Article, Radiology

Parenchymal lung injuries related to standard cardiopulmonary resuscitation

a b s t r a c t

Objectives: We analyzed Chest computed tomographic scan to evaluate parenchymal lung injury and its clinical significance in patients who received standard cardiopulmonary resuscitation and were resuscitated from cardiac arrest.

Methods: We enrolled nontraumatic out-of-hospital cardiac arrest patients older than 19 years who had been ad- mitted to the emergency department in cardiac arrest and successfully resuscitated after cardiopulmonary resus- citation. chest computed tomography was obtained immediately after return of spontaneous circulation (ROSC). To allocate the area of Lung contusion, we divided both hemithoraces into 3 regions longitudinally, and each part was subdivided into 4 segments except the lower part of the left lung. To stratify the severity of lung contusion, each segment was scored depending on the area of lung contusion. Oxygen index (OI) was measured at the time of ROSC, 24, 48, and 72 hours and 1 week after cardiac arrest.

Results: Lung contusion was developed in 37 (41%) patients and median lung contusion score (LCS) was 17 (12-26). Lung contusion was not associated with hospital mortality (P = .924) or length of Intensive care unit stay (P = .446). The OI at the time of ROSC was lower in patients with LCS greater than 23 than that in patients with LCS less than or equal to 23 (126 [93-224] vs 278 [202-367]; P = .008); however, the OI at the other timelines was not different be- tween patients with LCS greater than 23 and patients with LCS less than or equal to 23.

Conclusion: Extensive lung contusion is associated with a lower oxygenation index at the time of ROSC, but did not affect the resuscitation outcome.

(C) 2016


Repetitive compression of the sternal area is a standard method of ar- tificial circulation for resuscitating a cardiac arrest victim and was intro- duced more than 50 years ago [1]. Physical injuries of the thorax are unavoidable complications of chest compressions during cardiopulmo- nary resuscitation (CPR) because it requires exerting physical forces to a victim’s thorax [2-4]. A broad spectrum of complications can occur during chest compressions ranging from minor thoracic skeletal injuries to fatal thoracic Organ injuries [5,6]. Skeletal chest injuries including fracturing the sternum or ribs are common complications of chest compressions. In a recent autopsy report, chest compressions caused skeletal chest inju- ries in almost 9 of 10 victims with nonTraumatic cardiac arrests [7]. Current CPR guidelines recommend high-quality chest compressions, which include adequate chest compressions that are at least 5 cm in depth [8-10]. These recommendations might increase the risk of chest in- juries because chest compression with a depth of more than 5 cm is

* Corresponding author at: Department of Emergency Medicine, Wonju College of Med- icine, Yonsei University, 20 Ilsan-ro, Wonju 26426, Republic of Korea. Tel.: +82 33 741

1611; fax: +82 33 734 9994.

E-mail address: [email protected] (S.O. Hwang).

associated with skeletal chest injuries [11]. Skeletal chest injuries could be accompanied by parenchymal lung injury. The occurrence of paren- chymal lung injury might affect oxygenation and ventilation during postcardiac arrest care. However, little is known about parenchymal lung injury associated with chest compression during CPR. Most previous studies investigated complications after CPR using an autopsy or simple x- ray scan [12,13]. Simple radiologic examination of the chest has a limita- tion in evaluating the occurrence or extent of parenchymal lung injury after resuscitation. A chest computed tomographic (CT) scan, which pro- vides comprehensive images of the thoracic structures, is the best modal- ity to evaluate CPR-related acute thoracic complications.

We performed Chest CT scans to evaluate parenchymal lung injury and analyzed their clinical significance in patients resuscitated from cardiac arrest.


Study design

A retrospective cohort study was conducted to evaluate CPR-related lung contusion by analyzing chest CT scans in patients with nontraumatic

0735-6757/(C) 2016

out-of-hospital cardiac arrests who were resuscitated after CPR from July 2006 to March 2010. The study was approved by the Institutional Review Board of Wonju Severance Christian Hospital.

Setting and subjects

This study was conducted at a tertiary care emergency department of a university hospital in Wonju, Republic of Korea. We enrolled nontraumatic out-of-hospital cardiac arrest patients older than 19 years who had been admitted to the emergency department in cardiac arrest and successfully resuscitated after CPR. Patients who failed to achieve hemodynamic stabilization after return of spontaneous circula- tion (ROSC) or patients who had congestive heart failure or preexisting parenchymal pulmonary disease were excluded from the analysis.

Chest CT scan

A 64-channel multidetector CT scanner (Brilliance CT, Philips healthcare systems, Cleveland, OH) was used. Once hemodynamic stabi- lization was achieved, the patient underwent a chest CT scan within a day after resuscitation. The following imaging parameters were used: tube voltage, 120 kV; tube current, 150 mA; detector collimation, 64 x 0.625 mm; pitch, 0.891; gantry rotation time, 0.5 seconds; slice thickness, 5 mm; slice overlap, 0; and matrix, 512 x 512. All studies were performed using the deep inspiration breath hold technique. The CT images were displayed at a window width of 400 HU and window level of 40 HU.

Thoracic complication evaluation

Two emergency physicians and 1 radiologist who were blinded to the patients’ clinical data reviewed the chest CT images and evaluated the occurrence of complications including rib fracture, sternal fracture, lung contusion, pneumothorax, hemothorax, retrosternal hematoma, pneumomediastinum, hemomediastinum, pericardial effusion, and aor- tic injury. If there was a disagreement in the interpretation, then the 3 doctors had a discussion and decided the result.

Assessment of parenchymal lung injury (lung contusion)

Lung contusion was defined as nonsegmental and nonlobar patchy or diffuse density in distribution in lung window of chest CT. If these den- sities were not related to other thoracic injuries, subpleural sparing was considered as a clue for diagnosing lung contusion [14]. Solitary segmen- tal or lobar-distributed air space consolidation was defined as pneumonia, not lung contusion.

We divided the hemithorax into 3 regions (the upper, middle, and lower regions). Each region was subdivided into 4 segments (anteromedial, anterolateral, posteromedial, and posterolateral) except the lower region of the left lung; the lower region of the left lung was subdivided into 3 seg- ments (anterolateral, posteromedial, and posterolateral) due to its relatively

small lung volume. Therefore, both lungs were divided into 23 segments to indicate the area of lung contusion (ALC) and stratify its severity (Fig. 1).

To stratify the severity of lung contusion, each segment was scored de- pending on the ALC; where (1) ALC less than one-third of a segment, (2) ALS equal to or greater than one-third of a segment and less than two- thirds of a segment, and (3) ALC greater than or equal to two-thirds of a segment. The severity was calculated as the sum of the scores of each seg- ment, and it was defined as the lung contusion score (LCS). Severe lung contusion was defined as LCS greater than 23, which meant that the lung contusion was observed in more than one-third of the entire lung.

Definition of CPR-related acute respiratory distress syndrome

To evaluate a relation between CPR-related lung contusion and the de- velopment of Acute respiratory distress syndrome , we performed serial arterial blood gas analyses during the first 48 hours after admission. Acute respiratory distress syndrome was defined as a decrease in the oxy- gen index (OI; partial pressure of oxygen, arterial/fraction of inspired ox- ygen ratio) to lower than 300 mm Hg with a positive end-expiratory pressure within 5 cm H2O, which was not explained by heart failure or volume overload during the first 48 hours after resuscitation.

Data analysis

We analyzed data using SPSS 18.0 for Windows statistical software package (SPSS, Chicago, IL). Continuous data were presented as median with interquartile range and compared with the independent sample t test or Mann-Whitney U test, as appropriate. Nominal data were pre- sented as the percentage of frequency of occurrence and compared with the ?2 or Fisher exact test, as appropriate. The Pearson correlation coefficient was calculated to analyze whether 2 continuous variables were correlated.

Multivariate logistic regression analyses were performed to deter- mine whether age, sex, estimated collapse time, total CPR time, witness, bystander CPR, thoracic circumference, presence of rib fracture, or pres- ence of sternal fracture affected the development of lung contusion or ARDS. The resulting odds ratios (ORs) were presented with 95% confi- dence intervals (CIs). Odds ratios with CIs above 1 were considered as statistically significant. Statistical significance was noted if the P value was less than .05.


General characteristics

A chest CT scan was performed in 113 consecutive patients during the study period. Twenty-two patients were excluded in the analysis due to congestive heart failure (11 patients), incomplete data (6 pa- tients), and preexisting pulmonary disease (5 patients). Of the remain- ing 91 patients, 49 (54%) were male and the median age was 60 years (51-74 years). The total CPR duration was 15 minutes (8-24 minutes).

Fig. 1. Allocation of segments of lung contusion in a chest CT scan and frequencies of lung contusion observed in areas more than two-thirds of each segment. A, upper; B, middle; and C, lower.

Table 1

Baseline characteristics of the study patients


Lung contusion

No lung contusion

P value

(n = 91)

(n = 37)

(n = 54)

Age (y)

60 (51-74)

65 (53-75)

57 (51-72)


Male sex

49 (54)

19 (51)

30 (56)


Estimated collapse time (min)

25 (11-33)

25 (12-37)

25 (15-39)


Total CPR time (min)

15 (8-24)

15 (9-24)

15 (8-24)


Witnessed cardiac arrest

85 (93)

33 (89)

52 (96)


Bystander CPR

33 (36)

15 (41)

18 (33)


Thoracic circumference (cm)

87 (82-93)

88 (83-97)

86 (82-92)


Initial presenting rhythm–no. of patients

Ventricular fibrillation or tachycardia

18 (20)

6 (16)

12 (22)



42 (46)

19 (51)

23 (43)

Pulseless electrical activity

31 (34)

12 (32)

19 (35)

survival discharge–no. of patients

62 (68)

25 (68)

37 (69)


Good neurologic outcome (CPC 1-2)–no. of patients

12 (13)

4 (11)

8 (15)


Continuous variables were presented as median (interquartile range) and nominal variables were presented as frequency (%). Abbreviation: CPC, cerebral performance category.

Cardiac arrest was witnessed in 85 (93%) patients. Thirty-three (36%) patients received CPR from a bystander. The median thoracic circumfer- ence of the patients was 86.9 cm (81.7-92.5 cm). The initial presenting rhythm was ventricular fibrillation or pulseless ventricular tachycardia in 14 (13%) patients. Sixty-two (68%) patients were discharged alive and 12 (13%) patients had a good neurological outcome. Acute respira- tory distress syndrome was observed in 80 (87.9%) patients until 48 hours after resuscitation. There were no statistical differences in general characteristics between the patients with lung contusion and the pa- tients without lung contusion (Table 1).

Cardiopulmonary resuscitationrelated thoracic injuries

Rib fractures were diagnosed in 48 (53%) patients. Thirty-seven (41%) patients developed lung contusion, which was the second most frequent thoracic complication. Other CPR-related thoracic injuries were sternal fracture in 11 (12%) patients, hemothorax in 10 (11%) pa- tients, pneumothorax in 9 (10%) patients, retrosternal hematoma in 3 (3%) patients, hemomediastinum in 2 (2%) patients, and pericardial ef-

fusion in 1 (1%) patient.

Disagreement for diagnosing lung contusion occurred in 13 seg- ments of 10 patients. Three doctors discussed together and then 9 seg- ments were diagnosed as lung contusion and 4 segments were diagnosed as pneumonia. There were no disagreements for diagnosing the other injuries.

Cardiopulmonary resuscitationrelated lung contusion: distribution and severity

Thirty-seven (41%) patients developed lung contusion. The median number of segments with lung contusion was 13 (10-16). Among pa- tients with lung contusion, bilateral lung contusions were present in 35 (95%) patients. Anatomically, the posterolateral segment of the lower region of the left lung was most frequently involved (n = 35; 95%). The posteromedial and posterolateral segments of the middle re- gion of the left lung were also frequently involved (n = 34; 92%). The segments of lung contusion with ALC greater than or equal to 2 were lo- cated in the posterior half of both lungs. Lung contusions with ALC equal to 3 were most frequently observed in the posteromedial segment of the middle region of the left lung (n = 11; 30%) (Fig. 2).

The LCS range observed was from 1 to 58 among patients with lung contusion, and the median LCS was 17 (12-26) (Fig. 2).

Factors related to lung contusion and ARDS

Age, sex, total estimated collapse time, CPR duration, witness, by- stander CPR, presence of rib fracture, and presence of sternal fracture

were not related to CPR-related lung contusion and the development of ARDS (Table 2). The LCS did not correlate with age (r = 0.190; P =

.261), total estimated collapse time (r = 0.235; P = .201), CPR duration (r = 0.285; P = .097), thoracic circumference (r = 0.070; P = .681), or total number of rib fractures (r = 0.150; P = .375).

Severity of lung contusion and oxygen index

The OI at the time of ROSC was lower in patients with LCS greater than 23 than that in patients with LCS less than or equal to 23 (126 [93-224] vs 278 [202-367]; P = .008); however, the OI at 24, 48, and 72 hours and 1 week after ROSC was not different between patients with LCS greater than 23 and patients with LCS less than or equal to 23 (Fig. 3).


Lung contusion or LCS greater than 23 was not associated with hospital mortality (P = .924 and .539, respectively) or length of ICU stay (P = .446 and .203, respectively).


In this study, we found that parenchymal lung injury is a common complication of standard CPR and is related to the development of ARDS during early postresuscitation care. Interestingly, the presence and severity of parenchymal lung injury after CPR were not associated with survival.

Standard CPR may be complicated with various injuries of the paren- chymal organs as well as skeletal structures. It is well known that rib fractures are the most common CPR-related injury of the thorax [3- 5,7]. The lungs are the largest organs in the thorax. Repetitive exertions of force to the thorax that must be transmitted to the lung during chest compressions may result in parenchymal lung injury. Unlike thoracic

Fig. 2. Distribution of lung contusion scores in patients with lung contusion.

Table 2

Multivariate analysis of factors related to lung contusion and ARD.

Lung contusion



95% CI

P value


95% CI

P value








Male sex







Estimated collapse time (min)







Total CPR time (min)







Witnessed cardiac arrest







Bystander CPR







Thoracic circumference (cm)







Rib fracture







Sternal fracture







Lung contusion




skeletal injuries, parenchymal lung injuries associated with CPR were rarely addressed so far. Furthermore, clinical significance of parenchy- mal lung injury that developed after CPR has never been explored.

Cardiopulmonary resuscitation-related lung contusion was primar- ily observed in the posterior half of the lung. This feature was similarly observed in a previous CT-based study [15]. It may be postulated that lung contusions in the posterior region of the lung occur because an in- crease in the intrathoracic pressure generated by chest compressions exerts the area of the lung with a high hydrostatic pressure. During CPR, the patient lies in the supine position and blood is redistributed to the dependent part of the thorax, which might produce higher hydro- static pressure in posterior half of the lung than in the anterior half of the lung [16]. In addition, positive airway pressure generated by artifi- cial ventilation during CPR might be transmitted posteriorly due to the forceful chest compressions. Extrinsic compressions on the posterior part of the lung with a high hydrostatic pressure would make the lung parenchyma more vulnerable to injury.

It is important to evaluate and treat ARDS in critically ill patients be- cause it is closely related to long hospital stay and mortality [17,18]. In general, it is known that lung contusion from blunt chest trauma is strongly related to ARDS and outcomes [19]. In this study, lung contu- sion after CPR was associated with the development of ARDS, but it did not result in a poor outcome. This discrepancy in outcomes might be attributed to the mechanisms of lung contusions. Lung contusion from blunt chest trauma is mainly caused by a parenchymal lung lacer- ation secondary to sudden impact to the chest, whereas lung contusion secondary to CPR might be associated with fluctuations of the intratho- racic pressure caused by repetitive compressions of the thorax with a limited extent (less than 6 cm in depth) [20]. However, the exact mech- anism of lung contusion during CPR needs to be investigated.

In this study, we found that severe lung contusion that developed after CPR was associated with hypoxemia in the early phase of postresuscitation

care. Hypoxia in patients with ROSC after cardiac arrest is associated with a poor neurological outcome [21]. The CPR guidelines recommend avoiding hypoxia during postcardiac arrest care [22-24]. Physicians should pay close attention to the potential development of hypoxemia in the presence of lung contusion because optimal oxygenation is important during postresuscitation care.


This study had several limitations. First, this is a retrospective cohort study; therefore, it is possible that unforeseen confounders might have been included even though data related to cardiac arrest and resuscitation were prospectively collected based on Utstein reporting [25]. Second, a small sample size from 1 institution is a limitation and is insufficient to de- termine the factors related to outcomes or the development of ARDS. Third, there is a possibility that patients with pulmonary edema were included in the analysis. Although we excluded patients with any history or previous symptoms suggestive of congestive heart failure, postresuscitation myocar- dial dysfunction may occur during the early phase of postcardiac arrest.


Lung contusion is a common complication of standard CPR, which is associated with ARDS during the early phase of postcardiac arrest.


  1. Kouwenhoven WB, Jude JR, Knickerbocker GG. Closed-chest cardiac massage. JAMA 1960;173:1064-7.
  2. Buschmann CT, Tsokos M. Frequent and rare complications of resuscitation at- tempts. Intensive Care Med 2009;35:397-404.
  3. Hoke RS, Chamberlain D. Skeletal chest injuries secondary to cardiopulmonary re- suscitation. Resuscitation 2004;63:327-38.

    Fig. 3. Comparison of the oxygen index between the LCS less than 23 group and LCS greater than or equal to 23 group.

    Kim EY, Yang HJ, Sung YM, Cho SH, Kim JH, Kim HS, et al. Multidetector CT findings of skeletal chest injuries secondary to cardiopulmonary resuscitation. Resuscitation 2011;82:1285-8.

  4. Meron G, Kurkciyan I, Sterz F, Susani M, Domanovits H, Tobler K, et al. Cardiopulmo- nary resuscitation-associated major liver injury. Resuscitation 2007;75:445-53.
  5. Machii M, Inaba H, Nakae H, Suzuki I, Tanaka H. Cardiac rupture by penetration of fractured sternum: a rare complication of cardiopulmonary resuscitation. Resuscita- tion 2000;43:151-3.
  6. Rutty GN, Morgan B. Frequency and number of resuscitation related rib and sternum fractures are higher than generally considered. Resuscitation 2015;93:A1-2.
  7. Kleinman ME, Brennan EE, Goldberger ZD, Swor RA, Terry M, Bobrow BJ, et al. Part 5: adult basic life support and cardiopulmonary Resuscitation quality: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emer- gency cardiovascular care. Circulation 2015;132:S414-35.
  8. Perkins GD, Handley AJ, Koster RW, Castren M, Smyth MA, Olasveengen T, et al. European Resuscitation Council guidelines for resuscitation 2015: section 2. Adult basic life support and Automated external defibrillation. Resuscitation 2015;95:81-99.
  9. Song KJ, Kim J-B, Kim J, Kim C, Park SY, Lee CH, et al. Part 2. Adult basic life support: 2015 Korean guidelines for cardiopulmonary resuscitation. Clin Exp Emerg Med 2016;3:S10-6.
  10. Hellevuo H, Sainio M, Nevalainen R, Huhtala H, Olkkola KT, Tenhunen J, et al. Deeper chest compression–more complications for cardiac arrest patients? Resuscitation 2013;84:760-5.
  11. Krischer JP, Fine EG, Davis JH, Nagel EL. Complications of Cardiac resuscitation. Chest 1987;92:287-91.
  12. Himmelhoch SR, Dekker A, Gazzaniga AB, Like AA. Closed-chest cardiac resuscita- tion. A prospective clinical and pathological study. N Engl J Med 1964;270:118-22.
  13. Donnelly LF, Klosterman LA. Subpleural sparing: a CT finding of lung contusion in children. Radiology 1997;204:385-7.
  14. Cho SH, Kim EY, Choi SJ, Kim YK, Sung YM, Choi HY, et al. Multidetector CT and ra- diographic findings of lung injuries secondary to cardiopulmonary resuscitation. In- jury 2013;44:1204-7.
  15. Nyren S, Radell P, Lindahl SG, Mure M, Petersson J, Larsson SA, et al. Lung ventilation and perfusion in prone and supine postures with reference to anesthetized and me- chanically ventilated healthy volunteers. Anesthesiology 2010;112:682-7.
  16. Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med 2000; 342:1334-49.
  17. Croce MA, Fabian TC, Davis KA, Gavin TJ. Early and late acute respiratory distress syndrome: two distinct clinical entities. J Trauma 1999;46:361-6 [discus- sion 6-8].
  18. Miller PR, Croce MA, Kilgo PD, Scott J, Fabian TC. Acute respiratory distress syndrome in blunt trauma: identification of independent risk factors. Am Surg 2002;68:845-50 [discussion 50-1].
  19. Miller DL, Mansour KA. Blunt traumatic lung injuries. Thorac Surg Clin 2007;17: 57-61 [vi].
  20. Kilgannon JH, Jones AE, Shapiro NI, Angelos MG, Milcarek B, Hunter K, et al. Associ- ation between arterial hyperoxia following resuscitation from cardiac arrest and in- hospital mortality. JAMA 2010;303:2165-71.
  21. Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, et al. Part 8: post-cardiac arrest care: 2015 American Heart Association guidelines update for car- diopulmonary resuscitation and emergency cardiovascular care. Circulation 2015; 132:S465-82.
  22. Nolan JP, Soar J, Cariou A, Cronberg T, Moulaert VR, Deakin CD, et al. European Resus- citation Council and European Society of Intensive Care Medicine guidelines for post-resuscitation care 2015: section 5 of the European resuscitation council guide- lines for resuscitation 2015. Resuscitation 2015;95:202-22.
  23. Kim Y-M, Park KN, Choi SP, Lee BK, Park K, Kim J, et al. Part 4. Post-cardiac arrest care: 2015 Korean guidelines for cardiopulmonary resuscitation. Clin Exp Emerg Med 2016;3:S27-38.
  24. Peberdy MA, Cretikos M, Abella BS, DeVita M, Goldhill D, Kloeck W, et al. Recom- mended guidelines for monitoring, reporting, and conducting research on medical emergency team, outreach, and rapid response systems: an Utstein-style scientific statement: a scientific statement from the International Liaison Committee on Re- suscitation (American Heart Association, Australian Resuscitation Council, European Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa, and the New Zealand Resuscitation Council); the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiopulmonary, Perioperative, and Critical Care; and the Interdisciplinary Working Group on Quality of Care and Outcomes Research. Circulation 2007;116:2481-500.

Leave a Reply

Your email address will not be published. Required fields are marked *