The Effect of Electronic Medical Records on Nurses’ Job Satisfaction: A Multi-Year Analysis

To measure nurses’ rankings of their electronic medical records (EMRs) on their job satisfaction over time, a retrospective analysis of a set of cross sectional data from a survey conducted by the United States’ California Registered Nursing Board in 2008, 2010, 2012, 2014, and 2016. Approximately 4,500 nurses ranked the usefulness of their EMRs in each of the five years. The EMR rankings increased steadily between 2008 and 2016 but the changes are small and the rates of change are very slow, suggesting that the problems with EMRs have been difficult to solve. The results show EMRs have a large impact: a one category increase in EMR rankings increased job satisfaction by as much as or slightly more than one-third for hospital and non-hospital nurses. The size of the effects and their persistence over eight years imply a substantial loss from poorly designed EMRs, and one which could have been avoided had EMR designs more closely matched nurses’ day to day work. The reductions in job satisfaction and potential effects on burnout are losses to be added to the more widely measured losses in productivity and negative effects of EMRs on patient-provider relationships.


Introduction
The history of electronic medical records demonstrates the error of changing work environments without adequately compensating for the impact of the changes on the end users. Electronic Medical Records (EMRs) might, in concept, have simply been substitutes for paper records but they substantially changed the working environments of nurses and other health care professionals. The introduction of EMRs complicated nurses' workloads, interfered with patient communication, increased work-arounds, reduced productivity, and increased patient complications (Taylor & Bice, 2019;Bristol et al., 2018;Topaz et al., 2017;Walker-Czyz, 2016;Ratwani et al., 2015;Gardner & Sparnon, 2014;Furukawa et al., 2010;DesRoches et al., 2008).
Nurses' difficulties with EMRs have lessened over time as nurses adapted to EMR use and newly graduated nurses benefited from training in the use of EMRs during their schooling. Some of the changes over time can be glimpsed from repeated surveys of nurses.
The Black Book Market Research Organization conducts an ongoing national, proprietary survey of U.S. hospital nurses (Kent, 2018). The 2014 survey reported widespread dissatisfaction with EMRs that, at worst, caused some nurses to leave hospitals for other jobs (Millard, 2014). The 2018 results, however, show improvement. Disruptions in productivity from EMR use were cited by 69% of nurses in 2018 compared to 84% in 2016 (Black Book Complimentary Results, 2018). The percent of nurses reporting that EMRs interfered with workflow fell to 44% from 85% and obstacles to patient-provider relationships were cited by 80% compared to 90% 2016 (Black Book Complimentary Results, 2018).
Despite the signs of improvement and several years of experience with EMRs, change has been slow and significant problems remain. One would expect, therefore, that EMRs might have reduced nurses' job satisfaction over several years. High levels of job satisfaction are a key defense against the problems of attrition and burnout that beset the nursing profession (Kovner et al., 2014;Gilmartin, 2013;Irvine & Evans, 1995), but there is relatively little information on the effects of EMRs on job satisfaction (Lu et al., 2019Shin et al., 2021Moy et al., 2021).
One exception is a study of 371 Advanced Practice Registered Nurses, that found that EMR users were more than twice as likely to burnout as non-users (Harris et al., 2018). The Harris et al. results suggest that although EMRs are but one element in nurses' work flow, they have a large effect on burnout and job satisfaction.
During recessions, newly graduated nurses have difficulty finding jobs so the recession nursing workforce is somewhat older than during non-recessionary periods (Johnson et al., 2016). Older nurses in 2008-2009, however, were unlikely to have received formal training in the use of EMRs and nurses returning from non-nursing employments likely required some retraining in EMR use.
Our measures of the effects of EMRs on job satisfaction may, therefore, be influenced by the recession/recovery changes in the nursing workforce. Potential differences among the years are partially controlled by the inclusion of sociodemographic characteristics in the multivariate models that we estimate.

Method
In 2008 EMRs. We also test whether the impact of EMRs on job satisfaction differs between hospital and non-hospital settings.

Results
Hospital nurses' rankings of EMR usability increased from 2008-2016 (Tables 1a). The largest improvement occurred in the all systems work well category (18% in 2016 vs 10% in 2008) and in reductions in the two lowest categories. The most striking feature of the trend in rankings, however, is the very slow rate of improvement.
Remembering that EMRs were in place for approximately 97% of hospital nurses in 2008, the additional eight years of experience only increased the mean ranking of EMR usability from 2.7 to 2.9.
In words, the rank is systems problems affect my work, closely approaching systems generally helpful but with some flaws. The mean level of job satisfaction increased from 4.10 in 2008 to 4.14 in 2016, varying slightly among the individual years. Most of the change occurred in shifts from the dissatisfied and very dissatisfied categories to the satisfied group.
The trends for non-hospital nurses are similar but more muted (Table 1b) The interpretation of the descriptive data must recognize that the survey data are not drawn from a panel. Instead, each sample year is a different draw from a population of nurses whose composition changes over time. Thus, differences in rankings among the years are affected in unknown ways by differences in the nurses who responded to the survey.   Vol. 4, No. 3, 2021 Hospital nurses are younger and have shorter job tenures than non-hospital nurses. Although the absolute values differ, both groups have much shorter job tenures in 2008 than in subsequent years.
Among hospital nurses, job tenure is 9.7 years in (2008) (Table 2a). Since approximately 33.7 percent of EMR users were very satisfied in 2008 (Table 1a), the likelihood of nurses being very satisfied would increase to 45 percent (33.7+11.3). In 2010 and 2012, being very satisfied would increase from approximately 37% to more than 51% and from approximately 33% to slightly more than 47% in 2014 and 2016.  Vol. 4, No. 3, 2021 Note. all models include only those nurses with an EMR system (and hence, an EMR rating).
Significance levels: ***-significant at 1 percent level, **-significance at 5 percent level, *-significance at 10 percent level. "[ ]" marginal response as discussed in text: change in in the likelihood of going to the highest level of job satisfaction (from 4 to 5) for a unit increase in EMR functionality. Note. all models include only those nurses with an EMR system (and hence, an EMR rating).
Significance levels: ***-significant at 1 percent level, **-significance at 5 percent level, *-significance at 10 percent level. "[ ]" marginal response as discussed in text: change in in the likelihood of going to the highest level of job satisfaction (from 4 to 5) for a unit increase in EMR functionality.
Non-hospital nurses are generally more satisfied with their jobs than hospital nurses but increased EMR rankings would substantially increase their levels of satisfaction as well (  Vol. 4, No. 3, 2021 other States in many respects including mandated nurse to bed ratios in hospitals. Inferences from the data must be conditioned on these limits.

Conclusion
The nurses' evaluations of EMRs over the eight years covered by the survey trace the pattern of adaptation to the changes in nurses' work environment that were induced by the introduction of EMRs.
The EMR rankings increased steadily between 2008 and 2016 but the changes are small and the rates of change are very slow suggesting that the problems with EMRs have been difficult to solve. Recent national surveys for 2018 confirm the persistence of significant problems faced by nurses using EMRs.
What then, during this long and as yet incomplete period of adaptation, has been the effect on the job satisfaction of nurses?
The results show that EMRs have a large impact on nurses' job satisfaction even though EMRs are but one element in the nursing workplace. A one category increase in EMR rankings would have increased job satisfaction by as much as or slightly more than one-third for hospital and non-hospital nurses, varying among the years. Variations in the composition of the sample prohibit summarizing the potential losses in job satisfaction but the size of the effects and their persistence over eight years imply a substantial loss and one which could have been avoided had EMR designs more closely matched nurses' day to day work. There is evidence that the losses in job satisfaction from EMR use induced burnout and there is a voluminous literature on losses of productivity and other problems related to EMRs. The example is EMRs but the lesson is more general. Introducing new technologies into healthcare without careful consideration of how the changes affect the workplace environment guarantees a long, costly and inefficient process of adaptation.