Fareri Michael, VandeHei Matthew, Schnapp Benjamin, Jewell Corlin, Lasarev Michael R, Alexandridis Roxana, Resop Dana, Damewood Sara, Kuttab Hani I
University of Wisconsin-Madison, Department of Emergency Medicine, Madison, Wisconsin.
University of Wisconsin-Madison, Department of Biostatistics and Medical Informatics, Madison, Wisconsin.
West J Emerg Med. 2025 May 19;26(3):478-485. doi: 10.5811/westjem.21200.
Point-of-care ultrasound (POCUS) is integral to emergency medicine (EM) training. It is unclear how EM residents use POCUS and how these skills are maintained as they progress in residency training. The purpose of this study was to evaluate resident use of POCUS at various timepoints in EM training.
This was a retrospective cohort study of EM residents at a single, three-year training program between July 1, 2014-June 30, 2022. Residents were included if they had completed three consecutive years of training and an ultrasound rotation in their postgraduate year (PGY)-1. The following time points were assessed: PGY-1 rotation and 3-, 6-, 12-, 18-, and 24-months post-rotation. Number of scans, accuracy of interpretation, acceptability for credit, and percentage of technically limited studies (TLS) were collected at each point. We analyzed performance characteristics using mixed-effects binomial logistic regression with time as a fixed effect and resident as a random effect. Models were fit separately for each performance characteristic and likelihood ratio tests were performed to determine whether performance varied over time.
A total of 65 residents were included with a total of 13,229 exams performed during the study period. Cardiac and focused assessment with sonography in trauma examinations were performed most commonly. Overall accuracy of all exams during the examination period was 97.1% (95% confidence interval [CI] 96.2-98.0%), TLS was 14.5% (95% CI 9.7-20.6%), and acceptability was 82.9% (95% CI 76.3-88.2%). Trend over time (3, 6, 12, 18, and 24 months) found no differences in accuracy (P = 0.84), TLS (P = 0.20), or acceptability (P = 0.28). Further analyses by individual exam types also showed no significant differences in accuracy, acceptability, nor TLS.
Accuracy, acceptability, and percentage of technically limited scans did not significantly vary over time, suggesting that POCUS skills are maintained from PGY-1 rotation to each time point evaluated in this study.
床旁超声（POCUS）是急诊医学（EM）培训的重要组成部分。目前尚不清楚急诊住院医师如何使用POCUS，以及随着住院医师培训的推进，这些技能是如何保持的。本研究的目的是评估急诊住院医师在急诊培训不同时间点对POCUS的使用情况。
这是一项对2014年7月1日至2022年6月30日期间在一个为期三年的单一培训项目中的急诊住院医师进行的回顾性队列研究。如果住院医师完成了连续三年的培训以及研究生一年级（PGY-1）的超声轮转，则纳入研究。评估了以下时间点：PGY-1轮转时以及轮转后3、6、12、18和24个月。在每个时间点收集扫描次数、解读准确性、学分认可接受度以及技术受限研究（TLS）的百分比。我们使用混合效应二项逻辑回归分析性能特征，将时间作为固定效应，住院医师作为随机效应。针对每个性能特征分别拟合模型，并进行似然比检验以确定性能是否随时间变化。
共纳入65名住院医师，研究期间共进行了13229次检查。最常进行的是心脏和创伤超声重点评估检查。检查期间所有检查的总体准确性为97.1%（95%置信区间[CI]96.2 - 98.0%），TLS为14.5%（95%CI 9.7 - 20.6%），接受度为82.9%（95%CI 76.3 - 88.2%）。随时间推移（3、6、12、18和24个月），准确性（P = 0.84）、TLS（P = 0.20）或接受度（P = 0.28）均无差异。按个体检查类型进行的进一步分析也显示，准确性、接受度或TLS均无显著差异。
准确性、接受度以及技术受限扫描的百分比随时间没有显著变化，这表明从PGY-1轮转至本研究评估的每个时间点，POCUS技能都得以保持。
