Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, ECONOMICS AND FINANCE ( (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 11 December 2019

Health Information Technology

Summary and Keywords

The potential for health information technology (HIT) to reshape the information-intensive healthcare industry has been recognized for decades. Nevertheless, the adoption and use of IT in healthcare has lagged behind other industries, motivating governments to take a role in supporting its use to achieve envisioned benefits. This dynamic has led to three major strands of research. Firstly, the relatively slow and uneven adoption of HIT, coupled with government programs intended to speed adoption, has raised the issue of who is adopting HIT, and the impact of public programs on rates of adoption and diffusion. Secondly, the realization of benefits from HIT appears to be occurring more slowly than its proponents had hoped, leading to an ongoing need to empirically measure the effect of its use on the quality and efficiency of healthcare as well as the contexts under which benefits are best realized. Thirdly, increases in the adoption and use of HIT have led to the potential for interoperable exchange of patient information and the dynamic use of that information to drive improvements in the healthcare delivery system; however, these applications require developing new approaches to overcoming barriers to collaboration between healthcare organizations and the HIT industry itself. Intertwined through each of these issues is the interaction between HIT as a tool for standardization and systemic change in the practice of healthcare, and healthcare professionals’ desire to preserve autonomy within the increasingly structured healthcare delivery system. Innovative approaches to improve the interactions between professionals, technology, and market forces are therefore necessary to capitalize on the promise of HIT and develop a continually learning health system.

Keywords: health information technology, interoperability, technology adoption, healthcare quality, electronic health record, health economics


The potential for information technology to improve the delivery of healthcare has been evident since early applications of computing (Shortliffe, 1999). While other forms of medical technology have historically led to increased costs of healthcare (Cutler, Rosen, & Vijan, 2006), health information technology (HIT) has the potential to reduce the cost of healthcare while improving quality (Hillestad et al., 2005). HIT promises to improve efficiency and reduce redundant utilization of services by making patient information more readily available; by speeding adoption of best practices by presenting medical knowledge to providers when it is needed; and by decreasing opportunities for costly errors by standardizing the entry of information and orders (Orzag, 2008).

Despite the promised benefits of HIT, adoption has lagged behind other industries’ progress toward digitization (Jha et al., 2006). Competition and the design of reimbursement systems often encourage the adoption of new medical technologies, but incentives are not aligned to promote adoption of HIT, particularly in delivery systems in which system-wide efficiency can mean firm-level reduced revenue (Ash & Bates, 2005). Without financial incentives, healthcare providers historically found it challenging to justify the cost of implementation, especially because clinicians are not unified in support of HIT (Sprague, 2004). HIT can be challenging and time-consuming to use (Boonstra & Broekhuis, 2010) and fundamentally reshapes how clinicians practice: one study found that in 2016 over 50% of physicians’ time was spent interacting with a computer (Tai-Seale et al., 2017). And whereas the development of new medical technology is central to the mission of the medical profession, HIT can represent a challenge to professional autonomy and reinforce structures within which clinicians must practice (Freidson, 1974; Timmermans & Berg, 2010; Wears & Berg, 2005).

The push of potential benefits and pull of social forces slowing adoption and use of HIT have motivated public policy to increase adoption of HIT, and in turn motivates economic analysis of those policies and the effect of HIT on the delivery of care. However, several factors complicate the study of these policies and of HIT. Firstly, public programs have been broadly targeted, making it difficult to identify comparator groups and define clear effects of these policies (Adler-Milstein & Jha, 2017; Mennemeyer, Menachemi, Rahurkar, & Ford, 2016). Secondly, identifying the effect of HIT on care is difficult because HIT is not exogenously assigned but rather follows diffusion patterns driven by a wide range of organizational characteristics (Ash, 1997; Jones, Rudin, Perry, & Shekelle, 2014). Thirdly, variation in results from single-firm studies points toward the likelihood that organizational and implementation context are extremely important in determining the effect of HIT (Jones et al., 2014). Finally, adoption of IT in other industries has been subject to a “productivity paradox” in which anticipated benefits are either not achieved or only achieved after a long delay following implementation, making it harder to identify effects in the short run (David, 1990; Jones, Heaton, Rudin, & Schneider, 2012). Given these challenges, it is perhaps not surprising that despite a large body of research on HIT, there are more questions than answers about the broad, long-run effect of HIT on the delivery of healthcare.

The goal of this article is to survey key issues informing the economic analysis of HIT and policies intended to support its adoption. In particular, this article focuses on three topics: progress in adoption of HIT, including the logic for and effect of public policies around HIT; the anticipated benefits of HIT and related evidence; and the broader influence of HIT on the practice of medicine and the healthcare delivery system, with a focus on broad-based information sharing. The article concludes with a discussion of the future of HIT, highlighting the interplay between professions and changing technology necessary to bring about a new continuously learning system.

Defining Health Information Technology

HIT influences the delivery of healthcare in a variety of ways, in part because it is an umbrella term encompassing several intrinsically interlinked component technologies. Unpacking these technologies is necessary to conceptualize the effects of HIT on the quality and efficiency of care. While identifying a comprehensive definition of HIT or the tools that comprise it is challenging (in part because HIT continues to develop and change), five technologies consistently comprise the core of HIT.

  1. 1. The electronic health record (EHR) is the digital record of the patient’s medical history and can reasonably be equated to the paper chart. By maintaining digital records, providers should have easier access to information, the ability to track patients over time by discrete data points, and to use that data to drive other functions such as monitoring their patient population for necessary preventive screenings (Hillestad et al., 2005). Often, this technology is referred to as an electronic medical record (EMR), rather than EHR. In concept, these two terms are substantially different: an EMR is defined as the record kept by an individual practice, while the EHR includes information from beyond a single practice, incorporating the patient record from diverse sources including non-traditional healthcare providers (Garrett & Seidman, 2011). In practice, these distinctions can be difficult to make and the terms are not employed with sufficient regularity to maintain their intended meaning.

  2. 2. If the EHR is the electronic version of a paper chart, computerized provider order entry (CPOE) is the electronic version of the prescribing pad. By facilitating the legible recording of provider orders, and linking the order to a list of all potential orders, CPOE has the potential to reduce errors in orders for medications, laboratory tests, radiological examines, and other procedures (Institute of Medicine, 2001). Perhaps more importantly, CPOE provides a platform for two other core HIT tools. CPOE is often linked to e-Prescribing, which is “a prescriber’s ability to electronically send an accurate, error-free and understandable prescription directly to a pharmacy from the point-of-care” (Centers for Medicare and Medicaid Services, 2014), which can reduce errors in reading and interpreting prescriptions.

  3. 3. CPOE can also support clinical decision support (CDS) systems, which “are information systems designed to improve clinical decision making” (Garg et al., 2005). CDS is often built around the EHR and CPOE, leveraging patient information in the EHR to inform decisions at the point of ordering. CDS can be used for a broad range of purposes from the relatively simple to highly complex. Examples of key uses of CDS include recommending drug dosing by patient weight, checking for interactions between drugs and allergies, suggesting screening based on patient characteristics and clinical guidelines, or use of complex algorithms designed to recommend treatment courses (Kaushal, Shojania, & Bates, 2003).

  4. 4. While a large portion of HIT is focused on provider uses, technology can also facilitate changes in how patients engage with their healthcare through patient portals and personal health records. Portals provide patients a selected view of their providers’ EHR, and allow for secure messaging with providers, appointment scheduling, prescription refills, and other health-related activities (Ancker et al., 2011). Similarly, a personal health record is a patient-facing record of their healthcare drawn from multiple sources and used by patients to monitor their health and to share information with new providers (Kahn, Aulakh, & Bosworth, 2009).

  5. 5. Finally, health information exchange (HIE) is “the process of reliable and interoperable electronic health-related information sharing” (The National Alliance for Health Information Technology, 2008) between healthcare provider organizations. HIE can enable medical decision-making informed by all relevant patient information and reduce redundant tests and procedures through the automated sharing of results (Walker et al., 2005).

Growth of Support for HIT

With these definitions in mind, it is perhaps not surprising that there has been a high level of enthusiasm for the potential of HIT to address many of the lingering challenges in healthcare delivery. The value of HIT has seemed clear since the 1970s when the first demonstration projects showing the ability of HIT to improve care quality were conducted (McDonald, 1976; Yu et al., 1979). Despite this evidence, adoption of electronic systems was slow through much of the world. In the United States, only a few academic centers had developed EHRs by the start of the 1990s (Shortliffe, 1999), while in the United Kingdom, Denmark, New Zealand, and elsewhere, general practitioners began adopting EHRs but hospitals did not (Wachter, 2016). Nevertheless, evidence and excitement about using HIT to address the “non-perfectibility of man” continued to grow throughout the 1990s (McDonald, 1976; McDonald et al., 1999).

Enthusiasm for HIT likely peaked between 2000 and 2005, before HIT was truly widespread in any country (Jha, Doolan, Grandt, Scott, & Bates, 2008). One high-visibility document expressing support for HIT in this period was the United States’ Institute of Medicine’s (IoM) report on Crossing the Quality Chasm. In a previous report, the IoM estimated that medical errors resulted in nearly 100,000 deaths in the United States each year (Institute of Medicine, 2000). Crossing the Quality Chasm stood as an important call-to-arms to address this challenge and clearly identified HIT broadly as a tool for systemic reform, and in particular focused on the use of CPOE as an essential tool to address the high rate of errors (Institute of Medicine, 2001). In this mood, several research groups attempted to systematically estimate the potential benefits of HIT: one landmark study projected that nationwide adoption in the United States would save $41.8 billion annually, avoid 200,000 adverse drug events, and avoid tens of thousands of deaths annually through improved screenings (Hillestad et al., 2005). Another report, focused solely on the value of HIE, estimated even greater savings of $77.8 billion annually from nationwide adoption of HIE alone (Walker et al., 2005).

At the same time, the evidence supporting the potential for HIT to address structural problems appeared strong in some regards but exhibited fundamental weaknesses, with uneven attention to the range of HIT tools. A 2006 systematic review summarized 76 studies with empirical data on multifunctional systems that tested the effect of these systems on several outcomes. Of these 76 studies, 54 (71%) were conducted by four organizations, each of which had developed their own EHR system. This dominance leads to serious questions about the external validity of their findings; however, many of these studies were of high internal validity and leveraged randomized control trials. Studies included in this review demonstrated realization of several of the goals associated with HIT, most notably a reduction in medication errors and an increase in use of preventative measures. Notably, though this review attempted to include a broad range of technologies, it included no studies that focused on the effect of patient portals or health information exchange.

HIT-Focused Public Policies and Their Effects

Motivated by a sense of HIT’s enormous potential, policy-makers in the United States, Canada, the United Kingdom, Europe, and elsewhere moved to enact public policy to support HIT adoption. These policy efforts were diverse and the role of government varied. However, as described below, four key themes are consistent across countries. Firstly, the shape of each country’s HIT policy appears to be closely formed by the broader healthcare system, leading to approaches that match the environment. Secondly, those that began programs earlier (i.e., in the 1990s or earlier) experienced more success than programs started more recently, perhaps because these systems were given time to develop. Thirdly, despite variation in the programs designed to support adoption of HIT across countries, delays and cost overruns appear to be the rule rather than the exception, especially in programs aimed at hospital EHR use and HIE. Fourthly, regardless of success implementing EHRs, adoption of HIE and other advanced tools, such as using EHRs to monitor population health, remains an ongoing challenge. Six examples demonstrate this trend.1


The approach to EHR adoption in Canada mimicked the country’s general healthcare system: it relied largely on public funding intended to support private organizations. The federal government began support with an initial investment of $500 million to establish an independent corporation in 2000, followed by an additional $600 million investment in this corporation in 2003. Federal investment was complemented by efforts of individual provinces, leading to notable differences in progress across the country. In 2009, only 17% of Canadians lived in provinces with electronic records (Auditor General of Canada, 2010), and even fewer patients visited physicians using EHRs. It therefore appeared clear that the program was not going to meet its goal that 50% of patients have an electronic record in 2010 and 100% in 2016. One reason for this shortfall was low support from the clinical community, with low computer literacy among clinicians and low perceived benefits from the EHR (Chang & Gupta, 2015). By 2017, the federal government had invested a total of $2.1 billion to stimulate EHR adoption (Canada Health Infoway, 2017). While many providers used some electronic system, the sophistication of these systems appears to have been low. For instance, Canada lagged far behind the United States in establishing advanced EHRs: according to a prominent analytics firm that ranks hospitals along seven stages of health IT adoption, 38.8% of hospitals in the United States had reached stage 6 or 7, while only 1.7% of Canadian hospitals had (HIMSS Analytics EMRAM, 2017). Perhaps in recognition of this lag, investment in HIT was redoubled: the province of Alberta signed an agreement with Epic Systems to support EHR adoption with an expected total cost of $1.6 billion in 2017 (Gerein, 2017).


The French government made a substantial investment to support nationwide EHR connectivity in 2004, with a specific focus on facilitating data sharing and less directly on support for EHR use (Metzger, Durand, Lallich, Salamon, & Castets, 2012). Provider adoption of EHRs rose to 75%, although it is not clear that adoption was due to government support, and the functionality of those systems was reported to be quite low (Mossialos, Djordjevic, Osborn, & Sarnak, 2017). The government’s initiative to facilitate data sharing floundered amid privacy concerns. Initially policy-makers aimed to develop regional solutions for providing health records, but development of the program was stopped in 2007 over confidentiality concerns, leading to the launch of a national approach (Séroussi & Bouaud, 2017). The new program was designed as a centralized, highly secure system in 2009 in which patients opted in to participation. Once completed, the system will allow for centralized storage of patient information to any compatible EHR, potentially simplifying issues around interoperability; however, this requires providers to “push” information from their local EHR to the central system. Uptake and use of the system remained very low, slowed by low political support, resistance from healthcare professionals, and uncertainty from patients. For instance, program leaders set an ambitious timeline for increases that was dramatically missed: the government hoped to have 9 million patients enrolled in 2012; instead, 150,000 were enrolled (Séroussi & Bouaud, 2017).


Denmark has achieved near-universal adoption of EHRs by primary care physicians, driven by incentives beginning in the 1990s (Gray, Bowden, Johansen, & Koch, 2011). These incentives included faster payments to physicians and payments for email consultations, and were coupled with public sentiment supporting the importance of EHRs. Denmark successfully mandated use of EHRs by all physicians in 2004. Despite success encouraging adoption among primary care physicians, most data entered electronically was not structured and therefore hard to aggregate. Mirroring Denmark’s separation between privately employed primary care physicians and publicly owned hospitals, adoption of EHRs in hospitals occurred separately from physicians and was much slower among hospitals than physicians, with approximately 50% of hospital beds covered by EHRs in 2010 (Protti & Johansen, 2010). Nevertheless, Denmark remains among the most successful countries at encouraging adoption of HIT. It now faces a challenge of converting older systems toward more advanced use cases, such as using structured data to facilitate population health monitoring. Similarly, HIE in Denmark is facilitated by a National Health Portal and hospitals that routinely provide discharge summaries to physicians, but other types of data exchange remain limited. Finally, a recent attempt to use databases to monitor clinical quality was abandoned in 2015.

New Zealand

New Zealand is among the most successful countries in facilitating adoption of EHRs. Like Denmark, government support of EHRs began in the 1990s and was relatively soft. The government instituted national patient identifiers, adopted standards for electronic messaging, and required that claims were filed electronically. The government also made targeted investments in EHR development and grants to providers to purchase systems. Moreover, as in Denmark, adoption of EHRs by physicians outpaced adoption by hospitals. New Zealand providers have made repeated investments in improving their EHR system over time such that capabilities, not just level of adoption, remain advanced (Gray et al., 2011). However, inter-organizational HIE remains challenging and adoption of patient portals remains low (Mossialos et al., 2017).

United Kingdom

The first initiatives to encourage adoption of HIT began in 1975 with the implementation of subsidies to physicians in clinics. Because of this program’s long time horizon and significant clinical commitment to development of the underlying software, adoption of EHRs is nearly universal in the UK ambulatory practice. This experience starkly contrasts with a more recent initiative in hospitals. Once again mirroring the broader healthcare environment, public policy enacted in the United Kingdom in 2002 directly supported adoption of EHRs by hospitals through the existing trust structure in the National Health System (NHS) in the U.K.’s National Programme for Information Technology (NPfIT) (Comptroller and Auditor General, 2008). The ambitious program was plagued by major delays and cost overruns: while planners aimed to have all records electronic by 2010, a more recent estimate aims for completion by 2023 (Mossialos et al., 2017). The goal of the NPfIT was to create a central, fully integrated electronic health record including both ambulatory care physicians and hospitals (Wachter, 2016). Because many physicians had already adopted HIT, the program’s primary work was to establish functioning EHRs at 300 hospitals and connect these systems. To do so, the NPfIT contracted with several large commercial suppliers and committed £6.4 billion. Despite these outlays, the program eventually failed to implement EHRs in all NHS trusts. This failure has been attributed to several factors, including a focus on technology over necessary adaptive change and low engagement with clinicians. Clinicians viewed the program as a top-down, politically motivated rush for digitization, and many of the deadlines seemed unrealistic from the outset. Further, leadership of the program continually changed, decreasing buy-in from stakeholders.

United States

Public policies to support HIT adoption in the United States began in earnest in 2009 in the form of billions of dollars in funding supporting adoption by physicians and hospitals. The incentive program was tied into the two largest public insurers, Medicare and Medicaid, and provides financial rewards for physicians and hospitals that demonstrate “Meaningful Use” (MU) of the technology (Blumenthal & Tavenner, 2010). The program was designed to proceed toward progressively more difficult requirements to receive incentive payments, thereby encouraging both initial adoption and continued progress toward more valuable systems.

Given the sheer size of the program, researchers in the United States have spent considerable effort toward understanding its effect on adoption of HIT. It is clear that over the years that MU was in place, EHR use by both hospitals and physicians grew rapidly, such that by 2015, 88% of hospitals and 51% of physicians had adopted at least a basic EHR (see Figure 1, from Heisey-Grove & Patel, 2015; Henry, Pylypchuk, Searcy, & Patel, 2016). It is interesting to note that this adoption rate is the inverse of adoption trends in other countries, where physician practices lead hospitals in adoption.

Health Information Technology

Figure 1. Percent of non-federal acute care hospitals with adoption of EHR systems by level of functionality: 2008–2015.

Note: *Significantly different from previous year (p < 0.05).

Source: Henry, Pylypchuk, Searcy, & Patel (2016).

Despite this fast growth, the effect of MU on adoption remains contested because the counter-factual—what might have happened over those years without MU—was not observed. It is therefore challenging to prove that MU had a large stimulating effect. This challenge is compounded because MU is inclusive such that there is not a comparator group within the United States that could serve as a good control group. A review of the relevant literature indicates that the MU program had some effect on adoption of HIT, but the magnitude of that effect remains contested. Two studies provide compelling, though indirect, evidence of the effect of HITECH on adoption of HIT. A survey of physicians indicated that the program was instrumental in their decision to adopt HIT (Cohen, 2016). In addition, a study focused on the number of HIT-related job postings noted an 86% increase in postings following HITECH, indicating a stimulus effect on demand for the workforce necessary to enable HIT adoption (Schwartz, Magoulas, & Buntin, 2013).

Several studies have directly measured change in the number of providers that adopted HIT in the years following MU. However, in comparison with the less direct evidence cited above, methodological constraints make it challenging to draw clear evidence from this work. Two studies compared adoption between hospitals eligible for MU subsidies and hospitals that were not eligible, and found that eligible hospitals adopted HIT at a much faster rate (Adler-Milstein & Jha, 2017; Walker, Mora, Demosthenidy, Menachemi, & Diana, 2016). However, the strong conclusions of these studies should be tempered by the fact that eligible and ineligible hospitals are very different populations: eligible hospitals are primarily short-term acute care hospitals whereas non-eligible hospitals are primarily long-term care and psychiatric hospitals. Another study avoided this complication by looking within the population of eligible hospitals and using differences in the size of subsidies received by hospitals to estimate the effect of MU on adoption of EHRs (Dranove, Garthwaite, Li, & Ody, 2015). In doing so, the authors concluded that the subsidies had a positive, but ultimately small, effect on HIT adoption. However, because the size of subsidies is dependent on the number of Medicare discharges from the hospital, the resulting estimated effect may be an underestimate if the effect of larger monetary incentives on hospitals’ decisions was diluted by larger costs of adopting these systems. Although this study controls for some of the potential bias in these estimates by including hospital bed size, it likely misses others, such as the costs of implementing EHRs for additional specialties, units, and types of services. Finally, one study focused on physician adoption of HIT estimated that the rate of adoption was not statistically different from the expected rates generated by applying diffusion of technology curves derived from observations in other settings to physician adoption of EHRs (Mennemeyer et al., 2016). Application of this theoretical curve seems fraught: it is not clear that adoption of HIT should follow curves observed elsewhere given the distorted incentives in the U.S. healthcare system. In addition, the authors used a small number of data points (mean adoption by year) to evaluate the fit of the curve to available data, thereby providing limited power to assess differences. In sum, estimating the effect of the MU program on hospital and physician adoption of HIT has posed several challenges to investigators, yet overall it appears likely that the MU program did not fully crowd out private investment and had important stimulating effects.

HIT Adoption and Key Barriers

The challenges encountered by several countries’ hospital-based initiatives should not be surprising. Barriers to HIT adoption from the healthcare delivery system have been well documented and persistent resistance has slowed adoption in several nations. Low adoption rates and challenges encountered by incentive programs can be attributed to four key persistent barriers.

The first barrier to widespread HIT adoption has been financial: implementation and support of these systems is expensive and that cost is often not borne by the parties that most benefit from the system (Ash & Bates, 2005). Key costs come from hardware, software, training, facility space, and short-term reduced productivity (Slight, Quinn, Avery, Bates, & Sheikh, 2014). Costs to implement EHRs in physician offices vary but are often estimated as tens of thousands of dollars per physician, while large health systems’ implementations have run tens of millions of dollars (Jayanthi & Ellison, 2016; Fleming, Culler, McCorkle, Becker, & Ballard, 2011). It is, therefore, not surprising that despite incentives under MU, 73% of physicians that had not yet adopted HIT reported that cost was a major barrier (Jamoom, Patel, Furukawa, & King, 2014).

Secondly, HIT is disruptive, and successful integration into clinical care requires an organizational culture that can adapt to optimize its use (Ash & Bates, 2005). Organizational readiness for change—a combination of commitment and resource availability (Weiner, 2009)—is an important factor in implementation success. Yet with public incentives to adopt HIT, organizations with low capabilities to optimize its value may nevertheless move toward adoption. Relationships between management and clinical staff is a key cultural determinant of success but represents a Goldilocks dilemma: while support from upper management is a key driver of success (Lluch, 2011), issues of culture are particularly threatening when providers believe that administrators are forcing the system on them (Ford, Menachemi, Peterson, & Huerta, 2009). Furthermore, clinicians can become wary of adoption of these systems because of their potential to reshape power and autonomy within the care setting, potentially undermining their commitment to system implementation, design, and optimization (Ash, Sittig, Campbell, Guappone, & Dykstra, 2006).

Thirdly, HIT use is time-consuming and the efficiency of patient care declines immediately following implementation. It appears clear that there is a short-term loss of productivity directly following implementation of complex HIT. Evidence on the real, long-run effect of HIT on productivity remains mixed: some evidence indicates that HIT enhances productivity (Adler-Milstein & Huckman, 2013), while other evidence points toward decreased productivity, at least for some physicians (Bae & Encinosa, 2016; Howley, Chou, Hansen, & Dalrymple, 2014). While the long-term effect of EHRs on productivity remains unclear, among those who delayed adoption, there appears to be widespread perception that EHRs reduce productivity: 59% of physicians that had not yet adopted an EHR reported productivity loss as a major barrier to adoption (Jamoom et al., 2014). Despite this fear, as experience with an EHR increased, physician views became more positive (Jamoom, Heisey-Grove, Yang, & Scanlon, 2016). Whatever the overall effect on productivity, it appears clear that physicians are spending increasing hours online, such that more than half their time is spent in front of a computer (Tai-Seale et al., 2017).

Fourthly, concerns around the privacy and security of these systems have discouraged providers from taking on the responsibility of maintaining systems (Jamoom et al., 2014). Once paper records are made electronic, unauthorized remote access becomes possible. Small physician practices might justifiably doubt their ability to ensure the security of this information, leading them to avoid the risk. Security of large systems is also challenging, and the sheer volume of patient information makes them a more likely target. Several high-profile hacks into larger systems, including the NHS, have given credence to these concerns (Hern, 2017).

The importance of these barriers may seem to have waned in the wake of public programs that, whatever their flaws or inefficiencies, have ultimately seen increased adoption of HIT. However, many of the barriers that impeded adoption overall likely continue to slow adoption of specific tools within the broader electronic system that are not specifically mandated by government programs. These same barriers may also influence how and if HIT is used after adoption.

Evidence on HIT and Key Open Questions

Attempts to evaluate the effect of HIT have increased along with general support and public funding; here is a summary of recent systematic reviews of the evidence on HIT and an update on these findings with the most recent evidence from a targeted review.2

Despite the sheer volume of evidence on the efficacy of HIT, many questions remain unanswered. For instance, a 2014 systematic review of HIT on outcomes included 236 studies (Jones et al., 2014). This volume of evidence was sufficient for the review’s authors to conclude that “CDS generally results in improvements in the processes targeted by the decision support” and “CPOE reduces medication errors” (Jones et al., 2014). Yet these conclusions point toward the enormous gaps in the evidence: the extent to which these two functions affected outcomes was not known and varied substantially from study to study; the effect of other types of HIT, including the EHR generally, HIE, and patient portals, was not well evidenced; and the effect of HIT on other outcomes—especially productivity—was mixed and inconclusive, with a large range of effects. This trend—positive but hard to pin down overall findings, with limited evidence on several important types of HIT—was largely unchanged from earlier reviews that noted largely positive effects (Buntin, Burke, Hoaglin, & Blumenthal, 2011; Chaudhry et al., 2006).

A contemporary review with a narrower focus on patient outcomes found even less instructive results (Brenner et al., 2015). Thirty-six percent of the 69 included studies demonstrated positive findings, with the balance containing mixed, null, or negative results. Only 28% of the 18 randomized controlled trials (RCTs) showed positive results. Further, only 28% of all studies and 44% of RCTs were set at multiple centers, so that many of these studies likely suffer from a lack of generalizability. In addition, this review noted a limitation in the setting of care, with 86% of studies focused on inpatient care, 14% on outpatient, and only 1% on long-term care. An updated review focused on the years 2012–2017 found that the majority of studies included positive results; however, this review did not distinguish between studies with both positive and null results (i.e., mixed) and those with only positive results, making direct comparison with the most recent prior review challenging (Kruse & Beane, 2018); however, this categorization did mirror the approach taken by an earlier review that found that 92% of studies were positive overall (this study further categorized studies as either solely positive [62%] or mixed positive and null [30%]) (Buntin et al., 2011). Beyond these general reviews, some reviews focusing on specific use cases have found more consistently positive outcomes than more general reviews: one focused on CPOE in intensive care units found significant reductions in prescribing errors and a reduction in length of stay (Prgomet, Li, Niazkhani, Georgiou, & Westbrook, 2016), and another focused on cancer care found significant benefits from provider-used HIT, though not patient-used HIT (Tarver & Menachemi, 2015).

It is worth noting that the majority of reviewed studies focused on either simple pre-post designs, which attribute effects to HIT when other contemporary changes may be the underlying cause, or small-scale RCTs, which may have limited generalizability given the extent to which nuances seem to matter. A smaller body of literature has studied HIT adoption using methods that are more common in econometric evaluation using large-scale data. Unfortunately, most large-scale observational studies did not use quasi-experimental methods and instead used cross-sectional approaches that afford very limited causal inference.

Eleven studies in this review stand out as providing evidence of an effect of HIT that may represent a more average overall effect by including a large number of providers and by focusing on change over time, allowing for at least modest causal attribution. Ten of the 11 studies focused on hospital adoption of HIT, and included between two and four thousand hospitals. Of these 10 studies, three found only positive effects of HIT adoption (Lammers & McLaughlin, 2017; Lee & Dowd, 2013; Miller & Tucker, 2011), three found a combination of positive and null effects (Adler-Milstein, Everson, & Lee, 2015; McCullough, Casey, Moscovice, & Prasad, 2010; Parente & McCullough, 2009), three found both positive and negative effects (Appari, Eric Johnson, & Anthony, 2013; Himmelstein, Wright, & Woolhandler, 2010; Jones, Adams, Schneider, Ringel, & McGlynn, 2010), and one identified only negative or null effects (Agha, 2014). One study focused on adoption of HIT by approximately 600,000 physicians and changes in readmission rates and identified a beneficial effect (Lammers, McLaughlin, & Barna, 2016). Notably, the majority of negative effects were related to estimates of cost; however, one of the most careful and recent studies on cost found beneficial effects (Lee & Dowd, 2013).

Table 1. Review of the Effect of Health Information Technology


Study Design

Focal Providers

Outcome Result

Outcome Summary

Sample Size

Lammers & McLaughlin, 2017

Fixed Effects Model, N = 306 HRRs

Hospitals and Physicians

Markets with larger increases in physician and hospital EHR penetration experienced greater decreases in total Medicare, Medicare Part B, and acute care expenditures per beneficiary.


Lee & Dowd, 2013

Fixed effects, N = 294 hospitals


Predicts a decrease in cost of up to $1,550 of IT labor per bed, $27,909 of IT capital per bed, and $28,695 of all IT expenditure per bed.


Miller & Tucker, 2011

Panel Analysis, N = 3,764 hospitals


A 10% increase in births at hospitals with EMRs reduces neonatal mortality by 16 deaths per 100,000 live births. The cost of each baby’s life saved is estimated to be $531,000.


Adler‐Milstein et al., 2015

Fixed Effects Model, N = 9,328–11,363


Associated with better performance on process adherence and patient satisfaction, but not efficiency.


McCullough et al., 2010

Fixed Effects Model, N = 3,401 hospitals


CPOE and EHRs yielded improvements in 2 of 6 quality measures.


Parente & McCullough, 2009



EMRs have a small, positive effect on patient safety.


Jones et al., 2010

Difference-in-Differences, N = 2,021 hospitals


Basic EHR systems improved quality in heart failure, but advanced systems decreased quality in two measures.


Appari et al., 2013

Panel Analysis, N = 3,291 hospitals


Hospitals transitioning to more basic EHR systems improved process quality, but hospitals adopting advanced systems experienced quality declines.


Himmelstein et al., 2010

Multiple linear regression, N = ~4,000 hospitals


Hospital computing slightly improved process measures of quality but failed to reduce costs.


Agha, 2014

N = ~3,900 hospitals


HIT is associated with an increase in billed charges and there is no evidence of cost savings.


Lammers et al., 2016

Fixed Effects Model, N = 306 HRRs


EHR adoption was significantly associated with declines in ACSC admissions.


While these studies capture a large volume of providers, their empirical approach and ability to show causal effects is relatively weak: contemporary change among hospitals adopting HIT may cause the observed changes. In consequence, broad, causal estimates of the effect of HIT on outcomes remain limited because more advanced econometric techniques intended to identify exogenous assignment of HIT have not been identified or used. Given the public policy context in the United States, where most hospitals and physicians were eligible for incentive payments, it may be most instructive to look toward experience in other countries to identify causal effects. For instance, depending on the continued evolution of HIT implementation in the United Kingdom (Wachter, 2016), the NHS’s trust system may lead to essentially exogenous assignment of HIT implementation by geographic area providing an opportunity for large-scale, plausibly causal inference that has not yet been conducted.

Two of the reviews highlighted in Table 1 emphasized the importance of understanding the context of implementation in identifying the reason for the observed effects. Nevertheless, few reviewed studies described the context in detail, and only a few studies focused on context in succeeding years. Four studies used national-level data to identify hospitals with the greatest benefits from HIT: they found that HIT benefit was larger for more complex patients (McCullough, Parente, & Town, 2016), at academic hospitals (McCullough et al., 2010), was larger at hospitals that either did not employ physicians or employed few physicians (Everson, Lee, & Adler-Milstein, 2016), and had increased in more recent years (Adler‐Milstein et al., 2015).

This last finding is of particular interest. Studies of information technology in other industries have identified what is known as the “productivity paradox,” in which IT was not initially associated with gains in productivity (David, 1990; Triplett, 1999). It was only years after initial adoption that productivity attributable to IT appeared to develop and continue growth (Brynjolfsson & Hitt, 1998). These findings suggested either a firm-level or industry-level learning effect in which complementarities associated with achieving value from HIT take time to emerge. The gap between implementation and optimization of a HIT system appears vast and it may be that more benefits from HIT are yet to emerge.

Given the evolving state of HIT and its uses, identifying a reasonable average impact may not be possible, and even if one is found, it may not be meaningful for long. Instead, the existing literature points toward three essential areas of continued research. Firstly, are the benefits of HIT increasing over time as organizations and the field continue to learn? Secondly, what organizational and implementation factors influence the direction and magnitude of HIT impact? Thirdly, what are the effects of diverse types of HIT and in diverse care settings?

Challenges to HIE

In the United States, overall adoption of HIE has increased along with HIT such that most hospitals and ambulatory providers at least have the capability to exchange information (American Hospital Association, 2016). However, both the ability to exchange information with all relevant providers to cover the population of treated patients and the actual use of HIE in clinical practice appear to remain low (Devine et al., 2017). In one sense, this lag should not be surprising: the electronic exchange of information requires that it first be electronic, so that HIE must obviously follow HIT. However, key challenges to HIE have slowed its use beyond its logical lag behind other HIT technologies. This extended delay, and the potential benefits of HIE, led the U.S. government to focus on interoperability—that is, HIE integrated into the EHR—as one of two key priorities in 2017 (Rucker, 2017).

One set of challenges impeding HIE is primarily technical. Providers must develop exchanges with many outside providers to achieve broad-based HIE, and given the many EHRs and methods of exchange available, establishing broad HIE is challenging (Everson, 2017; Vest, Campion, Kaushal, & Investigators, 2013; Vest & Kash, 2016). Agreeing upon a technical standard and ensuring it is widely and efficiently used has also been challenging; often multiple versions of standards are developed before any one is widely implemented, and different vendors’ tweaks to standards appear to limit their universality (Imler, Vreeman, & Kannry, 2016). This moderate level of standardization can create challenges to using information even if it is exchanged because it is difficult to find unstandardized information or to integrate data into the EHR.

Perhaps more important than the technical hurdles to enabling HIE, social barriers seem aligned against HIE, which requires inter-organizational collaboration to implement a shared exchange system. In countries with private, competitive healthcare providers, establishing this collaboration has been challenging because providers can view patient data as a competitive advantage (Edwards, Hollin, Barry, & Kachnowski, 2009; Grossman, Kushner, & November, 2008). Distrust between providers has been a major hurdle for the most prominent attempts to establish HIE in the United States, regional health information organizations (RHIOs). RHIOs were designed to connect all providers in a geographic region and have had mixed experiences due to challenges bringing together stakeholders that do not always perceive exchange as a strategic imperative (Adler-Milstein, Lin, & Jha, 2016; Edwards et al., 2009). Provider hesitation is exacerbated because the value of these efforts, especially in their developmental stages, is uncertain (Adler-Milstein, Bates, & Jha, 2013; Kern, Barron, Abramson, Patel, & Kaushal, 2009; Ross, Schilling, Fernald, Davidson, & West, 2010; Vest et al., 2013). In response to weak participation in RHIOs, other approaches that do not force competitors to collaborate appear to be gaining traction (Everson, 2017; Vest et al., 2013). However, these efforts are unlikely to connect all relevant stakeholders, and their success depends on the EHR vendors’ commitment to design systems capable of sharing information. In some cases, vendors appear to have sought to block connectivity with their competitors (Adler-Milstein & Pfeifer, 2017; Everson & Adler-Milstein, 2016; The Office of the National Coordinator for Health Information Technology, 2015). This behavior prompted legislation aimed to penalize organizations found to engage in “information blocking” (21st Centuries Cures Act, 2015).

Even when the basic technical and collaborative challenges are overcome, valuable use of information exchanged electronically depends on careful design of the system and redesign of clinical workflows. Studies on the use of HIE have generally shown low levels of overall use (Rudin, Motala, Goldzweig, & Shekelle, 2014). One reason for low use is that providers are often forced to use a separate system, login, and search function to access information from the HIE; however, integration into the main EHR appears to be improving (Kaelber, Waheed, Einstadter, Love, & Cebul, 2013; Vest, Gamm, Ohsfeldt, Zhao, & Jasperson, 2012). Even when integration into the chart is accomplished, HIE information is often stored separately from information gathered from within the organization and not easily reconciled. HIE can also lead to an overwhelming volume of information; finding time to review outside records within patient care is often challenging. As an example, providers often complain of receiving 600-page summary of care records, which seem likely to offer little real value in a fast-paced clinical setting.

Adoption of HIE has lagged behind other forms of HIT due to an array of barriers. However, in recent years the gap between HIE and HIT in the United States appears to be closing. As overall adoption of HIE continues to increase, questions around whether all needed providers are connected and usefully sharing information will continue to be pressing.

Contemporary Evidence on HIE and Key Open Questions

The evidence for benefit from HIE is even less compelling than the evidence behind other forms of HIT. Three systematic reviews published in 2015 concluded that there was not substantial evidence of benefit across 34 total published studies (Hersh et al., 2015; Rahurkar, Vest, & Menachemi, 2015; Rudin et al., 2014). Despite the seemingly obvious benefits of HIE, it is likely that in some cases the challenges to HIE—low integration, poorly standardized data, and difficult clinical workflows—are sufficient to nullify the potential value of shared patient information.

In other cases, challenges to measuring HIE impact may be impeding identification of effects. Relatively few of the reviewed studies included a randomized control trial such that biases including unobserved patient characteristics could influence results. In non-RCT studies, it is challenging to identify logical comparison groups that would have benefited from HIE but did not receive it. Often HIE may not be used simply because there was no reason to use it, so that naïve assessments of the impact of HIE without clearly defined similar patient populations is problematic. In studies that did use a randomized approach, low HIE use translated into relatively small sample sizes for evaluating effects, especially since the outcomes measured in these studies are relatively rare. In addition, most HIE studies focus on one or a few healthcare organizations; in consequence, the literature has both limited internal and external validity.

While the overall literature on HIE has not produced consistent evidence of benefit (or of null effect), three studies on HIE’s impact on productivity have pointed toward significant benefit, and appear to be the best sign that HIE may be positively impacting care. Using a longitudinal, fixed effects design, one study found substantial reduction in the rate of imaging procedures performed in emergency departments (EDs) among 37 hospitals adopting HIE compared with 410 that did not (Lammers, Adler-Milstein, & Kocher, 2014). Two corroborating studies by Bailey et al. found that repeat patients presenting with headache to one of 15 EDs connected by HIE were less likely to receive neuroimaging if HIE was used; patients with back pain were similarly less likely to receive imaging when HIE was used (Bailey, Pope, et al., 2013; Bailey, Wan, et al., 2013).

While there is some promising evidence for specific use cases, overall the evidence supporting HIE adoption and use remains in its infancy. Whereas studies on CPOE and CDS have successfully demonstrated that there is at least some benefit, even this low bar has not been cleared by existing research on HIE. However, HIE continues to evolve, so its impact is likely to change as many features of the systems improve. Beyond these changes, organizational and implementation context and time to optimize HIE are likely to be important factors in the effect of HIE, just as they are in HIT. Better study design, more varied data, and simply more evidence will be useful to identify and increase the effect of HIE on outcomes.

The Professions: How Is HIT Used?

A key theme resonates throughout the discussion on the adoption and value of HIT: the success of technological initiatives depends on interactions between the technology and key healthcare professionals. Tension between the type of standardized work and evaluation facilitated by HIT and of professional expertise and autonomy is one of the core tensions across professional disciplines. Examples range from the use of standardized tests to evaluate students and teachers to the deployment of police officers according to statistical measurements of crime rather than community relationships. Like these efforts, HIT can be perceived as a threat to the profession or as a complement to tacit knowledge and expertise, while inevitably disrupting the work and roles of both nurses and physicians. Importantly, professionals’ perception of the value of technology can intersect with the values and beliefs inherent in the professions themselves. In consequence, making optimal use of HIT depends on an adaptive process in which the work and values of nurses and physicians adapt to the benefits HIT can offer, while HIT is perpetually refined toward the goals of clinicians.

Relative to physicians, nurses engage in greater levels of patient contact, usually care for fewer patients in a given day, and develop a more personal connection to their patients: the act of caring is consequentially central to the profession’s self-identification (Fox, Aiken, & Messikomer, 1990; May & Fleming, 1997). This ideal is complemented by the continuing evolution of a focus on evidence-based nursing care and an expansion of nurses’ scope of practice (Melnyk, Fineout-Overholt, Gallagher-Ford, & Kaplan, 2012). HIT can challenge nurses’ caring orientation if it places demands on the nurses to attend to the system and not the patient, especially if the structured and limited information captured by the system does not appear to capture their practice (Darbyshire, 2004; Stevenson, Nilsson, Petersson, & Johansson, 2010). Further, nurses often perceive themselves as responsible for entering data into systems, but others (physicians, management) as consumers, implying that the technology is a burden. Finally, by enforcing strict prescribing rules, HIT can limit the discretion of nurses and force them to work more directly with supervising providers in areas where, prior to HIT, prescribing rules might not be as closely adhered to. The challenge, then, for nursing professionals and designers of HIT systems alike is to identify ways in which technology can complement the humane caring that is central to nurses’ work (Barnard & Sandelowski, 2001). This may be facilitated by efforts that make the information captured by technology more reflective of the person receiving treatment and more useful to coordinated care between nurses and physicians (Almerud, Alapack, Fridlund, & Ekebergh, 2007).

Like nurses, physicians’ professional identity and daily work can be challenged by information technology. Physicians, as a proto-typical profession, are defined by a special claim over arcane knowledge of medical diagnosis and treatment (Freidson, 1988; Rueschemeyer, 1983). This claim in turn provides the basis for professional autonomy—that is, the authority of the profession to define correct medical care and of individual professionals as experts equipped to provide high-quality treatment (Freidson, 1974). HIT could promote challenges to this claim in two ways: colloquial stories of physicians bemoaning patients’ attempts to self-diagnose through the use of the internet are common and may represent a democratizing of medical knowledge, though with unclear ultimate effects. Perhaps more seriously, tensions exist between HIT’s promulgation of clinical guidelines through CDS and individual physicians’ claims to expertise (Walter & Lopez, 2008). These tools threaten to enforce homogeneity where greater allowances for differing professional opinion has been the norm. In part due to this tension, as well as poor design and implementation, CDS is often criticized as imprecise and unhelpful (Ash, Sittig, Campbell, Guappone, & Dykstra, 2007; Kuperman et al., 2007). Finally, as new sources of data and advanced informatics alter how new medical knowledge is developed, HIT could further threaten physicians’ professional status.

Beyond the relationship between HIT and knowledge domains, HIT also impacts physicians’ daily work. As previously mentioned, physicians now spend a large part of their day interacting with the computer, and the efficiency of computing remains an open question (Tai-Seale et al., 2017). Depending on design of the system and the way work is reallocated by adoption of HIT, this time may be or feel like misuse of physicians’ training, especially when HIT use is dictated by management or policy interventions. In consequence, physicians appear divided in support of HIT, simultaneously (and not homogeneously) supporting the potential for HIT to decrease errors and improve quality while fearing that these systems place unnecessary burdens on them (McAlearney, Chisolm, Schweikhart, Medow, & Kelleher, 2007; Poon et al., 2004).

Nurses and physicians are not powerless recipients of new technology, but instead work to shape the way technology impacts them and to set boundaries on its use. This ongoing dialectic is essential to develop well-functioning HIT; however, finding the right balance is challenging. Shifts in public policy in the United States serve as a useful example of this dynamic. Between 2016 and 2017, the third stage of the Meaningful Use program was converted into part of the Merit Incentive Payment System through the advancing care information (ACI) program. In some ways, the MU and ACI programs are similar: both reward physicians for adopting and using HIT and attesting to specific levels of use on several measures. Nevertheless, the shift from MU to ACI contains two key changes that highlight the role of physicians in successfully reshaping a regulatory program they felt was too burdensome into alignment with their goals. Firstly, some measures that specifically placed a burden on physicians, including requiring CPOE for non-medication orders and the implementation of CDS tools, were removed in the transition from MU Stage 3. Secondly, while MU required physicians to meet relatively high thresholds on a number of required criteria to receive incentive payments or avoid penalties, eligible physicians can now meet the requirements for ACI by achieving very low thresholds on some measures and then select which available measures they would like to target for high performance to achieve bonuses. Both of these changes reduce the requirements placed on physicians and return discretion to practitioners, for better or worse.

It is challenging to assess whether the shift from MU to ACI represents a loosening of overly prescriptive public policy or a step backward in support for robust HIT to improve quality. However, it is clear that this dialogue between the major institutions in healthcare—providers, management, and policy-makers—must continue if HIT is to fulfill promises of high value (Scott, 2000).

Toward the Learning Health System

Despite the inconsistent benefit of HIT documented to date, the broad conceptual imperative for its continued use and development appears undimmed. A common theme in projections to the future is that the development of the widespread HIT infrastructure over the last decade can be leveraged to support the translation of enormous and growing amounts of data and evidence to changes in clinical practice. Two important groups in the United States have championed this theme, The National Academies of Medicine (NAM) and the Agency for Healthcare Research and Quality (Bindman, 2017; Olsen, Aisner, & McGinnis, 2007). And on top of existing needs to translate an enormous body of evidence into practice, there is also the perceived need for more precise information on the best treatment of an individual patient, which has motivated policy-makers, researchers, and practitioners to consider new ways to tailor treatments to specific needs through additional, targeted evidence (Ashley, 2015; Chambers, Feero, & Khoury, 2016; Collins & Varmus, 2015).

Artificial intelligence (AI) provides a set of tools that can build upon a robust HIT infrastructure to process and help translate growing evidence and to generate more precise treatment information. AI is used to process healthcare data from a large body of patients to give physicians inferences for health risk alert and health outcome prediction (Neill, 2013). AI devices are generally categorized as either machine learning (ML) or natural language processing (NLP) (Jiang et al., 2017). ML utilizes structured data (e.g., imaging, genetic, and EP data) to create sophisticated algorithms that predict the likelihood of disease. NLP aims to turn unstructured data (e.g., clinical notes and medical journals) into machine-readable structured data, which can then be handled by ML. Both of these tools offer novel means to improve healthcare. However, there are enormous challenges shaping the application of AI to the enormous healthcare data infrastructure (Jiang et al., 2017). AI requires rich sources of data but (as discussed previously) incentives are aligned against sharing data in the healthcare environment. Algorithms need to be checked and reworked in an iterative human–computer team before entering the healthcare setting where they can assist physicians in predicting disease. Finally, national and international regulation is needed to assess the safety and efficacy of AI systems before they are widely used, and this work is only beginning.

The overarching work of turning data and existing evidence into practice and identifying targeted therapies that best benefit individuals has been described by the NAM and others as the development of a Learning Health System (LHS) (Friedman et al., 2014; Friedman, Wong, & Blumenthal, 2010; Krumholz, 2014). Though definitions vary, the LHS is envisioned as a combined socio-technical system leveraging HIT, AI, and insights from behavioral sciences to change how new medical knowledge is generated and disseminated. There appear to be four key components necessary for development of an LHS:

  • The first (and most relevant to this discussion) is the creation of a standardized IT infrastructure capable of aggregating usable information. While this objective continues to progress, barriers remain to making information available in a format from which insight can be drawn (McGinnis, Powers, & Grossmann, 2011).

  • The second requirement for an LHS is the development of analytic techniques to draw lessons from immense data, such as leveraging treatment data to assess the safety of a medication as observed over millions of prescriptions, or the subtle differences in types of diseases that may only be observed among large populations (as opposed to trials) but should guide proper treatment (Friedman et al., 2017). Similarly, in order for AI to realize its full potential in healthcare, physicians and computers must work well together and complement one another (Verghese, Shah, & Harrington, 2018). Proven effective models should be utilized for their predictive utility, but only as a counterpart to the clinician who interprets and decides on action. Arming clinicians with the strong predictive capabilities of AI alongside their own human intelligence could help bridge the gap between advances in HIT and improved health outcomes.

  • Thirdly, consistent with the discussion above, medical and nursing professionals must develop a culture that can utilize insights derived from new analytic methods in addition to the randomized trials and cases that have long stood as the standard of evidence in biomedicine. This will necessitate evolving medical education away from memorization of facts and accepting some reliance on computing to provide needed information, freeing clinicians to become more familiar with techniques for leveraging seemingly endless information to inform their treatment decisions (Krumholz, 2014).

  • Finally, organizational leaders and policy-makers must create an environment supportive of continued learning and capable of incentivizing investment in development of the best—not most expensive—treatment. Payment based on value (cost and quality) is core to this transformation but remains limited.

Achievement of a global LHS is not a near-term objective, but rather a long-term vision. Nevertheless, there are numerous incremental opportunities to advance toward that eventual goal. These smaller steps can be made within organizations as they explore new ways to aggregate information and identify areas of quality improvement. Several integrated systems and academic medical centers are leading the way in this innovation (Paulus, Davis, & Steele, 2008). The hope is that as these initiatives demonstrate success and the technological infrastructure allowing learning matures, the culture surrounding medical and nursing practice will also evolve toward a continual learning environment (Nelson et al., 2008). Simultaneously, policy-makers and groups responsible for paying for healthcare must adapt their approach to allow for, and even incentivize, continued improvement.

Over the past two decades, a variety of public initiatives have spurred increased adoption of HIT across the developed world with the goal of decreasing medical errors, increasing quality, and improving efficiency. Overall, the effect of information technology on healthcare has surely been positive and may be increasing. Nevertheless, based on the evidence through the early 21st century, the enormous potential envisioned for HIT has not been fully realized. Indeed, as thought leaders and informaticists have dreamed of future uses of the growing IT infrastructure, the gap between reality and potential has likely never been larger. It will require enormous effort from a variety of stakeholders, most importantly healthcare professionals, to truly transform healthcare using information technology; nevertheless, information technology remains essential to strategies to address the challenges healthcare systems face across the world.


The potential for HIT to revolutionize the delivery of healthcare has been understood for decades yet remains only partially realized. Since 2000, the use of HIT, spurred by varied policy initiatives, has grown markedly across several countries. In parallel, it appears that both the benefit from HIT and the evidence of that benefit is increasing, though econometric challenges make evaluation difficult. In contrast to the growth in HIT, electronic HIE between healthcare organizations remains uncommon and is slowed by conflicting incentive structures and incompatible technologies. Despite its promise for increased efficiency and quality, the empirical evidence of benefit from HIE remains limited. Taken together, further progress toward a healthcare system that incorporates information technology to systematically avoid errors, improve treatment, and learn from the enormous amount of available data will depend on change in the organization and practice of medicine and the development of new organizational capabilities.

Further Reading

Adler‐Milstein, J., Everson, J., & Lee, S. Y. D. (2015). EHR adoption and hospital performance: Time‐related effects. Health Services Research, 50(6), 1751–1771.Find this resource:

Adler-Milstein, J., Lin, S. C., & Jha, A. K. (2016). The number of health information exchange efforts is declining, leaving the viability of broad clinical data exchange uncertain. Health Affairs, 35(7), 1278–1285.Find this resource:

Adler‐Milstein, J., & Pfeifer, E. (2017). Information blocking: Is it occurring and what policy strategies can address it? The Milbank Quarterly, 95(1), 117–135.Find this resource:

Ash, J. S., Sittig, D. F., Campbell, E. M., Guappone, K. P., & Dykstra, R. H. (2007). Some unintended consequences of clinical decision support systems. Paper presented at the AMIA Annual Symposium Proceedings.Find this resource:

Bailey, J. E., Wan, J. Y., Mabry, L. M., Landy, S. H., Pope, R. A., Waters, T. M., & Frisse, M. E. (2013). Does health information exchange reduce unnecessary neuroimaging and improve quality of headache care in the emergency department? Journal of General Internal Medicine, 28(2), 176–183.Find this resource:

Buntin, M. B., Burke, M. F., Hoaglin, M. C., & Blumenthal, D. (2011). The benefits of health information technology: A review of the recent literature shows predominantly positive results. Health Affairs, 30(3), 464–471.Find this resource:

Dranove, D., Garthwaite, C., Li, B., & Ody, C. (2015). Investment subsidies and the adoption of electronic medical records in hospitals. Journal of Health Economics, 44, 309–319.Find this resource:

Everson, J., Lee, S.-Y. D., & Adler-Milstein, J. (2016). Achieving adherence to evidence-based practices: Are health IT and hospital–physician integration complementary or substitutive strategies?. Medical Care Research and Review, 73(6), 724–751.Find this resource:

Friedman, C., Rubin, J., Brown, J., Buntin, M., Corn, M., Etheredge, L., . . . Stead, W. (2014). Toward a science of learning systems: A research agenda for the high-functioning Learning Health System. Journal of the American Medical Informatics Association, 22(1), 43–50.Find this resource:

Grossman, J. M., Kushner, K. L., & November, E. A. (2008). Creating sustainable local health information exchanges: Can barriers to stakeholder participation be overcome? Center for Studying Health System Change, Washington, DC.Find this resource:

Krumholz, H. M. (2014). Big data and new knowledge in medicine: The thinking, training, and tools needed for a learning health system. Health Affairs, 33(7), 1163–1170.Find this resource:

Lammers, E. J., & McLaughlin, C. G. (2017). Meaningful use of electronic health records and medicare expenditures: Evidence from a panel data analysis of US health care markets, 2010–2013. Health Services Research, 52(4), 1364–1386.Find this resource:

Miller, A. R., & Tucker, C. E. (2011). Can health care information technology save babies? Journal of Political Economy, 119(2), 289–324.Find this resource:

Tai-Seale, M., Olson, C. W., Li, J., Chan, A. S., Morikawa, C., Durbin, M., . . . Luft, H. S. (2017). Electronic health record logs indicate that physicians split time evenly between seeing patients and desktop medicine. Health Affairs (Project Hope), 36(4), 655–662.Find this resource:

Vest, J. R., & Kash, B. A. (2016). Differing strategies to meet information‐sharing needs: Publicly supported community health information exchanges versus health systems’ enterprise health information exchanges. The Milbank Quarterly, 94(1), 77–108.Find this resource:


21st Century Cures Act, H.R. 34, 114th Cong. (2015).Find this resource:

Adler-Milstein, J., Bates, D. W., & Jha, A. K. (2013). Operational health information exchanges show substantial growth, but long-term funding remains a concern. Health Affairs, 32(8), 1486–1492.Find this resource:

Adler‐Milstein, J., Everson, J., & Lee, S. Y. D. (2015). EHR adoption and hospital performance: Time‐related effects. Health Services Research, 50(6), 1751–1771.Find this resource:

Adler-Milstein, J., & Huckman, R. S. (2013). The impact of electronic health record use on physician productivity. American Journal of Managed Care, 19(10 Spec. No.), SP345–352.Find this resource:

Adler-Milstein, J., & Jha, A. K. (2017). HITECH act drove large gains in hospital electronic health record adoption. Health Affairs, 36(8), 1416–1422.Find this resource:

Adler-Milstein, J., Lin, S. C., & Jha, A. K. (2016). The number of health information exchange efforts is declining, leaving the viability of broad clinical data exchange uncertain. Health Affairs, 35(7), 1278–1285.Find this resource:

Adler‐Milstein, J., & Pfeifer, E. (2017). Information blocking: Is it occurring and what policy strategies can address it? The Milbank Quarterly, 95(1), 117–135.Find this resource:

Agha, L. (2014). The effects of health information technology on the costs and quality of medical care. Journal of Health Economics, 34, 19–30.Find this resource:

Almerud, S., Alapack, R. J., Fridlund, B., & Ekebergh, M. (2007). Of vigilance and invisibility: Being a patient in technologically intense environments. Nursing in Critical Care, 12(3), 151–158.Find this resource:

American Hospital Association. (2016, February). Hospitals Advance information sharing, but external barriers to increased data exchange remain.Find this resource:

Ancker, J. S., Barrón, Y., Rockoff, M. L., Hauser, D., Pichardo, M., Szerencsy, A., & Calman, N. (2011). Use of an electronic patient portal among disadvantaged populations. Journal of General Internal Medicine, 26(10), 1117–1123.Find this resource:

Appari, A., Eric Johnson, M., & Anthony, D. L. (2013). Meaningful use of electronic health record systems and process quality of care: Evidence from a panel data analysis of US acute‐care hospitals. Health Services Research, 48(2 pt. 1), 354–375.Find this resource:

Ash, J. (1997). Organizational factors that influence information technology diffusion in academic health sciences centers. Journal of the American Medical Informatics Association, 4(2), 102–111.Find this resource:

Ash, J. S., & Bates, D. W. (2005). Factors and forces affecting EHR system adoption: Report of a 2004 ACMI discussion. Journal of the American Medical Informatics Association, 12(1), 8–12.Find this resource:

Ash, J. S., Sittig, D. F., Campbell, E., Guappone, K., & Dykstra, R. (2006). An unintended consequence of CPOE implementation: Shifts in power, control, and autonomy. Paper presented at the AMIA Annual Symposium Proceedings.Find this resource:

Ash, J. S., Sittig, D. F., Campbell, E. M., Guappone, K. P., & Dykstra, R. H. (2007). Some unintended consequences of clinical decision support systems. Paper presented at the AMIA Annual Symposium Proceedings.Find this resource:

Ashley, E. A. (2015). The precision medicine initiative: A new national effort. JAMA, 313(21), 2119–2120.Find this resource:

Auditor General of Canada. (2010). Electronic health records in Canada: An overview of federal and provincial audit reports.Find this resource:

Bae, J., & Encinosa, W. E. (2016). National estimates of the impact of electronic health records on the workload of primary care physicians. BMC Health Services Research, 16(1), 172.Find this resource:

Bailey, J. E., Pope, R. A., Elliott, E. C., Wan, J. Y., Waters, T. M., & Frisse, M. E. (2013). Health information exchange reduces repeated diagnostic imaging for back pain. Annals of Emergency Medicine, 62(1), 16–24.Find this resource:

Bailey, J. E., Wan, J. Y., Mabry, L. M., Landy, S. H., Pope, R. A., Waters, T. M., & Frisse, M. E. (2013). Does health information exchange reduce unnecessary neuroimaging and improve quality of headache care in the emergency department? Journal of General Internal Medicine, 28(2), 176–183.Find this resource:

Barnard, A., & Sandelowski, M. (2001). Technology and humane nursing care: (ir)reconcilable or invented difference? Journal of Advanced Nursing, 34(3), 367–375.Find this resource:

Bindman, A. B. (2017). The Agency for Healthcare Research and Quality and the development of a learning health care system. JAMA Internal Medicine, 177(7), 909–910.Find this resource:

Blumenthal, D., & Tavenner, M. (2010). The “meaningful use” regulation for electronic health records. New England Journal of Medicine, 2010(363), 501–504.Find this resource:

Boonstra, A., & Broekhuis, M. (2010). Barriers to the acceptance of electronic medical records by physicians from systematic review to taxonomy and interventions. BMC Health Services Research, 10(1), 231.Find this resource:

Brenner, S. K., Kaushal, R., Grinspan, Z., Joyce, C., Kim, I., Allard, R. J., . . . Abramson, E. L. (2015). Effects of health information technology on patient outcomes: A systematic review. Journal of the American Medical Informatics Association, 23(5), 1016–1036.Find this resource:

Brynjolfsson, E., & Hitt, L. M. (1998). Beyond the productivity paradox. Communications of the ACM, 41(8), 49–55.Find this resource:

Buntin, M. B., Burke, M. F., Hoaglin, M. C., & Blumenthal, D. (2011). The benefits of health information technology: A review of the recent literature shows predominantly positive results. Health Affairs, 30(3), 464–471.Find this resource:

Canada Health Infoway. (2017). Canada Health Infoway annual report 2016–2017.

Centers for Medicare and Medicaid Services. (2014, February 26). E-prescribing.

Chambers, D. A., Feero, W. G., & Khoury, M. J. (2016). Convergence of implementation science, precision medicine, and the learning health care system: A new model for biomedical research. JAMA, 315(18), 1941–1942.Find this resource:

Chang, F., & Gupta, N. (2015). Progress in electronic medical record adoption in Canada. Canadian Family Physician, 61(12), 1076–1084.Find this resource:

Chaudhry, B., Wang, J., Wu, S., Maglione, M., Mojica, W., Roth, E., . . . Shekelle, P. G. (2006). Systematic review: Impact of health information technology on quality, efficiency, and costs of medical care. Annals of Internal Medicine, 144(10), 742–752.Find this resource:

Cohen, M. F. (2016). Impact of the HITECH financial incentives on EHR adoption in small, physician-owned practices. International Journal of Medical Informatics, 94, 143–154.Find this resource:

Collins, F. S., & Varmus, H. (2015). A new initiative on precision medicine. New England Journal of Medicine, 372(9), 793–795.Find this resource:

Comptroller and Auditor General. (2008). The national programme for IT in the NHS: Progress since 2006. London, U.K.: Stationery Office Books.Find this resource:

Cutler, D. M., Rosen, A. B., & Vijan, S. (2006). The value of medical spending in the United States, 1960–2000. New England Journal of Medicine, 355(9), 920–927.Find this resource:

Darbyshire, P. (2004). “Rage against the machine?”: nurses’ and midwives’ experiences of using Computerized Patient Information Systems for clinical information. Journal of Clinical Nursing, 13(1), 17–25.Find this resource:

David, P. A. (1990). The dynamo and the computer: An historical perspective on the modern productivity paradox. The American Economic Review, 80(2), 355–361.Find this resource:

Devine, E. B., Totten, A. M., Gorman, P., Eden, K. B., Kassakian, S., Woods, S., . . . Hersh, W. R. (2017). Health information exchange use (1990–2015): A systematic review. eGEMs (Generating Evidence & Methods to improve patient outcomes), 5(1).Find this resource:

Dranove, D., Garthwaite, C., Li, B., & Ody, C. (2015). Investment subsidies and the adoption of electronic medical records in hospitals. Journal of Health Economics, 44, 309–319.Find this resource:

Edwards, A., Hollin, I., Barry, J., & Kachnowski, S. (2009). Barriers to cross: Institutional health information exchange, a literature review. Journal of Healthcare Information Management, 24(3), 22–34.Find this resource:

Everson, J. (2017). The implications and impact of 3 approaches to health information exchange: Community, enterprise, and vendor‐mediated health information exchange. Learning Health Systems, 1(2).Find this resource:

Everson, J., & Adler-Milstein, J. (2016). Engagement in hospital health information exchange is associated with vendor marketplace dominance. Health Affairs, 35(7), 1286–1293.Find this resource:

Everson, J., Lee, S.-Y. D., & Adler-Milstein, J. (2016). Achieving adherence to evidence-based practices: Are health IT and hospital-physician integration complementary or substitutive strategies?. Medical Care Research and Review, 73(6), 724–751.Find this resource:

Fleming, N. S., Culler, S. D., McCorkle, R., Becker, E. R., & Ballard, D. J. (2011). The financial and nonfinancial costs of implementing electronic health records in primary care practices. Health Affairs (Project Hope), 30(3), 481–489.Find this resource:

Ford, E. W., Menachemi, N., Peterson, L. T., & Huerta, T. R. (2009). Resistance is futile: But it is slowing the pace of EHR adoption nonetheless. Journal of the American Medical Informatics Association, 16(3), 274–281.Find this resource:

Fox, R. C., Aiken, L. H., & Messikomer, C. M. (1990). The culture of caring: AIDS and the nursing profession. The Milbank Quarterly, 68(Suppl. 2), 226–256.Find this resource:

Freidson, E. (1974). Professional dominance: The social structure of medical care. New Brunswick, NJ: Transaction Publishers.Find this resource:

Freidson, E. (1988). Profession of medicine: A study of the sociology of applied knowledge. Chicago, IL: University of Chicago Press.Find this resource:

Friedman, C., Rubin, J., Brown, J., Buntin, M., Corn, M., Etheredge, L., . . . Stead, W. (2014). Toward a science of learning systems: A research agenda for the high-functioning Learning Health System. Journal of the American Medical Informatics Association, 22(1), 43–50.Find this resource:

Friedman, C. P., Allee, N. J., Delaney, B. C., Flynn, A. J., Silverstein, J. C., Sullivan, K., & Young, K. A. (2017). The science of learning health systems: Foundations for a new journal. Learning Health Systems, 1(1), e10020.Find this resource:

Friedman, C. P., Wong, A. K., & Blumenthal, D. (2010). Achieving a nationwide learning health system. Science Translational Medicine, 2(57), 57cm29.Find this resource:

Garg, A. X., Adhikari, N. K., McDonald, H., Rosas-Arellano, M. P., Devereaux, P., Beyene, J., . . . Haynes, R. B. (2005). Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: A systematic review. JAMA, 293(10), 1223–1238.Find this resource:

Garrett, P., & Seidman, J. (2011, January 4). EMR vs EHR: What is the difference?

Gerein, K. (2017, September 29). Alberta Health Services signs $459-million deal for massive new technology system. Edmonton Journal.Find this resource:

Gray, B. H., Bowden, T., Johansen, I., & Koch, S. (2011). Electronic health records: An international perspective on “meaningful use” (Issues in International Health Policy No. Commwealth Fund pub. 1565 Vol. 28). The Commonwealth Fund.Find this resource:

Grossman, J. M., Kushner, K. L., & November, E. A. (2008). Creating sustainable local health information exchanges: Can barriers to stakeholder participation be overcome? Washington, DC: Center for Studying Health System Change.Find this resource:

Heisey-Grove, D., & Patel, V. (2015). Any, certified, and basic: Quantifying physician EHR adoption through 2014. Washington, DC: The Office of the National Coordinator for Health Information Technology.Find this resource:

Henry, J., Pylypchuk, Y., Searchy T., & Patel, V. (2016). Adoption of electronic health record systems among US non-federal acute care hospitals: 2008–2015. Washington, DC: The Office of the National Coordinator for Health Information Technology.Find this resource:

Hern, A. (2017, October 26). NHS could have avoided WannaCry hack with “basic IT security”, says report. The Guardian.Find this resource:

Hersh, W. R., Totten, A. M., Eden, K. B., Devine, B., Gorman, P., Kassakian, S. Z., . . . McDonagh, M. S. (2015). Outcomes from health information exchange: Systematic review and future research needs. JMIR Medical Informatics, 3(4), e39.Find this resource:

Hillestad, R., Bigelow, J., Bower, A., Girosi, F., Meili, R., Scoville, R., & Taylor, R. (2005). Can electronic medical record systems transform health care? Potential health benefits, savings, and costs. Health Affairs, 24(5), 1103–1117.Find this resource:

Himmelstein, D. U., Wright, A., & Woolhandler, S. (2010). Hospital computing and the costs and quality of care: A national study. The American Journal of Medicine, 123(1), 40–46.Find this resource:

HIMSS Analytics EMRAM. (2017, January 24). Electronic medical record adoption model.

Howley, M. J., Chou, E. Y., Hansen, N., & Dalrymple, P. W. (2014). The long-term financial impact of electronic health record implementation. Journal of the American Medical Informatics Association, 22(2), 443–452.Find this resource:

Hsiao, C.-J., & Hing, E. (2012). Use and characteristics of electronic health record systems among office-based physician practices: United States, 2001–2012. NCHS Data Brief, 111, 1–8.Find this resource:

Imler, T. D., Vreeman, D. J., & Kannry, J. (2016). Healthcare data standards and exchange. In Clinical Informatics Study Guide (pp. 233–253). Switzerland: Springer International Publishing.Find this resource:

Institute of Medicine. (2000). To err is human: Building a safer health system. Washington, DC: The National Academies Press.Find this resource:

Institute of Medicine. (2001). Crossing the quality chasm: A new health system for the 21st century. Washington, DC: The National Academies Press.Find this resource:

Jamoom, E., Heisey-Grove, D., Yang, N., & Scanlon, P. (2016). Physician opinions about EHR use by EHR experience and by whether the practice had optimized its EHR use. Journal of Health & Medical Informatics, 7(4), 1000240.Find this resource:

Jamoom, E., & Hing, E. (2015). Progress with electronic health record adoption among emergency and outpatient departments: United States, 2006–2011. NCHS Data Brief, 187, 1–8.Find this resource:

Jamoom, E. W., Patel, V., Furukawa, M. F., & King, J. (2014). EHR adopters vs. non-adopters: Impacts of, barriers to, and federal initiatives for EHR adoption. Healthcare (Amsterdam, The Netherlands), 2(1), 33–39.Find this resource:

Jayanthi, A., & Ellison, A. (2016, May 23). 8 hospitals’ finances hurt by EHR costs.

Jha, A. K., Doolan, D., Grandt, D., Scott, T., & Bates, D. W. (2008). The use of health information technology in seven nations. International Journal of Medical Informatics, 77(12), 848–854.Find this resource:

Jha, A. K., Ferris, T. G., Donelan, K., DesRoches, C., Shields, A., Rosenbaum, S., & Blumenthal, D. (2006). How common are electronic health records in the United States? A summary of the evidence. Health Affairs (Project Hope), 25(6), w496–507.Find this resource:

Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S., . . . Wang, Y. (2017). Artificial intelligence in healthcare: Past, present and future. Stroke and Vascular Neurology, 2(4), 230–243.Find this resource:

Jones, S. S., Adams, J. L., Schneider, E. C., Ringel, J. S., & McGlynn, E. A. (2010). Electronic health record adoption and quality improvement in US hospitals. The American Journal of Managed Care, 16(12 Suppl. HIT), SP64–71.Find this resource:

Jones, S. S., Heaton, P. S., Rudin, R. S., & Schneider, E. C. (2012). Unraveling the IT productivity paradox: Lessons for health care. New England Journal of Medicine, 366(24), 2243–2245.Find this resource:

Jones, S. S., Rudin, R. S., Perry, T., & Shekelle, P. G. (2014). Health information technology: An updated systematic review with a focus on meaningful use. Annals of Internal Medicine, 160(1), 48–54.Find this resource:

Kaelber, D., Waheed, R., Einstadter, D., Love, T., & Cebul, R. (2013). Use and perceived value of health information exchange: One public healthcare system’s experience. The American Journal of Managed Care, 19(10 Spec. No.), SP337–343.Find this resource:

Kahn, J. S., Aulakh, V., & Bosworth, A. (2009). What it takes: Characteristics of the ideal personal health record. Health Affairs, 28(2), 369–376.Find this resource:

Kaushal, R., Shojania, K. G., & Bates, D. W. (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.Find this resource:

Kern, L. M., Barron, Y., Abramson, E. L., Patel, V., & Kaushal, R. (2009). HEAL NY: Promoting interoperable health information technology in New York State. Health Affairs, 28(2), 493–504.Find this resource:

Krumholz, H. M. (2014). Big data and new knowledge in medicine: The thinking, training, and tools needed for a learning health system. Health Affairs, 33(7), 1163–1170.Find this resource:

Kruse, C. S., & Beane, A. (2018). Health information technology continues to show positive effect on medical outcomes: Systematic review. Journal of Medical Internet Research, 20(2), e41.Find this resource:

Kuperman, G. J., Bobb, A., Payne, T. H., Avery, A. J., Gandhi, T. K., Burns, G., . . . Bates, D. W. (2007). Medication-related clinical decision support in computerized provider order entry systems: A review. Journal of the American Medical Informatics Association, 14(1), 29–40.Find this resource:

Lammers, E. J., Adler-Milstein, J., & Kocher, K. E. (2014). Does health information exchange reduce redundant imaging? Evidence from emergency departments. Medical Care, 52(3), 227–234.Find this resource:

Lammers, E. J., & McLaughlin, C. G. (2017). Meaningful use of electronic health records and medicare expenditures: Evidence from a panel data analysis of US health care markets, 2010–2013. Health Services Research, 52(4), 1364–1386.Find this resource:

Lammers, E. J., McLaughlin, C. G., & Barna, M. (2016). Physician EHR adoption and potentially preventable hospital admissions among Medicare beneficiaries: Panel data evidence, 2010–2013. Health Services Research, 51(6), 2056–2075.Find this resource:

Lee, J., & Dowd, B. (2013). Effect of health information technology expenditure on patient level cost. Healthcare Informatics Research, 19(3), 215–221.Find this resource:

Lluch, M. (2011). Healthcare professionals’ organisational barriers to health information technologies: A literature review. International Journal of Medical Informatics, 80(12), 849–862.Find this resource:

McAlearney, A. S., Chisolm, D. J., Schweikhart, S., Medow, M. A., & Kelleher, K. (2007). The story behind the story: Physician skepticism about relying on clinical information technologies to reduce medical errors. International Journal of Medical Informatics, 76(11–12), 836–842.Find this resource:

McCullough, J. S., Casey, M., Moscovice, I., & Prasad, S. (2010). The effect of health information technology on quality in US hospitals. Health Affairs, 29(4), 647–654.Find this resource:

McCullough, J. S., Parente, S. T., & Town, R. (2016). Health information technology and patient outcomes: The role of information and labor coordination. The RAND Journal of Economics, 47(1), 207–236.Find this resource:

McDonald, C. J. (1976). Protocol-based computer reminders, the quality of care and the non-perfectibility of man. New England Journal of Medicine, 295(24), 1351–1355.Find this resource:

McDonald, C. J., Overhage, J. M., Tierney, W. M., Dexter, P. R., Martin, D. K., Suico, J. G., . . . Glazener, T. (1999). The Regenstrief medical record system: A quarter century experience. International Journal of Medical Informatics, 54(3), 225–253.Find this resource:

McGinnis, J. M., Powers, B., & Grossmann, C. (2011). Digital infrastructure for the learning health system: The foundation for continuous improvement in health and health care, workshop series summary. Washington, DC: National Academies Press (US).Find this resource:

May, C., & Fleming, C. (1997). The professional imagination: Narrative and the symbolic boundaries between medicine and nursing. Journal of Advanced Nursing, 25(5), 1094–1100.Find this resource:

Melnyk, B. M., Fineout-Overholt, E., Gallagher-Ford, L., & Kaplan, L. (2012). The state of evidence-based practice in US nurses: Critical implications for nurse leaders and educators. Journal of Nursing Administration, 42(9), 410–417.Find this resource:

Mennemeyer, S. T., Menachemi, N., Rahurkar, S., & Ford, E. W. (2016). Impact of the HITECH Act on physicians’ adoption of electronic health records. Journal of the American Medical Informatics Association, 23(2), 375–379.Find this resource:

Metzger, M.-H., Durand, T., Lallich, S., Salamon, R., & Castets, P. (2012). The use of regional platforms for managing electronic health records for the production of regional public health indicators in France. BMC Medical Informatics and Decision Making, 12, 28.Find this resource:

Miller, A. R., & Tucker, C. E. (2011). Can health care information technology save babies? Journal of Political Economy, 119(2), 289–324.Find this resource:

Mossialos, E., Djordjevic, A., Osborn, R., & Sarnak, D. (2017). International profiles of health care systems. The Commonwealth Fund.Find this resource:

Neill, D. B. (2013). Using artificial intelligence to improve hospital inpatient care. IEEE Intelligent Systems, 28(2), 92–95.Find this resource:

Nelson, E. C., Godfrey, M. M., Batalden, P. B., Berry, S. A., Bothe, A. E., McKinley, K. E., . . . Nolan, T. W. (2008). Clinical microsystems, part 1: The building blocks of health systems. Joint Commission Journal on Quality and Patient Safety, 34(7), 367–378.Find this resource:

Olsen, L., Aisner, D., & McGinnis, J. M. (2007). The learning healthcare system. Washington, DC: National Academies Press (US).Find this resource:

Orszag, P. R. (2008). Evidence on the costs and benefits of health information technology (A CBO Paper No. Pub. 2976). Washington, DC: Congressional Budget Office.Find this resource:

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

Paulus, R. A., Davis, K., & Steele, G. D. (2008). Continuous innovation in health care: Implications of the geisinger experience. Health Affairs, 27(5), 1235–1245.Find this resource:

Poon, E. G., Blumenthal, D., Jaggi, T., Honour, M. M., Bates, D. W., & Kaushal, R. (2004). Overcoming barriers to adopting and implementing computerized physician order entry systems in US hospitals. Health Affairs, 23(4), 184–190.Find this resource:

Prgomet, M., Li, L., Niazkhani, Z., Georgiou, A., & Westbrook, J. I. (2016). Impact of commercial computerized provider order entry (CPOE) and clinical decision support systems (CDSSs) on medication errors, length of stay, and mortality in intensive care units: A systematic review and meta-analysis. Journal of the American Medical Informatics Association, 24(2), 413–422.Find this resource:

Protti, D., & Johansen, I. (2010). Widespread adoption of information technology in primary care physician offices in Denmark: A case study. Issues in International Health Policy No. Commwealth Fund pub. 1397 Vol. 80. The Commonwealth Fund.Find this resource:

Rahurkar, S., Vest, J. R., & Menachemi, N. (2015). Despite the spread of health information exchange, there is little evidence of its impact on cost, use, and quality of care. Health Affairs, 34(3), 477–483.Find this resource:

Ross, S. E., Schilling, L. M., Fernald, D. H., Davidson, A. J., & West, D. R. (2010). Health information exchange in small-to-medium sized family medicine practices: Motivators, barriers, and potential facilitators of adoption. International Journal of Medical Informatics, 79(2), 123–129.Find this resource:

Rucker, D. (2017, June 15). Coming to ONC: One informaticist’s journey.

Rudin, R. S., Motala, A., Goldzweig, C. L., & Shekelle, P. G. (2014). Usage and effect of health information exchange: A systematic review. Annals of Internal Medicine, 161(11), 803–811.Find this resource:

Rueschemeyer, D. (1983). Professional autonomy and the social control of expertise. The Sociology of the Professions (pp. 38–58). London, U.K.: SAGE.Find this resource:

Schwartz, A., Magoulas, R., & Buntin, M. (2013). Tracking labor demand with online job postings: The case of health IT workers and the HITECH Act. Industrial Relations: A Journal of Economy and Society, 52(4), 941–968.Find this resource:

Scott, W. R. (2000). Institutional change and healthcare organizations: From professional dominance to managed care. Chicago, IL: University of Chicago Press.Find this resource:

Séroussi, B., & Bouaud, J. (2017). Adoption of a nationwide shared medical record in France: Lessons learnt after 5 years of deployment. AMIA Annual Symposium Proceedings, 2016, 1100–1109.Find this resource:

Shortliffe, E. H. (1999). The evolution of electronic medical records. Academic Medicine: Journal of the Association of American Medical Colleges, 74(4), 414–419.Find this resource:

Slight, S. P., Quinn, C., Avery, A. J., Bates, D. W., & Sheikh, A. (2014). A qualitative study identifying the cost categories associated with electronic health record implementation in the UK. Journal of the American Medical Informatics Association, 21(e2), e226–e231.Find this resource:

Sprague, L. (2004). Electronic health records: How close? How far to go? NHPF Issue Brief, 800, 1–17.Find this resource:

Stevenson, J. E., Nilsson, G. C., Petersson, G. I., & Johansson, P. E. (2010). Nurses’ experience of using electronic patient records in everyday practice in acute/inpatient ward settings: A literature review. Health Informatics Journal, 16(1), 63–72.Find this resource:

Tai-Seale, M., Olson, C. W., Li, J., Chan, A. S., Morikawa, C., Durbin, M., . . . Luft, H. S. (2017). Electronic health record logs indicate that physicians split time evenly between seeing patients and desktop medicine. Health Affairs (Project Hope), 36(4), 655–662.Find this resource:

Tarver, W. L., & Menachemi, N. (2015). The impact of health information technology on cancer care across the continuum: A systematic review and meta-analysis. Journal of the American Medical Informatics Association, 23(2), 420–427.Find this resource:

The National Alliance for Health Information Technology. (2008). Report to the Office of the National Coordinator for Health Information Technology on Defining Key Health Information Technology Terms.Find this resource:

The Office of the National Coordinator for Health Information Technology. (2015). 2015 Report to Congress on Health Information Blocking.

Timmermans, S., & Berg, M. (2010). The gold standard: The challenge of evidence-based medicine and standardization in health care. Philadelphia, PA: Temple University Press.Find this resource:

Triplett, J. E. (1999). The Solow productivity paradox: What do computers do to productivity? The Canadian Journal of Economics/Revue canadienne d’Economique, 32(2), 309–334.Find this resource:

Verghese, A., Shah, N. H., & Harrington, R. A. (2018). What this computer needs is a physician: Humanism and artificial intelligence. JAMA, 319(1), 19–20.Find this resource:

Vest, J. R., Campion, T. R., Kaushal, R., & Investigators, H. (2013). Challenges, alternatives, and paths to sustainability for health information exchange efforts. Journal of Medical Systems, 37(6), 9987.Find this resource:

Vest, J. R., Gamm, L. D., Ohsfeldt, R. L., Zhao, H., & Jasperson, J. S. (2012). Factors associated with health information exchange system usage in a safety-net ambulatory care clinic setting. Journal of Medical Systems, 36(4), 2455–2461.Find this resource:

Vest, J. R., & Kash, B. A. (2016). Differing strategies to meet information‐sharing needs: Publicly supported community health information exchanges versus health systems’ enterprise health information exchanges. The Milbank Quarterly, 94(1), 77–108.Find this resource:

Wachter, R. (2016). Making IT work: Harnessing the power of health information technology to improve care in England. London, U.K.: Department of Health.Find this resource:

Walker, D., Mora, A., Demosthenidy, M. M., Menachemi, N., & Diana, M. L. (2016). Meaningful use of EHRs among hospitals ineligible for incentives lags behind that of other hospitals, 2009–13. Health Affairs, 35(3), 495–501.Find this resource:

Walker, J., Pan, E., Johnston, D., Adler-Milstein, J., Bates, D. W., & Middleton, B. (2005). The value of health care information exchange and interoperability. Health Affairs (Project Hope), Suppl Web Exclusives, W5-10–W15-18.Find this resource:

Walter, Z., & Lopez, M. S. (2008). Physician acceptance of information technologies: Role of perceived threat to professional autonomy. Decision Support Systems, 46(1), 206–215.Find this resource:

Wears, R. L., & Berg, M. (2005). Computer technology and clinical work: Still waiting for Godot. JAMA, 293(10), 1261–1263.Find this resource:

Weiner, B. J. (2009). A theory of organizational readiness for change. Implementation Science, 4(1), 67.Find this resource:

Yu, V. L., Fagan, L. M., Wraith, S. M., Clancey, W. J., Scott, A. C., Hannigan, J., . . . Cohen, S. N. (1979). Antimicrobial selection by a computer: A blinded evaluation by infectious diseases experts. JAMA, 242(12), 1279–1282.Find this resource:


(1.) The literature comparing international experience in health information technology remains sparse, at least in English. Therefore, this section relies in several parts on informative studies conducted by the Commonwealth Foundation (Mossialos, Djordjevic, Osborn, & Sarnak, 2017; Gray, Bowden, Johansen, & Koch, 2011).

(2.) We conducted a search in PubMed using the most frequently adopted keywords to identify recent studies on the effect of HIT. We performed a narrow search meant to complement the most recently published systematic reviews. The search we executed used the following algorithm: (((Health information technology[Title] OR Electronic health record[Title] OR CPOE[Title] OR Computerized physician order entry[Title] OR electronic medical record[Title] OR Health Information[Title] OR Electronic medical record[Title] OR Personal health record[Title])) AND (Quality[Title] OR efficienc*[Title] OR cost[Title] OR outcome[Title] OR impact[Title] OR safety[Title] OR satisfaction[Title]))) NOT (imaging[Title] OR radiograph*[Title] OR “ct scan”[Title] OR mri[Title] OR magnetic resonance[Title] OR tomograph*[Title] OR imrt[Title] OR robot*[Title] OR vivo[Title] OR vitro[Title] OR situ[Title] OR simulat*[Title] OR driving[Title] OR driver*[Title] OR protein[Title] OR “probable high risk”[Title])) AND (Quality[Title] OR efficienc*[Title] OR cost[Title] OR outcome[Title] OR impact[Title] OR safety[Title] OR satisfaction[Title])) NOT (imaging OR radiograph* OR “ct scan” OR mri OR magnetic resonance OR tomograph* OR imrt OR robot* OR vivo OR vitro OR situ OR simulat* OR driving OR driver* OR protein OR “probable high risk”) Date: 8/13/13–3/6/2018.