Identifying and adapting interventions to reduce documentation burden and improve nurses’ efficiency in using electronic health record systems (The IDEA Study): protocol for a mixed methods study

The objective of this study is to engage nurses to understand their experiences using EHR systems and generate ideas on how to optimize their experiences with the goal of improving efficiency and reducing EHR-related burdens. In this context, efficiency is defined as the summative number of clicks (‘click burden’) in the system and the time required to complete certain tasks in the EHR. This work will ensure that the identified solutions are grounded in nurses’ perspectives and experiences and will address their EHR-related needs.

This study will be conducted at two hospital sites in Toronto, Canada. One site is Canada's largest mental health and addictions teaching hospital and academic centre. This organization offers care to individuals with mental health and addiction needs through virtual, inpatient and outpatient care, as well as a partial hospital/day treatment area, and a psychiatric emergency department. There are approximately 900 actively practicing nurses at the organization. Nursing staff work in all clinical areas of the organization and complete all documentation within the EHR system. Inpatient and outpatient clinical areas, excluding the emergency department, will be the setting in which the study will take place at this site.

The second site is an acute care hospital and academic health science centre with clinical areas including women’s and infants’ health, emergency medicine, cancer, complex orthopaedics, palliative care, diabetes, inflammatory bowel disease, geriatrics, and arthritis and autoimmune disease. This site employs more than 1200 nurses working in all areas of the organization and nurses complete the majority of their documentation within the EHR. The women’s and infants’ inpatient department will be the setting in which the study will take place at the second site. A list of study sites can be obtained from the corresponding author upon request.

Eligible participants in this study are 1) nurses (Registered Nurse, Registered Practical Nurse, Registered Psychiatric Nurse, or Nurse Practitioner); 2) employed in one of the two study settings; 3) providing direct patient care, and 4) documenting most of their notes in the EHR. Nurses at the first hospital site will be recruited through email communications sent to all nursing staff. The second hospital site will also conduct recruitment through email communications, but recruitment will be completed with a single inpatient program at the hospital (e.g., Women’s and Infants’ Health). If nurses are interested in taking part in the study, they will be able to contact the study team to join or obtain further information.

A Nursing Advisory Council (NAC) will be established and will include nurses from the relevant clinical areas at both hospital sites in which the study data will originate, who have differing levels of comfort using the EHR, and varied years of experience in nursing. The NAC will meet on a quarterly basis to provide guidance to the study team on the feasibility and applicability of research activities such as recruitment methods and data analysis. Both sites will use existing hospital structures to determine if nurses are interested in joining the NAC. Additionally, the study will be conducted collaboratively with the information technology/clinical informatics departments at each hospital to ensure that the proposed interventions are feasible and can directly inform current EHR optimization initiatives.

The research aims will be addressed through three phases of the study (Fig. 1) utilizing a mixed method design following Palinkas et al.’s taxonomy of designs in implementation research [40]. The taxonomy has three elements to study design including structure, function, and process [40]. For the purpose of this study, the Quan to QUAL structure will be used; this is defined by a sequential process of data collection and analysis that begins with quantitative data and is followed by qualitative data collection. The main purpose of the quantitative data collection and analysis is to test the hypothesis that nurses are spending an abundance of their time in the EHR, mainly on documentation. Based on Palinkas et al.’s taxonomy, the function of the study design will follow complementarity methods; this type of function uses both quantitative and qualitative data to address the set of aims of the study [40]. In this case, quantitative data will be used to evaluate various outcomes of the analytics platform and identify utilization patterns, and qualitative data will be used to provide a deeper understanding and contextualization of these results, as well as aid in the development of interventions. The study process will use the method of connection; meaning the quantitative data will be used to guide the qualitative focus group discussions [40].

This study will draw upon the Consolidated Framework for Implementation Research (CFIR), which outlines considerations for successful implementation projects [41]. The CFIR has been used successfully in the health informatics field to deliver actionable insights to drive optimization and implementation success. Specifically, the findings of Phase 2 will be mapped to the constructs of CFIR as shown in Table 1.

In addition to the CFIR, Sittig and Singh’s 8-dimensional Socio-technical Model will also be used for analysis during the study in Phase 2 (Table 2) [42, 43]. The Socio-technical Model will be used to build upon the five domains of the CFIR by considering the nuances of studying health information technologies and the complexities of healthcare systems [42].

Data used in Phases 1 and 2 of the study will be extracted from the nursing-specific EHR system’s analytics platform. This platform has previously been used to capture EHR utilization patterns for physicians at one of the study sites [20]. Prior to the initiation of Phase 2 focus groups, the NAC will be consulted to identify the subset of metrics to be validated from the complete list captured within the EHR analytics platform. The study team will document the raw time taken for each task within validation tests done at each site.

In Phase 2, a 12-month retrospective analysis of the nursing-specific EHR system’s analytics platform will be conducted to identify trends in utilization of the subset of metrics identified by the NAC. A list of all actively employed nurses (across the two study settings) for the duration of the 12-month period will be obtained. Since there is an aim to build interventions that improve the experience for nurses, all nurses will be included in the analysis who have used the EHR system during that period. As the metrics will be analyzed in aggregate (not at the individual level), nurses who are on leave (> 6 weeks) during the selected time period will not be contributing data to the analytics platform and will therefore not be included in the calculation for that period. Data will be collected on a monthly basis, consistent with one study site’s past analysis of physician data as well as studies by others that have completed similar analyses at comparable or larger scale (e.g., hospital network) [44, 45].

Following the analyses, an infographic summary of the results will be developed and used as a foundation for discussion in the first round of 60-min focus groups with nurses (Phase 2) [7, 46, 47]. The focus groups held during Phase 2 will be used to lead an open discussion about the trends in the analysis and to contextualize why the trends may be occurring. The second objective is for nurses to assess which processes associated with usage trends can be improved, the feasibility of improving the process, and the perceived impact it will make to their EHR experience. The focus groups will follow Kruger and Casey's methods [48]. A maximum variation sampling strategy will be used to recruit a variety of viewpoints and ensure that a diverse group of participants across sex, gender, age, and ethnicity are selected; however, each of the focus groups will consist of participants based on area of care (e.g., Women’s and Infants’ Health) and study site [49]. To achieve study objectives, approximately six focus groups with eight participants each will be conducted at each site. The estimated number of participants was determined using Palinkas et al.’s guidelines for meaningful sampling for qualitative data collection in mixed method implementation research [49]. To reach adequate participant enrolment, the study team will work with one of the sites equity and diversity offices to develop inclusive and effective recruitment materials as well as focus group guides.

In Phase 3, the purpose is to co-create a list of possible solutions to address the areas of improvement identified in Phase 2. A prioritization activity will be used to identify the top solutions. Data will be collected through a 90-min focus group using an online whiteboard platform (e.g., Miro) to allow the participants to collaborate in generating ideas. The focus group session will utilize service design methods incorporating brainstorming, affinity mapping, and an idea portfolio [17]. Through these methods, participants will brainstorm potential solutions, group similar solutions, and assess the solutions on their feasibility and impact. The focus group session will close with a discussion about the top-rated solutions and considerations for implementation.

Further details about data management processes used in this study, including security and storage, are available upon request from the corresponding author.

In Phase 1, raw and percent differences will be calculated to estimate the amount of discrepancy between the two sources of data. The tests run by the study team will be conducted multiple times for each workflow to ensure test–retest reliability of the findings. Based on previous studies, [12, 19, 50, 51], an estimated discrepancy of < 2 min, or 10%, is considered reasonable for validity. If a discrepancy higher than these thresholds is found, the study team will work with the EHR vendor to identify potential causes.

In Phase 2, analyses will be completed and individual users that fall significantly outside of typical ranges will be assessed to determine if they should be removed from the dataset (e.g., their role differs significantly from typical nursing practice). The metrics used for analysis will be determined by the NAC. Similar to previous work with physicians [20], the analysis will consist of descriptive statistics depicted over months to show trends and will include stratification by various user characteristics. The study team will explore the potential to perform a latent class analysis to cluster users into groups, which would be useful when identifying what set of interventions may be beneficial to sub-nursing population groups.

The focus groups will be audio recorded and transcribed verbatim and the results will be analyzed using qualitative content analysis [52‐54]—an approach commonly used by the study team. Specifically, directed content analysis will be used to deductively code the focus group data based on CFIR constructs and the inter-connected dimensions in Sittig and Singh’s seminal Socio-technical Model (Table 2) [42]. Inductive coding will be used for data that does not fit into the pre-existing list of factors, offering the opportunity to extend the model. Both inductive and deductive coding will be used within each domain to provide more specificity. Data will also be stratified by nursing practice and demographic variables to gain insights into the nuances and differences across characteristics.

Participants will be asked if they would like to receive any publications or outputs resulting from this work. This will be asked during the focus groups and a password -protected file containing email addresses will be kept for this purpose. Local nursing informatics groups and associations will also disseminate knowledge findings through national communication products such as webinars, newsletters, and open-access links as appropriate. A briefing note will be distributed to organizations to share knowledge and guide EHR improvements for broader nursing populations. This will include organizations such as nursing and informatics associations, local academic health sciences networks, nursing leadership networks, digital health agencies, EHR vendors, and local networks of Chief Information Officers, Chief Medical Information Officers, and Chief Nursing Information Officers.