Electronic Medical Records: Are Physicians Ready?

EXECUTIVE SUMMARY

The use of electronic medical records (EMR) in healthcare organizations will require substantial changes in the way physicians and their staff provide patient care. This study is the first part of a larger study assessing factors that influence successful implementation of EMR in ambulatory care settings. The purposes of this study were to identify specific attitudes or factors that should be targeted before implementating an EMR project, and demonstrate empirical support for a model of perceived usefulness of EMR. We found that computer experience, computer anxiety, and perceptions of organizational support predict the degree to which physicians and mid-level practitioners view the EMR effort positively. Strategies for the successful management of EMR implementation include engaging the physicians and practitioners in computer activities prior to implementation and providing strong organizational support before and during the redesign effort.

Acceptance of EMR by physicians and their support staff is essential if computerization is to be successful, yet anecdotal reports of resistance and negative attitudes are frequently reported. Empirical studies indicate that physicians have not yet embraced this technology. As part of strategic planning and deployment of a computerized patient record, attitudes of end-users must be assessed. Using an integrative framework from the job design literature and management information sciences, we propose that multiple factors influence attitudes toward EMR, offer a conceptual model of end-user acceptance, and present findings from an empirical test of our model.

COMPUTERS IN HEALTHCARE

Computerization in ambulatory healthcare is generally classified into four types of applications: (1) electronic medical records (EMR); (2) office or practice management; (3) medical literature integration; and (4) telecommunications. Although any of these can be used as stand-alone products, computerization in healthcare organizations is most effective when all applications are integrated, with the EMR forming the "hub" of the system in which all clinical, administrative, and financial data move electronically among users. The focus of this study is EMR.

With EMR, all elements of the patient's record exist in a computer file. In most contemporary software packages, different elements of the chart are accessible through graphic user interface methods that allow the user to move easily from one part of the record to another. In addition to having the ability to enter text, the user may navigate through the record by pointing to icons on the screen to generate progress notes and flowsheets, complete insurance forms, prescribe medications, or perform countless other clinical and administrative tasks.

An extension of EMR is the decision support system, which assists physicians in medical decision making. Computerized systems are designed to perform case analysis in the traditional sense by producing a differential diagnosis given a rich set of inputs and generating a therapeutic course of action (Middleton, Detmer, and Musen 1995). Diagnostic decision support systems integrate clinical findings (e.g., signs, symptoms, and test results) with disease profiles to produce probability-based pairings of findings and diseases, while therapeutic decision support permits the physician to generate a patient-specific, disease-specific treatment plan, including medication prescriptions, patient education, and diet therapy.

The Push Toward Electronic Medical Records

The increasing use of computers in healthcare has largely been the result of two forces: cost containment and quality improvement. Cost-containment efforts by employers and managed care organizations have led to increased productivity requirements. At the same time, physicians are being asked to improve the quality of their work, document their results, and benchmark against their peers. Measurement of patient satisfaction, compliance with health maintenance practices, and physician practice profiles are examples of current documentation requirements (Edelson 1995).

The EMR is viewed by many as a panacea for contemporary management problems. Although research on electronic records in ambulatory settings is limited, empirical studies of pharmaceutical practices have found that the EMR is effective at identifying drug interactions (Haumschild et al. 1987), generating physician pharmaceutical reminders (Rind et al. 1991), and monitoring the use of antibiotics in hospitals (Evans and Pestotnik 1994). EMR studies have also evaluated the use of electronic records to enhance preventive care. In a study of 49 physicians, EMR-generated reminders improved preventive services such as cholesterol measurement, mammography, and immunizations (Ornstein et al. 1991). A study that compared computer-based health maintenance tracking with a manual flowchart-based tracking system found that the former resulted in significantly higher provider compliance for 8 of 11 procedures (Frame et al. 1994).

In the current environment of integrated delivery systems, the connectivity attribute of the EMR enhances its value further. Marshall and Chin (1998) conducted a study of 497 clinicians in a large northwest HMO to investigate perceived benefits of the EMR. They reported that 82 percent of the respondents felt that the EMR improved their ability to care for patients with other providers.

In sum, a properly configured EMR system enables the routine generation and collection of data, provides decision support, and facilitates workflow. Productivity is gained from increased quality of work and fewer information handling errors, with numerous studies indicating cost savings as a result of the EMR (Hamdy et al. 1995; Tierney et al. 1993). Despite its apparent benefits, however, acceptance of the EMR is far from universal.

END-USER ACCEPTANCE

Little doubt exists that EMR will change the way physicians and other clinicians work. Before beginning such a significant job redesign, managers of EMR projects would be wise to assess the readiness of the principal end-users--that is, the physicians.

The job design literature suggests that characteristics of the job (Hackman and Oldham 1976) as well as cues in the work environment (Salancik and Pfeffer 1978) influence an individual's responses to the job redesign effort. Oldham and Hackman (1980) argue that …