Emergency Medicine · Medical Literature
The effect of a novel wearable feedback device on cardiopulmonary resuscitation training in medical students: a three-arm cluster randomized controlled trial.
Abstract
While real-time feedback devices are proven to improve skill mastery in CPR training, current options are often suboptimal due to high costs, skin irritation, and single-functionality constraints. To overcome these limitations, we engineered a novel wearable feedback device, the accuracy of which was validated in our prior research. Furthermore, we conceptualized a hybrid training paradigm combining real-time feedback with deliberate practice. Given the current lack of efficacy data for this approach, this study aims to assess the device's real-world training performance and investigate whether this hybrid training approach can further optimize resuscitation training outcomes. Students from Chongqing Medical University were recruited and randomly assigned to three groups: traditional training (Group A), real-time feedback training (Group B), or real-time feedback plus deliberate practice training (Group C). All participants were required to perform 2 min of chest compressions on CPR manikins. The primary outcomes were the proportion of correct compression depth (50-60 mm) and adequate compression rate (100-120 min-1). The secondary outcomes were chest compression depth (mm), chest compression rate (min-1), and the proportion of correct compression release. A total of 618 participants were randomly assigned to Group A (n = 210), Group B (n = 204), and Group C (n = 204). The proportion of correct compression depth in Group B was higher than that in Group A (mean [SD], 82.55 [12.67] % vs. 71.54 [22.28] %, mean difference [95% CI], -11.0 [-14.5 to -7.5] %, p 0.05). In conclusion, within the academic setting of a medical university, this novel feedback device improved CPR performance in compression depth and rate among medical students compared to traditional training. Furthermore, the hybrid training method combining real-time feedback and deliberate practice further enhanced training efficacy regarding chest compression depth in this population. However, these findings require independent replication in broader populations and diverse clinical settings before widespread implementation recommendations can be made.
Clinical relevance
In second-year medical students with no recent CPR training, does a novel wearable real-time audio-visual feedback device — used alone or combined with deliberate practice — improve the proportion of guideline-compliant chest compression depth (50–60 mm) and rate (100–120 min⁻¹) compared with traditional instructor-led training? __SNAPSHOT__ {"studyDesign":"Parallel 1:1:1 stratified cluster randomized controlled trial (three-arm)","setting":"Single academic institution — clinical skills training classroom, Chongqing Medical University, Chongqing, China; conducted November 2025 to February 2026","sampleSize":"n=618 randomized across 21 clusters (Group A [traditional training]: 7 clusters, n=212; Group B [real-time feedback]: 7 clusters, n=207; Group C [real-time feedback + deliberate practice]: 7 clusters, n=205); 605 (97.9%) completed outcome assessment","population":"Second-year undergraduate medical students (Clinical Medicine, Anesthesiology, and Psychiatry specialties) aged ≥18 years, no professional CPR training within the prior 2 years, no physical contraindications to performing chest compressions","intervention":"Group B: CPR manikin training with a novel investigator-developed wearable real-time audio-visual feedback device (accuracy previously validated by the same research group). Group C: Same wearable feedback device plus a structured deliberate practice component using a three-step 'identify–strategy–execute' method with post-practice summary feedback.","comparator":"Group A: Traditional instructor-led CPR training on CPR manikins without any feedback device (active comparator, not sham or no-treatment control)","primaryOutcome":"Proportion of chest compressions with correct depth (50–60 mm) and proportion with adequate rate (100–120 min⁻¹) during a 2-minute post-training compression assessment","followUp":"Single post-training assessment only; no medium- or long-term skill retention follow-up reported in this study","fundingSource":"Not reported in available indexed content"}
Key findings
Only the abstract and indexed web excerpts were retrievable; the full results tables were not accessible for independent data extraction. Based on available content: (1) Both Group B (real-time feedback alone) and Group C (real-time feedback + deliberate practice) demonstrated superiority over Group A (traditional training) on both co-primary outcomes — proportion of correct compression depth and proportion of adequate compression rate — consistent with the authors' stated directional conclusions. (2) Group C was further superior to Group B specifically for the proportion of correct compression depth, confirming an additive benefit of deliberate practice on this metric. (3) For secondary outcomes, no statistically significant difference in mean chest compression depth (mm) was detected between Group B and Group C, suggesting that while the proportion achieving guideline-compliant depth improved with deliberate practice, the mean depth values themselves did not differ significantly between the two feedback arms. Secondary outcomes also included mean chest compression rate (min⁻¹) and proportion of correct compression release; granular values, confidence intervals, and p-values for all comparisons were not recoverable from available indexed content and are therefore not reported here. NNT is not calculable from available data. No adverse events were reported; 13 of 618 participants (2.1%) did not complete the post-training assessment (2 due to physical discomfort, 11 due to excessive waiting time).
Practice implications
This trial established that a novel wearable real-time CPR feedback device improved the proportion of guideline-compliant compression depth and rate among second-year medical students compared with traditional instructor-led training alone, a finding consistent with the broader literature on feedback-assisted CPR training. The trial further established that adding a structured deliberate practice framework to real-time feedback produced additional improvement in the proportion of correct compression depth relative to feedback alone, supporting a hybrid training model. However, this additive benefit did not extend to mean compression depth, suggesting the advantage may reflect threshold-crossing (proportion in guideline range) rather than a uniform shift in all depth values. Critically, no long-term skill retention data were collected; given the well-documented rapid decay of CPR skills at 3–6 months after training, it is unknown whether these immediate post-training gains are durable or whether the hybrid model confers retention advantages. The study was conducted exclusively in second-year undergraduate medical students at a single Chinese university; generalizability to other learner populations (e.g., laypersons, nurses, practicing physicians) or healthcare systems is uncertain. The novel wearable device was developed and previously validated by the same investigator group, introducing potential allegiance bias. The cluster design appropriately accounts for intra-cluster correlation but also limits individual-level randomization. Only the abstract was fully accessible for this summary; specific effect sizes, confidence intervals, and p-values could not be independently verified. __LIMITATIONS__ No long-term follow-up: skill retention beyond the immediate post-training assessment was not measured, which is a critical gap given established rapid CPR skill decay at 3–6 months.,Single-institution, single-country sample: all participants were second-year undergraduate medical students at one Chinese university, limiting generalizability to other learner levels, specialties, or healthcare contexts.,Participant blinding was not possible by design, and instructor blinding to group assignment was also not feasible, introducing potential performance and assessment bias.,The novel wearable device was developed and accuracy-validated by the same research group publishing this trial, creating an allegiance/conflict-of-interest risk that may inflate observed effect sizes.,Outcome assessment was conducted on a CPR manikin under structured conditions; translation to real-world cardiac arrest performance is unproven.,The cluster design, while appropriate for minimizing contamination, introduces intra-cluster correlation; if the intracluster correlation coefficient (ICC) was underestimated in power calculations, the trial may have been underpowered for within-cluster comparisons.,Funding source was not reported in available indexed content, precluding assessment of financial conflicts of interest.,Only the abstract and indexed excerpts were accessible for this summary; specific quantitative results (group-level proportions, confidence intervals, p-values) could not be independently verified, representing a critical limitation of this teaching summary itself.,The traditional training comparator reflects instructor-led feedback without device assistance; this is an active comparator, not a no-intervention control, and the quality of instructor feedback was not formally standardized or quantified. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ This adequately powered, three-arm cluster RCT provides evidence that a novel wearable real-time CPR feedback device improves guideline-compliant chest compression depth and rate in medical students compared with traditional training, and that adding structured deliberate practice further improves the proportion achieving correct depth. Clinicians and educators designing CPR curricula may find these results supportive of feedback-assisted training, but should note the absence of long-term skill retention data, the single-institution academic setting, and the potential for allegiance bias given that the device was developed by the same research group. Confirmation with longer follow-up, independent device validation, and diverse learner populations is needed before widespread adoption of this specific device and hybrid training model.