Health IT and Patient Safety: Building Safer Systems for Better Care (2012)

2

Evaluating the Current State of Patient Safety and Health IT

Health IT creates new opportunities to improve patient safety that do not exist in paper-based systems. For example, paper-based systems cannot detect and alert clinicians of drug-drug interactions, whereas electronic clinical decision support systems can. As a result, the expectations for safer care may be higher in a health IT-enabled environment as compared to a paper-based environment. However, implementation of health IT products does not automatically improve patient safety. In fact, health IT can be a contributing factor to adverse events, such as the overdosing of patients because of poor user interface design, failing to detect life threatening illnesses because of unclear information displays, and delays in treatment because of the loss of data. Adverse events, such as these, have lead to serious injuries and death (Aleccia, 2011; Associated Press, 2009; Graham and Dizikes, 2011; Schulte and Schwartz, 2010; Silver and Hamill, 2011; U.S. News, 2011).

The way in which health IT is designed, implemented, and used can determine whether it is an effective tool for improving patient safety or a hindrance that threatens patient safety and causes patient harm (see Box 2-1). The implementation of health IT, particularly complex health IT products, may result in less efficient systems and not give clinicians the flexibility they need to deliver the safest care possible (Greenhalgh et al., 2008). Currently, the relationship between these unintended consequences and the design, implementation, and use are not well understood.

This chapter uses the literature and experiences of health professionals to evaluate the impact of health IT on patient safety. The first several sections of this chapter discuss the challenges faced by health IT researchers by

detailing the complexity of health IT and patient safety, the limitations in the literature to determine health IT’s impact on patient safety, and how the magnitude of harm is masked. The chapter then analyzes the literature to determine how individual components of health IT have impacted patient safety and how data from health IT can be leveraged to improve safety in different populations. Next, it describes how policy makers can learn from health IT experiences from abroad.

COMPLEXITY OF HEALTH IT AND PATIENT SAFETY

In general, health IT is not a specific product but is composed of components—such as computerized provider order entry (CPOE) systems and clinical decision support (CDS) systems—that are designed, implemented, and used differently by various vendors, health care settings, and users (Hayrinen et al., 2008). These differences can have dramatic effects on care processes including care design, workflow, and—ultimately—the quality and safety of the care delivered. When health IT is designed and implemented in a manner that complements how information is transferred between health professionals and patients, the reliability of patient information—and therefore patient safety—can increase (Dorr et al., 2007; Niazkhani et al., 2009; Shah et al., 2006). However, when health IT unexpectedly alters workflow, it has the potential to hinder clinicians’ abilities to communicate patient information (Niazkhani et al., 2009), and it may result in increased cognitive workload, clinicians ignoring computer- generated information, continued reliance on various traditional modes of communication, creation of unsafe workarounds, and more time spent dealing with health IT than with patient care (Ash et al., 2009). Several important factors regarding how health IT products are designed and implemented can have meaningful effects on the collection, storage, and transfer of information, as well as the utility of the product. Slight variations in these factors can have differing effects on how health IT impacts patient safety. Some of these factors include the following:

•   Decisions about implementation strategies (e.g., “big bang” versus incremental);

•   The degree to which users can configure their IT system and the approaches to such configurations;

•   Clinician training strategies;

•   Frontline use (e.g., the IT integration into and redesign of clinical workflow); and

•   Tools for analyzing and reporting results of care (e.g., quality improvement).

LIMITATIONS OF CURRENT LITERATURE TO DETERMINE HEALTH IT’S IMPACT ON PATIENT SAFETY

Like all studies regarding patient safety, studies focusing on health IT and patient safety are complex and subject to a variety of methodological challenges. To provide generalizable knowledge about the impact of health IT on patient safety, the interaction of the factors listed in the previous section (e.g., frontline use) needs to be understood. However, very few studies to date have done so, resulting in major gaps in our knowledge regarding how health IT affects safety. While most of the literature examining the effects of health IT has focused on quality and processes of care, studies regarding the impact of health IT on patient safety have been narrowly focused on a few specific aspects of care. Given that adverse events (events resulting in unintended harm to a patient from a medical intervention [IOM, 2004]) are multifaceted and diverse, much of the literature that does center on how health IT affects patient safety has focused on prevention of medication errors, identification of patients at high risk for adverse events, and avoidance of documentation errors. Although much of this evidence suggests that IT can be helpful in improving patient safety, a number of studies have failed to find a benefit (Black et al., 2011; Culler et al., 2006; Garg et al., 2005; Reckmann et al., 2009).

Many studies, including meta-analyses, offer strong evidence that computerization of prescribing can dramatically improve patient safety. These products were consistently correlated with lowering the frequency of medication errors and may be able to reduce preventable adverse drug events significantly (Kaushal et al., 2003; Shamliyan et al., 2008; Wolfstadt et al., 2008). However, the degree to which health IT can lower medication errors varies widely among the different computerized prescribing systems used (Nanji et al., 2011).

The evidence of similar impact outside of medication safety is much weaker (Bates and Gawande, 2003). Indeed, some systematic reviews conclude that the current literature is insufficient to establish any beneficial impact of health IT on patient safety and health outcomes (Black et al., 2011; Garg et al., 2005; Reckmann et al., 2009). More recently, new data have emerged, suggesting that health IT can introduce new patient safety challenges into the health care system (Magrabi et al., 2010, 2011). These studies are unable to accurately quantify the number of people harmed by health IT. This inability of the committee to quantify the harm makes it difficult to understand the tradeoffs between the potential safety benefits and harms caused by health IT.

The differing results found in the literature may be due to a variety of reasons. Among those reasons are the heterogeneous nature of health IT—including the differences in the products themselves, how they are

implemented, and how they are used across care settings. Most studies focused on health IT and patient safety examined care at a single medical center, often with homegrown IT systems. Aggregating many single-center studies, such as those common throughout the literature, does not necessarily lead to the same outcomes as having a few studies that are conducted in a broader array of clinical settings. However, systematic reviews have attempted to aggregate these studies and have done so in inconsistent ways, often choosing to include low-quality studies while failing to include higher- quality ones. Therefore, the committee could not point to any systematic reviews or studies as representing the most definitive evidence of the impact of health IT on patient safety (see Table B-1).

A major challenge in quantifying the harm that might result from health IT is the lack of data in this area. However, the absence of quantifiable evidence of health IT’s harm is not evidence that health IT does not create harm. It is clear that harm exists. The current literature does not sufficiently produce estimates on the harm that might result from health IT. For example, a recent study by Nanji et al. evaluated the frequency, types, and causes of errors associated with outpatient computer-generated prescriptions. The study evaluated 3,850 prescriptions and found 466 errors, involving almost 12 percent of the orders. Because the error rates varied widely between different computerized prescribing systems (from 5.1 to 37.5 percent), the authors strongly recommended that users evaluate the safety of each system (Nanji et al., 2011). However, the authors were not allowed to list which error rates and safety issues were associated with each particular system. Instead, the article prescribed a “vigorous vendor selection” process, which each potential user would have to go through in order to identify safety concerns of that system. Had the authors been allowed to identify specific systems with higher error rates, users could know which systems to avoid and could select systems with characteristics that would best fit their workflow and safety needs.

Studies with generic descriptions of health IT products and patient safety issues will be of little utility to users because health IT products— even those made by the same manufacturers—are heterogeneous, tailored to individual clinical settings, and have varying impacts on patient safety. Therefore, to assist users in selecting the safest health IT product for their unique clinical environment, studies need to be able to name specific health IT products, describe how those products have been implemented, and identify their impact on patient safety in different clinical environments. For example, as mentioned in Box 2-1, a Pittsburgh pediatric intensive care unit’s (ICU’s) implementation of a CPOE system resulted in higher mortality; however, several different hospitals were able to subsequently identify safety problems associated with Pittsburgh’s experience and implemented the same CPOE system with either no change or up to a 20 percent

decrease in hospital-wide patient mortality (Longhurst et al., 2010). In order to identify problems associated with the Pittsburgh implementation, a pediatric ICU in Seattle sent researchers to Pittsburgh’s facilities, met with their administrative and clinical leadership, and spoke with clinical staff. After months of correspondence, the Seattle pediatric ICU was able to determine why Pittsburgh’s implementation resulted in a higher mortality rate and was able to avoid such problems (Del Beccaro et al., 2006; Han et al., 2005). When selecting health IT products, many potential users do not have the time or the resources to spend months corresponding, visiting, and observing other hospitals. Users and researchers need to be encouraged to provide specific descriptions of safety problems associated with particular health IT products in order to provide potential users with credible data regarding which IT products are safer than others.

BARRIERS TO KNOWING THE MAGNITUDE OF THE HARM

When researchers, consumer groups, and users attempt to identify and share information on health IT features related to adverse events and patient safety risks, they can be faced with barriers created by market inefficiencies within health IT, such as lack of information available to cons umers and the inability of users to freely move between health IT products. For example, because the impact of health IT in each clinical environment is extremely diverse and highly dependent on the user’s specific clinical environment, it is difficult for clinicians to know how the myriad of different health IT products will affect patient safety. Additionally, because of the substantial costs and effort used in tailoring and integrating health IT products, users may not be able to readily switch products after discovering patient safety problems. Many health IT products can only be maintained by the manufacturer of that product, causing users to maintain service contracts with that manufacturer, regardless of whether that manufacturer addresses patient safety issues associated with its product. Even if users are willing to switch health IT products, there is no guarantee another product will achieve greater levels of patient safety, once integrated. These inefficiencies result in an inadequate understanding of how health IT impacts patient safety and leads users to select and make a long-term commitment to products that may not adequately complement their clinical environment.

To increase understanding of how health IT affects patient safety and allows users to make informed decisions, it is important that the health IT community share details, such as screenshots of risk-enhancing interfaces, descriptions of potentially unsafe processes, and other components of health IT products associated with adverse events. Some vendors allow users to share information through industry conferences, sponsored user

group meetings, blogs, and consultants that provide conduits for information about vendor experiences. However, the ability of users and researchers to share such information outside industry-controlled venues can be limited by nondisclosure clauses.

Nondisclosure clauses—commonly found in many types of commercial contracts and almost always included in software license agreements—are intended to protect licensors’ intellectual property interests, competitive edge, and liability from consumer misuse of their products¹ (Koppel and Kreda, 2009). The fear of violating nondisclosure clauses and intellectual property interests may discourage users from sharing health IT- related patient safety risks. Additionally, if users believe that hold-harmless clauses, which are placed in many vendor contracts, can shift the liability of unsafe health IT features solely to the user, they may fear that disclosing unsafe features may unfairly increase their risk of liability.

To adequately understand how health IT impacts patient safety, users and researchers need to be able to share information that may normally be protected by intellectual property rights or may expose users to unfair liability. Some vendors have expended considerable effort to ensure patient safety, but allowing the disclosure of patient safety issues may cause vendors to lose their competitive advantage. Thus, some vendors may impose or enforce such restrictions in ways that may conceal patient safety issues² (Koppel and Kreda, 2009). As long as vendors may restrict the release of information about safety issues through confidentiality clauses, intellectual property protections, and hold-harmless clauses, the health care community will be limited in its understanding of how health IT affects patient safety.

Because the nature of these legal issues limits publicly available information, very little evidence establishes their frequency of use or impact on users (Koppel and Kreda, 2009). However, the committee believes that these types of contractual restrictions limit transparency, which significantly contributes to the gaps in knowledge of health IT-related patient safety risks. Regardless of whether these barriers have actually been used to prevent reporting, the fear of legal action itself may prevent health professionals from sharing crucial health IT-related information with researchers, consumer groups, other users, and the government. As stated by the American Medical Informatics Association, such clauses should be considered unethical (Goodman et al., 2011).

¹ Personal communication, B. Leshine, LeClairRyan, April 20, 2011; personal communication, H. Levine, Blaszak, Block & Boothby, LLP, June 10, 2011.

² Personal communication, B. Leshine, LeClairRyan, April 20, 2011; personal communication, H. Levine, Blaszak, Block & Boothby, LLP, June 10, 2011.

IMPACT OF HEALTH IT COMPONENTS ON PATIENT SAFETY

The following sections examine how individual components of health IT affect patient safety. However, most health IT products are not a single component, but a complex system of health IT components, sometimes collectively referred to as electronic health records (EHRs). Although the definition of EHRs can vary substantially, there are generally four core components of an EHR: electronic clinical documentation (usually physician, nurse, and other clinician documentation), electronic prescribing (e.g., computerized provider order entry), results reporting and management (e.g., clinical data repository), and clinical decision support (DesRoches et al., 2008; Jha et al., 2006, 2009a, 2009b). Many EHRs also include bar- coding systems and patient engagement tools. The Office of the National Coordinator for Health Information Technology (ONC) defines an EHR as “a real-time patient health record with access to evidence-based decision support tools that can be used to aid clinicians in decision-making. The EHR can automate and streamline a clinician’s workflow, ensuring that all clinical information is communicated. It can also prevent delays in response that result in gaps in care. The EHR can also support the collection of data for uses other than clinical care, such as billing, quality management, outcome reporting, and public health disease surveillance and reporting” (HHS, 2004; ONC, 2009).

Although EHR and health IT are terms that are still evolving and are often interpreted differently, much of the evidence regarding the impact of EHRs on patient safety has focused on individual components of EHRs. The following sections explore the evidence for individual components and then discuss the evidence from studies that use the “EHR” as a general term.³ Because almost every component uses documentation and results review and management throughout their tasks (bar-coding, CPOE, and CDS systems all use documentation and results reporting and management in prescribing and delivering medication), this chapter will not address documentation results reporting and management individually. The section then looks at how current EHR systems can be leveraged to further improve patient safety. Table 2-1 summarizes the benefits and safety concerns commonly found in the literature.

³ Although there are many other components of health IT, the bulk of the literature has focused on the following components: EHR, CPOE systems, CDS systems, patient engagement tools, and bar-coding systems. Some other components not listed in this chapter include medication reconciliation systems and smartpumps; see Appendix B.

TABLE 2-1

Potential Benefits and Safety Concerns of Health IT Components

NOTE: Table 2-1 is not intended to be an exhaustive list of all potential benefits and safety concerns associated with health IT. It represents the most common potential benefits and safety concerns.

Computerized Provider Order Entry

CPOE is an electronic system that allows providers to record, store, retrieve, and modify orders (e.g., prescriptions, diagnostic testing, treatment, and radiology/imaging orders). The use of CPOE has varying degrees of impact on patient safety, depending on how well the CPOE system complements or improves provider workflow. The successful impact of a CPOE system on patient safety may also depend heavily on the change management approach employed by organizational leadership to prepare clinicians and recipients of the new workflow, as well as the decision support tools associated with it. Short-term benefits of CPOE systems commonly found

in studies include large increases in legible orders, shorter order turnaround times, lower relative risk of medication errors, and a higher percentage of patients who attain their treatment goals (Devine et al., 2010; Nam et al., 2007; Niazkhani et al., 2009). In the inpatient setting, a series of literature reviews and meta-analyses found that medication error rates fell (Kaushal et al., 2003; Shamliyan et al., 2008; Wolfstadt et al., 2008) due to the introduction of a CPOE system and most, though not all, studies suggest that the preventable adverse drug events (ADEs) rate decreases as well. Studies suggest that CPOE systems have a greater impact when designed for the specific needs of the hospital environment, workflow, and providers (Callen et al., 2010). For example, CPOE systems with order sets specifically designed for ICUs can increase efficiency and workflow (Ali et al., 2005).

Although the potential benefits of CPOE systems are well established, the harms that have been well articulated on a case-by-case basis have relatively little empirical basis behind them (Aleccia, 2011; Associated Press, 2009; Graham and Dizikes, 2011; Schulte and Schwartz, 2010; Silver and Hamill, 2011; U.S. News, 2011). The lack of data on harm is driven in large part, as described earlier, by practices that limit disclosure of health IT-related adverse drug events. Based on the existing information, it seems likely that, if these systems are either designed poorly or interface with clinicians in an ineffective manner, they can cause harm. Several experts have suggested that CPOE systems can have a number of potential adverse consequences, including increased ordering time, disruptions in workflow, new opportunities for errors (e.g., fragmented displays preventing a coherent view of patients’ medications, inflexible ordering formats generating wrong orders, separations of functions that facilitate double dosing, and incompatible orders), and increased relative risk of medication errors (Koppel et al., 2005; Niazkhani et al., 2009; Santell et al., 2009; Singh et al., 2009; Walsh et al., 2006; Weant et al., 2007).

Some of the variability in the impact of CPOE systems is likely due to differences in decision support systems that can detect potential errors and/or generate care suggestions. For example, a CPOE system was introduced to a pediatric ICU without a CDS and resulted in no significant change in the rate of potential adverse drug events. However, a significant reduction in potential adverse drug events was found after a CDS system was implemented (Kadmon et al., 2009). Further discussion regarding the impact of CDS on patient safety is examined in the next section.

Clinical Decision Support

CDS systems are also an important component of an EHR. They can monitor patient conditions, prescriptions, and treatment to provide evidence-based clinical suggestions to health professionals at the point of

care. The literature regarding the implementation of CDS is largely positive with respect to medication safety, though it is more mixed in domains such as chronic disease management. The majority of systematic reviews in this area report that most studies have demonstrated positive impacts on patient safety by improving practitioner performance and reducing the relative risk of medication errors, time to therapeutic stabilization, and risk of toxic drug levels (Ammenwerth et al., 2008; Conroy et al., 2007; Durieux et al., 2008; Garg et al., 2005; Georgiou et al., 2007). However, many of these reviews also stress that further research needs to be conducted to determine the full impact of CDS use on patient safety because many of the studies are often weak, differ substantially in their settings and design, and are inconsistent (Ammenwerth et al., 2008; Garg et al., 2005).

The inconsistency seen throughout the literature may be due to the differences in clinical settings, CPOE systems, CDS components, and workflow (Georgiou et al., 2007). Several studies have shown that the ability of CDS tools to perform even simple tasks, such as detecting drug-drug interactions, varies widely. For instance, when fictitious patients with simulated drug-drug interactions were entered in a CDS system, the results were disappointing. Using Leapfrog Group’s “flight simulator technology,” one study found that the mean scores for detecting simulated orders that would have led to serious adverse drug events for 62 hospitals was 43 percent (range 10 to 82 percent). The ability of the 62 hospitals to detect simulated medication errors that could result in fatalities was 53 percent (Metzger et al., 2010). In a similar study of 13 community and hospital pharmacies, six mock patients with a total of 37 drug-drug interactions were entered into each pharmacy’s CDS e-prescribing system. The ability of the CDS systems to detect drug-drug interactions significantly varied. The CDS systems’ sensitivity ranged from 0.15 to 0.94 and its specificity from 0.67 to 1.00, even among CDS systems designed by the same manufacturer (Abarca et al., 2006).

Although most of the literature suggests that CDS has had a positive overall impact on medication safety, a few studies have shown either no significant change with CDS use (Gandhi et al., 2005; Glassman et al., 2007; Gurwitz et al., 2005; Tierney et al., 2005) or negative consequences, such as increased patient mortality (see Box 2-1) (Han et al., 2005). However, these differences in impact may relate to variations in implementation and use of these systems, dissimilar designs, and how those CDS systems are used or differences in the organizations themselves (Del Beccaro et al., 2006). To create (or configure and populate) a CDS system to best improve patient safety, designers must consider hospitals’ different clinical environments and test management practices (Callen et al., 2010). Several examples of successful implementations indicate that CDS systems may be more effective at increasing medication safety when they are specifically tailored to the

clinical environment (Callen et al., 2010; Fitzgerald et al., 2011; Kadmon et al., 2009; Nies et al., 2010; Smith et al., 2006). For example, CDS use was found to be extremely effective at reducing management errors while resuscitating patients in an adult trauma center (Fitzgerald et al., 2011). Additionally, elderly patients were prescribed fewer nonpreferred medications after implementation of a CDS system specifically designed to alert and recommend clinicians of alternative treatments (Smith et al., 2006).

In addition to monitoring for potential medication errors, CDS systems can also suggest potential diagnoses and treatment, monitor patients’ conditions, determine whether a potential or actual adverse event may occur, and alert clinicians to potential adverse conditions. Alerts can come in the form of chimes, flashing lights, and/or popup windows that appear while health professionals are accessing patient EHRs or entering orders into a computer system. Many alert systems require health professionals to acknowledge the alert by clicking a button in the popup window before continuing treatment, while others may appear and disappear without interrupting the health professionals’ work. These systems can also help with surveillance and have been shown to be effective at diagnosing and alerting clinicians of adverse conditions (Claridge et al., 2009; Herasevich et al., 2009; Jha et al., 2008).

When implementing an alert system, success depends on how alerts impact workflow (Bates et al., 2003). If implemented correctly, alerts can improve patient safety. Alerts have been demonstrated to lower the rate of inappropriate medication prescriptions to select vulnerable populations, such as the elderly (Raebel et al., 2007). Flag alerts—reminders of patient diagnosis or conditions to clinicians who access patient EHRs—have been demonstrated to improve long-term treatment and increase the likelihood of achieving treatment goals (Agostini et al., 2007; Whitley et al., 2006). A retrospective analysis examining a diagnostic alarm system showed that the alarm system could detect and alert clinicians of critical events during anesthesia administration as effectively as anesthesiologists (Gohil et al., 2007).

Although patient safety can be improved by alerts, an improperly designed system may be ignored or even considered a nuisance to users (Phansalkar et al., 2010). In a retrospective cohort study of a large Veterans Affairs medical center and its five clinics, 10.2 percent of all alerts were unacknowledged and 6.8 percent of all alerts lacked timely follow-up (Singh et al., 2010). A controlled study of two medical departments in a French hospital showed that use of computer-generated alerts had no significant impact on the rate of inappropriate first prescriptions. Further analysis of the data showed that while the senior physicians made more inappropriate prescriptions with the alert system, residents made fewer inappropriate prescriptions, indicating that newer providers may be more adaptable to alert systems (Sellier et al., 2009).

Ineffectiveness of an alert system has been attributed to high rates of overrides and alert fatigue. If alerts are too numerous and are not representative of clinically significant conditions, they can overload clinical workflow and cause clinicians to ignore information that could prevent adverse events (Phansalkar et al., 2010). In an observational study, 25 percent of clinicians demonstrated signs of alert fatigue (van der Sijs et al., 2010). Several observational and retrospective studies found override rates between 80 and 98 percent (Lin et al., 2008; Shah et al., 2006; van der Sijs et al., 2009). Although high-severity alerts have a higher acceptance rate, they are still overridden more often than not (Isaac et al., 2009). One study was able to substantially reduce override rates of serious alerts by developing a tiered alert system, where alerts of less serious magnitude do not interrupt workflow. The more serious alerts caused a popup window to appear and forced the clinicians to acknowledge the alert (Shah et al., 2006). By limiting interruptions in workflow, an alert system can remind clinicians of important patient information without causing alert fatigue or high override rates.

Bar-Coding

The introduction of a bar-coding system to administer medication and verify patient identification has been strongly associated with significant reductions in relative risk of medication errors, including transcription, dispensing, and administration errors (Franklin et al., 2007; Poon et al., 2010). Although bar-coding has been shown to reduce medication errors, like any other technology, it can also create new opportunities for errors to occur. In a study observing and interviewing clinicians who use bar-coding systems to dispense and order medication, examples of multiple workarounds that could lead to patient harm were found. For example, clinicians who violate safe procedures and practices may scan medications and patient identification without visually checking to see if the medication, dosing, and patient are correct. Instead, clinicians depend on alarms or alerts to detect errors that may never have occurred if visually checked. An example of a workaround is that clinicians could attach patient identification bar-codes to another object instead of the patient, such as the patient’s bed, which may lead to patients receiving incorrect medications in cases where the technology was perceived to be an obstacle to providing care. The study also found that, instead of scanning an order and the medication and then dispensing the medication to the patient, clinicians sometimes scanned all the orders and medications of multiple patients at once. This would save clinicians time because they would not have to scan the orders and medications each time they administered the medications. However, this workaround could result in the clinician giving patients wrong medications (Koppel et

al., 2008). Despite the presence of these workarounds, the overall effect of bar-coding has been shown to substantially reduce the relative risk of medication errors, both at the point of care (Franklin et al., 2007; Poon et al., 2010) and in dispensing errors in the pharmacy (Poon et al., 2006).

Patient Engagement Tools

To date, much of the data on patient care reside on paper with little ability for patients and their families to access or use the information to improve their own health. Adoption of health IT by consumers is growing and includes a variety of tools that patients can use to engage in their care. These engagement tools are in varying stages of development and sophistication, with a growing number using smartphones as a common platform.

The literature regarding patient engagement tools generally does not focus on safety. Rather, the focus of most studies primarily examines the l evels of comfort patients have with patient engagement tools and how engaged they are when these tools are made available to them. However, some studies demonstrate that patient engagement tools reduce hospitalization rates in children, increase patients knowledge of treatment and illnesses, and increase clinician knowledge (Murray et al., 2009; Ngo-Metzger et al., 2010).

Electronic Health Records

The following section discusses studies that focus either on EHRs as a whole or on how multiple components have affected patient safety. Implementation of EHRs has been reported to increase providers’ perceptions of safety (Ferris et al., 2009), to lower infection rates (Parente and McCullough, 2009), and to reduce the number of documentation errors (Smith et al., 2009). While a review of the literature establishes that the use of EHRs improves process measures of quality of care in certain domains (e.g., preventions, specific chronic diseases), its impact on patient outcomes has been much more mixed (Einbinder and Bates, 2007).

The literature regarding EHR features, such as electronic documentation and results review and management, are also mixed. While health professionals perceive that these components can increase safety and efficiency (Ferris et al., 2009), they also expressed that features—such as copy and paste forward functions—can pose patient safety risks. One study found that the implementation of electronic vital sign documentation can reduce medical error rates in half (Gearing et al., 2006), while another study found that more than half of new aortic dilations discovered by computed tomography (CT) scan could not be found within patients’ EHRs (Gordon et al., 2009).

The effectiveness of an EHR system on patient safety is dependent

on the compatibility of that EHR with the individual needs of its users (Hayrinen et al., 2008). For example, EHR implementation can have differing effects on the flow of patient information (Benham-Hutchins and Effken, 2010). The Department of Veterans Affairs was able to demonstrate that an EHR could help coordinate care by providing a continuous flow of information among multiple clinicians (Litaker et al., 2005). Conversely, other studies in different clinical settings have found EHR implementation to have either no effect or a negative impact on workflow and patient outcomes (Benham-Hutchins and Effken, 2010; DesRoches et al., 2010). In a survey of an urban, university-based hospital, 84 percent of surveyed health professionals preferred verbal over electronic communication because they believed information contained in the EHR was unreliable. There, it was found that health professionals used nonlinear communication, combining several modes of communication to exchange patient information, including EHRs, paper notes, phone, and in-person verbal communication (Benham-Hutchins and Effken, 2010).

In general, EHRs have the potential to greatly increase patient safety, but the potential has not been realized consistently. For example, EHRs could include tools to help ensure that if a major issue such as an aortic aneurysm is detected, it is added to a problem list, or that problem and medication lists get updated more effectively. Research is needed to develop such tools, though early evidence suggests they have the potential to be highly effective (Wright et al., 2011). More broadly, additional research needs to be conducted on how various EHR designs affect different workflows and providers’ needs.

LEVERAGING EHR DATA TO IMPROVE SAFETY OF POPULATIONS

In addition to results reporting for individual patients, EHRs can be a rich source of data for the identification of care gaps and patient lists for monitoring and clinical action across populations. While the degree of harm to patients is unclear, the failure to follow up on laboratory results represents one of the leading causes of lawsuits in the outpatient setting (Gandhi et al., 2006). Reports have shown that many abnormal lab results had not been acted upon by the appropriate clinicians, leading to important delays in diagnosis and treatment to patients (Kravitz et al., 1997; Magid et al., 2010). Surveys demonstrate that physicians are dissatisfied with paper approaches to management of test results (Poon et al., 2004). Data, identifying care gaps, and patient lists for monitoring, and clinical action across populations can be extracted from EHRs. Inpatient “system lists” of patients provide real-time data to monitor and identify high-risk patients for falls, pain management, pressure ulcers, ventilator-acquired pneumonia,

and restraints. They can also be used to communicate test results directly to patients, which improves patient satisfaction (Matheny et al., 2007).

These population support tools have been shown to be effective at identifying gaps in care. Population support tools identified gaps in 32 evidence- based care recommendations for individual patients, groups of patients selected by a provider, or all patients on a primary care provider’s panel. One tool was shown to have improved primary care teams’ performance by up to 21 percent on preventive, monitoring, and therapeutic evidence-based recommendations (Zhou et al., 2011). A similar registry targeting females over the age 67 with a previous fracture along with follow-up activity showed a 13 to 44 percent improvement in patients receiving an evaluation and/or treatment for osteoporosis (Feldstein et al., 2007).

A powerful, more long-term impact was described in “The Best Medicine” (Begley, 2011). Begley describes the use of data from an EHR to find out which antihypertension drugs worked best if diuretics do not bring about the needed reduction in blood pressure (Begley, 2011). More proximate to patient safety is the early Kaiser Permanente recall of Vioxx. Here, Kaiser’s EHR data independently showed increased incidence of heart attack and stroke. Based on these data, Kaiser stopped use of Vioxx months prior to the Merck recall (Graham et al., 2005).

Finally, EHRs can be used to detect, document, analyze, track, and report patient safety problems, including both adverse events and errors. Initially, automated EHRs were used to detect adverse drug events in hospital patients (Classen et al., 1991; Jha et al., 1998). This type of automated surveillance was expanded to health care-associated infections (Evans et al., 1998) and has been used increasingly by hospitals as the routine and operational approach to detecting these infections from ICU central line-related bloodstream infections to surgical-site infections that occur long after discharge (Wright, 2008). Recently, commercial EHRs that allow broad use of real-time safety tracking systems have been expanded to detect global adverse events in hospitalized patients (Classen et al., 2011). These EHR databases could also be retrospectively data-mined to study the occurrence of harm to patients across the continuum of care.

LESSONS FROM THE INTERNATIONAL COMMUNITY: INTERNATIONAL COMPARISONS

There are a significant number of high-income countries and multinational programs that have made substantial progress in implementing health IT and improving patient safety, at least in the ambulatory care setting. These countries can serve as important lessons from these settings for health professionals and policy makers in the United States. There has been a series of multinational programs to improve patient safety in

which health IT has played a key role. For example, the World Health Organization’s (WHO’s) patient safety program has 13 specific patient safety action areas that focus on patient safety as a global health care issue. These action areas are aimed to “coordinate, disseminate, and accelerate improvements in patient safety worldwide” (WHO, 2011a). Although it focuses on a broad range of safety issues, “Action Area 8: Technology and Patient Safety,” most specifically, targets systemic and technical aspects to improve patient safety around the world by promoting personal health records (PHRs), automated prescribing systems, simulation training, and failsafe mechanisms in diagnostic tools, such as computerized radiographs (WHO, 2008, 2011a, 2011b).

On a similarly large scale, the European Union (EU) has funded specific eHealth initiatives (EU, 2010a) and the use of technology to improve the quality and safety of care delivered during disaster response efforts (EU, 2007). These programs focus on PHRs, patient guidance services, virtual physiological humans, and computer simulations. The EU is supporting several efforts using information and communication technologies to improve patient safety, focusing on the “development of advanced applications to improve risk assessment and patient safety” (EU, 2010b). In 2009 alone, the EU invested €28 million (EU, 2010b), including programs such as Patient Safety through Intelligent Procedures in Medication (PSIP), whose main aim is to develop computer applications and to educate providers and patients on how to prevent medication errors (PSIP, 2011). Additionally, the Safety for Robotic Surgery (SAFROS) project seeks to develop technologies for patient safety in robotic surgery (SAFROS, 2009).

Broad country comparison studies have been conducted on the use of health IT and its potential to improve patient safety. For example, an international cross-sectional study examined health IT’s functional capacity and quality of care delivered in Australia, Canada, Denmark, Germany, the Netherlands, New Zealand, the United Kingdom, and the United States. The study found that, when controlling for within country differences of specific health IT methods adopted and primary care physician (PCP) practice sizes, significant disparities exist in the quality of care delivered among practices with low IT capacity compared to those with high IT capacity. IT functional capacity was measured through a count of 14 different items (such as whether the clinician used an EHR, prescribed medicine electronically, and had a computerized system for patient reminders, prompts for potential drug interaction, and test results). Practices were deemed “low” if they had 2 or fewer of the 14 items and “high” if they had between 7 and 14 items (Davis et al., 2009).

Although the study focused on several outcomes, the specific safety outcome measured was whether a physician practice had a specific, docu

mented process for patient follow-up and analysis of adverse events. Thirty- eight percent of physicians had a documented process for all adverse events (ranging from 27 percent of physicians in low-capacity countries to 43 percent in high-capacity countries), while 17 percent of physicians had a process for adverse drug reactions online (ranging from 22 percent in low- capacity countries to 15 percent in high-capacity countries) (Davis et al., 2009). Approximately 50 percent of practices with low IT capacity reported no processes for following up on adverse events compared to 41 percent of practices with higher IT functionality. Researchers suggested that countries that support a stronger IT infrastructure are better suited to address coordination of care and safety issues, as well as to maintain satisfaction among the PCP community (Davis et al., 2009).

Other country-specific studies have been conducted, including a series of papers comparing the adoption of health IT among PCP offices in New Zealand and Denmark, two countries leading the way in the adoption of health IT over the past two decades (Protti et al., 2008a, 2008b, 2008c, 2008d, 2009). These studies suggest that it has been possible for many nations to adopt and use health IT in PCP practices without measurable, deleterious consequences on patient safety.

Although the United States has made significant strides in health IT over the past 20 years, it is clear that many other high-income nations are much further ahead in IT adoption, at least in the ambulatory setting. Despite the fact that these other nations have had a much greater experience with health IT, there is very little direct information on the impact of their investments on patient safety. The primary lesson for the United States is that it is possible to have widespread adoption of health IT without harming safety. What the optimal strategies are for doing so cannot be so easily gleaned by looking at these other nations.

CONCLUSION

Health IT has already been shown to improve medication safety. Although the evidence is mixed for areas outside of medication safety, both within the United States and abroad, the fact that several studies have improved patient safety with implementation of health IT leads the committee to believe that health IT has at least the potential to drastically improve patient safety in other areas of care. As with any new technology, health IT carries benefits and risks of new and greater harms. To fully capitalize on the potential that health IT may have on patient safety, a more comprehensive understanding of how health IT impacts potential harms, workflow, and safety is needed.

REFERENCES

Abarca, J., L. R. Colon, V. S. Wang, D. C. Malone, J. E. Murphy, and E. P. Armstrong. 2006. Evaluation of the performance of drug-drug interaction screening software in community and hospital pharmacies. Journal of Managed Care Pharmacy 12(5):383-389.

Agostini, J. V., Y. Zhang, and S. K. Inouye. 2007. Use of a computer-based reminder to improve sedative-hypnotic prescribing in older hospitalized patients. Journal of the American Geriatrics Society 55(1):43-48.

Aleccia, J. 2011. Nurse’s suicide highlights twin tragedies of medical errors. MSNBC.com. http://www.msnbc.msn.com/id/43529641/ns/health-health_care/t/nurses-suicide-highlights-twin-tragedies-medical-errors/#.Tph_EHI2Y9Y (accessed October 14, 2011).

Ali, N. A., H. S. Mekhjian, P. L. Kuehn, T. D. Bentley, R. Kumar, A. K. Ferketich, and S. P. Hoffmann. 2005. Specificity of computerized physician order entry has a significant effect on the efficiency of workflow for critically ill patients. Critical Care Medicine 33(1):110-114.

Ammenwerth, E., P. Schnell-Inderst, C. Machan, and U. Siebert. 2008. The effect of electronic prescribing on medication errors and adverse drug events: A systematic review. Journal of the American Medical Informatics Association 15(5):585-600.

Ash, J. S., D. F. Sittig, R. Dykstra, E. Campbell, and K. Guappone. 2009. The unintended consequences of computerized provider order entry: Findings from a mixed methods exploration. International Journal of Medical Informatics 78(Suppl 1):S69-S76.

Associated Press. 2009. Veterans given wrong drug doses due to glitch. MSNBC.com. http://www.msnbc.msn.com/id/28655104/#.Tph-b3I2Y9Z (accessed October 14, 2011).

Bates, D. W., and A. A. Gawande. 2003. Improving safety with information technology. New England Journal of Medicine 348(25):2526-2534.

Bates, D. W., G. J. Kuperman, S. Wang, T. Gandhi, A. Kittler, L. Volk, C. Spurr, R. Khorasani, M. Tanasijevic, and B. Middleton. 2003. Ten commandments for effective clinical decision support: Making the practice of evidence-based medicine a reality. Journal of the American Medical Informatics Association 10(6):523-530.

Begley, S. 2011. The best medicine: Cutting health costs with comparative effectiveness research. Scientific American. http://www.scientificamerican.com/article.cfm?id=the-best-medicine-july-11 (accessed August 5, 2011).

Benham-Hutchins, M. M., and J. A. Effken. 2010. Multi-professional patterns and methods of communication during patient handoffs. International Journal of Medical Informatics 79(4):252-267.

Black, A. D., J. Car, C. Pagliari, C. Anandan, K. Cresswell, T. Bokun, B. McKinstry, R. Procter, A. Majeed, and A. Sheikh. 2011. The impact of eHealth on the quality and safety of health care: A systematic overview. Public Library of Science Medicine 8(1).

Callen, J., R. Paoloni, A. Georgiou, M. Prgomet, and J. Westbrook. 2010. The rate of missed test results in an emergency department an evaluation using an electronic test order and results viewing system. Methods of Information in Medicine 49(1):37-43.

Claridge, J. A., J. F. Golob, A. M. A. Fadlalla, B. M. D’Amico, J. R. Peerless, C. J. Yowler, and M. A. Malangoni. 2009. Who is monitoring your infections: Shouldn’t you be? Surgical Infections 10(1):59-64.

Classen, D. C., S. L. Pestotnik, R. S. Evans, and J. P. Burke. 1991. Computerized surveillance of adverse drug events in hospital patients. Journal of the American Medical Association 266(20):2847-2851.

Classen, D. C., R. Resar, F. Griffin, F. Federico, T. Frankel, N. Kimmel, J. C. Whittington, A. Frankel, A. Seger, and B. C. James. 2011. “Global trigger tool” shows that adverse events in hospitals may be ten times greater than previously measured. Health Affairs 30(4):581-589.

Conroy, S., D. Sweis, C. Planner, V. Yeung, J. Collier, L. Haines, and I. C. K. Wong. 2007. Interventions to reduce dosing errors in children—a systematic review of the literature. Drug Safety 30(12):1111-1125.

Culler, S. D., A. Atherly, S. Walczak, A. Davis, J. N. Hawley, K. J. Rask, V. Naylor, and K. E. Thorpe. 2006. Urban-rural differences in the availability of hospital information technology applications: A survey of Georgia hospitals. Journal of Rural Health 22(3):242-247.

Davis, K., M. M. Doty, K. Shea, and K. Stremikis. 2009. Health information technology and physician perceptions of quality of care and satisfaction. Health Policy 90(2-3):239-246.

Del Beccaro, M. A., H. E. Jeffries, M. A. Eisenberg, and E. D. Harry. 2006. Computerized provider order entry implementation: No association with increased mortality rates in an intensive care unit. Pediatrics 118(1):290-295.

DesRoches, C. M., E. G. Campbell, S. R. Rao, K. Donelan, T. G. Ferris, A. Jha, R. Kaushal, D. E. Levy, S. Rosenbaum, A. E. Shields, and D. Blumenthal. 2008. Electronic health records in ambulatory care—a national survey of physicians. New England Journal of Medicine 359(1):50-60.

DesRoches, C. M., E. G. Campbell, C. Vogeli, J. Zheng, S. R. Rao, A. E. Shields, K. Donelan, S. Rosenbaum, S. J. Bristol, and A. K. Jha. 2010. Electronic health records’ limited successes suggest more targeted uses. Health Affairs 29(4):639-646.

Devine, E. B., R. N. Hansen, J. L. Wilson-Norton, N. M. Lawless, A. W. Fisk, D. K. Blough, D. P. Martin, and S. D. Sullivan. 2010. The impact of computerized provider order entry on medication errors in a multispecialty group practice. Journal of the American Medical Informatics Association 17(1):78-84.

Dorr, D., L. M. Bonner, A. N. Cohen, R. S. Shoai, R. Perrin, E. Chaney, and A. S. Young. 2007. Informatics systems to promote improved care for chronic illness: A literature review. Journal of the American Medical Informatics Association 14(2):156-163.

Durieux, P., L. Trinquart, I. Colombet, J. Nies, R. Walton, A. Rajeswaran, M. Rege-Walther, E. Harvey, and B. Burnand. 2008. Computerized advice on drug dosage to improve prescribing practice. Cochrane Database of Systematic Reviews (3):CD002894.

Einbinder, J. S., and D. W. Bates. 2007. Leveraging information technology to improve quality and safety. Yearbook of Medical Informatics 2007:22-29.

EU (European Union). 2007. Managing disasters and emergencies. http://ec.europa.eu/information_society/tl/qualif/safety/index_en.htm (accessed June 20, 2011).

EU. 2010a. Research in eHealth. http://ec.europa.eu/information_society/activities/health/research/index_en.htm (accessed June 20, 2011).

EU. 2010b. Risk assessment and patient safety. http://ec.europa.eu/information_society/activities/health/research/fp7ps/index_en.htm (accessed June 20, 2011).

Evans, R. S., S. L. Pestotnik, D. C. Classen, T. P. Clemmer, L. K. Weaver, J. F. Orme, J. F. Lloyd, and J. P. Burke. 1998. A computer-assisted management program for antibiotics and other antiinfective agents. New England Journal of Medicine 338(4):232-238.

Feldstein, A. C., W. M. Vollmer, D. H. Smith, A. Petrik, J. Schneider, H. Glauber, and M. Herson. 2007. An outreach program improved osteoporosis management after a fracture. Journal of the American Geriatrics Society 55(9):1464-1469.

Ferris, T. G., S. A. Johnson, J. P. T. Co, M. Backus, J. Perrin, D. W. Bates, and E. G. Poon. 2009. Electronic results management in pediatric ambulatory care: Qualitative assessment. Pediatrics 123(Suppl 2):S85-S91.

Fitzgerald, M., P. Cameron, C. Mackenzie, N. Farrow, P. Scicluna, R. Gocentas, A. Bystrzycki, G. Lee, G. O’Reilly, N. Andrianopoulos, L. Dziukas, D. J. Cooper, A. Silvers, A. Mori, A. Murray, S. Smith, Y. Xiao, D. Stub, F. T. McDermott, and J. V. Rosenfeld. 2011. Trauma resuscitation errors and computer-assisted decision support. Archives of Surgery 146(2):218-225.

Franklin, B. D., K. O’Grady, P. Donyai, A. Jacklin, and N. Barber. 2007. The impact of a closed-loop electronic prescribing and administration system on prescribing errors, administration errors and staff time: A before-and-after study. Quality & Safety in Health Care 16(4):279-284.

Gandhi, T. K., S. B. Bartel, L. N. Shulman, D. Verrier, E. Burdick, A. Cleary, J. M. Rothschild, L. L. Leape, and D. W. Bates. 2005. Medication safety in the ambulatory chemotherapy setting. Cancer 104(11):2477-2483.

Gandhi, T. K., A. Kachalia, E. J. Thomas, A. L. Puopolo, C. Yoon, T. A. Brennan, and D. M. Studdert. 2006. Missed and delayed diagnoses in the ambulatory setting: A study of closed malpractice claims. Annals of Internal Medicine 145(7):488-496.

Garg, A., N. Adhikari, H. McDonald, M. Rosas-Arellano, P. Devereaux, J. Beyene, J. Sam, and R. Haynes. 2005. Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: A systematic review. Journal of the American Medical Association 293(10):1223-1238.

Gearing, P., C. M. Olney, K. Davis, D. Lozano, L. B. Smith, and B. Friedman. 2006. Enhancing patient safety through electronic medical record documentation of vital signs. Journal of Healthcare Information Management 20(4):40-45.

Georgiou, A., M. Williamson, J. I. Westbrook, and S. Ray. 2007. The impact of computerised physician order entry systems on pathology services: A systematic review. International Journal of Medical Informatics 76(7):514-529.

Glassman, P. A., P. Belperio, A. Lanto, B. Simon, R. Valuck, J. Sayers, and M. Lee. 2007. The utility of adding retrospective medication profiling to computerized provider order entry in an ambulatory care population. Journal of the American Medical Informatics Association 14(4):424-431.

Gohil, B., H. Gholamhosseini, M. J. Harrison, A. Lowe, and A. Al-Jumaily. 2007. Intelligent monitoring of critical pathological events during anesthesia. Paper presented at Annual International Conference of the IEEE Engineering in Medicine & Biology Society, Lyon, France.

Goodman, K. W., E. S. Berner, M. A. Dente, B. Kaplan, R. Koppel, D. Rucker, D. Z. Sands, P. Winkelstein, and the AMIA Board of Directors. 2011. Challenges in ethics, safety, best practices, and oversight regarding HIT vendors, their customers, and patients: A report of an AMIA special task force. Journal of the American Medical Informatics Association 18(1):77-81.

Gordon, J. R. S., T. Wahls, R. C. Carlos, I. I. Pipinos, G. E. Rosenthal, and P. Cram. 2009. Failure to recognize newly identified aortic dilations in a health care system with an advanced electronic medical record. Annals of Internal Medicine 151(1):21-27.

Graham, D. J., D. Campen, R. Hui, M. Spence, C. Cheetham, G. Levy, S. Shoor, and W. A. Ray. 2005. Risk of acute myocardial infarction and sudden cardiac death in patients treated with cyclo-oxygenase 2 selective and non-selective non-steroidal anti-inflammatory drugs: Nested case-control study. The Lancet 365(9458):475-481.

Graham, J., and C. Dizikes. 2011. Baby’s death spotlights safety risks linked to computerized systems. Los Angeles Times. http://www.latimes.com/health/ct-met-technology-errors-20110627,0,2158183.story (accessed June 29, 2011).

Greenhalgh, T., K. Stramer, T. Bratan, E. Byrne, Y. Mohammad, and J. Russell. 2008. Introduction of shared electronic records: Multi-site case study using diffusion of innovation theory. British Medical Journal 337:a1786.

Gurwitz, J. H., T. S. Field, J. Judge, P. Rochon, L. R. Harrold, C. Cadoret, M. Lee, K. White, J. LaPrino, J. Erramuspe-Mainard, M. DeFlorio, L. Gavendo, J. Auger, and D. W. Bates. 2005. The incidence of adverse drug events in two large academic long-term care facilities. American Journal of Medicine 118(3):251-258.

Han, Y. Y., J. A. Carcillo, S. T. Venkataraman, R. S. B. Clark, R. S. Watson, T. C. Nguyen, H. Bayir, and R. A. Orr. 2005. Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system. Pediatrics 116(6):1506-1512.

Hayrinen, K., K. Saranto, and P. Nykanen. 2008. Definition, structure, content, use and impacts of electronic health records: A review of the research literature. International Journal of Medical Informatics 77(5):291-304.

Herasevich, V., M. Yilmaz, H. Khan, R. D. Hubmayr, and O. Gajic. 2009. Validation of an electronic surveillance system for acute lung injury. Intensive Care Medicine 35(6):1018-1023.

HHS (U.S. Department of Health and Human Services). 2004. Health IT strategic framework: Glossary of selected terms (accessed May 9, 2011).

IOM (Institute of Medicine). 2004. Patient safety: Achieving a new standard for care. Washington, DC: The National Academies Press.

Isaac, T., J. S. Weissman, R. B. Davis, M. Massagli, A. Cyrulik, D. Z. Sands, and S. N. Weingart. 2009. Overrides of medication alerts in ambulatory care. Archives of Internal Medicine 169(3):305-311.

Jha, A. K., G. J. Kuperman, J. M. Teich, L. Leape, B. Shea, E. Rittenberg, E. Burdick, D. L. Seger, M. V. Vliet, and D. W. Bates. 1998. Identifying adverse drug events. Journal of the American Medical Informatics Association 5(3):305-314.

Jha, A. K., T. G. Ferris, K. Donelan, C. DesRoches, A. Shields, S. Rosenbaum, and D. Blumenthal. 2006. How common are electronic health records in the United States? A summary of the evidence. Health Affairs 25(6):w496-w507.

Jha, A. K., J. Laguette, A. Seger, and D. W. Bates. 2008. Can surveillance system identify and avert adverse drug events? A prospective evaluation of a commercial application. Journal of the American Medical Informatics Association 15(5):647-653.

Jha, A. K., C. M. DesRoches, E. G. Campbell, K. Donelan, S. R. Rao, T. G. Ferris, A. Shields, S. Rosenbaum, and D. Blumenthal. 2009a. Use of electronic health records in U.S. hospitals. New England Journal of Medicine 360(16):1628-1638.

Jha, A. K., E. J. Orav, and A. M. Epstein. 2009b. Preventing readmissions with improved hospital discharge planning. New England Journal of Medicine 361(27):2637-2645.

Kadmon, G., E. Bron-Harlev, E. Nahum, O. Schiller, G. Haski, and T. Shonfeld. 2009. Computerized order entry with limited decision support to prevent prescription errors in a PICU. Pediatrics 124(3):935-940.

Kaushal, R., K. G. Shojania, and D. W. Bates. 2003. Effects of computerized physician order entry and clinical decision support systems on medication safety: A systematic review. Archives of Internal Medicine 163(12):1409-1416.

Koppel, R., and D. Kreda. 2009. Health care information technology vendors’ “hold harmless” clause. Journal of the American Medical Association 301(12):1276-1278.

Koppel, R., J. P. Metlay, A. Cohen, B. Abaluck, A. R. Localio, S. E. Kimmel, and B. L. Strom. 2005. Role of computerized physician order entry systems in facilitating medication errors. Journal of the American Medical Association 293(10):1197-1203.

Koppel, R., T. Wetterneck, J. L. Telles, and B. T. Karsh. 2008. Workarounds to barcode medication administration systems: Their occurrences, causes, and threats to patient safety. Journal of the American Medical Informatics Association 15(4):408-423.

Kravitz, R. L., J. E. Rolph, and L. Petersen. 1997. Omission-related malpractice claims and the limits of defensive medicine. Medical Care Research and Review 54(4):456-471.

Lin, C. P., T. H. Payne, W. P. Nichol, P. J. Hoey, C. L. Andreson, and J. H. Gennari. 2008. Evaluating clinical decision support systems: Monitoring CPOE order check override rates in the Department of Veterans Affairs’ computerized patient record system. Journal of the American Medical Informatics Association 15(5):620-626.

Litaker, D., C. Ritter, S. Ober, and D. Aron. 2005. Continuity of care and cardiovascular risk factor management: Does care by a single clinician add to informational continuity provided by electronic medical records? American Journal of Managed Care 11(11):689-696.

Longhurst, C. A., L. Parast, C. I. Sandborg, E. Widen, J. Sullivan, J. S. Hahn, C. G. Dawes, and P. J. Sharek. 2010. Decrease in hospital-wide mortality rate after implementation of a commercially sold computerized physician order entry system. Pediatrics 126(1):14-21.

Magid, D. J., S. M. Shetterly, K. L. Margolis, H. M. Tavel, P. J. O’Connor, J. V. Selby, and P. M. Ho. 2010. Comparative effectiveness of angiotensin-converting enzyme inhibitors versus P-blockers as second-line therapy for hypertension. Circulation: Cardiovascular Quality and Outcomes 3(5):453-458.

Magrabi, F., S. Y. W. Li, R. O. Day, and E. Coiera. 2010. Errors and electronic prescribing: A controlled laboratory study to examine task complexity and interruption effects. Journal of the American Medical Informatics Association 17(5):575-583.

Magrabi, F., M.-S. Ong, W. Runciman, and E. Coiera. 2011. Using FDA reports to inform a classification for health information technology safety problems. Journal of the American Medical Informatics Association 19(1):45-53.

Matheny, M. E., T. K. Gandhi, E. J. Orav, Z. Ladak-Merchant, D. W. Bates, G. J. Kuperman, and E. G. Poon. 2007. Impact of an automated test results management system on patients’ satisfaction about test result communication. Archives of Internal Medicine 167(20):2233-2239.

Metzger, J., E. Welebob, D. W. Bates, S. Lipsitz, and D. C. Classen. 2010. Mixed results in the safety performance of computerized physician order entry. Health Affairs 29(4):655-663.

Murray, E., J. Burns, T. S. See, R. Lai, and I. Nazareth. 2009. Interactive health communication applications for people with chronic disease. Cochrane Database of Systematic Reviews (1):CD004274.

Nam, H. S., S. W. Han, S. H. Ahn, J. Y. Lee, H. Y. Choi, I. C. Park, and J. H. Heo. 2007. Improved time intervals by implementation of computerized physician order entry-based stroke team approach. Cerebrovascular Diseases 23(4):289-293.

Nanji, K. C., J. M. Rothschild, C. Salzberg, C. A. Keohane, K. Zigmont, J. Devita, T. K. Gandhi, A. K. Dalal, D. W. Bates, and E. G. Poon. 2011. Errors associated with outpatient computerized prescribing systems. Journal of the American Medical Informatics Association 18(6):767-773.

Ngo-Metzger, Q., G. R. Hayes, Y. N. Chen, R. Cygan, and C. F. Garfield. 2010. Improving communication between patients and providers using health information technology and other quality improvement strategies: Focus on low-income children. Medical Care Research and Review 67(5):246S-267S.

Niazkhani, Z., H. Pirnejad, M. Berg, and J. Aarts. 2009. The impact of computerized provider order entry systems on inpatient clinical workflow: A literature review. Journal of the American Medical Informatics Association 16(4):539-549.

Nies, J., I. Colombet, E. Zapletal, F. Gillaizeau, P. Chevalier, and P. Durieux. 2010. Effects of automated alerts on unnecessarily repeated serology tests in a cardiovascular surgery department: A time series analysis. Health Services Research 10(1):70.

ONC (Office of the National Coordinator for Health Information Technology). 2009. Health IT terms: Glossary of selected terms related to Health IT. http://healthit.hhs.gov/portal/server.pt?open=512&mode=2&cached=true&objID=1256&PageID=15726 (accessed March 20, 2011).

Parente, S. T., and J. S. McCullough. 2009. Health information technology and patient safety: Evidence from panel data. Health Affairs 28(2):357-360.

Phansalkar, S., J. Edworthy, E. Hellier, D. L. Seger, A. Schedlbauer, A. J. Avery, and D. W. Bates. 2010. A review of human factors principles for the design and implementation of medication safety alerts in clinical information systems. Journal of the American Medical Informatics Association 17(5):493-501.

Poon, E. G., T. K. Gandhi, T. D. Sequist, H. J. Murff, A. S. Karson, and D. W. Bates. 2004. “I wish I had seen this test result earlier!”: Dissatisfaction with test result management systems in primary care. Archives of Internal Medicine 164(20):2223-2228.

Poon, E. G., J. L. Cina, W. Churchill, N. Patel, E. Featherstone, J. M. Rothschild, C. A. Keohane, A. D. Whittemore, D. W. Bates, and T. K. Gandhi. 2006. Medication dispensing errors and potential adverse drug events before and after implementing bar code technology in the pharmacy. Annals of Internal Medicine 145(6):426-434.

Poon, E. G., C. A. Keohane, C. S. Yoon, M. Ditmore, A. Bane, O. Levtzion-Korach, T. Moniz, J. M. Rothschild, A. B. Kachalia, J. Hayes, W. W. Churchill, S. Lipsitz, A. D. Whittemore, D. W. Bates, and T. K. Gandhi. 2010. Effect of bar-code technology on the safety of medication administration. New England Journal of Medicine 362(18):1698-1707.

Protti, D., T. Bowden, and I. Johansen. 2008a. Adoption of information technology in primary care physician offices in New Zealand and Denmark, part 1: Healthcare system comparisons. Informatics in Primary Care 16:183-187.

Protti, D., T. Bowden, and I. Johansen. 2008b. Adoption of information technology in primary care physician offices in New Zealand and Denmark, part 2: Historical comparisons. Informatics in Primary Care 16:189-193.

Protti, D., T. Bowden, and I. Johansen. 2008c. Adoption of information technology in primary care physician offices in New Zealand and Denmark, part 3: Medical record environment comparisons. Informatics in Primary Care 16:285-290.

Protti, D., T. Bowden, and I. Johansen. 2008d. Adoption of information technology in primary care physician offices in New Zealand and Denmark, part 4: Benefits comparisons. Informatics in Primary Care 16:291-296.

Protti, D., T. Bowden, and I. Johansen. 2009. Adoption of information technology in primary care physician offices in New Zealand and Denmark, part 5: Final comparisons. Informatics in Primary Care 17:17-22.

PSIP (Patient Safety through Intelligent Procedures in Medication). 2011. PSIP: Patient safety through intelligent procedures in medication. http://ec.europa.eu/information_society/activities/health/docs/projects/fp7/psip-factsheet.pdf (accessed June 25, 2011).

Raebel, M. A., J. Charles, J. Dugan, N. M. Carroll, E. J. Korner, D. W. Brand, and D. J. Magid. 2007. Randomized trial to improve prescribing safety in ambulatory elderly patients. Journal of the American Geriatrics Society 55(7):977-985.

Reckmann, M. H., J. I. Westbrook, Y. Koh, C. Lo, and R. O. Day. 2009. Does computerized provider order entry reduce prescribing errors for hospital inpatients? A systematic review. Journal of the American Medical Informatics Association 16(5):613-623.

SAFROS (Safety for Robotic Surgery). 2009. The SAFROS project. http://www.safros.eu/ (accessed June 27, 2011).

Santell, J. P., J. G. Kowiatek, R. J. Weber, R. W. Hicks, and C. A. Sirio. 2009. Medication errors resulting from computer entry by nonprescribers. American Journal of Health- System Pharmacy 66(9):843-853.

Schulte, F., and E. Schwartz. 2010. As doctors shift to electronic health systems, signs of harm emerge. Huffington Post Investigative Fund. http://huffpostfund.org/stories/2010/04/doctors-shift-electronic-health-systems-signs-harm-emerge (accessed October 13, 2010).

Sellier, E., I. Colombet, B. Sabatier, G. Breton, J. Nies, E. Zapletal, J. B. Arlet, D. Somme, and P. Durieux. 2009. Effect of alerts for drug dosage adjustment in inpatients with renal insufficiency. Journal of the American Medical Informatics Association 16(2):203-210.

Shah, N. R., A. C. Seger, D. L. Seger, J. M. Fiskio, G. J. Kuperman, B. Blumenfeld, E. G. Recklet, D. W. Bates, and T. K. Gandhi. 2006. Improving acceptance of computerized prescribing alerts in ambulatory care. Journal of the American Medical Informatics Association 13(1):5-11.

Shamliyan, T. A., S. Duval, J. Du, and R. L. Kane. 2008. Just what the doctor ordered. Review of the evidence of the impact of computerized physician order entry system on medication errors. Health Services Research 43(1 Pt 1):32-53.

Silver, J. D., and S. D. Hamill. 20111. Doctor, nurse disciplined by UPMC in kidney transplant case. Post-Gazette.com. http://www.post-gazette.com/pg/11147/1149429-114-0.stm#ixzz1amZZEw6Y (accessed October 14, 2011).

Singh, H., S. Mani, D. Espadas, N. Petersen, V. Franklin, and L. A. Petersen. 2009. Prescription errors and outcomes related to inconsistent information transmitted through computerized order entry: A prospective study. Archives of Internal Medicine 169(10):982-989.

Singh, H., E. J. Thomas, D. F. Sittig, L. Wilson, D. Espadas, M. M. Khan, and L. A. Petersen. 2010. Notification of abnormal lab test results in an electronic medical record: Do any safety concerns remain? American Journal of Medicine 123(3):238-244.

Smith, D. H., N. Perrin, A. Feldstein, X. H. Yang, D. Kuang, S. R. Simon, D. F. Sittig, R. Platt, and S. B. Soumerai. 2006. The impact of prescribing safety alerts for elderly persons in an electronic medical record—an interrupted time series evaluation. Archives of Internal Medicine 166(10):1098-1104.

Smith, L. B., L. Banner, D. Lozano, C. M. Olney, and B. Friedman. 2009. Connected care: Reducing errors through automated vital signs data upload. Computers, Informatics, Nursing 27(5):318-323.

Tierney, W. M., J. M. Overhage, M. D. Murray, L. E. Harris, X.-H. Zhou, G. J. Eckert, F. E. Smith, N. Nienaber, C. J. McDonald, and F. D. Wolinsky. 2005. Can computer-generated evidence-based care suggestions enhance evidence-based management of asthma and chronic obstructive pulmonary disease? A randomized, controlled trial. Health Services Research 40(2):477-497.

U.S. News. 2011. E-prescribing doesn’t slash errors, study finds. U.S. News and World Report. http://health.usnews.com/health-news/family-health/articles/2011/06/30/e-prescribing-doesnt-slash-errors--study-finds (accessed October 14, 2011).

van der Sijs, H., A. Mulder, T. van Gelder, J. Aarts, M. Berg, and A. Vulto. 2009. Drug safety alert generation and overriding in a large Dutch university medical centre. Pharmaco- epidemiology and Drug Safety 18(10):941-947.

van der Sijs, H., T. van Gelder, A. Vulto, M. Berg, and J. Aarts. 2010. Understanding handling of drug safety alerts: A simulation study. International Journal of Medical Informatics 79(5):361-369.

Walsh, K. E., W. G. Adams, H. Bauchner, R. J. Vinci, J. B. Chessare, M. R. Cooper, P. M. Hebert, E. G. Schainker, and C. P. Landrigan. 2006. Medication errors related to computerized order entry for children. Pediatrics 118(5):1872-1879.

Weant, K. A., A. M. Cook, and J. A. Armitstead. 2007. Medication-error reporting and pharmacy resident experience during implementation of computerized prescriber order entry. American Journal of Health-System Pharmacy 64(5):526-530.

Whitley, H. P., J. D. Fermo, and E. C. G. Chumney. 2006. 5-year evaluation of electronic medical record flag alerts for patients warranting secondary prevention of coronary heart disease. Pharmacotherapy 26(5):682-688.

WHO (World Health Organization). 2008. World alliance for patient safety: Forward programme 2008-2009. http://www.who.int/patientsafety/information_centre/reports/ Alliance_Forward_Programme_2008.pdf (accessed June 27, 2011).

WHO. 2011a. Patient safety: About us. http://www.who.int/patientsafety/about/en/index.html (accessed June 25, 2011).

WHO. 2011b. WHO patient safety—programme areas. http://www.who.int/patientsafety/about/programmes/en/index.html (accessed June 25, 2011). Wolfstadt, J. I., J. H. Gurwitz, T. S. Field, M. Lee, S. Kalkar, W. Wu, and P. A. Rochon. 2008. The effect of computerized physician order entry with clinical decision support on the rates of adverse drug events: A systematic review. Journal of General Internal Medicine 23(4):451-458.

Wright, A., J. Pang, J. C. Feblowitz, F. L. Maloney, A. R. Wilcox, H. Z. Ramelson, L. I. Schneider, and D. W. Bates. 2011. A method and knowledge base for automated inference of patient problems from structured data in an electronic medical record. Journal of the American Medical Informatics Association 18(6):859-867.

Wright, M.-O. 2008. Automated surveillance and infection control: Toward a better tomorrow. American Journal of Infection Control 36(3 Suppl 1):S1-S6.

Zhou, Y. Y., R. Unitan, J. J. Wang, T. Garrido, H. L. Chin, M. C. Turley, and L. Radler. 2011. Improving population care with an integrated electronic panel support tool. Population Health Management 14(1):3-9.

This page intentionally left blank.