Use and satisfaction with key functions of a common commercial electronic health record: a survey of primary care providers

The federal government is investing $27 billion over ten years through the Health Information Technology for Economic and Clinical Health Act to increase the adoption and use of electronic health records (EHR) in the United States to improve health outcomes [1]. The use of EHRs has the potential to improve quality and safety in health care, but these improvements will only occur if providers understand and use key functions regularly and effectively [2]. The Office of the National Coordinator has defined a minimum list of EHR functionalities, known as “meaningful use” criteria, intended to promote the effective use of the EHR to drive improvements in safety, quality and efficiency [3]. “Meaningful use” requires a variety of EHR functions, ranging from visit documentation to clinical decision support [1]. Commercial EHR systems such as Epic are increasingly being adopted because their off-the-shelf capabilities meet meaningful use criteria required to qualify for financial incentives [4]. However, there is relatively little evidence available about providers’ use of and satisfaction with key functions, particularly in commercial EHRs.

One prior study found considerable variability in physician use of EHR functions. However, this study included physicians using different EHR systems in various stages of EHR implementation, making it challenging to attribute suboptimal use of specific functions to provider preference rather than to differences in the design and technology of the EHR, or inexperience with using the system [5, 6]. Prior literature has also linked EHR use to high ratings of general satisfaction with the overall EHR system, medical practice, and career [7–9]. However, these studies did not examine providers’ satisfaction with specific key EHR functions, which may be important in understanding potential suboptimal use of embedded functions [10].

Therefore, the aims of this study were to determine: 1) how providers who practice in a mature EHR environment use key functions of a comprehensive commercial EHR system to perform common clinical tasks, with a particular interest in assessing use of structured documentation features which may improve efficiency and thoroughness; 2) their satisfaction with key functions, and 3) the time they spend on EHR activities outside of their clinic sessions.

Among a diverse group of PCPs practicing in clinics experienced with using a common commercial EHR system, we identified variability in providers’ use and satisfaction with many key EHR functions. A majority of providers were not regularly using or were not highly satisfied with a few core features of the EHR, including documentation of preventive services and the health maintenance module.

There was great variability in the extent to which users availed themselves of more structured electronic note writing tools (like templates) and most used a ‘mix and match’ approach of free text for the history and assessment, and templates for the exam. Whether this is because the exam is the easiest part of the note to template, or there is lack of awareness of or low opinions about the user friendliness of other EHR note writing functions is not known. It should be noted that the EHR encourages some “free text” writing for history and assessments and plan sections even within templates.

About half of the providers in our sample routinely used structured EHR documentation. While it remains uncertain whether there is a single best method to document notes, the use of structured documentation has been associated with improved performance on several quality measures compared to those using free text or dictating [12]. Additionally, using structured documentation such as templates (as opposed to free texting the note from scratch) could potentially be a time-saving benefit of the EHR. On the contrary, structured documentation may promote bloated notes with redundant or irrelevant information and cloning, the practice of copying and pasting past documentation into the current encounter note. Our findings suggest there is a generational influence on structured EHR documentation use, with both extremes of age having the lowest probabilities. For the oldest providers it may be related to a combination of a lack of awareness, training, or interest in engaging with the EHR. In contrast, the youngest providers may have unmet expectations for technological innovations in the EHR and as a result, may shy away from relatively inflexible templates with suboptimal user interfaces [17]. Family medicine providers were more likely to use structured documentation compared to their internal medicine counterparts, and may be due to differences in the way different specialties interact with the EHR.

While the EHR has been promoted as a useful tool to improve preventive health, there is mixed evidence for the effectiveness of EHR-based clinical decision support (CDS) reminders for improving cancer screening and vaccination rates [18]. In addition, prior research has not found an association between use of an EHR and improved performance on these process measures for health prevention [19–22]. We found that providers reported substantial variability in documentation practices for preventive services (cancer screening and vaccinations) and expressed considerable dissatisfaction with the structure and function of the existing health care maintenance module and its associated reminder system, which may have repercussions for EHR-based strategies to improve preventive health.

Our findings highlight a potential explanation for why the use of EHRs and CDS tools may be ineffective in improving concordance with delivery of recommended preventive care. CDS rules typically use results from structured data fields to make inferences about when certain preventive health measures are due. If providers are not consistently documenting results in structured fields in the EHR, such as the health maintenance module and the historical immunization section, then CDS tools will not have the necessary data to provide accurate and timely recommendations for providers.

There has been increased attention directed towards the ‘problem list’ as a key focus of meaningful use and driver of CDS. Our findings highlight some of the challenges that remain, because while most PCPs thought the problem list was important and kept it up-to-date, most erroneously thought listing a diagnosis in the ‘encounter’ section automatically carried over to the problem list, and half worried that the problem list was unreliable and inaccurate. This suggests that quality measurement and CDS strategies that rely solely on diagnoses in the problem list may be inaccurate. It also suggests that in order for the problem list to help promote meaningful use and be an effective driver of CDS more attention needs to be paid to understanding and allaying providers’ concerns on this issue.

In addition to quality of care, another important outcome to evaluate the success of EHRs is efficiency. Previous time and motion studies of providers using the EHR compared to paper charts found that time-efficiency was not likely to be achieved during the clinic visit with use of an EHR [23, 24]. Our study suggests that using an EHR can be considerably time-consuming for many providers even after the scheduled clinic session. Extrapolating our findings to full-time PCPs practicing nine half-day clinic sessions a week, nearly half of providers are spending greater than 9 extra hours a week beyond their scheduled clinic sessions completing documentation in the EHR and a third are spending at least 9 extra hours a week responding to electronic messages. This is a substantial after-hour “time tax” on providers who are already pressured to see more patients in light of waning reimbursements and may contribute to the high burnout risk seen in primary care [25].

Our study has several important limitations. First, though our sample was large and diverse, we surveyed PCPs using a single commercial EHR, so the generalizability to other systems besides Epic is unknown. We intentionally surveyed providers using a single system to better understand the use of key functions independent of differences in EHR design. In addition, Epic is considered to be a ‘best-of-breed’ comprehensive system with a dominant market share, accounting for one-quarter of providers practicing in medium to large-sized clinics that currently use an EHR system [11]. Therefore, lessons learned from this study may thus be readily applicable to a large proportion of primary care providers. In addition, many of the Epic features we asked about are similar to those in most commercial and home-grown EHRs, so our findings may have implications for the larger EHR community. Second, this study relied on providers’ self-reported use of the EHR. We attempted to mitigate possible recall bias by asking providers about their “typical” experiences rather than asking them to report their specific use patterns for previous clinic sessions. The providers included in this survey were all current users of the EHR in their daily clinical practices, helping to guard against inaccuracies in respondents’ historical accounts of their engagement with the system. Third, we used a web-based survey to ask providers about their experiences using an EHR system. This approach may have selected participants who were more likely to be technology-savvy and potentially use more advanced features of the EHR. However, our findings of suboptimal use and dissatisfaction for several key functions suggest that if we selected a greater number of EHR enthusiasts then we may have underestimated the perceived deficiencies of the EHR. In addition, we found that two-thirds of providers self-classified as early adopters of technology, a rate nearly identical in a study of providers by Kim et al. that employed both web-based and paper questionnaires [26].

The results of this study suggest that implementing a full-featured EHR that meets “meaningful use” criteria is insufficient and that we should shift our attention from adopting an EHR to optimizing provider use of key functions and improving providers’ time efficiency. Potential solutions to extract maximum benefit from the EHR include: refinement of EHRs to be more intuitive and user-friendly especially for common clinical tasks, development of more intelligent, clinically nuanced CDS, more effective provider training on use of advanced EHR functionalities, and greater attention to how to best use the EHR to improve the clinical workflow efficiency.

This study points to a sizeable gap between the ‘dream’ of policymakers’ vision for EHRs and the ‘reality’ of frontline PCPs’ current use of an industry-leading commercial system. The availability of certain functions that constitute “meaningful use” may be necessary but insufficient to ensure ‘meaningfully’ effective use of EHRs. Provider dissatisfaction with and suboptimal use of more advanced features of the EHR may in part explain the lack of association between EHR use and improved quality of care [19–22]. Further research is needed to both understand the proximal factors creating the gap between availability of functions and suboptimal use and how best to narrow the divide to improve quality and efficiency.