Canadian Journal of Anesthesia/Journal canadien d'anesthésie

, Volume 59, Issue 6, pp 578–585

Cardiac arrest in the operating room requiring prolonged resuscitation

Case Reports/Case Series

DOI: 10.1007/s12630-012-9698-4

Cite this article as:
Charapov, I. & Eipe, N. Can J Anesth/J Can Anesth (2012) 59: 578. doi:10.1007/s12630-012-9698-4

Abstract

Purpose

Prolonged cardiopulmonary resuscitation (CPR) is often associated with limited success and poor long-term outcomes. The purpose of this report is to present the case of a patient who suffered an unanticipated cardiac arrest in the operating room and survived following a prolonged period of CPR.

Clinical features

A previously healthy 53-yr-old male with inflammatory bowel disease was diagnosed with a perforated bowel and underwent emergency exploratory laparotomy under general anesthesia. Approximately two hours after induction of anesthesia, the patient experienced cardiac arrest, and for 55 min, he underwent CPR and defibrillation according to the Advanced Cardiac Life Support (ACLS) protocols. As the decision to terminate CPR was being considered, a return of spontaneous circulation was detected 56 min after the onset of cardiac arrest. The patient survived with no major organ failure or adverse neurological outcome. No definitive cause of cardiac arrest was diagnosed in the postoperative period. At the follow-up 14 months after the event, the patient had returned to the pre-arrest level of functioning. The results of our literature search showed that no upper limit for the duration of CPR has been defined. Good outcomes after prolonged CPR depend on the patient’s pre-arrest condition and the etiology of the cardiac arrest.

Conclusion

Perioperative cardiac arrests are rare events, and there is little evidence to suggest an upper limit for the duration of resuscitation. Unknown etiologies and the presence of good patient predictors may support the continuation of prolonged CPR with good outcomes.

Arrêt cardiaque en salle d'opération nécessitant une réanimation prolongée

Résumé

Objectif

Une réanimation cardiorespiratoire (RCR) prolongée est souvent associée à un succès limité et à un mauvais pronostic à long terme. L’objectif de ce rapport est de présenter le cas d’un patient ayant fait un arrêt cardiaque non anticipé en salle d’opération et ayant survécu après une RCR prolongée.

Caractéristiques cliniques

Un homme âgé de 53 ans, jusque-là en bonne santé à l’exception d’une maladie inflammatoire intestinale, a présenté une perforation intestinale et a dû subir une laparotomie exploratoire d’urgence sous anesthésie générale. Approximativement deux heures après l’induction de l’anesthésie, le patient a fait un arrêt cardiaque et il a reçu une RCR pendant 55 minutes avec défibrillation conformément aux protocoles de réanimation cardiorespiratoire avancée (ACLS). Alors que la décision de mettre un terme à la RCR était envisagée, un retour à une circulation spontanée a été détecte 56 minutes après la survenue de l’arrêt cardiaque. Le patient a survécu sans défaillance d’un organe essentiel ni évolution neurologique préjudiciable. Aucune cause certaine de l’arrêt cardiaque n’a pu être diagnostiquée au cours de la période postopératoire. Lors du suivi, 14 mois après l’événement, le patient avait retrouvé le niveau d’activité qui était le sien avant l’arrêt cardiaque. Nos recherches dans la littérature ont montré qu’aucune limite supérieure de durée de la RCR n’a été définie. Un bon pronostic après RCR prolongée dépend de l’état clinique du patient avant l’arrêt cardiaque et de l’étiologie de ce dernier.

Conclusion

Les arrêts cardiaques périopératoires sont des événements rares et il y a peu de données probantes pour proposer une limite supérieure à la durée de la réanimation. Les étiologies inconnues et l’existence de facteurs prédictifs favorables liés au patient peuvent être en faveur de la poursuite d’une RCR prolongée avec de bons résultats.

While considered rare, intraoperative cardiac arrests in the operating room under general anesthesia are potentially catastrophic events.1 Intraoperative cardiac arrests may be anticipated in certain surgical procedures, for example when associated with blood loss, anaphylaxis, or other complications.2 However, unexpected cardiac arrests with poorly defined etiology present a challenge, especially with regard to the duration of cardiopulmonary resuscitation (CPR).3 We present a case of intraoperative cardiac arrest which was treated with prolonged CPR.

The patient has provided written consent for the reporting and publication of this case.

Case report

A 53-yr-old male (American Society of Anesthesiologists physical status IIIE, weight 64 kg, height 166 cm) presented with a suspected perforated colon secondary to inflammatory bowel disease (IBD) and was scheduled to undergo an emergency exploratory laparotomy. The patient was well until two months before surgery when he presented with abdominal pain and bleeding per rectum and was diagnosed with IBD. He had no known allergies, no known cardiorespiratory disorders, and excellent exercise tolerance. Five days before surgery, he was admitted to hospital for treatment with intravenous fluids, antibiotics, and steroids, and he was given a transfusion of two units of packed red blood cells. A radiological examination on the day of surgery revealed free air under the diaphragm suggestive of bowel perforation.

At the preanesthetic examination, the patient was febrile, diaphoretic, and appeared unwell. Standard monitors were applied, including a five-lead electrocardiogram (ECG), a noninvasive blood pressure (BP) device, pulse oximetry, temperature, end-tidal CO2, oxygen and anesthetic agent analysis, and ventilatory pressure and volumes, and anesthesia was induced intravenously with fentanyl 200 μg, propofol 100 mg, and rocuronium 50 mg (Figure, available as electronic supplementary material [ESM]). After tracheal intubation, anesthesia was maintained with desflurane in an air-oxygen mixture with controlled ventilation. Approximately 15 min after induction, the patient’s BP decreased to 80/52 mmHg, and his heart rate (HR) increased to 127 beats·min−1 with associated ST changes on the ECG; a 2-mm ST elevation was seen in lead II, and a 1.3-mm depression was seen in V5. Phenylephrine 40 μg and metoprolol 2 mg were given intravenously with prompt resolution of the ECG abnormalities. Two units of packed red blood cells were rapidly transfused because the hemoglobin (Hb) was 71 g·L−1. The ECG abnormalities normalized, and the BP and HR returned to the preanesthetic baseline and remained stable.

Eighty minutes after induction, the patient was given hydromorphone 1 mg iv and ketamine 20 mg iv as a bolus, followed by an infusion of 0.1 mg·kg−1·hr−1. At 100 min, another dose of hydromorphone 1 mg iv was administered, and a lidocaine infusion 2 mg·kg−1iv, to be given over 12 min, was started. Six minutes later, ECG changes (ST elevation in II and ST depression in V5) were seen again, and the patient’s BP had decreased to 85/50 mmHg with a HR at 80 beats·min−1. The lidocaine was stopped after 60 mg had been administered. Phenylephrine 80 μg was given intravenously, restoring the patient’s BP to 96/72 mmHg and HR to 77 beats·min−1. However, ST segments continued to deteriorate. As a nitroglycerine infusion was prepared but not infused, the patient’s sinus rhythm changed to pulseless ventricular fibrillation (VF) six minutes after the onset of ST changes (112 min after induction). Up to this point, surgery had proceeded uneventfully with the surgeon about to excise the mobilized colon. A “resuscitation code” was called and another staff anesthesiologist was called to the operating room to assist. The surgery was halted and the surgeons started chest compressions without delay. One of the anesthesiologists checked the patient’s carotid pulse every minute throughout resuscitation.

Biphasic defibrillation was initiated (150 J followed by 200 J) and continued every three minutes. As VF continued, Advanced Cardiac Life Support (ACLS) guidelines for resuscitation were followed with the patient receiving epinephrine (total dose 9 mg), insulin (20 U iv), calcium chloride (2 g), 50% dextrose (50 mL), sodium bicarbonate (150 mEq), magnesium sulphate (7 g), and lidocaine (200 mg). Amiodarone 300 mg iv was given at approximately 30 min after the arrest and then again at 40 min after the arrest (Figure, available as ESM). The quality of the chest compressions was continuously assessed by palpation of the carotid pulse and was described at all times as “brisk upstroke with chest compressions”. Throughout CPR, the surgeon observed that bowel mucosa continued to be pink and well perfused. Given the nature of ST segment changes before the arrest, an urgent cardiology consultation was requested to consider coronary angiography and intervention while resuscitative efforts continued. The cardiologist recommended no change in management. At 41 min from the onset of the arrest, the patient continued to be in treatment-resistant VF. Biphasic defibrillation was repeated every four to five minutes and chest compressions were continued. At 43 min from the arrest, sinus rhythm changed to slow pulseless electrical activity (PEA). Atropine 3 mg iv was administered and the ACLS guidelines for slow PEA were followed for another ten minutes.

At 53 min after cardiac arrest, the decision to terminate resuscitation was being discussed; however, at 56 min after the arrest, while slow PEA persisted, a weak carotid pulse was felt for the first time. Sinus bradycardia improved with atropine to a normal sinus rhythm, and circulation returned 56 min after the onset of cardiac arrest. A review of the patient’s electronic anesthesia record after the event revealed that the recorded range (minimum-maximum) of end-tidal CO2 (EtCO2) doubled from 6-17 mmHg during the first 45 min of resuscitation to a consistent 21-30 mmHg during the last 11 min of CPR. The patient’s hemodynamic variables stabilized (BP 158/99 mmHg, HR 76 beats.min−1, oxygen saturation (SpO2) 93%, EtCO2 35 mmHg) with an infusion of norepinephrine. A cannula was inserted in the radial artery, and a central venous catheter was placed in the right internal jugular vein. A pulmonary artery catheter was floated within ten minutes of return of spontaneous return of circulation; first pulmonary arterial pressure (PAP) was 25/10 mmHg. Arterial blood gas analysis 35 min post-arrest revealed pH 7.0, PaCO2 48 mmHg, PaO2 150 mmHg, base excess 16 mmol·L−1, HCO3 14.3 mmol·L−1, oxygen saturation 98%, K 3.4 mmol·L−1, Na 134 mmol·L−1, Hb 99 g·L−1, international normalized ratio (INR) 3, and glucose 22.2 mmol·L−1.

The patient remained stable and ST segments normalized to baseline over ten minutes. There were no changes on the 12-lead ECG which would have warranted a cardiology intervention. The surgeons proceeded expeditiously to excise the already mobilized segment of perforated bowel and create a colostomy. The patient was then transferred to the intensive care unit (ICU) with an intubated trachea and ventilated lungs, and he remained stable (BP 115/60 mmHg and HR 80 beats·min−1) on an intravenous infusion of norepinephrine. The ICU team decided not to initiate cooling protocol for this patient.

In the ICU, bedside transthoracic echocardiography showed global hypokinesia. On postoperative day 1, the patient was weaned from inotropic support. A coronary angiography showed diffuse and mild coronary disease with no treatable lesions. A computed tomography scan revealed a small subsegmental pulmonary embolism. Three days after his initial surgery, the patient returned to the operating room to close the abdominal incision, and he was weaned from the ventilator uneventfully. He was formally evaluated by the ICU team and found to have a normal neurological examination. The patient was interviewed by the anesthesiologist prior to discharge and was found to have no recall of events and no apparent neurological deficits. He was discharged to the ward and then to home and normal level of activity with a well functioning colostomy. Fourteen months after his cardiac arrest, the patient underwent surgery for reversal of his colostomy. The patient has since returned to work and a normal level of activity. A neurologist did not perform a formal neurological assessment and no objective cognitive function testing was carried out in this patient.

Discussion

We have reported a case of unexpected intraoperative cardiac arrest in a patient with IBD undergoing emergency colon resection for a perforated bowel. The etiology of this cardiac arrest is unclear. The presence of anemia and ECG changes immediately after induction of anesthesia along with a reappearance of ECG abnormalities just prior to cardiac arrest makes reversible coronary vasospasm a possible primary mechanism of cardiac arrest. This opinion is supported by the postoperative coronary angiography. In our view, sepsis-induced cardiomyopathy was also a possibility considering the patient’s long septic illness preoperatively and the timing of his cardiac arrest at the end of a long surgical dissection of necrotic bowel tissue.4 A pharmacologically mediated sympathectomy by the combination of opioids, a beta-blocker, ketamine, a volatile agent, and lidocaine may have also contributed to the mechanism. With regard to management of cardiac arrest, the patient received immediate uninterrupted resuscitation guided by ACLS protocol. In retrospect, the first dose of amiodarone could have been given sooner than at the 30-min time point. An important sign was the upward capnography trend toward the end of the pulseless period—unrecognized at the time of CPR—possibly heralding return of spontaneous circulation. In our opinion, trends towards improvement of pulmonary blood flow suggested by capnography should encourage continued resuscitatory efforts.

We undertook a literature review with regard to the duration and outcomes of CPR. Databases (OVID, Medline, and Embase) were searched for the period 1990 to 2010 with the keywords: CPR and duration or cardiopulmonary resuscitation and duration or prolonged and CPR or prolonged and resuscitation. The results were limited to English language citations. The case reports were organized according to duration of resuscitation (Table 1), and they were summarized according to factors predicting outcomes of these efforts (Table 2) and criteria for continuing resuscitation (Table 3).
Table 1

Summary of shortest to longest cardiopulmonary resuscitation (CPR): a review of 19 case reports by increasing duration of CPR

Author (Reference No.)

Duration of CPR in minutes

Etiology

Age, Sex, Rhythm on Presentation

Case description

Long-term Outcome

Predictors of success

Cavallaro12

21

Pulmonary embolism

80 s, male, PEA

Witnessed, use of thrombolytic

Discharged home

Immediate CPR and ACLS, Reversible etiology

Nanda13

24

Hyperkalemia (DKA)

42, female, PEA

Witnessed

No neurological deficit

Reversible etiology

Chen14

35

Hyperkalemia (DKA)

42, male, asystole

Witnessed

No neurological deficit

Reversible etiology

Dolenska15

35

Cardiac causes and drug toxicity

35, female, VF

Intraoperative long QT intrinsic cardiac anomalies/ nasal surgery / cocaine use

No neurological deficit

Immediate CPR and ACLS, reversible etiology

Cavallaro12

45

Pulmonary embolism

80 s, male, PEA

Witnessed, use of thrombolytic

Discharged home

Immediate CPR and ACLS, Reversible etiology

Gabrielli16

54

Post ECT

80 s, male, asystole

Required pacing. Conscious during CPR.

Died 48 hr later of CVA

Immediate CPR and ACLS

Beckhoff 17

55

Myocardial infarction

32, male, asystole

Unwitnessed, Delay in CPR up to four minutes

Moderate neuropsychological deficit: one year post arrest able to drive, work half-time in his previous employment

Young physiological age

Cooper18

60

Myocardial infarction

56 male, VF → PEA

Unwitnessed, out of hospital CPR

Mostly complete neurological recovery: minor loss of short-term memory, some loss of “more sophisticated artistic and musical attributes”

Young physiological age

Archan19

60

Myocardial infarction

48, male, asystole

Unwitnessed, out of hospital CPR

No neurological deficit

Young physiological age

Marcus20

60

Lightning strike

14, male, asystole

Windsurfer struck by lightning, Immediate bystander CPR for ten minutes. Hypothermia (31.8C) during CPR. Fixed dilated pupils.

Extubated 48 hr post arrest. 3rd degree burns, MI, gastritis, depression. Back to school after ten months. No neurological deficit

Immediate BLS, hypothermia

Stockman21

84

Cardiac abnormality

38, female, VF

Sudden collapse, Family history of sudden death.

Colectomy due to colon to ischemia. No neurological deficit

Single organ failure

Braque22

90

Intraoperative local anesthetic toxicity

73, male, VT → VF → asystole

Intravascular local anesthetic toxicity following lumbar plexus block. ROSC: VT converted to SVT with clonidine.

No neurological deficit Discharged home four days after cardiac arrest.

Reversible etiology

Kelly23

97

Congenital heart disease

Newborn, VT

Cardiac arrest following arterial switch procedure in a newborn. After CPR required ECMO for 11 days.

Hydrocephalus. Developing normally.

Single organ failure/ ECMO

Lee24

140

Intraoperative rhabdomyolysis, hyperkalemia

11, female, asystole alternating with VT

Received succinylcholine and isoflurane for elective tonsillectomy. Post CPR required thoracotomy and CPB for 146 minutes. Fixed and dilated pupils during CPR.

No neurological deficit. Long ICU stay: hemodialysis for six weeks, re-operation for bleeding; hemothorax, DIC, eventual full recovery over ~ 20 weeks. Negative for MH.

Reversible etiology

Kelly23

176

Myocarditis

Four year-old, VT

ECMO for 9 days post CPR

No neurological deficit,

Single organ failure/ ECMO

Yu25

280

Myocarditis

27, female, VT → asystole,

ECMO for 9 days post CPR followed by heart transplant

No neurological deficit,

Single organ failure/ ECMO

Husby26

190

Hypothermia

51, male, asystole, rectal temperature 24°C

ETOH intoxication, sea immersion at 6°C for 40 min, brought ashore in asystole, CPR → CPB 120 min

No neurological deficit

Reversible etiology

Stoneham27

270

Hypothermia

30, male, VT → asystole

Suicide attempt, sea emersion, LOC and arrest when pulled ashore, 23°C hypothermia, 4.5 hr of CPR.

Extubated 48 hr post arrest. MI, depression. Discharge home five days post arrest. No neurological deficit.

Reversible etiology, ROSC, CPB

Jones28

315

Hypothermia hypoglycemia

46, male, VF

Homeless alcoholic found on the street in November with hypoglycemia. Mechanical CPR (“Thumper”) followed by CPB for 85 min with return of circulation. Opens eyes and obeys commands in ICU 13 hr after arrest.

Dies 48 hr after arrest from large intracerebral hemorrhage, renal and cardiovascular failure.

Initial success: hypothermia,

Demise: premorbid disease and hypoglycemia

BLS = basic life support; CPR = cardiopulmonary resuscitation with external cardiac massage; CPB = cardiopulmonary bypass; DIC = disseminated intravascular coagulation; DKA = diabetic ketoacidosis; ECMO = extracorporeal membrane oxygenation; ECT = electroconvulsive therapy; ETOH = alcohol; ICU = intensive care unit; MH = malignant hyperthermia; MI = myocardial infarction; LOC = loss of consciousness; PEA = pulseless electrical activity; ROSC = transient return of spontaneous circulation during CPR; SVT = sustained ventricular tachycardia; VT = ventricular tachycardia; VF = ventricular fibrillation

Table 2

Summary of outcome predictors for resuscitation from all-cause cardiac arrest

Predictors of successful outcome at the start of CPR: (success and good neurologic outcome)

• Witnessed arrest and VF/VT presentation: 60-88% survival

• Witnessed arrest and asystole / PEA presentation: 33% survival6

• In-hospital arrest in actively monitored area, e.g., OR, PACU7

• Arrest during daytime29

• Immediate start of BLS by bystander (within four minutes or less)29

• Early ACLS / a short ambulance response time: less than ten minutes29

• Immediate defibrillation if shockable rhythm (within three minutes)7,29

• Reversible cause, e.g., hypothermia, metabolic / electrolytes abnormalities, toxins / drug overdose

• Transient ROSC at any point of CPR9,10

Predictors of poor outcome at the start of CPR: (failure and/or major organ failure or death)

• Unwitnessed arrest and VF/VT presentation: 17% survival6,29

• Unwitnessed arrest and asystole / PEA presentation: 2-13% survival6,29

• Out of hospital arrest / location in hospital without active monitoring6,29

• Night time6

• Associated illness or nursing home residency with dependent status: DM, end stage organ failure, advanced cancer6

• Absent pupillary responses during resuscitation6,20

• Absent or low end-tidal CO26

BLS = basic life support; CPR = cardiopulmonary resuscitation with external cardiac massage; DM = diabetes mellitus; OR = operating room; PACU = postanesthesia care unit; PEA = pulseless electrical activity; ROSC = transient return of spontaneous circulation during CPR; VT = ventricular tachycardia; VF = ventricular fibrillation

Table 3

Summary of factors that support extending or limiting the duration of CPR

What would support extending CPR?

• Etiology of arrest: reversible cause, e.g., hypothermia, metabolic or electrolyte abnormality, toxins or medication overdose, uncertain etiology of arrest: time needed to make definitive diagnosis30

• CPR to bridge to extracorporeal circulation29

• Persistent VF10

• Temporary ROSC during CPR11

• Young physiological age: discrepancy between chronological and physiological age (controversial)7,10

• Good quality CPR (evidence of organ perfusion)11

What would support limiting duration of CPR?

• Absence of good predictors at the start of CPR or Presence of bad predictors at the start of CPR (see Table 2)

• Asystole for > 20 min10

• Irreversible cause9,10

• Long time to CPR11

• Poor functional pre-arrest state29

CPR = cardiopulmonary resuscitation with external cardiac massage; ROSC = transient return of spontaneous circulation during CPR; VF = ventricular fibrillation

The literature search revealed case reports of resuscitation after cardiac arrest which varied from 21 to 315 min (Table 1). The results of this search suggest that certain patient factors present before CPR is started may predict a positive outcome. The factors are summarized for CPR after all-cause cardiac arrest (Table 2). While presence of these factors may support prolonged resuscitation, the duration of such prolonged CPR seems to be poorly defined. It is important to emphasize that case reports of successful and unusually long resuscitations are biased because cases of prolonged resuscitation with poor outcomes may remain unreported or even undetected by our English language literature search. Nevertheless, the successful case reports do provide valuable information to help identify common features of successful resuscitations which could support prolonged resuscitatory efforts.

There are published guidelines which may be useful to guide resuscitation of in-hospital cardiac arrest in general. However, these recommendations may not be applicable to the perioperative setting. Perioperative cardiac arrests occur infrequently (estimated to be 0.69 per 10,000 anesthetics). While it is estimated that only one-third of these patients survive CPR, only 70% of these survivors are discharged home without major organ dysfunction.5 The duration of CPR also influences survival and outcomes.6 A review of all in-hospital cardiac arrests suggests that just under one-half of all cardiac arrests were immediately resuscitated successfully by 28 min of CPR; the survival rate decreases considerably beyond 28 min of CPR.6 The prognosis for long-term survival is also related to the duration of CPR; in the same series, no patients requiring CPR for longer than 15 min survived for more than six weeks. Another review of 14,720 all-cause cardiac arrests revealed that the probability of survival decreases rapidly if CPR lasts longer than ten minutes.7 It is important to accentuate that cases of perioperative cardiac arrest are underrepresented in this series. For CPR in critical care units, a more recent study reports a 26.7% survival to discharge, with mortality increasing with duration of CPR and only one-third of patients surviving CPR which lasted more than 45 min.8

Neither ACLS nor European Resuscitation Council (ERC) guidelines commit to a maximum duration of CPR before considering continued resuscitation futile, and both seem to conclude that rules for clinical decision-making cannot be recommended due to lack of prospective validation.9,10 Other guidelines suggest that patient context, setting of cardiac arrest, and promptness and quality of resuscitation are the most important factors to consider.11 After reviewing the literature, we are of the opinion that an appropriate duration of CPR is difficult to define, and the definition depends entirely on the context of patient presentation. While this case report may not necessarily reflect novelty or innovation in treatment, it may encourage others to persevere in the resuscitation of patients suffering a cardiac arrest of unknown etiology or patients with factors supporting prolonged resuscitation (summarized in Table 3).

In conclusion, this case presents prolonged resuscitation with survival, excellent neurological outcome, no major organ damage, and return to a normal level of functioning. Factors influencing such outcome were the patient’s young age, excellent functional status, and ultimately reversible etiology of cardiac arrest treated with immediate and good-quality CPR. The context of each case of cardiac arrest should be carefully considered to avoid CPR which is unnecessarily long and futile. There is currently no available evidence with regard to establishing duration of resuscitation efforts, particularly in the perioperative setting. Expert opinion is required to define what would be considered as “futile” duration of CPR.

Acknowledgements

We acknowledge Dr. David Ewing’s assistance with the conduct of patient resuscitation in the operating room. We also acknowledge the assistance of Ms. Alexandra Davis, a medical librarian at The Ottawa Hospital, who helped us with the literature search.

Conflicts of interest and sources of support

None declared.

Supplementary material

12630_2012_9698_MOESM1_ESM.pdf (381 kb)
Patient’s anesthesia record recreated from electronic data output of anesthesia monitor after cardiac arrest and resuscitation. (PDF 380 kb)

Copyright information

© Canadian Anesthesiologists' Society 2012

Authors and Affiliations

  1. 1.Department of Anesthesiology, The Ottawa Hospital (TOH)University of OttawaOttawaCanada