Electronic Health Record Systems

McDonald, Clement J.; Tang, Paul C.; Hripcsak, George

doi:10.1007/978-1-4471-4474-8_12

Clement J. McDonald MD³,
Paul C. Tang MD, MS⁴ &
George Hripcsak MD, MS⁵

134k Accesses
1 Citations

Abstract

After reading this chapter, you should know the answers to these questions:

Access provided by Autonomous University of Puebla. Download chapter PDF

Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

After reading this chapter, you should know the answers to these questions:

What is the definition of an electronic health record (EHR)?
How does an EHR differ from the paper record?
What are the functional components of an EHR?
What are the benefits of an EHR?
What are the impediments to development and use of an EHR?

1 What Is an Electronic Health Record?

The preceding chapters introduced the conceptual basis for the field of biomedical informatics, including the use of patient data in clinical practice and research. We now focus attention on the patient record, commonly referred to as the patient’s chart, medical record, or health record. In this chapter, we examine the definition and use of electronic health record (EHR) systems, discuss their potential benefits and costs, and describe the remaining challenges to address in their dissemination.

1.1 Purpose of a Patient Record

Stanley Reiser (1991) wrote that the purpose of a patient record is “to recall observations, to inform others, to instruct students, to gain knowledge, to monitor performance, and to justify interventions.” The many uses described in this statement, although diverse, have a single goal—to further the application of health sciences in ways that improve the well-being of patients, including the conduct of research and public health activities that address population health. A modern electronic health record (EHR) is designed to facilitate these uses, providing much more than a static view of events.

An electronic health record (EHR) is a repository of electronically maintained information about an individual’s health status and health care, stored such that it can serve the multiple legitimate uses and users of the record. Traditionally, the patient record was a record of care provided when a patient was ill. Health care is evolving to encourage health care providers to focus on the continuum of health and health care from wellness to illness and recovery. Consequently, we anticipate that eventually it will carry all of a person’s health related information from all sources over their lifetime. The Department of Veterans Affairs (VA) has already committed to keeping existing patient electronic data for 75 years. In addition, the data should be stored such that different views of those data can be presented to serve the many different uses described in Chap. 2.

The term electronic health record system (also referred to as a computer-based patient-record system) includes the active tools that are used to manage the information, but in common use, the term EHR can refer to the entire system. EHRs include information management tools to provide clinical reminders and alerts, linkages with knowledge sources for health care decision support, and analysis of aggregate data both for care management and for research. The EHR helps the reader to organize, interpret, and react to data. Examples of tools provided in current EHRs are discussed in Sect. 12.3.

1.2 Ways in Which an Electronic Health Record Differs from a Paper-Based Record

Compared to the historical paper medical record, whose functionality is constrained by its recording media, and the fact that only one physical copy of it exists—the EHR is flexible and adaptable (see also Sect. 2.3 in Chap. 2). Data may be entered in one format to simplify the input process and then displayed in many different formats according to the user’s needs. The entry and display of dates is illustrative. Most EHRs can accept many date formats, i.e. May 1, 1992, 1 May 92, or1/5/92, as input; store that information in one internal format, such as 1992-05-01; and display it in different formats according to local customs. The EHR can incorporate multimedia information, such as radiology images and echocardiographic video loops, which were never part of the traditional medical record. It can also analyze a patient’s record, call attention to trends and dangerous conditions and suggest corrective actions much like an airplane flight control computer. EHRs can organize data about one patient to facilitate his or her care or about a population of patients to assist management decisions or answer epidemiologic questions. When considering the functions of an EHR, one must think beyond the constraints of paper records. An EHR system can capture, organize, analyze, and display patient data in many ways.

Inaccessibility is a problem with paper records. They can only be in one place and with at most one user at one point in time. In large organizations, medical record departments often would sequester the paper medical record for days after the patient’s hospital discharge while the clinician completed the discharge summary and signed every form. Individual physicians may borrow records for their own administrative or research purposes, during which times the record will also be unavailable. In contrast, many users, including patients, can read the same electronic record at once. So it is never unavailable. With today’s secure networks, clinicians and patients can access a patient’s EHR from geographically distributed sites, such as the emergency room, their office, or their home. Such availability can also support health care continuity during disasters. Brown et al. (2007) found a “stark contrast” between the care VA versus non-VA patients obtained after Hurricane Katrina, because “VA efforts to maintain appropriate and uninterrupted care were supported by nationwide access to comprehensive electronic health record systems.” While EHR systems make data more accessible to authorized users, they also provide greater control over access and enforce applicable privacy policies as required by the Health Insurance Portability and Accountability Act (HIPAA) (see Chaps. 10 and 27).

The EHR’s content is more legible and better organized than the paper alternative and the computer can increase the quality of data by applying validity checks as data is being entered. The computer can reduce typographical errors through restricted input menus and spell checking. It can require data entry in specified fields, conditional on the value of other fields. For example, if the user answers yes to current smoker, the computer, guided by rules, could then ask how many packs per day smoked or how soon after awakening does the patient take their first smoke? So the EHR not only stores data but can also conditionally enforce the capture of certain data elements. This enforcement power should be used sparingly, however. As part of the ordering process, the computer can require the entry of data that may not be available (e.g., the height of a patient with leg contractures), and thus prevent the clinician from completing an important order (Strom et al. 2010); and overzealous administrators can ask clinicians to answer questions that are peripheral to clinical care and slow the care process.

The degree to which a particular EHR achieves benefits depends on several factors:

Comprehensiveness of information. Does the EHR contain information about health as well as illness? Does it include information from all organizations and clinicians who participated in a patient’s care? Does it cover all settings in which care was delivered (e.g., office practice, hospital)? Does it include the full spectrum of clinical data, including clinicians’ notes, laboratory test results, medication details, and so on?
Duration of use and retention of data. EHRs gain value over time because they accumulate a greater proportion of the patients’ medical history. A record that has accumulated patient data over 5 years will be more valuable than one that contains only the last month’s records.
Degree of structure of data. Narrative notes stored in electronic health records have the advantage over their paper counterparts in that they can be searched by word, although the success of such searches is subject to the wide variations in the author’s choice of medical words and abbreviations. Computer-supported decision making, clinical research, and management analysis of EHR data require structured data. One way to obtain such data is to ask the clinical user to enter information through structured forms whose fields provide dropdown menus or restrict data entry to a controlled vocabulary (see Chap. 7).
Ubiquity of access. A system that is accessible from a few sites will be less valuable than one accessible by an authorized user from anywhere (see Chap. 5).

An EHR system has some disadvantages. It requires a larger initial investment than its paper counterpart due to hardware, software, training, and support costs. Physicians and other key personnel have to take time from their work to learn how to use the system and to redesign their workflow to use the system. Although it takes time to learn how to use the system and to change workflows, clinicians increasingly recognize that EHR systems are important tools to assist in the clinical, regulatory, and business of practicing medicine.

Computer-based systems have the potential for catastrophic failures that could cause extended unavailability of patients’ computer records. However, these risks can be mitigated by using fully redundant components, mirrored servers, and battery backup. Even better is to have a parallel site located remotely with hot fail over, which means that a failure at the primary site would not be noticed because the remote site could support users with, at most, a momentary pause. Yet, nothing provides complete protection; contingency plans must be developed for handling brief or longer computer outages. Moreover, paper records are also subject to irretrievable loss, caused by, for example, human error (e.g. misfiling), floods, or fires.

2 Historical Perspective

The development of automated systems was initially stimulated by regulatory and reimbursement requirements. Early health care systems focused on inpatient charge capture to meet billing requirements in a fee-for-service environment.

The Flexner report on medical education was the first formal statement made about the function and contents of the medical record (Flexner 1910). In advocating a scientific approach to medical education, the Flexner report also encouraged physicians to keep a patient-oriented medical record. Three years earlier, Dr. Henry Plummer initiated the “unit record” for the Mayo Clinic (including its St. Mary’s Hospital), placing all the patient’s visits and types of information in a single folder. This innovation represented the first longitudinal medical record (Melton 1996). The Presbyterian Hospital (New York) adopted the unit record for its inpatient and outpatient care in 1916, studying the effect of the unit record on length of stay and quality of care (Openchowski 1925) and writing a series of letters and books about the unit record that disseminated the approach around the nation (Lamb 1955).

The first record we could find of a computer-based medical record was a short newspaper article describing a new “electronic brain” – to replace punched and file index cards and to track hospital and medical records (Brain 1956). Early development of hospital information systems (HIS)—that used terminals rather than punched cards for data entry—emerged around 1970 at varying degrees of maturity (Lindberg 1967; Davis et al. 1968; Warner 1972; Barnett et al. 1979). Weed’s problem-oriented medical record (POMR) (1968) shaped medical thinking about both manual and automated medical records. His computer-based version of the POMR employed touch screen terminals, a new programming language and networking—all radical ideas for the time (Schultz et al. 1971). In 1971, Lockheed’s hospital information system (HIS) became operational at El Camino Hospital in Mountain View, CA. Technicon, Inc. then propagated it to more than 200 hospitals (see also Chap. 14) (Coffey 1979).

Hospital-based systems provided feedback (decision support) to physicians, which affected clinical decisions and ultimately patient outcomes. The HELP system (Pryor 1988) at LDS Hospital, the Columbia University system (Johnson et al. 1991), the CCC system at Beth Israel Deaconess Medical Center (Slack and Bleich 1999), the Regenstrief System (Tierney et al. 1993; McDonald et al. 1999) at Wishard Memorial Hospital, and others (Giuse and Mickish 1996; Halamka and Safran 1998; Hripcsak et al. 1999; Teich et al. 1999; Cheung et al. 2001; Duncan et al. 2001; Brown et al. 2003) are long-standing systems that add clinical functionality to support clinical care, and set the stage for future systems.

The ambulatory care medical record systems emerged around the same time as inpatient systems but were slower to attract commercial interest than hospital information systems. COSTAR (Barnett et al. 1978; Barnett 1984), the Regenstrief Medical Record System (RMRS) (McDonald et al. 1975), STOR (Whiting-O’Keefe et al. 1985), and TMR (Stead and Hammond 1988) are among the examples. Costar and RMRS are still in use today. The status of ambulatory care records was reviewed in a 1982 report (Kuhn et al. 1984). There are now hundreds of vendors who offer ambulatory care EHRs, and a number of communities have begun to adopt EHRs on a broad scale for ambulatory care (Goroll et al. 2009; Menachemi et al. 2011). Morris Collen, who also pioneered the multiphasic screening system (1969), wrote a readable 500-page history of medical informatics (1995) that provides rich details about these early medical records systems, as does a three decade summary of computer-based medical record research projects from the U.S. Agency for Health Care Policy and Research (AHCPR, now called the Agency for Health Care Research and Quality (AHRQ)) (Fitzmaurice et al. 2002).

3 Functional Components of an Electronic Health Record System

As we explained in Sect. 12.1.2, an EHR is not simply an electronic version of the paper record. A medical record that is part of a comprehensive EHR system has linkages and tools to facilitate communication and decision making. In Sects. 12.3.1, 12.3.2, 12.3.3, 12.3.4, and 12.3.5, we summarize the components of a comprehensive EHR system and illustrate functionality with examples from systems currently in use. The five functional components are:

1.
Integrated view of patient data
2.
Clinician order entry
3.
Clinical decision support
4.
Access to knowledge resources
5.
Integrated communication and reporting support

3.1 Integrated View of Patient Data

Providing an integrated view of all relevant patient data is an overarching goal of an EHR. However, capturing everything of interest is not yet possible because: (1) Some patient data do not exist in electronic form anywhere, for example, the hand-written data in old charts. (2) Much of the clinical data that do exist in electronic form are sequestered in isolated external computer systems, for example, office practices, free-standing radiology centers, home-health agencies, and nursing homes that do not yet have operational links to a given EHR or each other. (3) Even when electronic and organizational links exist, a fully integrated view of the data may be thwarted by the difference in conceptualization of data among systems from different vendors, and among different installations of one vendor’s system in different institutions.

An integrated EHR must accommodate a broad spectrum of data types ranging from text to numbers and from tracings to images and video. More complex data types such as radiology images are usually delivered for human viewing—standards like DICOM^{Footnote 1} exist for displaying most of these complex data types, and JPEG^{Footnote 2} display of images is universally available for any kind of image (see also Chaps. 7 and 9). Figure 12.1 shows the VistA CPRS electronic health record system, which integrates a variety of text data and images into a patient report data screen including: demographics, a detailed list of the patient’s procedures, a DICOM chest x-ray image, and JPG photo of a skin lesion. Other tabs in the system provide links to: problems, medications, orders, notes, consults, discharge summary, and labs. An important challenge to the construction of an integrated view is the lack of a national patient identifier in the United States. Because each organization assigns its own medical record number, a receiving organization cannot directly file a patient’s data that is only identified by a medical record number from an external care organization. Linking schemes based on name, birth date and other patient characteristics must be implemented and monitored (Zhu et al. 2009).

The idiosyncratic, local terminologies used to identify clinical variables and their values in many source systems present major barriers to integration of health record data within EHRs. However, those barriers will shrink as institutions adopt code standards (Chap. 7) such as LOINC^{Footnote 3} for observations, questions, variables, and assessments (McDonald et al. 2003; Vreeman et al. 2010); SNOMED CT^{Footnote 4} (Wang et al. 2002) for diagnoses, symptoms, findings, organisms and answers; UCUM^{Footnote 5} for computable units of measure; and RxNorm^{Footnote 6} and RxTerms^{Footnote 7} for clinical drug names, ingredients, and orderable drug names. Federal regulations from CMS and ONC for Meaningful Use 2 (MU2) encourage or require the use of LOINC, RxNorm and SNOMED CT for various purposes. (Final Rule: CMS 2012; Final Rule: ONC 2012) (see also Chaps. 7 and 27). Now most laboratory instrument vendors specify what LOINC codes to use for each test result generated by their instruments.

Today, most clinical data sources and EHRs can send and receive clinical content as version 2.× Health Level 7 (HL7)^{Footnote 8} messages. Larger organizations use interface engines to send, receive, and, when necessary, translate the format of, and the codes within, such messages (see Chap. 7); Fig. 12.2 shows an example of architecture to integrate data from multiple source systems. The Columbia University Medical Center computerized patient record (CPR) interface depicted in this diagram not only provides message-handling capability but can also automatically translate codes from the external source to the preferred codes of the receiving EHR. And although many vendors now offer single systems that serve “all” needs, they never escape the need for HL7 interfaces to capture data from some systems, e.g., EKG carts, cardiology systems, radiology imaging systems, anesthesia systems, off-site laboratories, community pharmacies and external collaborating health systems. At least one high-capability open-source interface engine, Mirth Connect,^{Footnote 9} is now available. One of us, (CM), used it happily, for example, in a project that links a local hospital’s emergency room to Surescripts’ medication history database.^{Footnote 10}

3.2 Clinician Order Entry

One of the most important components of an EHR is order entry, the point at which clinicians make decisions and take actions, and the computer can provide assistance. Electronic order entry can improve health care at several levels. An electronic order entry system can potentially reduce errors and costs compared to a paper system, in which orders are transcribed manually from one paper form (e.g., the paper chart) to another (e.g., the nurse’s work list or a laboratory request form). Orders collected directly from the decision maker can be passed in a legible form to the intended recipient without the risk of transcription errors or the need for additional personnel. Order entry systems also provide opportunities to deliver decision support at the point where clinical decisions are being made. Most order entry systems pop up alerts about any interactions or allergies associated with a new drug order. But implementers should be selective about which alerts they present and which ones are interruptive, to avoid wasting provider time on trivial or low-likelihood outcomes (Phansalkar et al. 2012a, b). This capability is discussed in greater detail in the next section. Order entry systems can facilitate the entry of simple orders like “vital signs three times a day,” or very complicated orders such as total parenteral nutrition (TPN) which requires specification of many additives, and many calculations and checks to avoid physically impossible or dangerous mixtures and to assure that the prescribed goals for the number of calories and the amount of each additive are met. Figure 12.3 shows an example of a TPN order entry screen from Vanderbilt (Miller 2005b). Once a clinician order-entry system is adopted by the practice, simply changing the default drug or dosing based on the latest scientific evidence can shift the physician’s ordering behavior toward the optimum standard of care, with benefits to quality and costs. Because of the many potential advantages for care quality and efficiency, care organizations are adopting computerized physician order entry (CPOE) (Khajouei and Jaspers 2010).

3.3 Clinical Decision Support

Clinical trials have shown that reminders from decision support improve the care process (Haynes 2011; Damiani et al. 2010; Schedlbauer et al. 2009). The EHR can deliver decision support in batch mode at intervals across a whole practice population in order to identify patients who are not reaching treatment targets, are past due for immunizations or cancer screening, or have missed their recent appointments, to cite a few examples. In this mode, the practice uses the batch list of patients generated by decision support to contact the patient and encourage him or her to reach a goal or to schedule an appointment for the delivery of suggested care. This is the only mode that can reach patients who repeatedly miss appointments.

Decision support—especially related to prevention—is most efficiently delivered when the patient comes to the care site for other reasons (e.g., a regularly scheduled visit). In addition, many kinds of computer suggestions are best delivered during the physician order entry process. For example, order entry is the only point in the workflow at which to discourage or countermand an order that might be dangerous or wasteful. It is also a convenient point to offer reminders about needed tests or treatments, because they will usually require an order for their initiation.

The best way for the computer to suggest actions that require an order is to present a pre-constructed order to the provider who can confirm or reject it with a single key stroke or mouse click. It is best to annotate such suggestions with their rationale, e.g., “the patient is due for his pneumonia vaccine because he has emphysema and is over 65,” so the provider understands the suggestion.

Figure 12.4a, b show the suggestions of a sophisticated inpatient decision support system from Intermountain Health Care that uses a wide range of clinical information to recommend antibiotic choice, dose, and duration of treatment. Decision support from the system improved clinical outcomes and reduced costs of infections among patients managed with the assistance of this system (Evans et al. 1998; Pestotnik 2005). Vanderbilt’s inpatient “WizOrder” order entry (CPOE) system also addresses antibiotic orders, as shown in Fig. 12.5; it suggests the use of Cefepine rather than ceftazidine, and provides choices of dosing by indication.

Clinical alerts attached to a laboratory test result can include suggestions for appropriate follow up or treatments for some abnormalities (Ozdas et al. 2008; Rosenbloom et al. 2005). Physician order-entry systems can warn the physician about allergies (Fig. 12.6a) and drug interactions (Fig. 12.6b) before they complete a medication order, as exemplified by screenshots from Partner’s outpatient medical record orders.

Reminders and alerts are employed widely in outpatient care. Indeed, the outpatient setting is where the first clinical reminder study was performed (McDonald 1976) and is still the setting for the majority of such studies (Garg et al. 2005). Reminders to physicians in outpatient settings quadrupled the use of certain vaccines in eligible patients compared with those who did not receive reminders (McDonald et al. 1984b; McPhee et al. 1991; Hunt et al. 1998; Teich et al. 2000). Reminder systems can also suggest needed tests and treatments for eligible patients. Figure 12.7 shows an Epic system screen with reminders to consider ordering a cardiac echocardiogram and starting an ACE inhibitor—in an outpatient patient with a diagnosis of heart failure but no record of a cardiac echocardiogram or treatment with one of the most beneficial drugs for heart failure.

Though the outpatient setting is the primary setting for preventive care reminders, preventive reminders also can be influential in the hospital (Dexter et al. 2001). And reminders directed to inpatient nurses can improve preventive care as much or more than reminders directed to physicians (Dexter et al. 2004).

3.4 Access to Knowledge Resources

Most clinical questions, whether addressed to a colleague or answered by searching through text books and published papers, are asked in the context of a specific patient (Covell et al. 1985). Thus, an appropriate time to offer knowledge resources to clinicians is while they are writing notes or orders for a specific patient. Clinicians have access to a rich selection of knowledge sources today, including those that are publically available, e.g. the National Library of Medicine’s (NLM) PubMed and MedlinePlus, the Centers for Disease Control and Prevention’s (CDC) vaccines and international travel information, the Agency for Healthcare Research and Quality’s (AHRQ) National Guideline Clearinghouse, and those produced by commercial vendors such as UpToDate, Micromedex, and electronic textbooks, all of which can be accessed from any web browser at any point in time. Some EHR systems are proactive and present short informational nuggets as a paragraph adjacent to the order item that the clinician has chosen. EHRs can also pull literature, textbook or other sources of information relevant to a particular clinical situation through an Infobutton and present that information to the clinician on the fly (Del Fiol et al. 2012), an approach being encouraged by the CMS MU2 regulations (see Fig. 12.8) (Final Rule: CMS 2012).

3.5 Integrated Communication and Reporting Support

Increasingly, the delivery of patient care requires multiple health care professionals and may cross many organizations; thus, the effectiveness, efficiency, and timeliness of communication among such team members and organizations are increasingly important. Such communications usually focus on a single patient and may require a care provider to read content from his or her local EHR or from an external clinical system or to send information from his system to an external system. Therefore, communication tools should be an integrated part of the EHR system.

Ideally providers’ offices, the hospital, and the emergency room should all be linked together—not a technical challenge with today’s Internet, but still an administrative challenge due to organizational barriers. Connectivity to the patient’s home will be increasingly important to patient-provider communication: for delivery of reminders directly to patients (Sherifali et al. 2011), and for home health monitoring, such as home blood pressure (Earle 2011; Green et al. 2008), and glucose monitoring. The patient’s personal health record (PHR) will also become an important destination for clinical messages and test results (see Chap. 17). Relevant information can be “pushed” to the user via e-mail or pager services (Major et al. 2002; Poon et al. 2002) or “pulled” by users on demand during their routine interactions with the computer.

EHR systems can also help with patient handoffs, during which the responsibility for care is transferred from one clinician to another. Typically the transferring clinician delivers a brief verbal or written turn-over note to help the receiving clinician understand the patient’s problems and treatments. Figure 12.9 shows an example of a screen that presents a “turn-over report” with instructions from the primary physician, as well as relevant system-provided information (e.g., recent laboratory test results) and a “to-do” list, that ensures that critical tasks are not dropped (Stein et al. 2010). Such applications support communication among team members and improve coordination.

Although a patient encounter is usually defined by a face-to-face visit (e.g., outpatient visit, inpatient bedside visit, home health visit), provider decision making also occurs during patient telephone calls, prescription renewal requests, and the arrival of new test results; so the clinician and key office personnel should be able to respond to these events with electronic renewal authorizations, patients’ reports about normal test results, and back-to-work forms as appropriate. In addition, when the provider schedules a diagnostic test such as a mammogram, an EHR system can keep track of the time since the order was written and can notify the physician that a test result has not appeared in a specified time so that the provider can investigate and correct the obstacle to fulfillment.

EHRs are usually bounded by the institution in which they reside. The National Health Information Infrastructure (NHII) (NCVHS, 2001) proposed a future in which a provider caring for a patient could reach beyond his or her local institution to automatically obtain patient information from any place that carried data about the patient (see Chap. 13). Today, examples of such community-based “EHRs,” often referred to as Health Information Exchanges (HIE), serve routine and emergency care, public health and/or other functions. A few examples of long-existing HIEs are those in: Indiana (McDonald et al. 2005), Ontario, Canada (electronic Child Health Network),^{Footnote 11} Kentucky (Kentucky Health Information Exchange),^{Footnote 12} and Memphis (Frisse et al. 2008).^{Footnote 13} A study from this last system showed that the extra patient information provided by this HIE reduced resource use and costs (Frisse et al. 2011). The New England Health care Exchange Network (NEHEN)^{Footnote 14} has created a community-wide collaborative system for managing eligibility, preauthorization, and claim status information (Fleurant et al. 2011).

The Office of the National Coordinator (ONC) has developed two communication tools to support the Nationwide Health Information Network (NwHIN)^{Footnote 15} and health data exchange (see Chaps. 13 and 27). NwHIN Connect^{Footnote 16} is an HHS project designed for pulling information from any site within a national network of health care systems. It offers a sophisticated consenting system by which patients can control who can use and see their information, but has only been used in a few pairs of communicating institutions. NwHIN Direct ^{Footnote 17} is a much simpler approach that uses standard Web Email, domain name system (DNS) and public-private keys to push patient reports as encrypted email messages from their source (e.g. laboratory system) to clinicians and hospitals. It could also be used to link individual care organizations to an HIE. Microsoft, among others, has implemented NwHIN Direct.

Communication tools that support timely and efficient communication between patients and the health care team can enhance coordination of care and disease management, and eHealth applications can provide patients with secure online access to their EHR and integrated communication tools to ask medical questions or conveniently perform other clinical (e.g., renew a prescription) or administrative tasks (e.g., schedule an appointment) (Tang 2003).

4 Fundamental Issues for Electronic Health Record Systems

All health record systems must serve the same functions, whether they are automated or manual. From a user’s perspective, the major difference is the way data are entered into, and delivered from, the record system. In this section, we explore the issues and alternatives related to data entry and then describe the options for displaying and retrieving information from an EHR.

4.1 Data Capture

EHRs use two general methods for data capture: (1) electronic interfaces from systems, such as laboratory systems that are already fully automated, and (2) direct manual data entry, when no such electronic source exists or it cannot be accessed.

4.1.1 Electronic Interfaces

The preferred method of capturing EHR data is to implement an electronic interface between the EHR and the existing electronic data sources such as laboratory systems, pharmacy systems, electronic instruments, home monitoring devices, registration systems, scheduling systems, etc.

The creation of interfaces requires effort to implement as described under Sect. 12.3.1, but, once implemented they provide near-instant availability of the clinical data without the labor costs and error potential of manual transcription. Interfacing is usually easier when the organization that owns the EHR system also owns, or is tightly affiliated with, the source system. Efforts to interface with systems outside the organizational boundary can be more difficult. However, interfaces between office practice systems and major referral laboratories for exchanging laboratory test orders and results, and between hospitals and office practices to pharmacies for e-prescribing, are now relatively easy and quite common.

The above discussion about interfacing concerns data produced, or ordered, by a home organization. However, much of the information about a patient will be produced or ordered by an outside organization and will not be available to a given organization via any of the conventional interfaces described above. For example, a hospital-based health care system will not automatically learn about pediatric immunizations done in private pediatric offices, or public health clinics, around town. So, special procedures and extra work are required to collect all relevant patient data. The promotion of health information exchange stimulated by passage of the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 (see Chap. 27) and other information exchange mechanisms (e.g. NwHIN Direct) described in Sect. 12.3.5 will facilitate the capture of such information from any source (see Chaps. 7 and 13).

4.1.2 Manual Data Entry

Data may be entered as narrative free-text, as codes, or as a combination of codes and free text annotation. Trade-offs exist between the use of codes and narrative text. The major advantage of coding is that it makes the data “understandable” to the computer and thus enables selective retrieval, clinical research, quality improvement, and clinical operations management. The coding of diagnoses, allergies, problems, orders, and medications is of special importance to these purposes; using a process called auto complete, clinicians can code such items by typing in a few letters of an item name, then choosing the item they need from the modest list of items that match the string they have entered. This process can be fast and efficient when the computer includes a full range of synonyms for the items of interest, and has frequency statistics for each item, so that it can present a short list of the most frequently occurring items that match the letters the user has typed so far.

Natural-language processing (NLP) (see Chap. 8) offers hope for automatic encoding of narrative text (Nadkarni et al. 2011). There are many types of NLP systems, but in general, such systems first regularize the input to recognize sections, sentences, and tokens like words or numbers. Through a formal grammar or a statistical technique, the tokens are then mapped to an internal representation of concepts (e.g., specific findings), their modifiers (e.g., whether a finding was asserted as being present or denied, and the timing of the finding), and their relations to other concepts. The internal representation is then mapped to a standard terminology and data model for use in a data warehouse or for automated decision support.

4.1.3 Physician-Entered Data

Physician-gathered patient information requires special comment because it presents the most difficult challenge to EHR system developers and operators. Physicians spend about 20 % of their time documenting the clinical encounter (Gottschalk and Flocke 2005; Hollingsworth et al. 1998). And the documentation burden has risen over time, because patient’s problems are more acute, care teams are larger, physicians order more tests and treatments, and billing regulatory bodies demand more documentation.

Many believe that clinicians themselves should enter all of this data directly into the EMR under the assumption that the person who collects the data should enter it. This tactic makes the most sense for prescriptions, orders, and perhaps diagnoses and procedure codes, whose immediate entry during the course of care will speed service to the patient and provide crucial grist for decision support. Direct entry by clinicians may not be as important for visit notes because the time cost of physician input is high and the information is not a pre-requisite to the check-out process.

Physicians’ notes can be entered into the EHR via one of three general mechanisms: (1) transcription of dictated or written notes, (2) clinic staff transfer or coding of some or all of the data by clinicians on a paper encounter form, and (3) direct data entry by physicians into the EHR (which may be facilitated by electronic templates or macros). Dictation with transcription is a common approach for entering narrative information into EHRs. If physicians dictate their reports using standard formats (e.g., present illness, past history, physical examinations, and treatment plan), the transcriptionist can maintain a degree of structure in the transcribed document via section headers, and the structure can also be delivered as an HL7 CDA document (Ferranti et al. 2006).

Some practices have employed scribes (a variant on the stenographers of old) to some of the physicians’ data entry work (Koshy et al. 2010), and CMS’s MU2 regulation (Final Rule: CMS 2012) allows credentialed medical assistants to take on this same work. Speech recognition software offers an approach to “dictating” without the cost or delay of transcription. The computer translates the clinician’s speech to text automatically. However, even with accuracy rates of 98 %, users may have to invest important amounts time to find and correct these errors.

Some dictation services use speech recognition to generate a draft transcription, which the transcriptionist corrects while listening to the audio dictation, thus saving transcriptionist time; others are exploring the use of natural language processing (NLP) to auto-encode transcribed text, and employ the transcriptionist to correct any NLP coding errors (see Chap. 8).

The second data-entry method is to have physicians record information on a structured encounter form, from which their notes are transcribed or possibly scanned (Downs et al. 2006; Hagen et al. 1998). One system (Carroll et al. 2011) uses paper turn-around documents to capture visit note data in one or more steps. First, the computer generates a child-specific data-capture form completed by the mother and the nursing staff. The computer scans the completed form (Fig. 12.10a), reads the hand-entered numeric data (top of form), check boxes (middle of form) and the bar codes (bottom of form), and stores them in the EHR. Next, the computer generates a physician encounter form that is also child-specific. The physician completes this form (Fig. 12.10b) and the computer processes it the same way it processed the nursing form.

The third alternative is the direct entry of data into the computer by care providers. This alternative has the advantage that the computer can immediately check the entry for consistency with previously stored information and can ask for additional detail or dimensions conditional on the information just entered. Some of this data will be entered into fields which require selection from pre-specified menus. For ease of entry, such menus should not be very long, require scrolling, or impose a rigid hierarchy (Kuhn et al. 1984). A major issue associated with direct physician entry is the physician time cost. Studies document that structured data entry consumes more clinician time than the traditional record keeping (Chaudhry et al. 2006), as much as 20s per SNOMED CT coded diagnoses (Fung et al. 2011)—which may be a function of the interface terminology used (or not used), and a small study suggests that the EHR functions taken together may consume up to 60 min of the physician’s free time per clinic day (McDonald and McDonald 2012). So, planners must be sensitive to these time costs. In one study, the computer system was a primary cause of clinician dissatisfaction (Edgar 2009) and their reason for leaving military medicine.

The use of templates and menus can speed note entry, but they can also generate excessive boilerplate and discourage specificity, i.e., it is easier to pick an available menu option than to describe a finding or event in detail. Further, with templates, the user may also accept default values too quickly so notes written via templates may not convey as clear a picture of the patient’s state as a note that is composed free-form by the physician and may contain inaccurate information.

Free-form narrative entry—by typing, dictation, or speech recognition—allows the clinician to express whatever they deem to be important. When clinicians communicate, they naturally prioritize findings and leave much information implicit. For example, an experienced clinician often leaves out “pertinent negatives” (i.e., findings that the patient does not have but that nevertheless inform the decision making process) knowing that the clinician who reads the record will interpret them properly to be absent. The result is usually a more concise history with a high signal-to-noise ratio that not only shortens the data capture time but also lessens the cognitive burden on the reading clinician. Weir and colleagues present compelling evidence about these advantages, especially when narrative is focused and vivid, and emphasize that too much information interferes with inter-provider communication (Weir et al. 2011).

Most EHRs let physicians cut and paste notes from previous visits and other sources. For example, a physician can cut and paste parts of a visit note into a letter to a referring physician and into an admission note, a most appropriate use of this capability. However, this cutting and pasting capability can be over-used and cause ‘note bloat.’ In addition, without proper attention to detail, some information may be copied that is no longer pertinent or true. In one study, 58 % of the text in the most recent visit notes duplicated the content of a previous note (Wrenn et al. 2010), although of course some repetition from note to note can be appropriate.

Tablets and smart phones provide new opportunities for data capture by clinical personnel including physicians. The University of Washington (Hartung et al. 2010) has developed a sophisticated suite of open source tools called the Open Data Interface (ODI) that includes form design and deployment to smart phones as well as delivery of captured data to a central resource. Data capture can be fast, and physicians and health care assistants in some third-world countries are using these tools eagerly. Figure 12.11 shows four screen shots from a medical record application of ODI. The first (Fig. 12.11a) is the patient selection screen. After choosing a patient, the user can view a summary of the patient’s medical record. Scrolling is usually required to view the whole summary. Figure 12.11b, c show screen shots of two portions of the summary. Users can choose to see the details of many kinds of information. Figure 12.11d shows the details of a laboratory test result. ODI ties into the OpenMRS project (Were et al. 2011), which has also been adopted widely in developing countries.

The long-term solution to data capture of information generated by clinicians is still evolving. The current ideal is the semi-structured data entry, which combines the use of narrative text fields and formally structured fields that are amenable to natural language processing combined with structured data entry fields where needed. With time and better input devices, direct computer entry will become faster and easier. In addition, direct entry of some data by patients will reduce the clinician’s data entry (Janamanchi et al. 2009).

4.1.4 What to Do About Data Recorded on Paper Before the Installation of the EHR

Care organizations have used a number of approaches to load new EHR systems with pre-existing patient data. One approach is to interface the EHR to available electronic sources—such as a dictation service, pharmacy systems, and laboratory information systems—and load data from these sources for 6–12 months before going live with the EHR. A second approach is to abstract selected data, e.g., key laboratory results, the problem lists, and active medications from the paper record and hand enter those data into the EHR prior to each patient’s visit when the EHR is first installed. The third approach is to scan and store 1–2 years of the old paper records. This approach does solve the availability problems of the paper chart, and can be applied to any kind of document, including handwritten records, produced prior to the EHR installation. Remember that these old records will have to be labeled with the patient ID, date information, and, preferably, the type of content (e.g., laboratory test, radiology report, provider dictation, and discharge summary, or, even better, a precise name, such as chest x-ray or operative note) and this step requires human effort. Optical Character Recognition (OCR) capability is built into most document scanners today, and converts typed text within scanned documents to computer understandable text with 98–99 % character accuracy.

4.1.5 Data Validation

Because of the chance of transcription errors with the hand entry of data, EHR systems must apply validity checks scrupulously. A number of different kinds of checks apply to clinical data (Schwartz et al. 1985). Range checks can detect or prevent entry of values that are out of range (e.g., a serum potassium level of 50.0 mmol/L—the normal range for healthy individuals is 3.5–5.0 mol/L). The computer can ask the users to verify results beyond the absolute range. Pattern checks can verify that the entered data have a required pattern (e.g., the three digits, hyphen, and four digits of a local telephone number). Computed checks can verify that values have the correct mathematical relationship (e.g., white blood cell differential counts, reported as percentages, must sum to 100). Consistency checks can detect errors by comparing entered data (e.g., the recording of cancer of the prostate as the diagnosis for a female patient). Delta checks warn of large and unlikely differences between the values of a new result and of the previous observations (e.g., a recorded weight that changes by 100 lb in 2 weeks). Spelling checks verify the spelling of individual words.

4.2 Data Display

Once stored in the computer, data can be presented in numerous formats for different purposes without further entry work. In addition, computer-stored records can be produced in novel formats that are unavailable in manual systems.

Increasingly, EHRs are implemented on web browser technology because of the ease of deployment to any PC or smart device (including smart phone and tablets; see Chap. 14) so health care workers (e.g., physicians on call) can view patient data off-site. Advanced web security features such as Transport Layer Security (TLS) (NIST 2005)—a revised designation for Secure Sockets Layer (SSL)—can ensure the confidentiality of any such data transmitted over the Internet.

Here, we discuss a few helpful formats. Clinicians need more than just integrated access to patient data; they also need various views of these data: in chronologic order as flowsheets or graphs to highlight changes over time, and as snapshots that show a computer view of the patients’ current status and their most important observations.

4.2.1 Timeline Graphs

A graphical presentation can help the physician to assimilate the information quickly and draw conclusions (Fafchamps et al. 1991; Tang and Patel 1994; Starren and Johnson 2000). An anesthesia system vendor provides an especially good example of the use of numbers and graphics in a timeline to convey the patient’s state in form that can be digested at a glance (Vigoda and Lubarsky 2006). Sparklines—“small, high resolution graphics embedded in a context of words, numbers, images” (Tufte 2006), which today’s browsers and spreadsheets can easily generate—provide a way to embed graphic timelines into any report. One study found that with sparklines, “physicians were able to assess laboratory data faster … enable more information to be presented in a single view (and more compactly) and thus reduce the need to scroll or flip between screens” (Bauer et al. 2010). The second column of the flowsheet in Fig. 12.12a displays sparklines that include all of the data points for a given variable. The yellow band associated with those sparklines highlights the reference range. Clicking on one or more sparklines produces a pop-up that displays a standard graph for all of the selected variables. The user can expand the timeline of this graph to spread out points that are packed too closely together as shown in Fig. 12.12b.

4.2.2 Timeline Flowsheets

Figure 12.13a shows an integrated view of a flowsheet of the radiology impressions with the rows representing different kinds of radiology examinations and the columns representing study dates. Clicking on the radiology image icon brings up the radiology images, e.g., the quarter resolution chest X-ray views in Fig. 12.13b. An analogous process applies to electrocardiogram (ECG) measurements where clicking on the ECG icon for a particular result brings up the full ECG tracing in Portable Document Format (PDF) form. Figure 12.14 shows the popular pocket rounds report that provides laboratory and nursing measurements as a very compact flowsheet that fits in a white coat pocket (Simonaitis et al. 2006).

Flowsheets can be specialized to carry information required to manage a particular problem. A flowsheet used to monitor patients who have hypertension (high blood pressure) for example might contain values for weight, blood pressure, heart rate, and doses of medications that control hypertension as well as results of laboratory tests that monitor complications of hypertension, or the medications used to treat it. Systems often permit users to adjust the time granularity of flowsheets on the fly. An ICU user might view results at minute-by-minute intervals, and an out-patient physician might view them with a month-by-month granularity.

4.2.3 Summaries and Snapshots

EHRs can highlight important components (e.g., active allergies, active problems, active treatments, and recent observations) in clinical summaries or snapshots (Tang et al. 1999b). Figure 12.15 from Epic’s product shows an example that presents the active patient problems, active medications, medication allergies, health maintenance reminders, and other relevant summary information. These views are updated automatically with any new data entry so they are always current. In the future, we can expect more sophisticated summarizing strategies, such as automated detection of adverse events (Bates et al. 2003b) or automated time-series events (e.g., cancer chemotherapy cycles). We may also see reports that distinguish abnormal changes that have been explained or treated from those that have not, and displays that dynamically organize the supporting evidence for existing problems (Tang and Patel 1994; Tang et al. 1994a). Ultimately, computers should be able to produce concise and flowing summary reports that are like an experienced physician’s hospital discharge summary.

4.2.4 Dynamic Search

Anyone who has reviewed a patient’s chart knows how hard it can be to find a particular piece of information. From 10 % (Fries 1974) to 81 % (Tang et al. 1994b) of the time, physicians do not find patient information that has been previously recorded in a paper medical record. Furthermore, the questions clinicians routinely ask are often the ones that are difficult to answer from perusal of a paper-based record. Common questions include whether a specific test has ever been performed, what kinds of medications have been tried, and how the patient has responded to particular treatments (e.g., a class of medications) in the past. Physicians constantly ask these questions as they flip back and forth in the chart searching for the facts to support or refute one in a series of evolving hypotheses. Search tools (see Sect. 12.4.3) help the physician to locate relevant data. The EHR can then display these data as specialized presentation formats (e.g., flowsheets or graphics) to make it easier for them to draw conclusions from the data. A graphical presentation can help the physician to assimilate the information quickly and to draw conclusions (Fafchamps et al. 1991; Tang et al. 1994a; Starren and Johnson 2000).

4.3 Query and Surveillance Systems

The query and surveillance capabilities of computer-stored records have no counterpart in manual systems. Medical personnel, quality and patient safety professionals, and administrators can use these capabilities to analyze patient outcomes and practice patterns. Public health professionals can use the reporting functions of computer-stored records for surveillance, looking for emergence of new diseases or other health threats that warrant medical attention.

Although these functions of decision support on the one hand, and query surveillance systems, on the other, are different, their internal logic is similar. In both, the central procedure is to find records of patients that satisfy pre-specified criteria and export selected data when the patient meets those criteria. Surveillance queries generally address a large subset, or all, of a patient population; the output is often a tabular report of selected raw data on all the patient records retrieved or a statistical summary of the values contained in the records. Decision support generally addresses only those patients under active care; its output is an alert or reminder message (McDonald 1976). Query and surveillance systems can be used for clinical care, clinical research, retrospective studies, and administration.

4.3.1 Clinical Care

A query can also identify patients who are due for periodic screening examinations such as immunizations, mammograms, and cervical Pap tests and can be used to generate letters to patients or call lists for office staff to encourage the preventive care. Query systems are particularly useful for conducting ad hoc searches such as those required to identify and notify patients who have been receiving a recalled drug. Such systems can also facilitate quality management and patient safety activities. They can identify candidate patients for concurrent review and can gather many of the data required to complete such audits.

4.3.2 Clinical Research

Query systems can be used to identify patients who meet eligibility requirements for prospective clinical trials. For example, an investigator could identify all patients seen in a medical clinic who have a specific diagnosis and meet eligibility requirements while not having any exclusionary conditions. These approaches can also be applied in real time. At one institution, the physician’s work station was programmed to ask permission to invite the patient into a study, when that physician entered a problem that suggested the patient might be a candidate for a local clinical trial. If the physician gave permission, the computer would send an electronic page to the nurse recruiter who would then invite the patient to participate in the study. It was first applied to a study of back pain (Damush et al. 2002).

4.3.3 Quality Reporting

Query systems can also play an important role in producing quality reports that are used for both internal quality improvement activities and for external public reporting. And, although it would be difficult for paper-based records to incorporate patient-generated input, and would require careful tagging of data source, an EHR could include data contributed by patients (e.g., functional status, pain scores, symptom reports). These patient-reported data may be incorporated in future quality measures. With the changing reimbursement payment models focusing more on outcomes measures instead of volume of transactions, generating efficient and timely reports of clinical quality measures will play an increasingly important role in management and payment.

4.3.4 Retrospective Studies

Randomized prospective studies are the gold standard for clinical investigations, but retrospective studies of existing data have contributed much to medical progress (See Chap. 11). Retrospective studies can obtain answers at a small fraction of the time and cost of comparable prospective studies.

EHR systems can provide many of the data required for a retrospective study. They can, for example, identify study cases and comparable control cases, and provide data needed for statistical analysis of the comparison cases (Brownstein et al. 2007). Combined with access to discarded specimens, they also offer powerful approaches to retrospective genome association studies that can be accomplished much faster and at cost magnitudes lower than comparable prospective studies (Kohane 2011; Roden et al. 2008).

Computer-stored records do not eliminate all the work required to complete an epidemiologic study; chart reviews and patient interviews may still be necessary. Computer-stored records are likely to be most complete and accurate with respect to drugs administered, laboratory test results, and visit diagnoses, especially if the first two types of data are entered directly from automated laboratory and pharmacy systems. Consequently, computer-stored records are most likely to contribute to research on a physician’s practice patterns, on the efficacy of tests and treatments, and on the toxicity of drugs. However, NLP techniques make the content of narrative text more accessible to automatic searches (see Chap. 8).

4.3.5 Administration

In the past, administrators had to rely on data from billing systems to understand practice patterns and resource utilization. However, claims data can be unreliable for understanding clinical practice because the source data are coarse and often entered by non-clinical personnel not directly involved with the care decisions. Furthermore, relying on claims data as proxies for clinical diagnoses can produce inaccurate information and lead to inappropriate policymaking (Tang et al. 2007). Medical query systems in conjunction with administrative systems can provide information about the relationships among diagnoses, indices of severity of illness, and resource consumption. Thus, query systems are important tools for administrators who wish to make informed decisions in the increasingly cost-sensitive world of health care. On the other hand, the use of EHR data for billing and administration can produce incentives for clinicians to steer their documentation to optimize payment and resource allocation, potentially making that documentation less clinically accurate. It may therefore be best to base financial decisions on variables that are not open to interpretation.

5 Challenges Ahead

Although many commercial products are labeled as EHR systems, they do not all satisfy the criteria that we defined at the beginning of this chapter. Even beyond matters of definition, however, it is important to recognize that the concept of an EHR is neither unified nor static. As the capability of technology evolves, the function of the EHR will expand. Greater involvement of patients in their own care, for example, means that personal health records (PHRs) should incorporate data captured at home and also support two-way communication between patients and their health care team (see also Chap. 17). The potential for patient-entered data includes history, symptoms, and outcomes entered by patients as well as data uploaded automatically by home monitoring devices such as scales, blood pressure monitors, glucose meters, and pulmonary function devices. By integrating these patient-generated data into the EHR, either by uploading the data into the EHR or by linking the EHR and the PHR, a number of long-term objectives can be achieved: patient-generated data may in some circumstances be more accurate or complete, the time spent entering data during an office visit by both the provider and the patient may be reduced, and the information may allow the production of outcomes measures that are better attuned to patients' goals. One caveat in this vision is the perception that this may lead to a deluge of data that the provider will never have time to sort through yet will be legally responsible for. A review of current products would be obsolete by the time that it was published. We have included examples from various systems in this chapter, both developed by their users and commercially available, to illustrate a portion of the functionality of EHR systems currently in use.

The future of EHR systems depends on both technical and nontechnical considerations. Hardware technology will continue to advance, with processing power doubling every 2 years according to Moore’s law (see Chap. 1). Software will improve with more powerful applications, better user interfaces, and more integrated decision support. New kinds of software that support collaboration will continue to improve; social media are growing rapidly both inside and outside of health care. For example, as both providers and patients engage increasingly in social media, new ways to capture data, share data, collaborate, and share expertise may emerge. Perhaps the greater need for leadership and action will be in the social and organizational foundations that must be laid if EHRs are to serve as the information infrastructure for health care. We touch briefly on these challenges in this final section.

5.1 Users’ Information Needs

We discussed the importance of clinicians directly using the EHR system to achieve maximum benefit from computer-supported decision making. On the one hand, organizations that require providers to enter all of their order, notes, and data directly into the EHR will gain substantial operational efficiency. On the other hand, physicians will bear the time costs of entering this information and may lose efficiency. Some balance between the organization’s and providers’ interests must be found. This balance is easiest to reach when physicians have a strong say in the decision.

Developers of EHR systems must thoroughly understand clinicians’ information needs and workflows in the various settings where health care is delivered. The most successful systems have been developed either by clinicians or through close collaborations with practicing clinicians.

Studies of clinicians’ information needs reveal that common questions that physicians ask concerning patient information (e.g., Is there evidence to support a specific patient diagnosis? Has a patient ever had a specific test? Has there been any follow up because of a particular laboratory test result?) are difficult to answer from the perusal of the paper-based chart (Tang et al. 1994b)). Regrettably, most clinical systems in use now cannot easily answer many of the common questions that clinicians ask. Developers of EHR systems must have a thorough grasp of users’ needs and workflows if they are to produce systems that help health care providers to use these tools efficiently to deliver care effectively.

5.2 Usability

An intuitive and efficient user interface is an important part of the system. Designers must understand the cognitive aspects of the human and computer interaction and the professional workflow if they are to build interfaces that are easy-to-learn and easy-to-use (see Chap. 4). Improving human–computer interfaces will require changes not only in how the system behaves but also in how humans interact with the system. User interface requirements of clinicians entering patient data are different from the user interfaces developed for clerks entering patient charges. Usability for clinicians means fast computer response times, and the fewest possible data input fields. A system that is slow or requires too much input is not usable by clinicians. The menus and vocabularies that constrain user input must include synonyms for all the ways health professionals name the items in the vocabularies and menus, and the system must have keyboard options for all inputs and actions because switching from mouse to keyboard steals user time. To facilitate use by busy health care professionals, health care applications developers must focus on clinicians’ unique information needs. What information the provider needs and what tasks the provider performs should influence what and how information is presented. Development of human-interface technology that matches the data-processing power of computers with the cognitive capability of human beings to formulate insightful questions and to interpret data is still a rate-limiting step (Tang and Patel 1994). For example, one can imagine an interface in which speech input, typed narrative, and mouse-based structured data entry are accepted and seamlessly stored into a single data structure within the EHR, with a hybrid user display that shows both a narrative version of the information and a structured version of the same information that highlights missing fields or inconsistent values.

5.3 Standards

We alluded to the importance of standards earlier in this chapter, when we discussed the architectural requirements of integrating data from multiple sources. Standards are the focus of Chap. 7. Here, we stress the critical importance of national standards in the development, implementation, and use of EHR systems (Miller and Gardner 1997b). Health information should follow patients as they interact with different providers in different care settings. Uniform standards are essential for systems to interoperate and exchange data in meaningful ways. Having standards reduces development costs, increases integration, and facilitates the collection of meaningful aggregate data for quality improvement and health policy development. The HIPAA legislation has mandated standards for administrative messages, privacy, security, and clinical data. Regulations based on this legislation have already been promulgated for the first three of these categories.^{Footnote 18} Incentives provided by the HITECH Act (see Chaps. 7 and 27) stimulated a number of efforts including a report by the ONC HIT Standards Committee (Health IT Standards Committee 2011) and Meaningful Use 2 (MU2) federal regulations (Final Rules: CMS 2012; Final Rule: ONC 2012) defining message and vocabulary standards for clinical data and encouraging EHR vendors and users to adopt them (see Sect. 12.3.1).^{Footnote 19} The US Department of Health and Human Services (HHS) maintains the current status of its HITECH programs on their Web site.^{Footnote 20}

5.4 Privacy and Security

Privacy and security policies and technology that protect individually identifiable health data are important foundational considerations for all applications that store and transmit and display health data. HIPAA established key regulations, and HITECH enhanced them, to protect the confidentiality of individually identifiable health information. With appropriate laws and policies computer-stored data can be more secure and confidential than those data maintained in paper-based records (Barrows and Clayton 1996).

5.5 Costs and Benefits

The Institute of Medicine (IOM) declared the EHR an essential infrastructure for the delivery of health care, and the protection of patient safety (IOM Committee on Improving the Patient Record 2001). Like any infrastructure project, the benefits specifically attributable to infrastructure are difficult to establish; an infrastructure plays an enabling role in all projects that take advantage of it. Early randomized controlled clinical studies showed that computer-based decision-support systems reduce costs and improve quality compared with usual care supported with a paper medical record (Tierney et al. 1993; Bates et al. 1997, 2003b; Classen et al. 1997), and recent meta-analyses of health information technology have demonstrated quality benefits (Buntin et al. 2011; Lau et al. 2010); however, Romano and Stafford (2011) did not find any “consistent association between EHRs and CDS and better quality.”

Because of the significant resources needed and the significant broad-based potential benefits, the decision to implement an EHR system is a strategic one. Hence, the evaluation of the costs and benefits must consider the effects on the organization’s strategic goals, as well as the objectives for individual health care (Samantaray et al. 2011). Recently, the federal government and professional organizations have both expressed interest in Open Source options for EHR software (Valdes 2008).

5.6 Leadership

Leaders from all segments of the health care industry must work together to articulate the needs, to define the standards, to fund the development, to implement the social change, and to write the laws to accelerate the development and routine use of EHR systems in health care. Because of the prominent role of the federal government in health care—as a payer, provider, policymaker, and regulator—federal leadership to create incentives for developing and adopting standards and for promoting the implementation and use of EHRs is crucial. Recently, Congress and the administration have acted to accelerate the adoption and meaningful use of health information technology based on some of the foundational research done in the informatics community (see Chap. 27). Technological change will continue to occur at a rapid pace, driven by consumer demand for entertainment, games, and business tools. Nurturing the use of information technology in health care requires leaders who promote the use of EHR systems and work to overcome the obstacles that impede widespread use of computers for the benefit of health care.

Questions for Discussion

1.
What is the definition of an EHR? What, then, is an EHR system? What are five advantages of an EHR over a paper-based record? Name three limitations of an EHR.
2.
What are the five functional components of an EHR? Think of the information systems used in health care institutions in which you work or that you have seen. Which of the components that you named do those systems have? Which are missing? How do the missing elements limit the value to the clinicians or patients?
3.
Discuss three ways in which a computer system can facilitate information transfer between hospitals and ambulatory care facilities, thus enhancing continuity of care for previously hospitalized patients who have been discharged and are now being followed up by their primary physicians.
4.
Much of medical care today is practiced in teams, and coordinating the care delivered by teams is a major challenge. Thinking in terms of the EHR functional components, describe four ways that EHRs can facilitate care coordination. Describe two ways in which EHRs are likely to create additional challenges in care coordination.
5.
How does the health care financing environment affect the use, costs, and benefits of an EHR system? How has the financing environment affected the functionality of information systems? How has it affected the user population?
6.
Would a computer scan of a paper-based record be an EHR? What are two advantages and two limitations of this approach?
7.
Among the key issues for designing an EHR system are what information should be captured and how it should be entered into the system. Physicians may enter data directly or may record data on a paper worksheet (encounter form) for later transcription by a data-entry worker. What are two advantages and two disadvantages of each method? Discuss the relative advantages and disadvantages of entry of free text instead of entry of fully coded information. Describe an intermediate or compromise method.
8.
EHR data may be used in clinical research, quality improvement, and monitoring the health of populations. Describe three ways that the design of the EHR system may affect how the data may be used for other purposes.
9.
Identify four locations where clinicians need access to the information contained in an EHR. What are the major costs or risks of providing access from each of these locations?
10.
What are three important reasons to have physicians enter orders directly into an EHR system? What are three challenges in implementing such a system?
11.
Consider the task of creating a summary report for clinical data collected over time and stored in an EHR system. Clinical laboratories traditionally provide summary test results in flowsheet format, thus highlighting clinically important changes over time. A medical record system that contains information for patients who have chronic diseases must present serial clinical observations, history information, and medications, as well as laboratory test results. Suggest a suitable format for presenting the information collected during a series of ambulatory-care patient visits.
12.
The public demands that the confidentiality of patient data must be maintained in any patient record system. Describe three protections and auditing methods that can be applied to paper-based systems. Describe three technical and three nontechnical measures you would like to see applied to ensure the confidentiality of patient data in an EHR. How do the risks of privacy breaches differ for the two systems?

5.6 Suggested Readings

Barnett, G. O. (1984). The application of computer-based medical-record systems in ambulatory practice. New England Journal of Medicine, 310(25), 1643–1650. This seminal article compares the characteristics of manual and automated ambulatory patient record systems, discusses implementation issues, and predicts future developments in technology.

Bates, D. W., Kuperman, G. J., Wang, S., et al. (2003). Ten commandments for effective clinical decision support: Making the practice of evidence-based medicine a reality. Journal of the American Medical Informatics Association, 10(6), 523–530. The authors present ten very practical tips to designers and installers of clinical decision support systems.

Berner, E. S. (Ed.). (2010). Clinical decision support systems, theory and practice: Health informatics series (3rd ed.). New York: Springer. This text focuses on the design, evaluation, and application of Clinical Decision Support systems, and examines the impact of computer-based diagnostic tools both from the practitioner’s and the patient’s perspectives. It is designed for informatics specialists, teachers or students in health informatics, and clinicians.

Collen, M. F. (1995). A history of medical informatics in the United States, 1950–1990. Indianapolis: American Medical Informatics Association, Hartman Publishing. This rich history of medical informatics from the late 1960s to the late 1980s includes an extremely detailed set of references.

Gauld, R., & Goldfinch, S. (2006). Dangerous enthusiasms: E-government, computer failure and information system development. Dunedin: Otago University Press. Gauld and Goldfinch describe a number of large-scale information and communications technology (ICT) projects with an emphasis on health information systems, emphasizes the high failure rates of mega projects that assume they can create a design denovo, build from the design and deploy successfully. It also highlights the advantages of starting with more modest scopes and growing incrementally based on experience with the initial scope.

Institute of Medicine (IOM) Roundtable on Value and Science-Driven Health Care. (2011). Digital infrastructure for the learning health system: The foundation for continuous improvement in health and health care – workshop series summary. Washington, DC: National Academy Press. This report summarizes three workshops that presented new approaches to the construction of advanced medical record system that would gather the crucial data needed to improve the health care system.

Kuperman, G. J., Gardner, R. M., & Pryor, T. A. (1991). The HELP system. Berlin/Heidelberg: Springer-Verlag GmbH and Co. K. The HELP (Health Evaluation through Logical Processing) system was a computerized hospital information system developed by the authors at the LDS Hospital at the University of Utah, USA. It provided clinical, hospital administration and financial services through the use of a modular, integrated design. This book thoroughly documents the HELP system. Chapters discuss the use of the HELP system in intensive care units, the use of APACHE and APACHE II on the HELP system, various clinical applications and inactive or experimental HELP system modules. Although the HELP system has now been retired from routine use, it remains an important example of several key issues in EHR implementation and use that continue in the commercial systems of today.

Osheroff, J., Teich, J., Levick, D., et al. (2012). Improving outcomes with clinical decision support: An implementers guide (2nd ed.). Scottsdale: Scottsdale Institute, AMIA, AMDIS and SHM. This text provides guidance on using clinical decision support interventions to improve care delivery and outcomes in a hospital, health system or physician practice. The book also presents considerations for health IT software suppliers to effectively support their CDS implementer clients.

Walker, J. M., Bieber, E. J., & Richards, F. (2006). Implementing an electronic health record system. London: Springer. This book provides rich details, including the process plans, for implementing an EHR in a large provider setting. It is a great resource for anyone trying to learn about EHR deployments, covering topics related to preparation, support, and implementation.

Weed, L. L. (1969). Medical records, medical evaluation and patient care: The problem-oriented record as a basic tool. Chicago: Year Book Medical Publishers. In this classic book, Weed presents his plan for collecting and structuring patient data to produce a problem-oriented medical record.

Notes

1.
Digital Imaging and Communications in Medicine, http://dicom.nema.org/ (Accessed 1/2/2013).
2.
JPEG from Wikipedia, the free encyclopedia, http://en.wikipedia.org/wiki/JPEG (Accessed 1/2/2013).
3.
Logical Observation Identifiers Names and Codes (LOINC®). http://loinc.org/ (Accessed 1/2/2013).
4.
SNOMED Clinical Terms® (SNOMED CT®). http://www.ihtsdo.org/snomed-ct/ (Accessed 1/2/2013).
5.
The Unified Code for Units of Measure. http://unitsofmeasure.org/ (Accessed 1/2/2013).
6.
RxNorm Overview. http://www.nlm.nih.gov/research/umls/rxnorm/overview.html (Accessed 1/2/2013).
7.
RxTerms. https://wwwcf.nlm.nih.gov/umlslicense/rxtermApp/rxTerm.cfm (Accessed 1/2/2013).
8.
Health Level Seven International, http://www.hl7.org/ (Accessed 1/2/2013).
9.
Mirth Corporation Community Overview. http://www.mirthcorp.com/community/overview. (Accessed 1/2/2013).
10.
Surescripts: The Nation’s e-Prescription Network http://www.surescripts.com/ (Accessed 1/2/2013).
11.
eCHN electronic Child Health Network. http://www.echn.ca/ (Accessed 1/2/2013).
12.
Kentucky Health Information Exchange Frequently Asked Questions. http://khie.ky.gov/Pages/faq.aspx?fc=010 (Accessed 1/2/2013).
13.
MidSoutheHealth Alliance. http://www.midsoutheha.org (Accessed 1/2/2013).
14.
New England Health care Exchange Network (NEHEN). www.nehen.net (Accessed 1/2/2013).
15.
http://www.healthit.gov/policy-researchers-implementers/nationwide-health-information-network-nwhin (Accessed 1/3/2013).
16.
http://www.healthit.gov/policy-researchers-implementers/connect-gateway-nationwide-health-information-network (Accessed 1/3/2013).
17.
Office of the National Coordinator for Health Information Technology. Direct Project http://directproject.org/ (Accessed 1/2/2013).
18.
http://www.hhs.gov/ocr/privacy/hipaa/administrative/index.html (Accessed 1/2/2012).
19.
http://www.healthit.gov/sites/default/files/standards-certification/HITSC_CQMWG_VTF_Transmit_090911.pdf (Accessed 1/3/2012).
20.
http://www.healthit.gov/policy-researchers-implementers/health-it-rules-regulations (Accessed 1/3/2012).

References

Barnett, G. O. (1984). The application of computer-based medical-record systems in ambulatory practice. The New England Journal of Medicine, 310(25), 1643–1650.
PubMed CAS Google Scholar
Barnett, G. O., Winickoff, R., Dorsey, J. L., Morgan, M. M., & Lurie, R. S. (1978). Quality assurance through automated monitoring and concurrent feedback using a computer-based medical information system. Medical Care, 16(11), 962–970.
PubMed CAS Google Scholar
Barrows, R. C., Jr., & Clayton, P. D. (1996). Privacy, confidentiality, and electronic medical records. Journal of the American Medical Informatics Association: JAMIA, 3(2), 139–148.
PubMed Google Scholar
Bates, D. W., Spell, N., Cullen, D. J., et al. (1997). The costs of adverse drug events in hospitalized patients. Journal of the American Medical Association, 277(4), 307–311.
PubMed CAS Google Scholar
Bates, D. W., Ebell, M., Gotlieb, E., Zapp, J., & Mullins, H. C. (2003a). A proposal for electronic medical records in U.S. primary care. Journal of the American Medical Informatics Association: JAMIA, 10(1), 1–10.
PubMed Google Scholar
Bates, D. W., Evans, R. S., Murff, H., et al. (2003b). Detecting adverse events using information technology. Journal of the American Medical Informatics Association: JAMIA, 10(2), 115–128.
PubMed Google Scholar
Bauer, D. T., Guerlain, S., & Brown, P. J. (2010). The design and evaluation of a graphical display for laboratory data. Journal of the American Medical Informatics Association: JAMIA, 17(4), 416–424.
PubMed Google Scholar
‘Brain’ to store medical data. (1956, October 24). The New York Times.
Google Scholar
Brown, S. H., Lincoln, M. J., Groen, P. J., et al. (2003). VistA–U.S. Department of veterans affairs national-scale HIS. International Journal of Medical Informatics, 69(2–3), 135–156.
PubMed Google Scholar
Brownstein, J. S., Sordo, M., Kohane, I. S., & Mandl, K. D. (2007). The tell-tale heart: Population-based surveillance reveals an associaton f rofecoxib and celecoxib with myocardial infarction. PLoS One, 2(9), e840.
PubMed Google Scholar
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 (Millwood), 30(3), 464–471.
Google Scholar
Carroll, A. E., Biondich, P. G., Anand, V., Dugan, T. M., Sheley, M. E., Xu, S. Z., & Downs, S. M. (2011). Targeted screening for pediatric conditions with the CHICA system. Journal of the American Medical Informatics Association: JAMIA, 18(4), 485–490.
PubMed Google Scholar
Chaudhry, B., Wang, J., Wu, S., Maglione, M., Mojica, W., Roth, E., Morton, S. C., et al. (2006). Systematic review: Impact of health information technology on quality, efficiency, and costs of medical care. Annals of Internal Medicine, 144(10), 742–752.
PubMed Google Scholar
Cheung, N. T., Fung, K. W., Wong, K. C., et al. (2001). Medical informatics – the state of the art in the hospital authority. International Journal of Medical Informatics, 62, 113–119.
PubMed CAS Google Scholar
Classen, D. C., Pestotnik, S. L., Evans, R. S., Lloyd, J. F., & Burke, J. P. (1997). Adverse drug events in hospitalized patients. Excess length of stay, extra costs, and attributable mortality. Journal of the American Medical Association, 277(4), 301–306.
PubMed CAS Google Scholar
Coffey, R. (1979). How medical information systems affect costs: the El Camino experience. Washington, D.C.: Public Health Service. Publication DHEW-PHS-80-3626.
Google Scholar
Collen, M. F. (1969). Value of multiphasic health checkups. The New England Journal of Medicine, 280(19), 1072–1073.
PubMed CAS Google Scholar
Collen, M. F. (1995). A history of medical informatics in the United States, 1950 to 1990. Bethesda: American Medical Informatics Association.
Google Scholar
Covell, D., Uman, G., & Manning, P. (1985). Information needs in office practice: are they being met? Annals of Internal Medicine, 103, 596–599.
PubMed CAS Google Scholar
Damiani, G., Pinnarelli, L., Colosimo, S. C., Almiento, R., Sicuro, L., Galasso, R., Sommello, L., & Ricciardi, W. (2010). The effectiveness of computerized clinical guidelines in the process of care: A systematic review. BMC Health Services Research, 10, 2.
PubMed Google Scholar
Damush, T. M., Weinberger, M., Clark, D. O., Tierney, W. M., Rao, J. K., Perkins, S. M., & Verel, K. (2002). Acute low back pain self management intervention for urban primary care patients: Rationale, design, and predictors of participation. Arthritis and Rheumatism, 47, 372–379.
PubMed Google Scholar
Davis, L. S., Collen, M. F., Rubin, F. L., & Van Brunt, E. E. (1968). Computer stored medical record. Computer Biomedical Research, 1, 452–469.
CAS Google Scholar
Del Fiol, G., Huser, V., Strasberg, H. R., et al. (2012). Implementations of the HL7 context-aware knowledge retrieval (“infobutton”) standard: Challenges, strengths, limitations, and uptake. Journal of Biomedical Informatics, 45(4), 726–735.
PubMed Google Scholar
Dexter, P. R., Perkins, S., Overhage, J. M., et al. (2001). A computerized reminder system to increase the use of preventive care for hospitalized patients. The New England Journal of Medicine, 345(13), 965–970.
PubMed CAS Google Scholar
Dexter, P. R., Perkins, S. M., Maharry, K. S., Jones, K., & McDonald, C. J. (2004). Inpatient computer-based standing orders vs. physician reminders to increase influenza and pneumococcal vaccination rates. Journal of the American Medical Association, 292(19), 2366–2371.
PubMed CAS Google Scholar
Downs, S. M., Biondich, P. G., Anand, V., Zore, M., & Carroll, A. E. (2006). Using Arden syntax and adaptive turnaround documents to evaluate clinical guidelines. AMIA Annual Symposium Proceedings, 2006, 214–218.
Google Scholar
Duncan, R. G., Saperia, D., Dulbandzhyan, R., et al. (2001). Integrated web-based viewing and secure remote access to a clinical data repository and diverse clinical systems. Proceedings of the AMIA Fall Symposium, 2001, 149–153.
Google Scholar
Earle, K. (2011). In people with poorly controlled hypertension, self-management including telemonitoring is more effective than usual care for reducing systolic blood pressure at 6 and 12 months. Evidence-Based Medicine, 16(1), 17–18.
PubMed Google Scholar
Edgar, E.P. (2009). Physician retention in Army Medical Department. Strategic Research Project, U.S. Army War College, Carlisle Barracks PA 17013–5050. Available at: http://handle.dtic.mil/100.2/ADA499087
Evans, R. S., Pestotnik, S. L., Classen, D. C., Clemmer, T. P., Weaver, L. K., Orme, J. F., Lloyd, J. F., & Burke, J. P. (1998). A computer-assisted management program for antibiotics and other antinfective agents. The New England Journal of Medicine, 338(4), 232–238.
PubMed CAS Google Scholar
Fafchamps, D., Young, C. Y., & Tang, P. C. (1991). Modelling work practices: Input to the design of a physician’s workstation. In Proceedings of the 15th annual symposium on Computer Applications in Medical Care (pp. 788–792). Washington, D.C.: CAMC.
Google Scholar
Ferranti, J. M., Musser, R. C., Kawamoto, K., & Hammond, W. E. (2006). The clinical document architecture and the continuity of care record: A critical analysis. Journal of the American Medical Informatics Association: JAMIA, 13(3), 245–252.
PubMed Google Scholar
Final Rule: Centers for Medicare and Medicaid Services (CMS), HHS. Medicare and Medicaid Programs; Electronic Health Record Incentive Program—Stage 2, 77(171) Fed. Reg. 53968 (Sept. 4, 2012) (amending 42 C.F.R. § 412, 413, and 495).
Google Scholar
Final Rule: Office of the National Coordinator (ONC) for Health Information Technology, Department of Health and Human Services. Health Information Technology: Standards, Implementation Specifications, and Certification Criteria for Electronic Health Record Technology, 2014 Edition; Revisions to the Permanent Certification Program for Health Information Technology, 77(171) Fed. Reg. 54163 (Sept. 4, 2012) (amending 45 C.F.R. § 170).
Google Scholar
Fitzmaurice, J. M., Adams, K., & Eisenberg, J. M. (2002). Three decades of research on computer applications in health care: Medical informatics support at the agency for healthcare research and quality. Journal of the American Medical Informatics Association: JAMIA, 9(2), 144–160.
PubMed Google Scholar
Fleurant, M., Kell, R., Love, J., Jenter, C., Volk, L. A., Zhang, F., Bates, D. W., & Simon, S. R. (2011). Massachusetts e-Health Project increased physicians’ ability to use registries, and signals progress toward better care. Health Affairs (Millwood), 30(7), 1256–1264.
Google Scholar
Flexner, A. (1910). Medical education in the united states and Canada: A report to the Carnegie foundation for the advancement of teaching. New York: The Carnegie Foundation for the Advancement of Teaching. OCLC 9795002.
Google Scholar
Fries, J. F. (1974). Alternatives in medical record formats. Medical Care, 12(10), 871–881.
PubMed CAS Google Scholar
Frisse, M. E., King, J. K., Rice, W. B., et al. (2008). A regional health information exchange: Architecture and implementation. American Medical Informatics Association Annual Symposium Proceedings, 2008, 212–216.
Google Scholar
Frisse, M. E., Johnson, K., Nian, H., et al. (2011). The financial impact of health information exchange on emergency department care. Journal of the American Medical Informatics Association: JAMIA, 19(3), 328–333.
PubMed Google Scholar
Fung, K. W., Xu, J., Rosenbloom, S. T., et al. (2011). Testing three problem list terminologies in a simulated data entry environment. American Medical Informatics Association Annual Symposium Proceeds, 2011, 445–454.
Google Scholar
Garg, A. X., Adhikari, N. K. J., McDonald, H., et al. (2005). Effects of computerized clinical decision support systems on practitioner performance and patient outcomes: A systematic review. Journal of the American Medical Association, 293(10), 1223–1238.
PubMed CAS Google Scholar
Giuse, D. A., & Mickish, A. (1996). Increasing the availability of the computerized patient record. Proceedings of the AMIA Annual Fall Symposium, 1996, 633–637.
Google Scholar
Goroll, A. H., Simon, S. R., Tripathi, M., Ascenzo, C., & Bates, D. W. (2009). Community-wide implementation of health information technology: The Massachusetts eHealth collaborative experience. Journal of the American Medical Informatics Association: JAMIA, 16(1), 132–139.
PubMed Google Scholar
Gottschalk, A., & Flocke, S. A. (2005). Time spent in face-to-face patient care and work outside the examination room. Annals of Family Medicine, 3(6), 488–493.
PubMed Google Scholar
Green, B. B., Cook, A. J., Ralston, J. D., et al. (2008). Effectiveness of home blood pressure monitoring, web communication, and pharmacist care on hypertension control: A randomized controlled trial. Journal of the American Medical Association, 299(24), 2857–2867.
PubMed CAS Google Scholar
Hagen, P. T., Turner, D., Daniels, L., & Joyce, D. (1998). Very large-scale distributed scanning solution for automated entry of patient information. TEPR Proceedings (Toward an ElectronicPatient Record), 1, 228–232.
Google Scholar
Halamka, J. D., & Safran, C. (1998). CareWeb: A web-based medical record for an integrated healthcare delivery system. Proceedings of Medinfo 1998, Part 1, 36–39.
Google Scholar
Hartung, C., Anokwa, Y., Brunette, W., Lerer, A., Tseng, C., & Borriello, G. (2010). Open data kit: Tools to build information services for developing regions. International Conference on Information and Communication Technologies and Development (ICTD2010) Proceedings. http://www.gg.rhul.ac.uk/ict4d/ictd2010/papers/ICTD2010%20Hartung%20et%20al.pdf. Accessed 2 Jan 2012.
Haynes, R.B. (Ed.) (2011). Computerized clinical decision support systems: how effective are they? [article collection] Implementation Science, 6, 87–108. Articles available at: http://www.implementationscience.com/series/CCDSS. Accessed 2 Jan 2012.
Health Information Technology for Economic and Clinical Health (HITECH) Act, Title XIII of Division A and Title IV of Division B of the American Recovery and Reinvestment Act of 2009 (ARRA).(2009).
Google Scholar
Health IT Standards Committee. (2011). Recommendations to ONC on the assignment of code sets to clinical concepts [data elements] for use in quality measures. [Letter]. Retrieved from http://www.healthit.gov/sites/default/files/standards-certification/HITSC_CQMWG_VTF_Transmit_090911.pdf. Accessed 3 Jan 2012.
Hollingsworth, J. C., Chisholm, C. D., Giles, B. K., Cordell, W. H., & Nelson, D. R. (1998). How do physicians and nurses spend their time in the emergency department? Annals of Emergency Medicine, 31, 87–91.
PubMed CAS Google Scholar
Hripcsak, G., Cimino, J. J., & Sengupta, S. (1999). WebCIS: Large scale deployment of a web-based clinical information system. Proceedings of the Annual AMIA Symposium, 1999, 804–808.
Google Scholar
Hunt, D. L., Haynes, R. B., et al. (1998). Effects of computer-based clinical decision support systems on physician performance and patient outcomes: A systematic review. JAMA : The Journal of the American Medical Association, 280(15), 1339–1346.
CAS Google Scholar
Institute of Medicine Committee on Improving the Patient Record. (2001). Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, D.C.: National Academy Press. Available at: http://www.iom.edu/Reports/2001/Crossing-the-Quality-Chasm-A-New-Health-System-for-the-21st-Century.aspx. Accessed 2 Jan 2012.
Janamanchi, B., Katsamakas, E., Raghupathi, W., & Gao, W. (2009). The state and profile of open source software projects in health and medical informatics. International Journal of Medical Informatics, 78(2009), 457–472.
PubMed Google Scholar
Johnson, S.B., Friedman, C., Cimino, J.J., Clark, T., Hripcsak, G., & Clayton, P.D. (1991). Conceptual data model for a central patient database. Proceedings of the Symposium on Computer Applications in Medical Care, 381–385.
Google Scholar
Khajouei, R., & Jaspers, M. W. (2010). The impact of CPOE medication systems’ design aspects on usability, workflow and medication orders: A systematic review. Methods of Information in Medicine, 49(1), 3–19.
PubMed CAS Google Scholar
Kohane, I. S. (2011). Using electronic health records to drive discovery in disease genomics. Nature Reviews Genetics, 12(6), 417–428.
PubMed CAS Google Scholar
Koshy, S., Feustel, P. J., Hong, M., & Kogan, B. A. (2010). Scribes in an ambulatory urology practice: Patient and physician satisfaction. Journal of Urology, 184(1), 258–262.
PubMed Google Scholar
Kuhn, I.M., Wiederhold, G., Rodnick, J.E., et al. (1984). Automated ambulatory medical record systems in the U.S. In Blum, B. (Ed.), Information Systems for Patient Care (pp.199–217), New York: Springer. Original 1982 report available at: http://infolab.stanford.edu/TR/CS-TR-82-928.html. Accessed 2 Jan 2012.
Lamb, A. (1955). The Presbyterian Hospital and the Columbia-Presbyterian Medical Center, 1868–1943: A history of a great medical adventure. New York: Columbia University Press.
Google Scholar
Larsen, R. A., Evans, R. S., Burke, J. P., Pestotnik, S. L., Gardner, R. M., & Classen, D. C. (1989). Improved perioperative antibiotic Use and reduced surgical wound infections through use of computer decision analysis. Infection Control and Hospital Epidemiology, 10(7), 316–320.
PubMed CAS Google Scholar
Lau, F., Kuziemsky, C., Price, M., & Gardner, J. (2010). A review on systematic reviews of health information system studies. Journal of the American Medical Informatics Association: JAMIA, 17(6), 637–645.
PubMed Google Scholar
Lindberg, D. (1967). Collection, evaluation, and transmission of hospital laboratory data. Methods of Information in Medicine, 6(3), 97–107.
PubMed CAS Google Scholar
Major, K., Shabot, M. M., & Cunneen, S. (2002). Wireless clinical alerts and patient outcomes in the surgical intensive care unit. American Surgery, 68, 1057–1060.
Google Scholar
McDonald, C. J. (1976). Protocol-based computer reminders, the quality of care and the nonperfectibility of man. The New England Journal of Medicine, 295(24), 1351–1355.
PubMed CAS Google Scholar
McDonald, M. H., & McDonald, C. J. (2012). Electronic medical records and preserving primary care physicians’ time: Comment on “electronic health record-based messages to primary care providers”. Archives of Internal Medicine, 172(3), 285–287.
PubMed Google Scholar
McDonald, C.J., Bhargava, B., Jeris, D.W. (1975). A clinical information system (CIS) for ambulatory care. Proceedings AFIPS National Computing Conference, Anaheim.
Google Scholar
McDonald, C.J., Wiederhold, G., Simborg, D., Hammond, W.E., Jelovsek, F., Schneider, K. (1984a). A discussion of the draft proposal for data exchange standards for clinical laboratory results. Proceedings of the 8th Annual Symposium on Computer Applications in Medical Care, 406–413.
Google Scholar
McDonald, C. J., Hui, S. L., Smith, D. M., et al. (1984b). Reminders to physicians from an introspective computer medical record. A two year randomized trial. Annals of Internal Medicine, 100(1), 130–138.
PubMed CAS Google Scholar
McDonald, C. J., Overhage, J. M., Tierney, W. M., et al. (1999). The regenstrief medical record system: A quarter century experience. International Journal of Medical Informatics, 54(3), 225–253.
PubMed CAS Google Scholar
McDonald, C. J., Huff, S. M., Suico, J. G., et al. (2003). LOINC, a universal standard for identifying laboratory observations: A 5-year update. Clinical Chemistry, 49(4), 624–633.
PubMed CAS Google Scholar
McDonald, C. J., Overhage, J. M., Barnes, M., Schadow, G., Blevins, L., Dexter, P. R., & Mamlin, B. W. (2005). The Indiana network for patient care: A working local health information infrastructure (LHII). Health Affairs (Millwood)., 24(5), 1214–1220.
PubMed Google Scholar
McPhee, S. J., Bird, J. A., Fordham, D., Rodnick, J. E., & Osborn, E. H. (1991). Promoting cancer prevention activities by primary care physicians: Results of a randomized, controlled trial. Journal of the American Medical Association, 266(4), 538–544.
PubMed CAS Google Scholar
Melton, L. J., 3rd. (1996). History of the Rochester epidemiology project. Mayo Clinic Proceedings, 71, 266–274.
PubMed Google Scholar
Menachemi, N., Powers, T. L., & Brooks, R. G. (2011). Physician and practice characteristics associated with longitudinal increases in electronic health records adoption. Journal of Healthcare Management, 56(3), 183–198.
PubMed Google Scholar
Miller, R. A., & Gardner, R. M. (1997b). Recommendations for responsible monitoring and regulation of clinical software systems. Journal of the American Medical Informatics Association: JAMIA, 4, 442–457.
PubMed CAS Google Scholar
Miller, R. A., Gardner, R. M., Johnson, K. B., & Hripcsak, G. (2005a). Clinical decision support and electronic prescribing systems: A time for responsible thought and action. Journal of the American Medical Informatics Association: JAMIA, 12(4), 403–409.
PubMed Google Scholar
Miller, R. A., Waitman, L. R., Chen, S., & Rosenbloom, S. T. (2005b). The anatomy of decision support during inpatient care provider order entry (CPOE): empirical observations from a decade of CPOE experience at Vanderbilt. Journal of Biomedical Informatics, 38(6), 469–485.
PubMed Google Scholar
Nadkarni, P. M., Ohno-Machado, L., & Chapman, W. W. (2011). Natural language processing: An introduction. Journal of the American Medical Informatics Association: JAMIA, 18(5), 544–551.
PubMed Google Scholar
National Committee on Vital and Health Statistics. (2001) Information for health: A strategy for building the National Health Information Infrastructure. Report and Recommendations From the National Committee on Vital and Health Statistics. Available at http://www.ncvhs.hhs.gov/nhiilayo.pdf. Accessed 17 Dec 2012.
National Institutes of Standards and Technology (NIST). (2005). Guidelines for the selection and use of transport layer security (TLS) implementations (NIST Special Publication 800–52). Gaithersburg: U.S. Department of Commerce. http://csrc.nist.gov/publications/nistpubs/800-52/SP800-52.pdf
Openchowski, M. W. (1925). The effect of the unit record system and improved organization on hospital economy and efficiency. Archives of Surgery, 10(3), 925–934.
Google Scholar
Pestotnik, S. L. (2005). Expert clinical decision support systems to enhance antimicrobial stewardship programs: Insights from the Society of Infectious Diseases Pharmacists. Pharmacotherapy, 25(8), 1116–1125.
PubMed Google Scholar
Phansalkar, S., Desai, A. A., Bell, D., et al. (2012a). High-priority drug-drug interactions for use in electronic health records. Journal of the American Medical Informatics Association: JAMIA, 19(5), 735–743.
PubMed Google Scholar
Phansalkar, S., van der Sijs, H., Tucker, A. D., et al. (2012b). Drug-drug interactions that should be non-interruptive in order to reduce alert fatigue in electronic health records. Journal of the American Medical Informatics Association: JAMIA, 20(3), 489–493.
PubMed Google Scholar
Poon, E. G., Kuperman, G. J., Fiskio, J., & Bates, D. W. (2002). Real-time notification of laboratory data requested by users through alphanumeric pagers. Journal of the American Medical Informatics Association: JAMIA, 9, 217–222.
PubMed Google Scholar
Reiser, S. (1991). The clinical record in medicine. Part 1: Learning from cases. Annals of Internal Medicine, 114(10), 902–907.
PubMed CAS Google Scholar
Roden, D. M., Pulley, J. M., Basford, M. A., Bernard, G. R., Clayton, E. W., Balser, J. R., & Masys, D. R. (2008, September). Development of a large-scale de-identified DNA biobank to enable personalized medicine. Clinical Pharmacology and Therapeutics, 84(3), 362–369.
PubMed CAS Google Scholar
Romano, M. J., & Stafford, R. S. (2011). Electronic health records and clinical decision support systems: Impact on national ambulatory care quality. Archives of Internal Medicine, 171(10), 897–903.
PubMed Google Scholar
Rosenbloom, S. T., Geissbuhler, A. J., Dupont, W. D., Giuse, D. A., Talbert, D. A., Tierney, W. M., Plummer, W. D., Stead, W. W., & Miller, R. A. (2005). Effect of CPOE User Interface Design on User-Initiated Access to Educational and Patient Information during Clinical Care Journal of the American Medical Informatics Association: JAMIA, 12(4), 458–473.
PubMed Google Scholar
Samantaray, R., Njoku, V. O., Brunner, J. W., Raghavan, V., Kendall, M. L., & Shih, S. C. (2011). Promoting electronic health record adoption among small independent primary care practices. The American Journal of Managed Care, 17(5), 353–358.
PubMed Google Scholar
Schedlbauer, A., Prasad, V., Mulvaney, C., Phansalkar, S., Stanton, W., Bates, D. W., & Avery, A. J. (2009). What evidence supports the use of computerized alerts and prompts to improve clinicians’ prescribing behavior? Journal of the American Medical Informatics Association: JAMIA, 16(4), 531–538.
PubMed Google Scholar
Schultz, J. R., Cantrill, S. V., & Morgan, K. G. (1971). AFIPS Conference Proceedings, 38, 239–264.
Google Scholar
Schwartz, R. J., Weiss, K. M., & Buchanan, A. V. (1985). Error control in medical data. MD Computing, 2(2), 19–25.
PubMed CAS Google Scholar
Sherifali, D., Greb, J. L., Amirthavasar, G., Hunt, D., Haynes, R. B., Harper, W., Holbrook, A., Capes, S., Goeree, R., O’Reilly, D., Pullenayegum, E., & Gerstein, H. C. (2011). Effect of computer-generated tailored feedback on glycemic control in people with diabetes in the community: A randomized controlled trial. Diabetes Care, 34(8), 1794–1798.
PubMed Google Scholar
Simonaitis, L., Belsito, A., Warvel, J., Hui, S., & McDonald, C. J. (2006). Extensible stylesheet language formatting objects (XSL-FO): a tool to transform patient data into attractive clinical reports. AMIA Annual Symposium Proceedings, 2006, 719–723.
Google Scholar
Slack, W. V., & Bleich, H. L. (1999). The CCC system in two teaching hospitals: A progress report. International Journal of Medical Informatics, 54(3), 183–196.
PubMed CAS Google Scholar
Starren, J., & Johnson, S. B. (2000). An object-oriented taxonomy of medical data presentations. Journal of the American Medical Informatics Association: JAMIA, 7(1), 1–20.
PubMed CAS Google Scholar
Stead, W. W., & Hammond, W. E. (1988). Computer-based medical records: The centerpiece of TMR. MD Computing, 5(5), 48–62.
PubMed CAS Google Scholar
Stein, D.M., Vawdrey, D.K., Stetson, P.D., & Bakken, S. (2010). An analysis of team checklists in physician signout notes. American Medical Informatics Association Annual Symposium Proceedings, 767–771.
Google Scholar
Strom, B. L., Schinnar, R., Aberra, F., Bilker, W., Hennessy, S., Leonard, C. E., & Pifer, E. (2010). Unintended effects of a computerized physician order entry nearly hard-stop alert to prevent a drug interaction: A randomized controlled trial. Archives of Internal Medicine, 170(17), 1578–1583.
PubMed Google Scholar
Tang, P.C. (2003). Key Capabilities of an Electronic Health Record System (Letter Report). Committee on Data Standards for Patient Safety. Board on Health Care Services, Institute of Medicine.
Google Scholar
Tang, P. C., & Patel, V. L. (1994). Major issues in user interface design for health professional workstations: Summary and recommendations. International Journal of Biomedical Computing, 34(104), 130–148.
Google Scholar
Tang, P. C., Annevelink, J., Suermondt, H. J., & Young, C. Y. (1994a). Semantic integration in a physician’s workstation. International Journal of Bio-Medical Computing, 35(1), 47–60.
PubMed CAS Google Scholar
Tang, P.C., Fafchamps, D., Shortliffe, E.H. (1994b). Traditional medical records as a source of clinical data in the outpatient setting. Proceedings of the Annual Symposium Computer Applications in Medical Care, 575–579.
Google Scholar
Tang, P. C., LaRosa, M. P., & Gorden, S. M. (1999a). Use of computer-based records, completeness of documentation, and appropriateness of documented clinical decisions. Journal of the American Medical Informatics Association: JAMIA, 6(3), 245–251.
PubMed CAS Google Scholar
Tang, P. C., Marquardt, W. C., Boggs, B., et al. (1999b). NetReach: Building a clinical infrastructure for the enterprise. In J. M. Overhage (Ed.), Fourth annual proceedings of the Davies CPR recognition symposium (pp. 25–68). Chicago: McGraw-Hill.
Google Scholar
Tang, P. C., Ralston, M., Arrigotti, M. F., Qureshi, L., & Graham, J. (2007). Comparison of methodologies for calculating quality measures based on administrative data versus clinical data from an electronic health record system: Implications for performance measures. Journal of the American Medical Informatics Association: JAMIA, 14, 10–15.
PubMed CAS Google Scholar
Teich, J. M., Merchia, P. R., Schmiz, J. L., et al. (2000). Effects of computerized physician order entry on prescribing practices. Archives of Internal Medicine, 160(18), 2741–2747.
PubMed CAS Google Scholar
Tierney, W. M., Miller, M. E., Overhage, J. M., & McDonald, C. J. (1993). Physician inpatient order writing on microcomputer workstations: Effects on resource utilization. Journal of the American Medical Association, 269(3), 379–383.
PubMed CAS Google Scholar
Tufte, E. (2006). Beautiful evidence. Cheshire: Graphics Press. ISBN 978-0-9613921-7-8.
Google Scholar
Valdes, I. (2008). Free and open source software in healthcare 1.0. American Medical Informatics Association Open Source Working Group White Paper. Available at: http://www.scribd.com/doc/14109414/AMIA-Free-and-Open-Source-Software-in-Healthcare-10. Accessed 2 Jan 2012.
Vigoda, M. M., & Lubarsky, D. A. (2006). Failure to recognize loss of incoming data in an anesthesia record-keeping system may have increased medical liability. Anesthesia and Analgesia, 102, 1798–1802.
PubMed Google Scholar
Vreeman, D. J., McDonald, C. J., & Huff, S. M. (2010). Representing patient assessments in LOINC®. American Medical Informatics Association Annual Symposium Proceedings, 2010, 832–836.
Google Scholar
Warner, H. R. (1972). A computer-based patient information system for patient care. In G. A. Bekey & M. D. Schwartz (Eds.), Hospital information systems (pp. 293–332). New York: Marcel Dekker.
Google Scholar
Weir, C. R., Hammond, K. W., Embi, P. J., et al. (2011). An exploration of the impact of computerized patient documentation on clinical collaboration. International Journal of Medical Informatics, 80(8), e62–e71.
PubMed Google Scholar
Were, M. C., Shen, C., Tierney, W. M., Mamlin, J. J., Biondich, P. G., Li, X., Kimaiyo, S., & Mamlin, B. W. (2011). Evaluation of computer-generated reminders to improve CD4 laboratory monitoring in sub-Saharan Africa: A prospective comparative study. Journal of the American Medical Informatics Association: JAMIA, 18(2), 150–155.
PubMed Google Scholar
Whiting-O’Keefe, Q. E., Simborg, D. W., Epstein, W. V., & Warger, A. (1985). A computerized summary medical record system can provide more information than the standard medical record. Journal of the American Medical Association, 254(9), 1185–1192.
PubMed Google Scholar
Wrenn, J. O., Stein, D. M., Bakken, S., & Stetson, P. D. (2010). Quantifying clinical narrative redundancy in an electronic health record. Journal of the American Medical Informatics Association: JAMIA, 17(1), 49–53.
PubMed Google Scholar
Zhu, V. J., Overhage, M. J., Egg, J., Downs, S. M., & Grannis, S. J. (2009). An empiric modification to the probabilistic record linkage algorithm using frequency-based weight scaling. Journal of the American Medical Informatics Association: JAMIA, 16(5), 738–745.
PubMed Google Scholar
Barnett, G. O., Justice, N. S., Somand, M. E., et al. (1979). COSTAR – a computer-based medical information system for ambulatory care. Proceedings of the Institute of Electrical and Electronics Engineers (IEEE), 67, 1226–1237.
Google Scholar
Teich, J. M., Glaser, J. P., Beckley, R. F., et al. (1999). The Brigham integrated computing system (BICS): Advanced clinical systems in an academic hospital environment. International Journal of Medical Informatics, 54(3), 197–208.
PubMed CAS Google Scholar
Brown, S.H., Fischietti, L.F., Graham, G., Bates, J., et al. (2007). Use of electronic health records in disaster response: The experience of department of veterans affairs after hurricane katrina. American Journal of Public Health, 97, S136–S141.
PubMed Google Scholar
Ozdas, A., Miller, R. A., & Waitman, L. R. (2008). Care Provider Order Entry (CPOE): A perspective on factors leading to success or to failure. Yearbook of Medical Informatics, 2007, 128–137. Review. Erratum in: Yearbook of Medical Informatics. 2008:19.
Google Scholar
Barnett, G. O., Justice, N. S., Somand, M. E., et al. (1979). COSTAR – a computer-based medical information system for ambulatory care. Proceedings of the Institute of Electrical and Electronics Engineers (IEEE), 67, 1226–1237.
Google Scholar

Download references

Author information

Authors and Affiliations

Lister Hill National Center for Biomedical Communications, National Library of Medicine, National Institutes of Health, 8600 Rockville Pike 38a 7n707, Bethesda, 20894, MD, USA
Clement J. McDonald MD (Office of the Director)
Internal Medicine, Palo Alto Medical Foundation, 2350 W. El Camino Real, Mountain View, 94040, CA, USA
Paul C. Tang MD, MS
Department of Biomedical Informatics, Columbia University Medical Center, 622 W 168th St, VC5, New York, 10032, NY, USA
George Hripcsak MD, MS

Authors

Clement J. McDonald MD
View author publications
You can also search for this author in PubMed Google Scholar
Paul C. Tang MD, MS
View author publications
You can also search for this author in PubMed Google Scholar
George Hripcsak MD, MS
View author publications
You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Clement J. McDonald MD .

Editor information

Editors and Affiliations

Departments of Biomedical Informatics at Columbia University and Arizona State University, New York, New York, USA
Edward H. Shortliffe
Bethesda, Maryland, USA
James J. Cimino

Rights and permissions

Reprints and permissions

Copyright information

About this chapter

Cite this chapter

McDonald, C.J., Tang, P.C., Hripcsak, G. (2014). Electronic Health Record Systems. In: Shortliffe, E., Cimino, J. (eds) Biomedical Informatics. Springer, London. https://doi.org/10.1007/978-1-4471-4474-8_12

Download citation

DOI: https://doi.org/10.1007/978-1-4471-4474-8_12
Published: 07 November 2013
Publisher Name: Springer, London
Print ISBN: 978-1-4471-4473-1
Online ISBN: 978-1-4471-4474-8
eBook Packages: MedicineMedicine (R0)

Publish with us

Policies and ethics