favorable attitudes with respect to system efficiency (i.e. the time taken to use system versus handwritten prescriptions) also were higher system utilizers. Not surprisingly, physicians that reported feeling comfortable using the system had written more electronic prescriptions and, conversely, those preferring that the practice had stuck with hand-written prescriptions had written fewer. Interestingly, respondent utilization was not associated with their expressed satisfaction with ease of use or system capabilities since even high utilizers felt that the system should be easier to use with more features. Utilization was also not associated with the amount of prior computer experience or with clinical practice satisfaction. Furthermore, there was no association between respondent age, gender, or level of training and utilization.
4. Discussion
Our survey six months following implementation showed that system adoption and degree of utilization varied substantially among physicians in this practice. Furthermore, the variation in actual use was strongly associated with physician attitude toward issues of system efficiency and effect on quality, but not with prior computer experience, level of training, or satisfaction with their primary care practice. Additionally, there was a strong association between self-reported technology adoption and measured technology usage in this physician group.
Several inpatient studies have reported that physician satisfaction was strongly correlated with perceived ease of use, efficiency and impact on quality of care. In a study of physician attitudes before and three months after use of a clinical information system, Brown found that the greatest negative attitudinal change was with respect to dashed expectations that computer use would be time-saving. In the ambulatory setting, Overhage et al. studied physician computer order entry and found that it required more time than paper-based entry, despite elimination of duplicative and administrative tasks. Nevertheless, physicians became more efficient with experience, believed that care was improved, and wanted to continue usage. On the other hand, Gadd and Penrod reported that physicians who preferred jettisoning an EMR outnumbered those who felt it was worth the time and effort when surveyed six months post-implementation. These physicians felt that the EMR had a detrimental effect on physician efficiency as well as patient privacy and rapport. However, in a separate and smaller post-EMR implementation survey, the same authors reported on two groups of physicans that would choose to keep the EMR despite similar usability and efficiency concerns. Wilson et al. Studied satisfaction with computerized order entry systems at an outpatient and an inpatient military facility and also reported a strong association with perceived ease of use and efficiency as well as impact on quality of care These study results indicate that a major reason for dissatisfaction has been the perception that computer order entry is less efficient in the short run than the paper-based method it replaced despite acknowledged longer term benefits. This finding may reflect implementation of less user friendly technology than is currently available as well as an unrealistic expectation that use of electronic records would be an immediate time-saver. In our study, most physicians still felt it was quicker to write a prescription by hand at the six month post-implementation survey but almost all felt the computer would be quicker eventually. Consistent with this, physicians with the most computer prescribing experience were more likely to favor the computer's efficiency (though were not generally more satisfied overall with system capabilities or ease of use). It clearly takes time for physicians to get ‘up to speed’ even with a fairly simple prescription expert system implementation, but it appears that physicians are willing to invest time to learn such a system if the perceived longer term benefits are sufficiently large.
Previous research has differed on the effect of physician computer background on electronic records or orders implementation projects. In surveys of physician attitudes not directly linked to an actual systems implementation, both Detmer and Friedman and Dansky found that respondent computer experience and sophistication was associated with more favorable attitudes toward medical computing. Similarly, Van der Meijden et al. Evaluated anticipatory attitudes toward an inpatient electronic records system prior to implementation and found that computer experience was associated with a favorable attitude among the study physicians. In hospital-based studies following computer systems implementations, both Gardner and Tierney found that computer experience or literacy did not correlate with physician acceptance of or attitude toward actual use of such systems. Brown studied the change in intern attitudes and anxiety with respect to use of an inpatient clinical information system. Although computer attitude and anxiety was significantly correlated with prior computer experience at baseline, the strength of the association substantially diminished following three months of system usage. In the present study, we found no significant association of prior computer experience and self-rated skills with physician use of or attitude toward the prescription expert system. Our results and review of the literature suggests that prior computer experience may diminish anticipatory anxiety and enhance predisposing attitudes toward computer systems, but that such experience does not significantly affect physician use of or response to clinical computing systems.
Most previous studies have involved system implementations where user choice was not an option, hence adoption was uniform and therefore unstudied. In a setting where use of an electronic order entry or record system is not required, it is possible to examine user characteristics associated with adoption. In their studies of outpatient decision support or guidelines systems, Hetlevik et al. Demonstrated that initial physician interest and training do not necessarily translate into system adoption. Both showed physician-reported enthusiasm for the computer-based interventions, but demonstrated little actual use in practice due to the exigencies of busy clinical schedules and the need for efficiency. In the system implementation described in our report, 86% of the physicians were actively utilizing the prescription expert system by the end of the six month observation period. The early adopters were in the minority and included those self-described as involved in the system implementation and oversight process as well as other technophiles (in the role of system ‘champions’ as described by Miller ). Subsequent utilizers bought into the system's impact on quality, felt that it did not impede efficiency, and generally thought it had a favorable impact on patient rapport. Interestingly, system utilizers were not more likely to feel it had all the capabilities they desired (in fact, they tended to be more critical of system capabilities than non-utilizers) and were not significantly more satisfied with its ease of use.
A significant finding of the current study is that physician self-reported use correlated well with actual measured system utilization. Findings were quite similar regardless of whether we used the measured or self-reported usage variable in the reported analyses. Since most studies of physician attitudes and computer usage rely on survey self-reported behavior only our results suggest that such self-reports may be accurate reflections of actual physician systems use.
There are several caveats that must be applied to the findings of this study. First, it refers to a single practice, also the site of an internal medicine training program. As such, the participants were mainly physicians in training whose adoption behaviors and attitudes may differ from those of physicians in other settings. Additionally, the survey was conducted at six months post-implementation only. Accordingly, we can not determine if the attitudes expressed were potentially causative or consequences of the computer adoption behavior that was reported and measured. Although we have drawn comparisons with findings from other technology implementation evaluations (including physician order entry systems and full EMRs), our study concerned a prescription expert system only. The complexity of such an implementation is lower and the physician attitudinal and behavioral response might differ from that toward other computer-based interventions.
In conclusion, physician adoption of a prescription expert system varied substantially and was associated with attitudes toward system efficiency and quality effects but not with physician computer experience. We believe that our results and those of prior studies of physician attitudes toward clinical computing systems suggest both practical strategies and pitfalls in the implementation process. These include taking care to manage expectations, especially with regard to the learning curve and time burden of any new system, rather than following a natural tendency to hype it and thereby raise expectations to unrealistic levels. Physicians in our and other reports that expected early efficiencies tended to be disappointed and frustrated. Instead, we believe that EMR implementation strategies and tactics should be framed within the context of clear quality of care improvements that will justify physician interest and effort. Belief that such outcomes would result from the prescription expert system were nearly universal among the physician subjects of this report. Finally, prior computer experience does not appear to be a significant determinant of success provided that adequate training and technical assistance is made available. As physician interaction with clinical computing systems inevitably becomes more prevalent, research to gain additional insight into easing the transition from paper is clearly needed.
上一页 [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] ... 下一页 >>
医生处方系统论文+源程序 第2页下载如图片无法显示或论文不完整，请联系qq752018766
