Cardiac & Resuscitation, Exclusives

Improving the Effectiveness of CPR Chest Compressions in Neonates and Infants

Compliance of advanced pediatric life support instructors with current international recommendations during simulated infant CPR is poor, the author writes. (Photo/Adam Mason)

More than fifty percent of pediatric arrests occur in children younger than one year old and more commonly in males (~60%).1,2 Fortunately, only about 0.1% of term infants and 5% of preterm infants receive cardiopulmonary resuscitation (CPR) chest compressions in the delivery room.3 Infants who receive chest compressions have a high incidence of mortality (41%) and short-term neurological morbidity (e.g., 57% hypoxic-ischemic encephalopathy and seizures). Moreover, newborns who received chest compressions and epinephrine but had no signs of life at 10 minutes following birth have 83% mortality, with 93% of survivors experiencing moderate-to-severe neurological disability.4


The poor prognosis associated with receiving chest compressions in the delivery room raises questions as to whether improved CPR methods specifically tailored to the newborn could improve outcomes.3 However, the incidence of infants who need CPR at birth is rare and in general unexpected, and therefore randomized clinical trials of alternative CPR methods have not been performed. Therefore, continuing efforts should be made to improve conventional CPR techniques for both neonates and older infants.

Twenty-two certified advanced pediatric life support instructors performed two-minute continuous two thumb (TT) encircling hands method chest compressions and two finger (TF) chest compressions on an instrumented infant CPR manikin.5 Overall compliance with all four performance targets of compression rate, compression depth, compression duty cycle and compression release force (100% complete recoil = zero release force) was less than 1% for both techniques (P= .14). When performing chest compressions at a 130/min rate, a compression/decompression occurs every 0.46 seconds. Performing chest compressions at a 130/min rate with a 50% duty cycle means that the duration of the downstroke of compression (0.23 seconds) equals the duration of the upstroke of decompression (0.23 seconds). At a 60% duty cycle, the duration of downstroke is longer (0.276 seconds) than the duration of upstroke (0.184 seconds), which significantly affects coronary perfusion because the heart is circulated during compression diastole in which a shorter decompression duration has adverse consequences.

It was concluded that compliance of advanced pediatric life support instructors with current international recommendations during simulated infant CPR is poor. As evidenced by this study, TT-CPR is a low-momentum prolonged compression technique that much more adversely affects CPR efficacy than TF-CPR. While mean compression rates were equivalent (128 vs 131 per minute) and mean compression depth was poorer during TF-CPR (26 mm) versus TT-CPR (33 mm) (P< .001; target, ≥ 36.7 mm), mean duty cycles (which significantly affects coronary perfusion because the heart is circulated during compression diastole in which an inadequate decompression duration has adverse consequences) were close to 50% with TF-CPR (53%) but poor with TT-CPR (61%), and release force (which affects complete recoil) was four times greater with TT-CPR than TF-CPR (0.8 versus 0.2 kg (P< .001; target, < 2.5 kg). Therefore, TF-CPR may be more recommendable than TT-CPR. 

However, when using TF-CPR at a 130/min rate, the middle finger rather than the index finger should be used to apply all the force because it is the longest and strongest finger and because it can be used with the index for support, which strengthens it.6 Using a rate of less than 130/min may not be fast enough to generate effective depth chest compression because while higher rates of compression increase the depth of compression, lower rates of compression decreases the depth of compression.7 To position the middle finger over the lower third of the sternum adequately above the xiphoid process, left-handed resuscitators should perform TF-CPR from the left side of the patient and right-handed resuscitators should do the opposite.

Based on evidence,8 to achieve correct location of the dominant middle finger used to apply all of the force of compression adequately above the xiphoid process, from the right side of the patient, CPR providers need to run the right ring finger along the lower costal margin to locate the end of the bony sternum (or tip of xiphoid process) and place the lower edge of that finger along the xiphisternum (or slightly above the xiphoid tip), which safely positions the middle finger next to it 1 finger-width above the xiphisternum (or sterno-xiphoid junction) and the index finger used only for support two finger widths above the xiphisternum.

Left-handed CPR providers need to do the opposite from the left side of the patient. It can be postulated that the reason why mean compression depth was poorer during TF-CPR (26 mm) versus TT-CPR (33 mm), may be because the index finger, which is directly aligned with the radial side (or weak side) of the heel of the hand, which only applies 1/3 the force of chest compression,9 was predominantly used to apply the force of compression.10 While the TF method produced chest compression forces of as much as 23.7 ± 2.9 N with the left index finger and 22.7±2.1 N with the right index finger, only as much as 16.5±1.9 N was applied with the left middle finger and only 20.4±3.8 N with the right middle finger. Therefore, a conscious effort must be made to use TF-CPR such that one applies all the force with one’s dominant middle finger supported by the index finger next to it, which strengthens it.

However, initially applying a lower force using one’s dominant index finger can be used to one’s advantage to perform neonatal CPR more safely. Chest compressions performed at 1/3 anteroposterior (AP) diameter or 1/2 AP diameter may be too deep in some neonates. Although one analysis of computed tomographic images of neonates predicted overcompression or lack of adequate residual chest depth in 49 of 54 patients with 1/2 AP compression depth but not in patients with 1/4 or 1/3 AP compression depth,11 another analysis revealed that with 33% compression depth, one quarter of patients would experience maximal or overcompression of the mediastinum, completely flattening the heart within the anatomic space available between the sternum and the spine.12

Moreover, even with only 1/4 AP compression depth, adequate compression was predicted in close to 50% (25/54) of the patients.11 To correctly position the dominant index finger from the right side of the patient, CPR providers need to run the right middle finger along the lower costal margin to locate the tip of xiphoid processand place the lower edge of thatfinger slightly above the xiphoid tip, which positions the index finger one finger-width above the xiphisternum. Chest compression is then performed with only the index finger supported by the thumb, which strengthens it. Left-handed CPR providers need to do the opposite.

Resuscitators likely should initially provide compressions with the index finger at a rate of 130 per minute (a metronome should be used to guide the rate), which as evidenced above, could produce a mean compression depth of 26 mm (or approximately 1/4 AP diameter since 26/110.1 mm = .236). The quality of CPR during in-hospital arrestsshould be monitored continuously by measuring cerebral blood oxygenation in the frontal lobe of the brain using near-infrared spectroscopy with a sensor placed on the forehead.13 If this compression depth is found to be inadequate, one should switch to using a correctly located dominant middle finger, which increases the depth of compression because more force may be applied when doing so.

Because paramedics never perform CPR on neonates in the delivery room, CPR in all infants should be performed using only the dominant middle finger. It can be postulated that similar to adults, the reason for highly ineffective chest compression in infants is the majority of the force of compression is applied over the cranial side of the sternum in 50% of all arrests.14 Studies in manikins comparing TT-CPR versus TF-CPR using the techniques described above are needed.


  1. Young KD, Seidel JS. Pediatric cardiopulmonary resuscitation: A collective review. Ann Emerg Med. 1999;33:195–205.
  2. Sirbaugh PE, Pepe PE, Shook JE, et al. A prospective, population-based study of the demographics, epidemiology, management, and outcome of out-of-hospital pediatric cardiopulmonary arrest. Ann Emerg Med. 1999;33:174–184.
  3. Schmölzer GM. Chest Compressions During Sustained Inflation During Cardiopulmonary Resuscitation in Newborn Infants Translating Evidence From Animal Studies to the Bedside. JACC Basic Transl Sci. 2019;4:116-121.
  4. Harrington DJ, Redman CW, Moulden M, et al. The long-term outcome in surviving infants with Apgar zero at 10 minutes: a systematic review of the literature and hospital-based cohort. Am J Obstet Gynecol 2007;196:463.e1–5.
  5. Martin PS, Kemp AM, Theobald PS, Maguire SA, Jones MD. Do chest compressions during simulated infant CPR comply with international recommendations? Arch Dis Child 2013;98:576-81.
  6. Gach MR. How to apply pressure to acupressure points. [accessed 23.03.16].
  7. Chung TN, Kim SW, You JS, Cho YS, Chung SP, Park I. A higher chest compression rate may be necessary for metronome-guided cardiopulmonary resuscitation. Am J Emerg Med. 2012;30:226-30.
  8. Clements F, McGowan J. Finger position for chest compressions in cardiac arrest in infants. Resuscitation. 2000;44:43-6.
  9. Baubin M, Kollmitzer J, Pomaroli A, et al. Force distribution across the heel of the hand during simulated manual chest compression. Resuscitation 1997;35:259-63.
  10. Dellimore K, Heunis S, Gohier F, et al. Development of a diagnostic glove for unobtrusive measurement of chest compression force and depth during neonatal CPR. Conf Proc IEEE Eng Med Biol Soc. 2013;2013:350-3.
  11. Meyer A, Nadkarni V, Pollock A, et al. Evaluation of the Neonatal Resuscitation Program’s recommended chest compression depth using computerized tomography imaging. Resuscitation. 2010;81:544-8.
  12. Babbs CF, Meyer A, Nadkarni V. Neonatal CPR: room at the top–a mathematical study of optimal chest compression frequency versus body size. Resuscitation 2009;80:1280-4.
  13. Yagi T, Kawamorita T, Kuronuma K, et al. Usefulness of a New Device to Monitor Cerebral Blood Oxygenation Using NIRS During Cardiopulmonary Resuscitation in Patients with Cardiac Arrest: A Pilot Study. Adv Exp Med Biol. 2020;1232:323-329.
  14. Rottenberg EM. A simple method of performing maximally effective CPR chest compressions. Resuscitation. 2019;138:213-214.