We conducted a prospective, randomized crossover trial to evaluate CPR performance using two compression techniques for infant patients (Soar et al., 2015). Our institutional review board approved the study (West Texas A&M University, approval number 202111004) and all participants provided informed consent. We recruited healthcare provider trainees (athletic training or nursing students) as our rescuers. All participants had recently completed an infant resuscitation training program and were certified in infant Basic Life Support to ensure familiarity with CPR procedures (Patel et al., 2023).

Thirty-five participants were recruited based on an a priori analysis for a two-test difference between means with a predicted effect size of 0.5, alpha of 0.05, beta of 0.20, and a power of 0.80. Each participant acted as their control, performing CPR using the two-finger and two-thumb encircling techniques in random order. The allocation sequence for the order of techniques was generated using a randomizer and concealed until each trial began to prevent bias. Participants were instructed in both techniques per AHA guidelines before testing, and a brief practice on the manikin was allowed to standardize technique performance.

CPR was performed on a high-fidelity infant CPR manikin equipped with feedback sensors for compression depth, rate, recoil, and ventilation (e.g., Resusci® Baby QCPR with SkillReporter). The manikin was placed on a firm surface at 90 cm in height as recommended in resuscitation guidelines. For each trial, participants delivered continuous cycles of 30 compressions and two breaths (30:2 ratio) for 2 minutes using the assigned technique, reflecting a single-rescuer scenario. A bag-valve device was used for ventilation. Participants were instructed to target a compression depth of at least 4 cm (per current guidelines) and a rate of 100–120 compressions per minute during compressions (Patel et al., 2023). No metronome was provided; participants relied on their training to maintain the rate. The manikin’s software recorded objective CPR quality metrics, including: average compression depth (mm), proportion of compressions achieving ≥4 cm depth, average compression rate (per minute), compression fraction (percentage of time spent compressing during the 2-minute interval), and adequacy of recoil (percentage of compressions with full recoil). CPR continued for a total of 24 minutes to simulate the scene and transport time to a hospital for an emergency (Ashburn et al., 2020; Cornwell et al., 2000; McCoy et al., 2013).

Continuous heart rate monitoring was conducted using a validated wearable heart rate monitor (Polar Electro, Kempele, Finland), capturing minimum heart rate (MinHR), maximum heart rate (MaxHR), and the heart rate difference (HRDiff). Oxygen consumption parameters, including VO₂min, VO₂max, and VO₂diff, were measured using a portable metabolic analyzer to assess the metabolic demands of each technique. After completing both trials, participants completed a brief questionnaire, rating their experience with each technique. We used a 10-point visual analog scale to assess perceived exertion/fatigue, as well as any hand or thumb pain experienced while performing compressions. Participants also indicated which technique they preferred or found easier for maintaining high-quality CPR (Meaney et al., 2013).

Data analysis was performed using SPSS Statistics (version 27.0; IBM Corp., Armonk, NY, USA). The primary outcome variables were compression depth (DepthMM and DepthPCT) and chest recoil (RecoilPCT). Secondary outcomes included heart rate parameters, oxygen consumption, compression rate, and ratings of perceived exertion on a 1 to 10 scale, as indicated by the participant. Pairwise comparisons of estimated marginal means between the two techniques were conducted using the least significant difference method. Results are expressed as mean differences ± standard error (SE) with corresponding 95% confidence intervals (CI) (Sutton et al., 2011). A two-sided p-value < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS Version 27.0.

Participants maintained appropriate compression rates with both techniques. All rescuers stayed within the AHA-recommended 100–120/min range regardless of technique (American Heart Association, 2020). The chest compression fraction (time actively compressing during the 2-minute scenario) was similar between techniques. Importantly, we observed no significant differences in rate or recoil. Nearly all compressions achieved full recoil in both methods. These results indicate that either technique did not detrimentally affect the rhythm or consistency of compressions.

All participants could deliver breaths for both techniques in the 30:2 CPR cycles. In the two-finger trials, rescuers typically kept one hand on the forehead to maintain an open airway, while in the two-thumb trials, they had to reposition their hands to give breaths. Despite this, the average pause time was not statistically significant. The number of ventilations delivered over 2 minutes was the same for both techniques. All delivered breaths generated a visible chest rise, indicating successful ventilation. Thus, using either technique did not impede ventilation delivery in our simulated single-rescuer scenario.

Heart rates were higher in participants during two-finger method than in the two-thumb method by an average of 3.2 ± 1.5 beats per minute, indicating a more taxing procedure. VO2 differences also showed more oxygen consumption indicating the two-finger technique is more metabolically taxing, though it was not statistically significant (p = 0.08). Self-reported fatigue and discomfort scores favored the two-thumb technique. On the 10-point exertion scale, participants rated fatigue during CPR at 6.8 ± 1.9 for the two-finger method, compared to 4.7 ± 1.8 for the two-thumb method (p < 0.001, lower scores = less fatigue). Similarly, hand/thumb pain was reported as moderate with two-finger compressions (5.3 ± 2.4 out of 10) but minimal with two-thumb compressions (2.9 ± 2.1, p < 0.001). Most participants (87%) preferred the two-thumb encircling technique after experiencing both, citing better control and less strain. A summary of the primary pairwise comparisons is presented in Table 1.

We found that two-thumb compressions achieved a clinically and statistically significant difference with an average of ~5–6 mm greater depth than two-finger compressions. Achieving adequate depth is vital, as prior studies have linked deeper compressions with better perfusion and outcomes in pediatric cardiac arrest (Patel et al., 2023). The depth advantage of the encircling technique observed in our study mirrors that reported in previous research. For example, Millin et al. (2020) noted a 5.61 mm greater depth with two-thumb compressions across pooled studies. Cioccari et al. (2021) similarly documented enhanced depth and compression quality with the two-thumb method in a simulated infant cardiac arrest. By increasing the contact surface area and employing both thumbs, rescuers can generate higher pressure more uniformly on the sternum, likely explaining the superior depth. Additionally, the encircling grip provides better thoracic support and stability, which may improve the mechanics of each compression and reduce energy wasted on chest wall distortion (Chang et al., 2020 a–b).

Importantly, the two-thumb method’s deeper compressions translated into more compressions meeting current guideline targets (≥4 cm). In clinical terms, more compressions will likely effectively generate blood flow. In our simulation, over 90% of compressions with TTET reached the recommended depth, compared to ~69% with TFT. This magnitude of improvement aligns with prior findings of roughly 30–40% more adequate compressions when using two thumbs versus two fingers in infants (Millin et al., 2020). Such a difference could be lifesaving in resuscitation, as more consistent perfusion during CPR increases the chances of ROSC.

We also observed that compression rate and recoil were maintained with the two-thumb technique. Some providers express concern that switching hand position might alter their pacing or that encircling the chest could impede full recoil, but our data did not show any detrimental effect. Both techniques allowed participants to stay within the recommended rate range, and complete recoil was achieved virtually equally. This suggests that with proper training, rescuers can use the two-thumb approach without sacrificing other aspects of CPR quality.

A key question has been whether a single rescuer can effectively manage ventilations while performing two-thumb compressions. Our study adds to the evidence that it is feasible: there was no significant difference in ventilation delivery between the techniques in our 30:2 scenario. This finding echoes the conclusions of Cioccari et al. (2021) and others, who reported that the two-thumb technique can be employed by a single rescuer “without interfering with ventilation” (Cioccari et al., 2021). Moreover, our participants could execute a quick hand reposition: thumbs were lifted from the chest, and one hand was immediately moved to open the airway, a maneuver practiced during the orientation. These results indicate that with training and technique, a lone rescuer does not need to avoid the two-thumb method out of concern for ventilation difficulty.

It is worth noting that current AHA guidelines remain cautious on this point, continuing to recommend the two-finger technique for solo rescuers primarily to facilitate rapid ventilation and airway control (Dellarocca, 2021). Our data, in line with recent studies, suggest that these concerns can be mitigated. The slight increase in hands-off time for breaths when using TTET did not significantly impact overall CPR performance in our simulation. Nonetheless, in an actual resuscitation, any added pause could be critical; therefore, rescuers should weigh their ability to transition quickly. Maintaining an open airway during compressions (e.g., using a jaw-thrust or having equipment like a pocket mask ready) could further minimize delays if employing the two-thumb technique alone.

With the two-thumb encircling technique, participants experienced less fatigue and hand pain. This is consistent with anecdotal reports and prior research that suggest the two-finger method can be strenuous over time (Cioccari et al., 2021). With two fingers, the force is concentrated on a small area of the rescuer’s fingertips, often leading to faster muscle fatigue and discomfort in the fingers or hand. In contrast, two-thumb compressions allow the rescuer to use the strength of both arms and shoulders more directly, distributing force across the hands and thumbs. In our study, the average self-reported exertion was significantly lower for TTET, and nearly 90% of rescuers preferred it. Similarly, a recent 2025 trial by Solecki et al. observed that a novel two-thumb–based technique improved rescuer endurance (lower-rated perceived exertion) compared to the traditional two-finger approach (Solecki et al., 2025). Reduced fatigue is beneficial for the rescuer’s comfort and can also translate to sustained high-quality compressions over longer durations of CPR. If a single rescuer becomes too fatigued, compression depth and rate will inevitably suffer. Thus, a technique that prolongs effective performance is highly valuable, especially in out-of-hospital settings where a lone rescuer must perform CPR for several minutes before help arrives.

This study has some limitations. First, it was conducted on a manikin model, which, while high-fidelity, cannot perfectly replicate an actual infant’s anatomical and physiological response. Manikin studies may overestimate compression depth and do not capture potential differences in clinical outcomes. As noted in prior systematic reviews, results observed in simulations may not fully translate to real infants (Millin et al., 2020). Second, our scenario was relatively short (2 minutes of CPR per technique) and controlled. In a true resuscitation, fatigue over time, the urgency of the situation, and other factors (like patient movement or transport conditions) could influence technique effectiveness (Jo et al., 2017; Lee et al., 2018; Lee et al., 2022; Wu et al., 2020). We also did not measure long-term outcomes such as actual ROSC or survival, as those cannot be assessed in a simulation.

Further research is needed to confirm these findings in clinical settings. Large observational studies or trials (where feasible) during actual infant CPR could assess whether two-thumb compressions lead to better patient outcomes (e.g., higher ROSC rates or neurologically intact survival) than two-finger compressions (Chang et al., 2020a, b, c, d, e). Additionally, studies should explore the integration of techniques: for example, at what point (in terms of rescuer fatigue or patient size) should a single rescuer switch from two-finger to two-thumb or even to the heel of one hand. Investigations into training methods are also necessary—teaching lay rescuers the two-thumb method and testing their performance could inform community CPR guidelines (Reynolds et al., 2020). The development of assistive devices or feedback tools that encourage proper depth without compromising ventilation might also stem from this line of research. Ultimately, ongoing updates to resuscitation guidelines by bodies like the AHA and International Liaison Committee on Resuscitation (ILCOR) will likely incorporate the growing evidence base, potentially adjusting recommendations if two-thumb compressions demonstrate advantages.