Implementation of a geriatric assessment SmartPhrase: A multi‐institutional pilot study

As the population ages, geriatric assessment (GA) is an increasingly important clinical skill. The GA also encompasses many geriatrics competencies recommended for medical students and internal medicine (IM) and family medicine (FM) residents.1, 2 The geriatric 5Ms offer a succinct framework for key components of a GA3 and are increasingly used in trainee education.4 Note templates in the electronic health record (EHR), called "SmartPhrases" in the Epic EHR, have the potential to improve clinical care, documentation, and education. We created a 5Ms-based SmartPhrase with the goal of improving trainees' familiarity with the GA. We developed a 5Ms-based GA SmartPhrase, incorporating feedback from an informal resident focus group at the University of Pennsylvania (Penn); this original version was used at Penn and Cornell. The SmartPhrase was adjusted per institutional preferences and trainee feedback at Yale; this version was used at Yale. All institutions use Epic EHR. The SmartPhrase (Supplement S1) was implemented on two Acute Care for Elders (ACE) unit rotations (Penn and Yale) and an outpatient Geriatrics rotation (Cornell). ACE unit rotations lasted 2 weeks and the outpatient rotation lasted 1 week as part of a four-week geriatrics block. Trainees included IM, FM, and medicine-pediatrics residents in post-graduate years (PGY) 1 through 4; PGY-1 preliminary residents; and medical students. Trainee composition varied by institution. At the rotation start, trainees received an institution-specific orientation to GA and the SmartPhrase. Orientations were led by geriatric fellows or faculty. Each included an overview of the 5Ms, an introduction to the SmartPhrase, and advice about how to incorporate the SmartPhrase during the rotation, though the delivery format differed (Supplement S2). We conducted voluntary surveys for trainees and involved teaching faculty and geriatrics fellows. Trainees received a pre-survey at rotation start, a post-survey at rotation completion, and a follow-up survey at either 3 months post-rotation or at the end of the academic year (AY). Faculty and fellows received one survey at the AY's end. Trainee pre- and post-survey responses were compared with SAS software v9.4 using unpaired Chi-square or Fischer exact tests when appropriate and a two-tailed significance level of 0.05. Free-text responses were independently coded by three authors, who then met to resolve discrepancies, create a joint codebook, and assign broader themes to the qualitative data. All local Institutional Review Boards deemed the study exempt. Across all three sites, 137 trainees were introduced to the GA SmartPhrase and 101 trainees completed the pre-survey (73.7% response rate). Response rates varied by site (Cornell 17/18; Penn 20/41; Yale 64/78). Respondents included medical students (10.9%), PGY-1 residents (33.7%), and PGY-2 and PGY-3 residents (55.4%). Seventy-six trainees completed the post-survey (55.5% response rate). Response rates varied by site (Cornell 15/18; Penn 22/41; Yale 39/78). Respondents included medical students (11.8%), PGY-1 residents (23.7%), and PGY-2 and PGY-3 residents (64.5%). Forty-eight trainees completed the follow-up survey (35.0% response rate). Response rates varied by site (Cornell 11/18; Penn 15/41; Yale 22/78). Respondents included medical students (6.3%), PGY-1 residents (20.8%), and PGY-2 and PGY-3 residents (72.9%). Respondents reported a significant increase in knowledge of and comfort with GA as well as having a framework for GA (Figure 1A, p < 0.001). Results were consistent across sites and trainee levels (Supplement S3). Most respondents shared that using the GA SmartPhrase impacted clinical practice (Figure 1B). The majority of respondents agreed that using the SmartPhrase was helpful when caring for older adults (94.5%) and planned to use the SmartPhrase in the future (72.9%). Qualitative analysis of post-survey free-text responses suggested that the SmartPhrase facilitated the performance of the GA by providing a user-friendly, thorough, and structured framework with prompts for questions. Respondents felt the SmartPhrase helped with geriatrics learning, patient-centered care, and recognizing the 5Ms. Respondents identified challenges, including a negative impact on workflow. Responses suggested better integrating the SmartPhrase into the rotation curriculum and interdisciplinary workflow of the ACE units (Supplement S4). In the follow-up survey, eight trainee respondents (21.6%) at the inpatient sites indicated using the SmartPhrase in at least one non-geriatrics-specific setting, including other inpatient rotations (10.8%), clinic (16.2%), and home visits (2.7%). At the outpatient site, seven trainee respondents (63.6%) indicated using the SmartPhrase outside of the outpatient geriatrics rotation, including the ACE unit rotation (54.5%), other inpatient rotations (36.4%), and clinic (45.5%). Sixteen faculty and fellows completed the faculty survey (50% response rate). All respondents observed trainees using the GA SmartPhrase and 94% agreed that it is an effective teaching tool. Qualitative analysis of free-text responses suggested that the SmartPhrase facilitated trainee performance of the GA by providing a framework and enhancing consistency and thoroughness, while also standardizing notes and presentations on rounds. Respondents found that the SmartPhrase helped trainees recognize the 5Ms and provide geriatric-specific care. Respondents reported challenges with SmartPhrase completion including inconsistent use, incomplete information, challenges obtaining collateral, and unclear plans for follow-up if incomplete. Respondents also noted a negative impact on workflow, plus a need to enhance trainee education around GA (Supplement S5). In this multi-institutional pilot study, we effectively integrated a GA SmartPhrase into inpatient and outpatient geriatrics rotations, and they remain in use. This was associated with significant improvement in trainee self-reported knowledge of and comfort with performing a GA. Most trainees self-reported an impact on clinical practice, and several trainees reported use beyond their geriatrics rotation, indicating a transfer of knowledge and applicability of the SmartPhrase in other settings. Furthermore, trainees and faculty described complementary benefits of using the SmartPhrase, as well as opportunities for improvement. Strengths of our study include its multi-institutional nature and perspectives from both trainees and faculty. Our findings are limited by possible non-responder bias and differences in SmartPhrases and orientations at each site. Additionally, we were not able to evaluate the change in trainee knowledge and comfort with the GA as a result of each geriatrics rotation, independent of the SmartPhrase. Based on our experience and results, we note several key elements to successfully implement a GA SmartPhrase for educational purposes. First, integration into the rotation curriculum helps to emphasize the importance of the GA and provide basic guidance on the use of the SmartPhrase. Second, SmartPhrase implementation should be tailored to the local needs and context of each institution. Third, the workflow may be impacted by the SmartPhrase itself and by the time it takes to perform and discuss a GA. This should be considered in the design and implementation of a SmartPhrase, and team members may need to embrace flexibility in their personal and teamwork processes. Finally, faculty buy-in is critical for the reinforcing use and providing guidance on the nuances of performing and applying the GA. At the time of the study, JXZ was funded as a Geriatric Medicine Education Fellow by the Connecticut Older Adult Collaboration for Health 4M (COACH 4M) grant. JXZ and JO were supported by a Geriatric Workforce Enhancement Program funded by the Health Resources and Services Administration (HRSA). EBF was supported by a generous donation from the Rosanne H. Silbermann Foundation. The COACH 4M grant, HRSA, and the Rosanne H. Silbermann Foundation had no role in the development of this work or the preparation of this manuscript. All authors met the criteria for authorship stated in the Uniform Requirements for Manuscripts Submitted to Biomedical Journals. Creation of SmartPhrase: JXZ, EPS, RZC, RKM. Acquisition of learners and data: JXZ, EPS, EBF, RZC, JO, LS, RM, RKM. Analysis and interpretation of data: JXZ, EPS, EBF, XZ, DX. Preparation of manuscript: JXZ, EPS. We thank the trainees and faculty members who participated in the implementation of the GA SmartPhrase project. We are especially grateful to Brian Eiss, MD for his efforts in implementing the SmartPhrase at Cornell, to Mark Simone, MD and James Lai, MD for their support of the SmartPhrase on Penn and Yale's ACE units, respectively, and to the Penn and Yale geriatric medicine fellows who provided valuable input regarding the SmartPhrase: Colin Brien, MBBS; Dhruv Jani, DO; Ravneet Randhawa, MD; Julianna Ricco, MD; Rebecca Russell, MD; and Matthew Van Dongen, MD. The authors have no conflicts. At the time of the study, JXZ was funded as a Geriatric Medicine Education Fellow by the Connecticut Older Adult Collaboration for Health 4 M (COACH 4 M) grant. JXZ and JO were supported by a Geriatric Workforce Enhancement Program funded by the Health Resources and Services Administration (HRSA). EBF was supported by a generous donation from the Rosanne H. Silbermann Foundation. The COACH 4 M grant, HRSA, and the Rosanne H. Silbermann Foundation had no role in the development of this work or the preparation of this manuscript. Supplement S1. Geriatric assessment (GA) SmartPhrase examples from each site. Supplement S1a. Instructions for how to create GA SmartPhrase in Epic. Supplement S1b. Downloadable GA SmartPhrase examples. Supplement S2. GA orientation materials from each study site. Supplement S3. Self-reported impact of GA SmartPhrase by trainee level and training site. Supplement S4. Trainee post-survey qualitative themes and representative free-text quotes. Supplement S5. Faculty survey qualitative themes and representative free-text quotes. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.