings from other data collected in the larger study, for example, from unstructured observation, brief interviews and meetings with hospital staff, job-task analysis, provider surveys, and usability evaluation. An iterative process of data review and comment was used to determine the final set of contributing factors and generate a list of proposed solutions to decrease the risk of duplicate ordering errors.


Data were collected on 630 patients, 45 658 medication orders, and 4147 patient-days pre-CPOE, and 625 patients, 32 841 medication orders, and 4013 patient-days post-CPOE. The number of duplicate medication orders increased after CPOE implementation (pre: 48 errors, 1.16 errors/100 patient-days; post: 167 errors, 4.16 errors/100 patient-days; p<0.0001; table 1). The distribution of the categories of duplicate orders also changed. The frequency of duplicate orders due to identical orders and same medication orders increased after CPOE implementation, from 0.36 to 1.72 errors/100 patient-days and 0.31 to 1.87 errors/100 patient-days, respectively, while the number of therapeutic class duplicate errors changed relatively little, from 0.48 to 0.57 errors/100 patient-days (p<0.0001 for the comparison).

Table 1
Duplicate ordering errors characteristics

Duplicate order themes

Six themes emerged from review of the duplicate orders. These themes related to: (1) whether one or two providers were involved in the duplicate ordering, (2) the time of day that duplicate orders were entered, (3) the amount of time that passed between the first and second duplicate order, (4) the category of duplicate ordering error (eg, same therapeutic class orders), (5) duplicate orders for which no duplicate alert fired to notify the provider, and (6) the design and content of duplicate order alerts presented to providers.

Number of providers involved in duplicate orders and time of order entry

Many same medication and identical order duplicate orders (72 occurrences, 43% of duplicate order errors) were placed within 1 h of each other by different providers, that is, by attending and resident physicians, physician assistants, and pharmacists. The timing of these orders suggests that many were entered during patient rounds (38 occurrences, 23% of duplicate order errors), typically a time when the interdisciplinary ICU team meets to discuss patient care. Our observations of team rounds pre-CPOE confirmed that one team member held the paper chart and wrote orders. Post-CPOE implementation, multiple team members were working on various tasks simultaneously in a loosely coordinated manner on different computers.

Also, some duplicate orders were placed by the same user within the same ordering session (11 occurrences, 7% of duplicate order errors). This means that the same user placed two or more duplicate medication orders during a single order entry session, received a duplicate alert when signing off on these orders, and overrode the alert. Other same provider duplicate orders were associated with the use of order sets—either due to the use of two order sets both of which contained identical or similar orders, or due to the use of a single order set preceded or followed by a separate order.

Duplicate error categories and missed duplicate alerts

Identical medication order duplicates were entered around morning change of shift for electrolyte replacement (12 occurrences, 7% of duplicate order errors). In the ICUs, lab results are routinely received early in the morning and the overnight on-call team is paged to write electrolyte replacement orders. A few hours later, the day staff arrive and review the lab results; they may enter a duplicate order for electrolyte replacement. If the first order has already been completed (ie, administered and documented), the second (duplicate) order does not trigger an alert because the duplicate-checking algorithm does not check new orders against orders that have been completed (due to the design of the Medi-Span medication database and the algorithms used to search for duplicate orders). In addition, the medication database's design prevents duplicate order alerts from being triggered when same medication orders are for different routes of administration.

Same therapeutic class duplicate orders tended to be orders for a change in therapy without cancellation of the existing order. For example, when a patient was transitioned from an IV proton-pump inhibitor to an oral H2 blocker, no duplicate alert was triggered. Duplicate orders also occurred when medications were ordered even though orders already existed for the same or similar medication to be mixed with and administered in total parenteral nutrition (TPN), particularly, electrolyte replacement, insulin, and H2 blockers.

Design of duplicate alerts

The design and content of alerts also contributed to their being overridden. For example, a patient who had undergone a cardiac stent procedure had an order entered for aspirin 81 mg p.o. daily. Five minutes later a different physician ordered aspirin 324 mg p.o. daily. The second provider received the duplicate alert content shown in box 1. The alert is structured to show the type of alert, for example, drug–drug interaction, the drug classification category (in bold), and the other order(s) present for which the new order is a suspected duplication. The row with the word ‘level’ in it for each alert is information extraneous to the ordering provider's information needs.

Box 1

Content of complex duplicate alert


    • Level: Major Reason: –
    • Other Orders: heparin 25 000 units+D5W 250 ml infusion

Duplicate therapy alert

    • Level: No Abuse/Dependency Potential Reason: –
    • Other Orders: acetaminophen (TYLENOL) 160 mg/5 ml oral soln 960 mg aspirin chew tab 81 mg
    • Level: No Abuse/Dependency Potential Reason: –
    • Other 0rders: aspirin chew tab 81 mg clopidogrel (PLAVIX) tab 75 mg
    • Level: No Abuse/Dependency Potential Reason: –
    • Other Orders: aspirin chew tab 81 mg

Duplicate medication order alert

    • Level: Reason: –
    • Other Orders: aspirin chew tab 81 mg

Box 1 shows three different alerts triggered by the order for aspirin: (1) a drug–drug interaction alert with heparin (first alert), (2) a duplicate therapy alert (ie, same therapeutic class) for acetaminophen, aspirin (which is listed twice, first as an antiplatelet agent and then as a salicylate), and clopidogrel, and (3) a duplicate medication order alert for aspirin. The duplicate medication order alert is the true positive alert for the duplicate order for aspirin. The duplicate therapy alert for the salicylate drug class (aspirin) is also valid but provides no additional information for the user and therefore is redundant. The other duplicate therapy alerts and drug–drug (interaction) alerts are false-positive alerts for medications intentionally used together for demonstrated therapeutic benefit (either for pain control or management of acute cardiac ischemia).17 The provider may be used to seeing these false alerts and override all the alerts without noticing the duplicate alert triggered by the two aspirin orders. The display increases the risk that the user will not recognize true positive alerts in that (1) all of the medication alerts are presented without prioritization or other clarifying structure and (2) extraneous information is presented that is not helpful to the user (eg, the information in the ‘level’ row in several alerts).

Contributing factors

We identified 12 contributing factors across the five elements of the SEIPS work system (table 2). Nine of these factors related to the CPOE technology or the duplicate alerts in the CDS system, three related to organizational factors, three related to people (the individual providers or care teams), one related to tasks, and one was related to the environment.

Table 2
Contributing factors associated with duplicate ordering errors after CPOE/CDS implementation and proposed solutions

Contributing factors included: (1) ordering practices and computer availability; (2) teamwork, communication, and hand-offs; (3) CDS and medication database design, for example, confusing alert content, high false-positive rate, and CDS algorithms missing duplicate orders; (4) CPOE design, for example, difficulty reviewing existing orders; and (5) organizational decisions regarding local CDS design, for example, order set defaults. Table 2 provides examples.

Provider perception of duplicate alert usefulness

Fifty-one ordering providers (41% response rate of 124 providers surveyed) responded to the 3-month post-implementation questionnaire and 53 responded to the 1-year post-implementation questionnaire (56% response rate of 95 providers surveyed). In both rounds, ordering providers were neutral about the usefulness of duplicate medication alerts for identifying problems with medication orders (3 months post implementation: mean 4.12 on a 7-point scale (SD 1.79); 1 year post-implementation: 4.55 (SD 1.65); p=0.205 for comparison) (table 3). Providers were similarly neutral about the usefulness of duplicate alerts for helping them correct a problem with a medication order (3 months post-implementation: mean 3.88 (SD 1.7); 1 year post-implementation: 4.53 (SD 1.73); p=0.057 for comparison).

Table 3
Ordering provider perceptions of duplicate order alerts


Duplicate medication orders increased significantly after the implementation of an EHR with CPOE. This occurred despite CDS designed to identify duplicate orders. Our in-depth study of duplicate errors before and after the implementation of CPOE with CDS highlights the remaining vulnerabilities in the system. Duplicate medication errors can increase after CPOE implementation if the multiple factors contributing to the risk of these errors are not anticipated or cannot be remedied before implementation.

CDS is an important feature of health IT. In the future, its effectiveness will depend not only upon the design and implementation of the CDS functionality but also upon consideration of changes to the work system in which it is implemented. The SEIPS model shows us that changing one part of the work system, such as the addition of CPOE, will change all parts of the work system.15 These changes can be predicted before CPOE implementation by using techniques such as proactive risk assessment.18 19 Once a change is implemented, it is incumbent on the organizations and health IT manufacturers to monitor effects and evaluate the new or persisting hazards. Duplicate medication ordering errors are an example of anticipatable events based on the contributing factors identified in this study. Error recovery systems separate from the CDS systems remain important mechanisms to ensure that medication errors, especially duplicate errors, are caught and do not reach the patient. Heightened physician, nurse, and pharmacist awareness of new errors or unexpected worsening of existing errors after technology implementation is important to ensure that team members remain vigilant for potential ordering errors even though CDS has been implemented to combat the problem.

Our analysis of the contributing factors shows that the causes of most of these duplicate ordering errors were multifactorial (like most other error types). The CPOE design limited the visibility of previously ordered and administered medications, limiting the situation awareness of the ordering providers.20 CPOE and the other components that together comprise the EHR's order-management system must be designed to inform users—at the point of ordering, reviewing, modifying, and administering medications—of the forms and total amount of the medication(s) being ordered that the patient has received and is scheduled to receive. In addition, potentially additive medications (eg, anticoagulants) must be highly visible. The CDS content and interface need to be optimally designed to support the providers in identifying and correcting these duplicate errors.21

Furthermore, the performance of the CPOE manufacturer's duplication-checking algorithms was limited by the design of the medication database. For example, orders for oral and intravenous forms of a medication were not identified as the same medication in the database and therefore were not identified as representing potentially duplicate orders. These false negatives are dangerous to the patient and increase provider distrust of alerts. Additional contributors to error are also important. With the introduction of new information systems, changes in teamwork and communication patterns must be anticipated proactively, monitored, and managed actively. To improve the situation awareness of individual clinicians as well as the team as a whole, any difficulty in accessing and integrating the patient's past, current, and planned future medication use imposed by the order-management system will necessitate more effective communication about order status, particularly around the time of hand-offs.20 Although policies are a relatively weak form of hazard control, clear definition of team roles with regard to order entry is likely to decrease duplication errors.22 For example, rounding teams might designate a single person (physician or pharmacist) to enter medication orders on rounds, with a supervising physician (the attending physician or senior resident) reviewing orders at the conclusion of rounds.

An even more fundamental issue is that organizations make important safety decisions through their purchase and configuration of CPOE. One example of such decisions concerns the design of order sets consistent with evidence-based healthcare and human factors engineering principles. This is an area where substantial research is needed. Order sets can be programmed to have specific orders automatically selected to be ordered when the order set is used. This may decrease the omission of evidence-based interventions (eg, aspirin for patients with an acute heart attack) in patient care but may also be a contributory factor in duplicate medication ordering errors as the provider may have less awareness of what they are ordering. Further research should assess the safety and effectiveness of order set defaults: is it safe and effective that orders are defaulted orders as selected on the order set with the user warned if they are duplicative or potentially additive?23

While we were able to ascertain the frequency of occurrence for some of the contributing factors or duplicate ordering themes, these data do not indicate the importance of the contributing factor; other factors like potential for patient harm and the ability to detect and correct the error before it reaches the patient should also be considered. Moreover, multiple contributing factors likely contribute to each of the duplicate medication order events, thus fixing only one of the factors may not prevent the error from occurring in the future. This study evaluated the design and implementation of a single EHR, yet the persistence or increase in duplicate ordering errors has been noted with other EHRs. However, the CPOE and CDS design issues noted in this study may not be seen with other EHRs.

We propose solutions to mitigate the factors associated with the risk of duplicate medication ordering errors. The solutions conform to the principles of designing systems to (1) support communication and collaboration among the healthcare team, (2) enhance error recovery by improving team and individual situation awareness, particularly around shift changes and other hand-offs, (3) improve teamwork with regards to order entry, and (4) improve the usability and functionality of order entry screens and medication alerts.

  1. Manufacture medication databases that
    • Identify the same medication, regardless of combination, formulation, or route of delivery;
    • Identify potentially additive medications regardless of medication class; and
    • Identify medication combinations (including doses) that could potentially be additive but have been demonstrated to be safe and effective, for example, the combination of aspirin 81–625 mg once a day with clopidogrel 75 mg a day.
  2. Manufacture EHR algorithms for checking for potentially duplicate and additive medication orders.
  3. Apply human factors design principles to make the patient's medication record—both recently administered and planned—more accessible and comprehensible during order planning, entry, review, and administration.
  4. Apply human factors design principles to create more accessible and comprehensible alerts.
  5. Select, implement, and maintain health IT products and services for both safety and effectiveness.
  6. Develop context appropriate policies for order entry. For example:
    • Review all recent and planned medications before entering or fulfilling orders.
    • Identify one person on a rounding team to enter orders during patient rounds and a supervising physician to review orders immediately after rounds.
  7. Require hand-offs (for example, at change of shift and patient transfer) to include communication of recent and planned medication orders.
  8. Review and optimize protocols for care processes likely to create duplicate or additive orders, for example, electrolyte replacement in patients on TPN, anticoagulation, or verbal orders.

To prevent or mitigate an increase in duplicate medication orders will take the concerted efforts of multiple actors including policy makers, medication database manufacturers, EHR manufacturers, healthcare organizations, healthcare teams, and individual clinicians. At the societal level, the availability of safe and effective medication databases and EHRs requires the inclusion of safety into certification requirements and the inclusion of additional software systems (ie, medication databases) into certification schemes. Health IT manufacturers will need to enact the solutions listed above that are relevant to them. Healthcare organizations will need to include the above solutions in their efforts to maximize the benefits and decrease the risks when they implement information technologies.24 Two demonstrated risks of CPOE are particularly relevant to medication safety: increased duplicate errors and alert fatigue.25 The EHR manufacturers must perform usability evaluations of CPOE and design the technology with the contexts of use in mind. Healthcare organizations must work with manufacturers to inform them about those contexts, ensuring that alerts encompass all predictable, relevant situations. Healthcare organizations should provide competency-based EHR training and feedback to users (as well as their supervisors) regarding the frequency of duplicate order entry. Of course, this reporting should not be seen as an adequate substitution for the more effective solutions above. Finally, definitions of healthcare professionalism will need to be extended to include the management of electronic patient information.26


We have documented an increase in duplicate medication order entry and described the potential contributing factors in one EHR implementation. Based on these findings, we have recommended multiple improvement activities, particularly continuous improvement of the social systems of healthcare (care-delivery organizations, care teams, and clinicians) and the health IT that is meant to service them. Systems factors such as teamwork, communication, organizational decision-making, and user-centered design of health IT represent important areas for further research.


Funding: This study was supported by grants R01-HS15274 and K08-HS17014 from the Agency for Healthcare Research and Quality and grant 1UL1RR025011 from the Clinical and Translational Science Award (CTSA) program of the National Center for Research Resources, NIH. These agencies had no other role in the design, conduct, and reporting of this research.

Competing interests: None.

Ethics approval: The Geisinger Health IRB and the University of Wisconsin Health Sciences IRB approved this study.

Provenance and peer review: Not commissioned; externally peer reviewed.


1. Ammenwerth E, Schnell-Inderst P, Machan C, et al. The effect of electronic prescribing on medication errors and adverse drug events: a systematic review. J Am Med Inform Assoc 2008;15:585–600 [PMC free article] [PubMed]
2. Bates DW. Computerized physician order entry and medication errors: finding a balance. J Biomed Inform 2005;38:259–61 [PubMed]
3. Carter JH. Electronic Health Records a Guide for Clinicians and Administrators. 2nd edn Philadelphia, PA: American College of Physicians Press, 2008
4. Kuperman GJ, Bobb A, Payne TH, et al. Medication-related clinical decision support in computerized provider order entry systems: a review. J Am Med Inform Assoc 2007;14:29–40 [PMC free article] [PubMed]
5. Shah NR, Seger AC, Seger DL, et al. Improving override rates for computerized prescribing alerts in ambulatory care. AMIA Annu Symp Proc 2005;2005:1110. [PMC free article] [PubMed]
6. Bedouch P, Allenet B, Grass A, et al. Drug-related problems in medical wards with a computerized physician order entry system. J Clin Pharm Ther 2009;34:187–95 [PubMed]
7. Koppel R, Metlay JP, Cohen A, et al. Role of computerized physician order entry systems in facilitating medication errors. JAMA 2005;293:1197–203 [PubMed]
8. Colpaert K, Decruyenaere J. Computerized physician order entry in critical care. Best Pract Res Clin Anaesthesiol 2009;23:27–38 [PubMed]
9. George D, Austin-Bishop N. Error rates for computerized order entry by physicians versus nonphysicians. Am J Health Syst Pharm 2003;60:2250–2 [PubMed]
10. Lin CP, Payne TH, Nichol WP, et al. Evaluating clinical decision support systems: monitoring CPOE order check override rates in the Department of Veterans Affairs' Computerized Patient Record System. J Am Med Inform Assoc 2008;15:620–6 [PMC free article] [PubMed]
11. Bates DW, Teich JM, Lee J, et al. The impact of computerized physician order entry on medication error prevention. J Am Med Inform Assoc 1999;6:313–21 [PMC free article] [PubMed]
12. Hoonakker PL, Carayon P, Walker JM. Measurement of CPOE end-user satisfaction among ICU physicians and nurses. Appl Clin Inform 2010;1:268–85 [PMC free article] [PubMed]
13. Hoonakker PL, Cartmill RS, Carayon P, et al. Development and psychometric qualities of the SEIPS survey to evaluate CPOE/EHR implementation in ICUs. Int J Healthc Inf Syst Inform 2011;6:51–69 [PMC free article] [PubMed]
14. Karsh BT, Holden R, Escoto K, et al. Do beliefs about hospital technologies predict nurses' perceptions of quality of care? A study of task-technology fit in two pediatric hospitals. Int J Hum Comput Interact 2009;25:374–89
15. Carayon P, Schoofs Hundt A, Karsh BT, et al. Work system design for patient safety: the SEIPS model. Qual Saf Health Care 2006;15(Suppl I):i50–8 [PMC free article] [PubMed]
16. Smith MJ, Carayon-Sainfort P. A balance theory for job design for stress reduction. Int J Ind Ergon 1989;4:67–79
17. Van der Sijs H, Aarts J, Vulto A, et al. Overriding of drug safety alerts in computerized physician order entry. J Am Med Inform Assoc 2006;13:138–47 [PMC free article] [PubMed]
18. Carayon P, Faye H, Hundt AS, et al. Patient safety and proactive risk assessment. In: Yuehwern Y, editor. , ed. Handbook of Healthcare Delivery Systems. New York, NY: Taylor & Francis, 2010;12:1–15
19. Wetterneck TB, Skibinski KA, Roberts TL, et al. Using failure mode and effects analysis to plan implementation of smart i.v. pump technology. Am J Health Syst Pharm 2006;63:1528–38 [PubMed]
20. Endsley MR, Bolte B, Jones DG. Designing for Situation Awareness. An Approach to User-Centered Design. Boca Raton, FL: Taylor & Frances Group, 2003
21. Phansalkar S, Edworthy J, Hellier E, et al. A review of human factors principles for the design and implementation of medication safety alerts in clinical information systems. J Am Med Inform Assoc 2010;17:493–501 [PMC free article] [PubMed]
22. Dykstra R. Computerized physician order entry and communication: reciprocal impacts. Proc AMIA Symp 2002:230–4 [PMC free article] [PubMed]
23. Halpern SD, Ubel PA, Asch DA. Harnessing the power of default options to improve health care. N Engl J Med 2007;357:1340–4 [PubMed]
24. Walker JM, Carayon P, Leveson N, et al. EHR safety: the way forward to safe and effective systems. J Am Med Inform Assoc 2008;15:272–7 [PMC free article] [PubMed]
25. Patterson ES, Doebbeling BN, Fung CH, et al. Identifying barriers to the effective use of clinical reminders: bootstrapping multiple methods. J Biomed Inform 2005;38:189–99 [PubMed]
26. Wachter RM, Pronovost PJ. Balancing “no blame” with accountability in patient safety. N Engl J Med 2009;361:1401–6 [PubMed]

Articles from Journal of the American Medical Informatics Association : JAMIA are provided here courtesy of American Medical Informatics Association