Cerebral oximetry monitoring during unexpected cardiopulmonary arrest and tension pneumothorax

EDITOR: Cerebral oximetry (rSO2) has had mixed reviews in its use as an accurate tool for determining the level of cerebral oxygenation [1,2]. We present the case of an elderly male who presented for skin grafting in whom cerebral pulse oximetry was monitored. A 66-yr-old male presented for skin grafting to his left shin area. The area had become blackened and necrotic as a result of a blockage to a previous femoro-distal bypass graft. His graft was found to be blocked on Duplex scanning and he was scheduled for a semi-elective femoro-distal bypass graft. His medical background included previous myocardial infarction with subsequent three vessel coronary artery bypass graft. He was also hypertensive and had a 50-yr smoking history. He was assessed as an ASA IV. He had an episode of melaena on the day of his admission and his surgery was postponed until a cause could be found. An oesophago–gastro duodenoscopy revealed gastric erosions. He was transfused with packed red blood cells to bring his haemoglobin to a level of 9.1g dL−1. Three days later he underwent a left re-do femoro-distal graft procedure. This was carried out under a general anaesthetic with etomidate induction and maintained with remifentanil and desflurane. The procedure lasted 7.5 h and was uneventful. During a further procedure 2 weeks later under general anaesthesia using a laryngeal mask airway, his breathing became laboured and he became hypotensive. Vasoconstrictors were administered and he required endotracheal intubation and intermittent positive pressure ventilation. It was noted on the anaesthetic chart that this was assumed to be an episode of bronchospasm. He was extubated uneventfully at the end of the procedure. Two weeks later he underwent a third debridement of his left shin, which was performed uneventfully under a spinal anaesthetic. One week later he was scheduled to have skin grafting to his left shin as the final component of his treatment. It was decided that the procedure would again be carried out under spinal anaesthesia. The cerebral oximeter was attached at the beginning of the procedure and the rSO2 was found to be 40%. This value is at the lower level of the normal range. Values for non-invasive blood pressure (130/80 mmHg), electrocardiograph (85 beats min−1, sinus rhythm) and pulse oximetry (SPO2 98%) monitoring were normal. A 16-G cannula was inserted into the dorsum of the left hand and Hartmann's solution infused. Spinal anaesthesia was administered in the sitting position. Using a 25-G Whitacre needle at the level of L3/L4, bupivacaine 0.5% 2.4 mL was administered using an aseptic technique. The height of the block was noted as T4 using ethyl chloride spray for cold sensation and pinprick testing for pain. Ten minutes after the spinal anaesthetic was administered, mild hypotension was noted and 9 mg of ephedrine administered with a return to normotension. Surgery commenced and proceeded uneventfully with minimal blood loss. However, 25 min into the procedure the patient complained that he felt unwell and ‘thought that he was going to die’. His breathing became laboured and he started to become bradycardic and hypotensive. About 600 μg of atropine was administered with no effect and cardiopulmonary arrest ensued. The rSO2 fell promptly to less than 20 (Fig. 1). One milligram of epinephrine was administered and cardiopulmonary resuscitation (CPR) initiated and the patient's trachea was intubated. After 3 min of CPR, a peripheral pulse was detected and a blood pressure of 170/80 mmHg obtained along with a heart rate of 130. rSO2 reading immediately rose to 65% – much increased from the start of surgery. However, once again BP and rSO2 began to fall and his lungs became difficult to ventilate. Chest auscultation revealed decreased air entry over both left and right lung fields. A diagnosis of a tension pneumothorax was made. Two 14-G canula were inserted at the level of the second intercostal space in the mid-clavicular line on both sides. The needle was withdrawn from the left and a ‘whoosh’ heard. Both NIBP and rSO2 immediately improved (Fig. 1). A left-sided chest drain was inserted followed by a right-sided chest drain along with an arterial line, central venous catheter and urinary catheter. Surgery was completed and the patient awaited transfer to the intensive care unit (ICU). A low-dose epinephrine infusion running at 0.05 μg kg−1 min−1 was started to maintain a mean arterial pressure between 80–90 mmHg, the patient being sedated with a 1% propofol and morphine infusion.Figure 1.: Changes in rSO2 before during and after cardiopulmonary arrest and the accompanying interventions. CPR: cardiopulmonary resuscitation.The patient remained in the operating theatre for 5 h until an ICU bed became available. After 4 h, with BP, temperature, acid–base and fluid balance observations stable, the patient was extubated. rSO2 readings during this post-arrest period were consistently greater than 60, some 50% greater than the pre-arrest picture. He was transferred to the ICU where he remained stable and was discharged the following morning. He underwent one further procedure 2 weeks later to apply the remaining skin grafts without anaesthesia. His rSO2 was again monitored and found to be back to 40%. He continued to make a complete recovery and was discharged home 2 weeks later. Discussion The Somanetics INVOS® cerebral oximeter has been in use for more than 10 yr [3] as a tool for assessing cerebral oxygenation. It has been used during carotid artery surgery [4], cardiac arrest situations [5] and during cardiopulmonary bypass surgery [6]. However, their use in clinical practice remains controversial not least in part due to interpretation of the information provided. It has been demonstrated that it is possible to obtain rSO2 readings from cadavers that are higher than in healthy alive individuals [7]. The explanation for these findings [5] is that near infrared spectroscopy reflects the balance between regional oxygen supply and demand. In dead or infarcted non-metabolizing brain, saturation may be near normal because of sequestered cerebral venous blood in capillaries and venous capacitance vessels. However, in regionally or globally ischaemic, but still metabolising brain, rSO2 decreases because oxygen supply is insufficient to meet metabolic demand. Studies have also shown [8] during isocapnic hypoxia in healthy persons, cerebral oxygenation as estimated by near infrared spectroscopy precisely tracks changes in measured jugular venous oxygen saturation within individuals, but the relation exhibits a wide range of slopes and intercepts. It was therefore felt that the usefulness of the device is limited to situations in which tracking trends in cerebral oxygenation would be acceptable. This was certainly true in our case where a pre-arrest reading for rSO2 of 40% was replaced with one of 65% post-arrest. It was at least reassuring to believe that post-arrest cerebral oxygenation was not impaired and indeed had improved on the pre-arrest picture. This being manifest upon extubation of the patient who demonstrated no neurological deficit. During this case, the use of rSO2 provided useful information. This was not only in the pre-arrest period where an rSO2 reading of 40% reflected the patient's poor physical condition but also during the procedure, where BP, SPO2 and etCO2 fell in parallel with rSO2 before and during the arrest and tension pneumothorax and correspondingly rose following successful resuscitation.