Chest wall injuries following cardiopulmonary resuscitation

Resuscitation 63 (2004) 339–343

Short communication

Chest wall injuries following cardiopulmonary resuscitation Catherine J. Blacka,∗ , Anthony Busuttilb , Colin Robertsonc a

Department of Pathology, Level 2, The Royal Infirmary of Edinburgh, 51 Little France Crescent, Edinburgh EH164SA, UK Department of Pathology, Forensic Medicine Section, The University of Edinburgh, Teviot Place, Edinburgh EH89AG, UK Department of Accident and Emergency Medicine, The Royal Infirmary of Edinburgh, 51 Little France Crescent, Edinburgh EH164SA, UK b

c

Received 20 February 2004; received in revised form 11 June 2004; accepted 13 July 2004

Abstract The forensic records were reviewed of 1823 deaths referred to Edinburgh City Mortuary for autopsy over a 15-month period, 2000–2001; 499 cases (343 males, 156 females) that received CPR prior to death were studied. Rib fractures were found in 29%, sternal fracture in 14%, and 11% of cases showed external chest wall bruising or abrasion. More females sustained rib fractures than males (37% versus 26%; P < 0.05). There was no significant gender difference for sternal fracture (females 17%, males 12%; P = 0.051). The incidence of rib fractures increased with age (P < 0.001). There was no significant difference in the number of left or right ribs fractured (P = 0.631). This study incorporates all cases of in and out-of-hospital CPR and does not discriminate for the CPR provider or technique employed, therefore, providing a current and representative overview of the incidence of rib and sternal fractures in non-survivors of CPR. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Cardiopulmonary resuscitation (CPR); Rib; Sternum; Fracture; Autopsy

Resumo Foram revistos os registos forenses de 1823 mortes referidas para aut´opsia no Instituto da cidade de Edimburgo durante um per´ıodo de 15 meses, 2000–2001; 499 casos (343 masculinos, 156 femininos) foram reanimados com CPR antes da morte. Foram encontradas fracturas costais em 29%, fracturas esternais em 14%, e 11% dos casos evidenciaram abras˜oes e escoriac¸o˜ es externas da parede tor´acica. Evidenciaramse mais fracturas costais nas mulheres do que nos homens (37% versus 26%; P < 0.05). N˜ao existiu diferenc¸a significativa entre os sexos no que se refere a fracturas esternais (feminino 17%, masculino 12%; P = 0.051). A incidˆencia de fracturas costais aumentou com a idade (P < 0.001). N˜ao existe diferenc¸a significativa no n´umero de fracturas costais esquerdas os direitas (P = 0.631). Este estudo incorporou todos os casos de CPR intra e extra-hospitalar e n˜ao descrimina quem realiza as manobras de reanimac¸a˜ o nem a t´ecnica utilizada, contudo, fornece uma perspectiva corrente e representativa da incidˆencia de fracturas costais e esternais numa populac¸a˜ o n˜ao seleccionada. © 2004 Elsevier Ireland Ltd. All rights reserved. Palavras chave: Reanimac¸a˜ o cardio-pulmonar; Costais; Esternais; Fracturas; Aut´opsia

Resumen Revisamos los datos de autopsia de 1823 fallecidos, referidos para autopsia al servicio de anatom´ıa patol´ogica de Edimburgo, en un periodo de 15 meses, en 2000 y 2001; se estudiaron 499 casos (343 varones, 156 mujeres) que recibieron reanimaci´on cardiopulmonar (CPR) antes de morir. Se encontraron fracturas costales en 29%, fractura esternal en 14%, y en 11% de los casos se encontr´o abrasi´on o equimosis de la pared tor´acica externa. Mas mujeres presentaron fracturas costales que varones (37% versus 26%; P < 0.05). No hubo diferencia significativa para fractura esternal seg´un sexo (mujeres 17%, varones 12%; P = 0.051). La incidencia de fracturas costales aumento con la edad

∗

Corresponding author. Tel.: +44 131 242 7136; fax: +44 131 242 7146. E-mail address: [email protected] (C.J. Black).

0300-9572/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2004.07.005

340

C.J. Black et al. / Resuscitation 63 (2004) 339–343

(P < 0.001). No hubo diferencia entre el numero de costillas fracturadas a izquierda o derecha (P = 0.631). Este estudio incorpora todos los casos de CPR intra y extra hospitalaria y no discrimina por tipo de proveedor de CPR o t´ecnica empleada, por lo tanto, proporciona una visi´on actual y representativa de la incidencia de fracturas costales y esternales en poblaci´on no seleccionada. © 2004 Elsevier Ireland Ltd. All rights reserved. Palabras clave: Preanimaci´on cardiopulmonar (RCP); Costilla; Estern´on; Fractura; Autopsia

1. Introduction Cardiopulmonary resuscitation (CPR) was first described by Kouwenhoven and Jude [1] and was introduced into clinical practice in the 1960s. CPR is widely practised both by lay persons in the community (bystander CPR) and by trained advanced life support (ALS) providers, most often in ambulance and hospital settings. The complications of CPR can be divided into those affecting the chest wall, e.g. fracture of the ribs and sternum [2,3], injuries to internal organs especially the liver [4], spleen [5], lung and heart [6,7] and those secondary to intubation and ventilation such as pneumonia [8], tracheal injury [9], gastric rupture and pneumoperitoneum [10,11]. The most commonly reported complications are fractures of the ribs and sternum. Rarely these may cause injury of sufficient severity to contribute to death, e.g. pneumothorax or cardiac laceration due to rib or sternal fracture [12,13]. Previous studies of the incidence of skeletal injuries have focused either on in-hospital or out-of-hospital cardiac arrests. This study does not discriminate for the level of CPR provision, whether the locus was in- or out-of-hospital or for the CPR technique employed. This study therefore provides a current and overall estimation of the incidence of chest wall injuries in non-survivors of CPR as assessed by autopsy.

2. Materials and methods The forensic records of 1823 deaths referred to The Forensic Medicine Section, The University of Edinburgh for autopsy over 15 months in 2000–2001 were reviewed. These were sudden unexpected deaths, uncertified deaths or deaths due to non-natural causes; most deaths were in the first two categories. Autopsies (Rokitansky–Aschoff technique) were carried out on instruction of the Procurator Fiscal by Consultant grade Pathologists directly, or under their supervision by junior colleagues. A standard protocol was used for reporting the autopsy. Cases were included for analysis where the forensic case notes indicated there had been attempts at CPR, age 18 years and over and no historical or physical evidence that trauma (e.g. road traffic accident) was the cause of death. A total of 499 cases (343 males, 156 females) were eligible for inclusion in the study. Data obtained included demographic data, cause of death, the presence, loca-

tion (right and/or left) and numbers of rib fracture, presence of sternal fracture and of chest wall bruising or abrasion. Data is presented as total numbers or as a percentage of the population, and summarised as mean ± standard error of the mean (SEM). Statistical analysis included Mann–Whitney U-test (T value) for continuous variables, Chi-square test (X value) for categorical variables and the Spearman rank sum correlation (C value) for the relationship between variables. During the study period manual CPR techniques only were used exclusively for out-of-hospital CPR. For inhospital CPR, manual CPR was used except in the Accident and Emergency Department where mechanical CPR with a ThumperTM device (Michigan Instruments) was also used. No patient was treated with ACD-CPR in or out of hospital during the study period.

3. Results The raw data are summarised in Tables 1–4. Of the 499 cases, 343 were male and 156 were female. The age range was 18–98 years (61.9 ± 0.8 years) and the age distribution summarised in Table 2. The commonest cause of death was cardiovascular disease (n = 323), followed by respiratory disease (n = 68). Of the 499 cases 29% had rib fractures, 14% had sternal fractures and 11% showed external chest wall bruising or abrasion. Significantly more females than males sustained rib fractures (37% versus 26%; X2 = 5.31, P < 0.05, power = 0.634). There was a trend for females sustaining more sternal fractures than males, but no significant gender effect (17% female versus 12% male; X2 = 3.8, P = 0.051, power = 0.484). The population with rib fractures was significantly older than those without rib fractures (67.5 ± 1.3 years versus 59.0 ± 0.9 years; T = 45 487, P < 0.001). There was a positive association between advancing age and the incidence of rib fractures (C = 0.641, P < 0.001). In 97 of the 146 cases of rib fracture, the report stated the number of ribs on the left and right side that were fractured; in the remainder the presence of “multiple” rib fractures was noted. In these 97 cases, there was no significant difference in the number of left (4.0 ± 0.2) and right (4.1 ± 0.2) ribs fractured (T = 9269, P = 0.631).

C.J. Black et al. / Resuscitation 63 (2004) 339–343

341

Table 1 CPR population and incidence of chest wall injury

Age CPR population (years) Rib fractures (%) Age (years) Sternal fracture (%) Age (years) Chest wall bruising/abrasion (%)

Male (n)

Female (n)

All (n)

61.1 ± 1.0 (343) 26 (89) 67.5 ± 1.3 12 (41) 69.0 ± 2.0 11 (38)

63.6 ± 1.5(156) 37 (57) 71.1 ± 1.9 17 (27) 70.7 ± 3.2 11 (17)

61.9 ± 0.8 (499) 29 (146) 68.9 ± 1.1 14 (68) 69.0 ± 1.8 11 (55)

4. Discussion From the earliest days of human CPR it was recognised that the procedure, while potentially lifesaving, also carried intrinsic risks of injury. Following the first report of rib fractures from the Massachusetts General Hospital in 1961 [14] many studies have reported the complications and hazards of CPR (summarised in Table 5). Approximately 12–15% of patients receiving CPR survive to hospital discharge [15,16]. It is important to minimise any adverse effects of CPR to reduce morbidity and mortality and also to recognise the frequency with which such injuries occur. In the majority of clinical situations CPR is performed manually, but mechanical cardiopulmonary assist devices have been used for over 30 years. More recently a device that Table 2 Age distribution of CPR population (n = 499) Age range (years)

Male (n)

Female (n)

All (n)

18–30 31–50 51–70 >71

18 73 147 105

10 29 46 71

28 102 193 176

Table 3 Number of ribs fractured Male n Number of right ribs fractured Number of left ribs fractured

60 4.0 ± 0.3 3.8 ± 0.3

Female 37 4.5 ± 0.4 4.3 ± 0.4

All 97 4.1 ± 0.2 4.0 ± 0.2

Table 4 Cause of death in CPR population Cause of death

Number of cases (Total 499)

Cardiovascular Respiratory Drug/alcohol Gastrointestinal Mechanical asphyxia Cerebrovascular disease Neurologicala Metabolic Miscellaneousb Unascertained Iatrogenic

323 68 35 19 16 9 8 7 7 5 2

a b

Includes raised intracranial. Includes sepsis, multi-organ failure, hypothermia.

allows active compression and decompression phases during CPR (ACD-CPR) has been used [17]. Following manual CPR the reported incidence of rib fractures varies from 3.5 to 89% [2,3,14,18–29,31], and for sternal fractures 2–31% [2,3,18,21–25,28,31]. There is a lack of consensus regarding the incidence of injuries with the use of ACD-CPR with rib fractures reported between 5.7 and 93% [25–28,30] and sternal fractures 10–93% [25,28,30]. Consequently comparative studies do disagree as to whether ACD-CPR results in a higher rate of chest wall injuries when compared with standard CPR [25–28]. The mechanical cardiopulmonary device (ThumperTM ) is thought not to add to the incidence of rib fractures [24]. Of interest, in our study 14 cases were identified with evidence of ThumperTM use of which nine had rib fractures and two sternal fractures; sternal fracture was not observed in the absence of rib fracture. A prospective study is required to explore this finding further. One source of discrepancy among previous studies reporting CPR injuries is that the gender distribution of the sample population is not always stated. There are also inconsistent reports of the correlation between fracture incidence and age [2,19,24]. The difficulty in evaluating the overall incidence of CPR injuries is compounded by the fact that reports employ different methods of fracture assessment; these have included external examination, autopsy and, as necessary for the examination of CPR survivors, chest radiography (see Table 5). Conventional chest radiographs provide limited views of rib and sternal fractures, indeed a recent study confirms that autopsy by a trained forensic pathologists is a more accurate method of detecting chest wall injuries (31). With the limitations of examination of CPR survivors, isotope scanning can detect rib fractures following CPR that were unsuspected from examination of a chest radiograph [32] and these fractures may have a characteristic scintigraphic appearance [33]. Our study reports the incidence of chest wall injuries following CPR in a cohort assessed by forensic autopsy. We believe that the inclusion of all cases of CPR, regardless of the qualification of CPR provider, is justified since previous studies indicate there is no difference in the incidence of adverse effects when bystander and ALS CPR are compared [29]; also not only does bystander CPR improve survival, but even incorrectly performed CPR does not have a significantly adverse effect on outcome [34]. Our incidence of rib and sternal fractures, 29 and 14% respectively, sits within the mid-range of values in published reports (Table 5). Females were more at risk of sustaining

342

C.J. Black et al. / Resuscitation 63 (2004) 339–343

Table 5 Published studies of chest wall injuries after CPR Author

Study mode

Method CPR

n

Gender M/F (n)

Analysis method

Rib fractures (%)

Sternal fractures (%)

Baringer et al. [14] Clark [18] Himmlehoch et al. [19] Jackson and Greendyke [20] Paaske et al. [2] Patterson et al. [21] Nagel et al. [22] Powner et al. [3] Bedell and Fulton [23] Krischer et al. [24] Rabl et al. [25]

P R P R R R P P P P R

Autopsy Autopsy records Autopsy Autopsy records Autopsy records Autopsy records Autopsy/external Autopsy Autopsy Autopsy Autopsy records

P

NS

NS

Mauer et al. [27]

P

Baubin et al. [28]

P P P

Lederer et al. [31]

P

NS NS NS NS NA NS M = 20, F = 17 NS NS M = 13, F = 16

Clin exam Clin exam CXR CXR Autopsy

Baubin et al. [30] Oschatz et al. [29]

46 19 52 57 323 235 655 70 130 705 25 16 258 254 114 106 114 106 20 15 37 59 96 19 19

NS NS NS NS M = 224, F = 99 NS NS NS M = 82, F = 48 NS NS

Plaisance et al. [26]

Standard Standard Standard Standard Standard Standard Standard/ThumperTM Standard Standard Standard/ThumperTM Standard ACD Standard ACD Standard ACD Standard ACD Standard ACD ACD on cadavers Standard (bystander) Standard (ALS) Standard (bystander/ALS) Standard (bystander/ALS)

33 58 47 35 M = 48, F = 36 17 34 19 11 31.6 28 93 19.1 14.1 3.5 8.5 13.2 5.7 55 87 M = 70, F = 77 8 8 32 89

NA 26 NA NS 22 2 22.2 9 8 21.1 16 93 NA NA NA NA NA NA 30 93 M =10, F = 53 NA NA 26 32

Autopsy CXR CXR CXR Autopsy

NA: not assessed; Clin exam: clinical examination; P: prospective; M: male; R: retrospective; F: female; NS: not stated; CXR: chest radiograph.

rib fractures than males; this has been reported previously [2] and is likely to be related to factors such as osteoporosis and age. A study of ACD-CPR on cadavers reported a greater incidence of sternal fractures in females [30]. We report only a positive trend towards more females showing sternal fracture than males, but none of the cases in our study had undergone ACD-CPR. The positive correlation we report between age and the incidence of rib fracture is in agreement with previous studies on standard CPR [2,24] and ACD-CPR [30]. It is therefore unsurprising that our study reports that the population with rib fractures following CPR is significantly older than those that did not sustain such an injury. The total number of ribs fractured on both the left and right sides is fewer in our study than previously reported [2]. Although in our study 146 cases sustained rib fractures, only the numbers of ribs fractured could be analysed where this was stated (n = 97); in the remainder “multiple fractures” were listed in the autopsy notes (n = 49). It is therefore possible our figure is an underestimate. The retrospective design of this study has potential limitations. The information in the forensic case notes did not always include the details of the CPR procedures performed. It is therefore possible that resuscitative measures may not always have involved chest compression, i.e. in situations of respiratory arrest. The inclusion of such cases in this study

may result in an underestimate of the incidence of secondary injuries. For similar reasons cases may have been excluded from this study where resuscitation measures occurred but were not documented. The effect of the duration of CPR on the incidence of injuries is unclear; one study comments that the incidence of rib fractures does not increase with the duration of CPR [19] but another reports a positive relationship [24]. In this study we were unable to provide information as to the duration of the chest compressions and therefore cannot comment on the possibility that prolonged resuscitation may have influenced the results.

5. Conclusions This study shows a notable incidence of rib and sternal fractures after CPR. In view of the morbidity and potential mortality associated with chest wall injuries, it is important to explore these findings further. A prospective study incorporating both in- and out-hospital arrests with appropriate monitoring of the arrest protocol, follow-up to discharge of CPR survivors, access to medical records and autopsy where permissible following unsuccessful resuscitation, is required to address and clarify the points raised.

C.J. Black et al. / Resuscitation 63 (2004) 339–343

Conflict of interest The authors Catherine J. Black, Anthony Busuttil and Colin Robertson declare no conflict of interests.

Acknowledgements This study was funded by The Resuscitation Council, UK. The authors would like to thank Dr Tim Squires for administrative assistance and Dr Simone L. Meddle for help with preparation of the manuscript.

References [1] Kouwenhoven WB, Jude JR. Closed chest cardiac massage. J Am Med Assoc 1960;173:1064–7. [2] Paaske F, Hart Hansen JP, Koudahl G, Olsen J. Complications of closed chest cardiac massage in forensic autopsy material. Dan Med Bull 1968;15:225–30. [3] Powner DJ, Holcombe PA, Mello LA. Cardiopulmonary resuscitation-related injuries. Crit Care Med 1984;12:54–5. [4] Gillies M, Hogarth I. Liver rupture after cardiopulmonary resuscitation during peri-operative cardiac arrest. Anaesthesia 2001;56(4):387–8. [5] Stallard N, Findlay G, Smithies M. Splenic rupture following cardiopulmonary resuscitation. Resuscitation 1997;53(2):171–3. [6] Sokolove PE, Willis-Shore J, Panacek EA. Exsanguination due to right ventricular rupture during closed-chest cardiopulmonary resuscitation. J Emerg Med 1997;23(2):161–4. [7] Fosse E, Lindberg H. Left ventricular rupture following external chest compression. Acta Anaesth Scan 1996;40(4):502–4. [8] Rello J, Valles J, Juber P, et al. Lower respiratory tract infections following cardiac arrest and cardiopulmonary resuscitation. Clin Infect Dis 1995;21:310–4. [9] Wagner A, Roeggla M, Hirschl MM, Roeggla G, Schreiber W, Sterz F. Tracheal rupture after emergency intubation during cardiopulmonary resuscitation. Resuscitation 1995;30(3):263–6. [10] Offerman SR, Holmes JF, Wisner DH. Gastric rupture and massive pneumoperitoneum after bystander cardiopulmonary resuscitation. J Emerg Med 2001;21(2):137–9. [11] Parke TR. Unexplained pneumoperitoneum in association with basic cardiopulmonary resuscitation efforts. Resuscitation 1993;26(2):177–81. [12] Noffsinger AE, Blisard KS, Balko MG. Cardiac laceration and pericardial tamponade due to cardiopulmonary resuscitation after myocardial infarction. J Forensic Sci 1991;36:1760–4. [13] Machii M, Inaba H, Nakae H, Suzuki I, Tanaka H. Cardiac rupture by penetration of fractured sternum: a rare complication of cardiopulmonary resuscitation. Resuscitation 2000;43:151–3. [14] Baringer JR, Salzman EW, Jones WA, Friedlich AL, External cardiac massage. N Eng J Med 1961;265:62–5.

343

[15] Schneider AP, Nelson DJ. In-hospital cardiopulmonary resuscitation: a 30 years review. J Am Board Fam Pract 1993;6:91–101. [16] Brindely PG, Markland DM, Mayers I, Kutsogiannis DJ. Predictors of survival following in-hospital adult cardiopulmonary resuscitation. Can Med Assoc J 2002;167:343–8. [17] Lurie KG, Lindo C, Chin J. CPR: The P stands for plumber’s helper. J Am Med Assoc 1990;264:1661. [18] Clarke DT. Complications following closed-chest cardiac massage. J Am Med Assoc 1962;181:337. [19] Himmelhoch AR, Dekker MD, Gazzaniga AM, Like AA. Closedchest cardiac resuscitation. N Eng J Med 1964;270:118–22. [20] Jackson CT, Greendyke RM. Pulmonary and cerebral fat embolism after closed chest cardiac massage. Surg Gynecol Obstet 1965;120:25–7. [21] Patterson RH, Burns WA, Janotta FS. Complications of external cardiac resuscitation: a retrospective review and survey of the literature. Med Ann DC 1974;43:389. [22] Nagel EL, Fine EG, Krischer JP, Davis JH. Complication of CPR. Crit Care Med 1981;9:424. [23] Bedell SE, Fulton EJ. Unexpected findings and complications at autopsy after cardiopulmonary resuscitation. Arch Int Med 1986;146:1725. [24] Krischer JP, Fine EG, Davis JH, Nagel EL. Complications of cardiac resuscitation. Chest 1987;92(2):287–91. [25] Rabl W, Baubin M, Broinger G, Scheithauer R. Serious complications from active compression-decompression cardiopulmonary resuscitation. Int J Legal Med 1996;109:84–9. [26] Plaisance P, Adnet F, Vicaut E, et al. Benefits of active compressiondecompression CPR as a prehospital advanced cardiac life support: a randomised multicenter study. Circulation 1997;95:955– 61. [27] Mauer D, Schneider T, Dick W, Wilhelm A, Elich D, Mauer M. Active compression-decompression resuscitation: a prospective, randomised study in a two-tiered EMS system with physicians in the field. Resuscitation 1997;33:125–34. [28] Baubin M, Sumann G, Rabl W, Eibl G, Wenzel V, Mair. P. Increased frequency of thorax injuries with ACD-CPR. Resuscitation 1999;43:33–8. [29] Oschatz E, Wunderbaldinger P, Sterz F, et al. Cardiopulmonary resuscitation performed by bystanders does not increase adverse effects as assessed by chest radiography. Anaesth Analg 2001;93:128–33. [30] Baubin M, Rabl W, Pfeiffer KP, Benzer A, Gilly H. Chest injuries after active compression-decompression cardiopulmonary resuscitation (ACD-CPR) in cadavers. Resuscitation 1999;43:9–15. [31] Lederer W, Mair D, Rabl W, Baubin M. Frequency of rib and sternal fractures associated with out-of-hospital cardiopulmonary resuscitation is underestimated by conventional chest X-ray. Resuscitation 2004;60:157–62. [32] Martin F, Setoain X, Ortega M, Mateos JJ, Muxi A, Pons F. Rib fractures after cardiopulmonary resuscitation. Rev Esp Med Nucl 2001;20:392. [33] Schroeder J, Stevens JS. CPR-induced rib fractures. Characteristic scintigraphic appearance. Clin Nucl Med 1993;18:717. [34] Van Hoeyweghen RJ, Bossaert LL, Mullie A, Calle P, Buylaert WA, Delooz H. Quality and efficiency of bystander CPR. Belgian Cerebral Resuscitation Study Group. Resuscitation 1993;26:47–52.