Background: Health information technology (HIT) systems were created with the intent to improve efficiency and streamline clinical workflows. However, evidence suggests HIT systems are a significant source of clinician burnout (Adler-Milstein 2020). A thorough understanding of clinicians’ frustrations is necessary to develop solutions to improve HIT systems. In doing so, efforts from the clinical informatics team and software engineers can allocate time and effort to improving clinicians’ highest-priority workflow inefficiencies.
Purpose: The project purpose is to develop an effective process for surveying hospitalists and identifying the current and most significant HIT-related sources of frustration for hospitalists. Once identified, the issues will be presented to the hospital informatics department, vendor liaison(s), and key stakeholders with supporting evidence of high priority issues for hospitalists.
Description: A focus group of 10 hospitalists from a large academic medical center was formed. A heuristic-type of workflow analysis was performed (i.e. usability principle violations were identified and then ranked based on severity). An online survey was sent daily via text message to the focus group during their rounding shifts for 2 weeks. The survey identified specific daily sources of HIT-related frustrations within the hospitalists’ clinical workflow. The survey elicited 68 responses. After 2 weeks, the frequency of recurrent issues increased, indicating a saturation point for unique responses, and the survey was stopped. The list of responses was then reviewed by the hospitalist focus group lead, and multiple clinical informaticists, including informatics leadership. The responses were organized, and duplicates were removed, resulting in 38 unique issues. Next, a second online survey was distributed to hospitalists asking them to rank the severity of frustration for each of the 38 issues using a Likert scale. The hospitalists ranked 71% of the HIT issues as causing at least a moderate or above level of frustration, with 44% of those classified as “very” or “extremely” frustrating (i.e. average ranking ≥4).
Conclusions: This process successfully identified common HIT-related problems and inefficiencies that are sources of high levels of frustration for hospitalists. The first survey (identifying the issues) was sent via text at various times during the day allowing hospitalists to respond when least invasive to their workflow. In addition, the daily text provided a reminder, and the survey was short. All focus group members also answered the second survey (ranking the issues), which was also sent via text message over a 1-week period, with 2 reminders. This workflow analysis could be repeated at regular intervals, based on institutional needs and resources. This iterative approach is important because the most frustrating issues change over time based on variables such as EHR updates, workflow changes, and clinician experience. Repeating this process also reinforces ongoing issues that require more urgent attention from clinical informatics leadership. This process could be easily implemented by other hospitalist groups. These important issues can then be incorporated into the hospitals’ formal HIT optimization plans and upgrade processes to improve hospitalist workflow, decrease frustrations, and ultimately reduce burnout.