130 - Pediatric Trainees’ Perceptions and Documentation of the use of Shared Decision-Making: If you didn’t document it, did it even happen?

Publication Number: 4127.130

Karen Joanie Campoverde Reyes, Yale-New Haven Children's Hospital, New Haven, CT, United States; David Chartash, Yale School of Medicine, New Haven, CT, United States; Victoria Dabrowski, University of Connecticut School of Medicine, Southington, CT, United States; Hayly A. Caruso, Providence College, North Haven, CT, United States; Sarah M. Van, Yale School of Medicine, New Haven, CT, United States; Beatriz A. Ruiz, Yale-New Haven Children's Hospital, New Haven, CT, United States; Zachary J. Kravetz, Yale Pediatrics, Meriden, CT, United States; Paul L. Aronson, Yale School of Medicine, New Haven, CT, United States; Amanda R. Quijano, Boston Children's Hospital, Weston, MA, United States

Background: Shared decision-making (SDM) is increasingly recognized as essential in pediatrics, yet little is known about real-world SDM use by pediatric residents and documentation in practice.

Objective: To assess pediatric residents' perception and documentation of SDM use in patient encounters.

Design/Methods: In this single-institution cross-sectional study using convenience sampling, we included emergency department (ED) and inpatient encounters of patients < 6 years-old with a legal caregiver present whose preferred language was English, Spanish, Haitian, Pashto, Arabic or Portuguese. We excluded ED ESI levels 1/2 and intensive care unit or hematology/oncology admissions. At the end of each encounter, Pediatric residents completed brief, multiple-choice surveys assessing their perceptions of SDM use, caregiver's health literacy, caregiver's desired involvement in SDM, and caregiver and resident satisfaction. Survey responses were linked to deidentified caregiver and patient demographic and clinical data and to EMR documentation using a unique study identifier for each encounter. We analyzed documentation of SDM use within the trainee and attending sections of the encounter-specific EMR using previously validated and published natural language-processing techniques developed from pediatric and adult ED documentation within the same institution. Chi-squared analysis was used to test associations between SDM use and documentation and trainee demographics. Cohen's kappa was used for measures of agreement.

Results: 101 patient encounters were included in analysis. Residents' data reflected a broad distribution of training years and race/ethnic/gender identities representative of the institution's residency program (Table 1). Residents reported variable use of the three queried elements of SDM (Fig 1), with all three elements only being used in a third of encounters. There was minimal trainee documentation of SDM (8%) with poor agreement with reported SDM use and trainee documentation of SDM (k = -0.04). While there was more frequent attending documentation of SDM (34%) (Fig 1), there was still significant lack of agreement between reported SDM use and attending documentation of SDM (k = -0.16). There was no association between resident PGY year or demographic factors and their perceived SDM use or documentation.

Conclusion(s): While trainees report substantial use of SDM, neither they nor attending physicians consistently documented it, and documentation was independent of perceived use. These findings offer a potential opportunity for EMR-based interventions aimed at standardizing SDM use and documentation.

Resident Trainee Demographics
Table 1: Resident trainee cohort demographics as categorized for analysis of any association with trainee use or documentation of SDM. PGY year, Gender, Race, and Ethnic identity were all self-reported by the trainee.

Trainee Perception and EMR Documentation of SDM in Patient Encounters
Figure 1: Percentage of patient encounters for which each of three queried elements of SDM were used: 1. Caregiver was asked if they wanted to be included in the medical decision-making process, 2. Multiple care options were offered to the caregiver, and 3. The caregiver's values and preferences were included in the medical decision-making process), as well as the percentage of patient encounters for which all three elements were reported to be used and for which PCC and SDM language was found in provider documentation. Percentages are reported as perceived to be used by the trainee provider, and as documented by the provider. *Cohen's Kappa for agreement between trainee reporting of all three elements used and either trainee or attending documentation.