Sample Size of Trials Investigating the Impact of Point-of-Care Ultrasound-Guided Strategies on Patient Outcomes: A Systematic Review.

Abstract

Point-of-care ultrasound (POCUS) is increasingly utilized for bedside diagnosis and management in diverse clinical contexts. However, the design of randomized controlled trials (RCTs) evaluating the impact of POCUS-guided strategies on clinical outcomes presents significant challenges. This study aims to explore the assumptions underlying sample size estimation in POCUS-guided trials and assess the adequacy of sample sizes in published trials through a systematic review. We performed a sample size analysis considering varying rates of POCUS-induced management changes and plausible effect sizes on binary and continuous patient-centered outcomes. Additionally, a systematic review of PubMed was conducted to identify RCTs comparing POCUS-guided management to usual care, extracting data on planned and actual sample sizes and justifications for sample size decisions. Sample size estimations revealed a substantial dependence on the proportion of participants experiencing management changes due to POCUS findings. For example, achieving adequate power in a trial with a moderate effect size requires over 1000 participants if POCUS alters management in 50% of cases. Our review included 25 RCTs, with a median sample size of 206 participants (interquartile range 122-250). Only 68% of trials reported sample size justifications, and 41% failed to meet planned recruitment targets, primarily due to recruitment challenges and other logistical barriers. Most trials investigating POCUS-guided strategies are underpowered, underscoring the need for realistic sample size estimations that consider the rate of POCUS-induced management changes and anticipated effect sizes. Future trials should incorporate pilot phases and innovative designs to optimize feasibility and power.