Article, Pediatrics

A new chest compression depth indicator would increase compression depth without increasing overcompression risk

a b s t r a c t

Purpose: Adequate chest compression depth is critical for effective cardiopulmonary resuscitation. pediatric resuscitation guidelines recommend that CC be at least one-third of the anterior-posterior (AP) chest diameter or approximately 4 cm in infants and 5 cm in children. We aimed to find a better indicator of CC depth that maximizes CC depth while also minimizing injury.

Basic procedures: Chest computed tomographic images of patients aged 8 years and younger were measured for external diameter (ED) (AP distance from skin to skin) and internal diameter (AP distance between internal surface of anterior chest wall and anterior surface of vertebral body) at the midway of the lower half of the sternum. Compressible depth was defined as 1 cm short of internal diameter. We determined that up to a 10% estimated risk of overcompression is acceptable and approximated a quantile regression line for the 10th percentile of compressible depth on ED. After rounding coefficients, we used its equation as a new indicator. Main findings: A total of 426 images were analyzed. The new indicator had a slope of 0.5 and an intercept of -1.9 cm (1 fingerbreadth). Compared to one-third ED, the new indicator would provide deeper CC with average difference of 1.9 mm (95% confidence interval, 1.6-2.2 mm) without increasing the risk of overcompression (both 4.9%). Chest compression of 4/5 cm would provide deeper CC compared to the new indicator (difference, 3.5 mm; 95% confidence interval, 2.7-4.1 mm); however, its overcompression risk was too high (31.5%).

Principal conclusion: Chest compression of one-half ED minus 1 fingerbreadth maximizes CC depth without increas- ing overcompression in pediatric population.

Introduction

Adequate chest compression is critical in both adult and Pediatric cardiopulmonary resuscitation (CPR) [1-3]. To achieve the goal, maximizing CC depth without increasing the risk of overcompression would be important. However, it is unclear whether current guidelines are the best ones to facilitate the goal. They recom- mend that compressions be at least one-third of the anterior-posterior (AP) chest diameter or approximately 4 cm in infants and 5 cm in chil- dren [4,5]. Their scientific background is based on the result of recent studies where the safety of previous CC depth, one-third to one-half

? Prior presentations: None.

?? Funding sources: None.

? Conflicts of interest: All authors do not have any conflicts of interest.

* Corresponding author at: Department of Emergency Medicine, Seoul National University, Bundang Hospital, 166 Gumi-ro, Bundang-gu, Seongnam-si, Gyeonggi-do, 463-707, Republic of Korea. Tel.: +82 31 787 7578; fax: +82 31 787 4055.

E-mail address: [email protected] (J. Kim).

¹ Contributed equally as the first author.

² Contributed equally as the corresponding author.

AP chest diameter, was tested using anthropometric measurement of pediatric population [6-8]. However, even the studies did not try to develop a new optimized CC indicator using their data. Although we now know that one-third AP diameter is safe target, we are not still sure about whether it is the best option for providing adequate CC depth in pediatric population where patients’ body sizes are hugely varied.

Our main study question was whether there would be any room for improvement in current depth indicators. We postulated that, by more closely approximating the anthropometric data, we may find a new in- dicator of depth goal that will help us to maximize CC depth while min- imizing the potential injury from overzealous CC. Therefore, the primary goal of this study was to find a better indicator for minimum CC depth in a young (<= 8 years old) pediatric population. The secondary goal was to provide comparative analysis data about the potential harm and benefit (increase in CC depth) of different CC depth indicators in the population.

Materials and methods

This was a retrospective study that analyzed chest computed tomo- graphic (CT) scans of infants and young children (<= 8 years old) that had been taken at an academic teaching hospital. The institutional review

http://dx.doi.org/10.1016/j.ajem.2015.08.014

computed tomographic measurements a”>Table 1

Patient characteristics

Age	n	Male (%)	Height (cm, IQR)	Weight (kg, IQR)	ED (cm, IQR)	ID (cm, IQR)
b3 mo	31	17 (54.8)	53.0 (50.5-57.3)	4.4 (3.1-5.4)	9.0 (8.5-10.0)	4.9 (4.4-5.1)
3 mo to 1 y	45	26 (57.8)	69.0 (66.0-73.0)	8.3 (7.0-9.6)	10.7 (10.1-11.5)	5.4 (4.8-5.8)
1	68	47 (69.1)	82.0 (78.6-86.0)	11.0 (9.8-12.3)	11.5 (11.0-11.9)	6.0 (5.5-6.4)
2	44	21 (47.7)	89.9 (88.0-91.8)	12.5 (11.8-13.6)	11.9 (11.5-12.5)	6.0 (5.5-6.5)
3	46	26 (56.5)	98.1 (95.0-104.0)	15.6 (13.5-17.0)	12.7 (11.8-13.2)	6.5 (5.8-6.9)
4	49	27 (55.1)	105.9 (101.1-110.3)	17.8 (15.4-19.5)	13.3 (12.7-14.2)	6.5 (6.2-7.4)
5	43	21 (48.8)	113.5 (110.0-119.9)	19.3 (17.0-22.6)	13.5 (12.5-14.5)	6.8 (6.3-7.1)
6	43	25 (58.1)	118.4 (115.1-123.4)	22.0 (18.7-24.0)	14.3 (13.3-14.7)	7.3 (6.5-7.8)
7	30	17 (56.7)	125.0 (118.6-131.6)	26.3 (22.4-30.5)	14.7 (13.9-15.6)	7.6 (6.9-8.2)
8	27	16 (59.3)	127.3 (123.7-132.9)	25.0 (23.4-31.0)	14.8 (13.6-16.4)	7.9 (7.1-8.5)

board at the study hospital approved the analysis and provided a waiver of consent.

Patients and data collection

The inclusion criterion was Chest CT scan images of patients aged 8 years and younger that had been taken at the study hospital from its opening in May 2003 until August 2013. Exclusion criteria were as follows: (1) any documented chest wall deformity, including both pectus excavatum and carinatum; (2) any other congenital anomaly syndrome, regardless of known chest wall deformity; and (3) significant chest wall trauma. The patients’ age at the timing of the CT scan, sex, height, and body weight were gathered from chart reviews.

Computed tomographic measurements and definition of depth-related variables

All chest CT scans of the eligible patients were reviewed using the picture archiving and communication system. External diameter (ED) (AP distance from skin to skin) and internal diameter (ID) (AP distance between internal surface of anterior chest wall and anterior surface of vertebral body) were measured at the midway of the lower half of the sternum by drawing a perpendicular line. The measuring method was the same as that of the previous study by Braga et al [7], except that the axial level of the measurement was the midsternum in the previous study. Secondary variables derived from the measurements were as follows: (1) compressible depth, 1 cm short of ID, as previously adopted in other studies [7,8]; (2) presence of the risk of overcompression, positive if CC depth suggested by an indicator of interest exceeds compressible depth; and (3) remaining depth, depth of potentially remaining space if a chest is compressed to the point as indicated by an indicator of interest.

Statistical analysis

We determined that up to a 10% risk of overcompression is accept- able and approximated a quantile regression line for the 10th percentile of the compressible depth on ED. After rounding coefficients of the fitted regression line, we used its equation as a new indicator of CC depth. To evaluate its potential implication for the resuscitation of infants and young children, we compared it with the conventional indicators in terms of the aspects of estimated risk of overcompression, remaining depth, and overall Depth of CC.

Continuous variables are reported using mean and SD, unless there was significant nonnormality when the median and interquartile range (IQR) were used. Paired t test or Wilcoxon Signed Rank Test were used as appropriate to test the significance of the difference of various depth-related measurements. ?² Test was performed to test the significance of the difference of the proportion of patients who were at risk for overcompression. The association between age category and relative frequency of patients at risk for overcompression was test- ed with the Cochrane-Armitage trend test, whereas the association

between age category and various depth-related measurements was tested with Spearman ?. All analyses were performed using R package version 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Nine hundred thirty-one chest CT scans were initially identified as meeting the inclusion criteria. After excluding 475 scans based on the exclusion criteria, 453 CT scans underwent the measurement process. During the process, 30 scans were additionally excluded, and a total of 426 scans were ultimately eligible for analysis. Table 1 summarizes the height; body weight; and measurements of ED, ID, and compressible depth.

Fig. 1 illustrates the linear relationship between ED and compress- ible depth. Simple Linear regression analysis showed that the predicted compressible depth was found to be 0.526 x ED - 1.257 cm (R² = 0.758; P b .001, not shown in the figure). To develop a CC depth indica- tor with a 10% risk of overcompression, we approximated a quantile regression line for the 10th percentile of compressible depth on ED. The regression line had a slope of 0.523 and an ED intercept of - 1.928 cm. As for the 25th, 50th, and 75th percentile lines, the slopes were 0.529, 0.532, and 0.538, respectively, and the intercepts were -1.599, -1.295, and -1.040 cm, respectively.

Fig. 1. Scatterplot demonstrating the relationship between ED and compressible depth. Black solid line: 10% quantile regression line; black dashed line: 50% quantile regression line; gray solid line: one-third of ED; gray dashed line (oblique): one-half of ED; gray dashed lines (horizontal): 4 cm and 5 cm absolute compression depth.

Fig. 2. Estimated risk of overcompression in varying age groups.

We devised a simple equation for the CC depth indicator by rounding the coefficients of the 10th percentile fitted line. It had a slope of 0.5 and an intercept of -1.9 cm, which happen to be the same length of 1 fingerbreadth. We calculated the estimated risk of overcompression when various CC depth goal indicators were used (Fig. 2). Our new indi- cator was found to have 21 (4.9%) of 426 chance of overcompression, which was the same with that of one-third ED (21/426; 4.9%) but significantly lower compared to that of the 4 cm for infants and 5 cm for children indicator (4/5 cm indicator; 134/426; 31.5%; P b .001). The estimated risks of overcompression of the new indicator was not associated with age category (P trend = .929), whereas that of conventional indicators were significantly associated with age category (P trend = .014 and P trend <= .001, respectively) with higher risk in younger age groups. The 4/5 cm indicator showed particularly high risk for overcompression in infants and Younger children (age b 3 years) with almost half of them (92/188; 48.9%) were at risk. For other CC depth goals such as one-fourth and one-half ED, the overall estimated risks of overcompression were 1 (0.2%) of 426 and 407

(95.3%) of 426, respectively.

Fig. 3 illustrates the estimated remaining compressible depth when the conventional (one-third ED and 4/5 cm) and the new indicator were used. With conventional depth indicators, there were increasing trends of remaining depth as age increased (Spearman ? = 0.412 and 0.517, both P b .001). In contrast, there was no such trend between remaining depth and age when the new depth indicator was used (Spearman ? = 0.035; P = .468). chest compression depth of the new indicator was compared to those of conventional depth indicators, and

their differences are plotted in Fig. 4. Compared to one-third ED indicator, the new indicator would provide significantly deeper CC depth with average difference of 1.9 mm (95% confidence interval, 1.6-2.2; P b .001). However, its CC depth was shallower than that of 4/5 cm indicator with average difference of 3.5 mm (95% confidence

interval, 2.7-4.1; P b .001).

Discussion

In the current study, we devised and tested a new indicator, which is one-half ED minus 1 fingerbreadth (1.9 cm). It would provide signifi- cantly deeper CC compared to one-third ED while maintaining the same low risk of overcompression (4.9%). As the new indicator was de- veloped closely following human anthropometric measurements, its risk of overcompression was evened out over pediatric age range, which was in contrast to other conventional indicators where they were significantly affected by the patients’ age. Therefore, we suggest that if a single indicator for minimum CC depth is recommended for pediatric population, our new indicator, which is one-half ED minus 1 fingerbreadth (1.9 cm), might be a reasonable choice.

There have been several changes in the recommended depth of CC

for Pediatric cardiac arrest patients since the first development of na- tional guideline in the United States in 1966 (Table 2) [9-15]. In the first guideline published in 1966, a relative depth indicator, one-fifth of AP diameter, was recommended. Absolute depth indicators were adopted in the next guideline published in 1974, where 1/2 to 3/4 inches (13-19 mm) for infant and 3/4 to 1 1/2 inches (19-38 mm) for children

Fig. 3. Remaining compressible depth in varying age groups.

Fig. 4. Difference of CC depth between current depth indicators and the new indicator.

were recommended. The recommended depths were somewhat in- creased to 1/2 to 1 inches (13-25 mm) for infants and 1 to 1 1/2 inches (25-38 mm) for children in the 1980 guideline and were maintained for 20 years until 2000 where a relative depth indicator, 1/3 to 1/2 of AP di- ameter for both infant and children, reappeared because of expert con- sensus opinion [16]. For 10 years after the change, the relative depth indicator was maintained, and it changed slightly to current form in the 2010 guideline because of recent CT-based studies reporting one-half of AP diameter might be too deep [6,7].

Recently, Sutton et al [3] reported that deeper CC (2010 American Heart Association guideline-compliant CC depth, >= 51 mm) is associ- ated with better survival in children older than 1 year. Although most of the study patients were aged 8 years or older (70/78; 89.7%), we now have a clinical evidence for pursuing deeper CC in pediatric population [17]. However, CC that is too deep might lead

to patient injury and, therefore, balancing these 2 goals for the best outcome would be important. Currently, there is no consensus re- garding which level of overcompression risk is acceptable in the pe- diatric population. If the 75th percentile regression line of this study is any indication, which had a slope of 0.538 and intercept of - 1.040, the upper margin of compressible depth should not be deeper than one-half AP diameter minus 1 cm.

One interesting finding in the current study was that 4/5 cm CC depth would cause overcompression frequently (almost up to half of cases) in younger age groups (age b 3 years). Braga et al [7] reported a similar finding, where 45% of noninfant children aged 1 to 3 years were at risk for overcompression. As guidelines recommend to com- press the chest “approximately,” not “at least,” 4 cm in infants and 5 cm in children, this might not be a serious problem. However, using 2 different indicators might cause confusion. We think that our new

Table 2

Change history of national Guideline recommendations on CC for pediatric cardiac arrest patients

Year	Age group	Anatomical location	Locating method	Depth	Rate	Ratio	compression-only CPR
1966	Infant	Midsternum	Not specified	1/5 of AP diameter	100 per minute	5:1	Not allowed
	Children
1974	Infant	Midsternum	Not specified	1/2 to 3/4 in (13-19 mm)	80-100 per minute	5:1	Not allowed
	Children			3/4 to 1 1/2 in (19-38 mm)
1980	Infant	Midsternum	Nipple line	1/2 to 1 in (13-25 mm)	100 per minute	5:1	Not allowed
	Children		Not specified	1 to 1 1/2 in (25-38 mm)	80 per minute
1986	Infant	Not specified	One finger’s width	1/2 to 1 in (13-25 mm)	At least 100 per minute	5:1	Not allowed
			below nipple line
	Children		One finger’s width above substernal notch^a	1 to 1 1/2 in (25-38 mm)	80-100 per minute
1992	Infant	Lower third	One finger’s width	1/2 to 1 in (13-25 mm)	At least 100 per minute	5:1	Not allowed
		of sternum	below nipple line
2000	Children Infant	Lower half	One finger’s width above substernal notch^a One finger’s width	1 to 1 1/2 in (25-38 mm) 1/3 to 1/2 of AP diameter^b	100 per minute At least 100 per minute	5:1	Not allowed
		of sternum	below nipple line
2005	Children Infant	Lower half	Not specified Just below nipple line	1/3 to 1/2 of AP diameter^b	100 per minute 100 per minute	30:2	Better
	Children	of sternum	Not specified			(15:2 with	than nothing
						2 rescuers)
2010	Infant	Lower half	Just below nipple line	At least 1/3 of AP diameter	At least 100 per minute	30:2	Better
		of sternum		or about 1 1/2 in (4 cm)		(15:2 with	than nothing
	Children		Not specified	At least 1/3 of AP diameter		2 rescuers)
				or about 2 in (5 cm)

^a With middle finger on the substernal notch, index finger is placed next to the middle finger and then the heel of the same hand is placed next to point where the index finger was located.

^b The 2000 guideline stated that this will correspond to a depth of 1/2 to 1 inches (13-25 mm) in infant and 1 to 1 1/2 inches (25-38 mm) in children with a caveat that these measurements are not precise.

indicator, which would provide deeper CC with balanced overcompression risk profile, might be a good candidate for a unified CC depth goal.

This study has many limitations. First, this was a retrospective, ob-

servational study that used the CT scans of infants and young children (aged <= 8 years). In this age group, undergoing CT scanning is a quite un- usual event, which may indicate significant underlying comorbidities. Although we excluded patients with any documented chest wall defor- mity, congenital anomaly syndrome, or significant chest wall trauma, there is the possibility of selection bias. Second, because ventilatory sta- tus could not be controlled, the measurements of the depth parameters might not correctly represent those of cardiac arrest patients. Third, the level of acceptable risk was too arbitrary. To estimate the level that will provide the best outcome, we need to know the relative benefit of overcompression compared to the possible harm from intrathoracic injury, of which we have no knowledge. Fourth, molding of the chest during CPR, which might disproportionately reduce ID, was not considered. Last but most important, it is unknown whether the theoretical difference of CC depth (1.9 mm) would lead to any meaning- ful improvement in outcome. As a first step to test the idea, we are currently planning a Mannequin study.

In conclusion, CC of half of AP diameter minus 1 fingerbreadth (1.9 cm) can provide deeper CC without increasing risk of overcompression compared to one-third ED. The 4/5 cm CC depth has significant risk of overcompression in younger age group (age b 3 years) and should be used with caution. Further research comparing various depth targets is warranted to improve the outcome of pediatric cardiac arrest patients.

References

AJEM