The effect of physician order entry and computerized nursing documentation on how emergency department providers spend their time

Abstract

Study objectives: We obtain objective measures of the change in time spent on various tasks by providers in an emergency department (ED) before and after major changes in the ED information system. Nursing documentation, computerized physician order entry, and results retrieval functionality were added to the existing ED information system in a large urban academic ED. Before this implementation, the information system was used primarily as an electronic tracking board. Laboratory and radiology orders were entered by nurses by a separate computerized ordering system, and laboratory test results were retrieved by a query in the system used for ordering or by querying a third system, an enterprise-wide clinical data repository. With this implementation, physicians now enter all orders directly into the ED information system via desktop computers. Results now return to the ED information system, triggering a change in the color of icons on the tracking board to notify users of the availability of results. Before the implementation, nurses entered some triage information into the ED information system, but most of their documentation was performed on paper. With the implementation, essentially all nursing documentation is performed in the ED information system via desktop computers in work areas and in patient rooms. Methods: Time-motion studies were performed using handheld computers with specially designed data collection software. The data collection software allows simultaneous timing of 2 “activities” to allow assignment of time to 2 task categories (eg, using the computer while speaking to a consultant by telephone). Nurse and physician providers were observed during approximately 4-hour observation sessions. The time spent in various tasks during user sessions was aggregated in various ways to allow comparison between the preimplementation sessions and the postdocumentation sessions. Means of the aggregated times were calculated for the pre- and postprovider groups, and 95% confidence intervals (CIs) were calculated. Results: During the preimplementation phase, 35 provider sessions were observed (7 attending physicians, 6 residents, 5 physician's assistants/nurse practitioners [PA/NP], 17 nurses). During the postimplementation phase, 34 provider sessions were observed (6 attending physicians, 5 residents, 6 PA/NP, 17 nurses). Physician-type providers (attending physicians, residents, PA/NPs) spent 15.4% of their time, with 95% CIs of 11.8% to 19.0% on computers in the preimplementation sessions (Pre) compared with 26.0% (CI 21.8% to 30.1%) in the postimplementation phase (Post). On paper documentation, they spent 17.9% (95% CI 13.7% to 22.1%) of their time Pre and 17.9% (95% CI 14.3% to 21.5%) Post. The computer was often used in conjunction with paper documentation. Adjusting for the simultaneous use of computer and paper, time spent in one or both was 32.5% (95% CI 27.2% to 37.8%) Pre and 41.7% (95% CI 35.9% to 47.5%) Post. Time spent in direct patient care was 28.8% (95% CI 22.7% to 34.9%) Pre and 26.9% (95% CI 23.6% to 30.3%) Post. The percentage of time spent by nursing providers using a computer was 10.1% (95% CI 7.8% to 12.5%) Pre and 26.0% (95% CI 23.2% to 28.9%) Post. Time spent on paper documentation was 17.1% (95% CI 14.4% to 19.9%) Pre and 1.7% (95% CI 0.8% to 2.6%) Post. Combining these, time spent either on a computer or paper documentation was 26.8% (95% CI 22.6% to 30.9%) Pre and 27.6% (95% CI 24.4% to 30.9%) Post. Time spent in direct patient care was 41.0% (95% CI 34.9% to 46.7%) Pre and 39.2% (95% CI 33.9% to 44.5%) Post. Time spent documenting while in the patient room (either on paper or on computer) was 0.2% (95% CI 0.0% to 0.4%) Pre and 2.1% (95% CI 0.5% to 3.6%) Post. Conclusion: Physician and nurse providers experienced significant changes in the way they spent their time with the implementation of physician order entry and computerized nursing documentation. The overall impact on the percentage of nursing time spent documenting was not significant, although there was a significant shift from time spent with paper to time spent with the computer. Physicians also experienced a significant increase in the time they spend on computers, but the time spent on paper or computers did not change significantly. The percentage of time spent on direct patient care did not change significantly for either nurses or physicians.