Closed Claims Case Review Case 2023-05: A Little too Late – A Case from the Crypt 2023

A 50-60-year-old ASA 3 male, with a BMI of 41.3, past medical history significant for coronary artery disease (CAD), post-right coronary artery (RCA) stent placement, obstructive sleep apnea (OSA) treated with continuous positive airway pressure (CPAP), hypertension treated with metoprolol, and hyperlipidemia, was “cleared” by a cardiologist for colonoscopy at an ambulatory center. The anesthesiologist noted the patient’s airway was Mallampati 2-3 and a thick neck. Monitored anesthesia care (MAC) was planned. ECG, BP, and pulse oximetry monitoring without end-tidal carbon dioxide (EtCO₂) were applied, and the patient was positioned in the left lateral decubitus position with O₂ administered via nasal prongs at 3-4 L/min. A bolus of propofol 80 mg was administered followed by an additional 20 mg. The patient’s SpO₂ fell to 94%, which was managed with a jaw lift maneuver, and the colonoscopy started. The patient’s respiratory rate decreased to 6, and SpO₂ decreased from 94% to 86%. An oral airway was inserted, and the respiratory status continued to deteriorate. Three minutes after the start of the colonoscopy, the procedure was aborted as the patient was determined to be apneic. The patient was repositioned to the supine position and ventilated with O₂ with an Ambu™ bag mask system. SpO₂ rose to 94%. During ventilation with the Ambu™ bag, the patient was noted to be in ventricular tachycardia (VT). Lidocaine 100 mg X2 was given. Ventricular tachycardia deteriorated into ventricular fibrillation (VF), and cardiopulmonary resuscitation (CPR) began. Twelve minutes after the apnea event and 15 minutes after the first episode of desaturation, the patient was intubated with SpO₂ 92% and confirmation by auscultation of breath sounds. Again, EtCO₂ was not utilized. The patient was transferred to emergency medical services (EMS) for transport to a higher level of care. EMS records showed that they questioned the anesthesiologist about the endotracheal tube placement. The anesthesiologist was adamant that the endotracheal tube was correctly positioned in the trachea. After moving the patient to the ambulance, EMS could not demonstrate the colorimetric presence of CO₂ during ongoing CPR. They removed the endotracheal tube and inserted a laryngeal mask airway (LMA), which showed low CO₂. The patient’s spontaneous circulation was not restored, and the patient was pronounced dead in the ER. An autopsy by the medical examiner noted severe CAD with numerous contributing factors. Listed cause of death was hypertensive and atherosclerotic cardiovascular disease. An endotracheal tube or LMA was not noted on the autopsy.

Patient factors, procedural factors, and systems factors are considered in evaluating the claims in a case. The relevant patient risk factors include Class III obesity and history of obstructive sleep apnea. OSA patients are sensitive to sedating medications, and the multimodal use of drugs for deep sedation increased risk of OSA. Though this patient seemed to only receive propofol, his CPAP was not instituted during the case, which increased his likelihood of experiencing obstruction and apneic events. This patient also has significant CAD with RCA stent. No documentation of additional cardiology studies was available.

The procedural risk factors are the endoscopy procedure itself and the potential of field avoidance of an unsecured airway. The incidence of a vasovagal reaction is around 16.5% and is usually associated with painful intubation of a sigmoid colon with a tight mesentery. Neurocardiogenic syncope can lead to asystole, especially in the already susceptible heart (Intern Med 2012;51:891-3). Overall, 30% of patients experience a rhythm disturbance, 7% have ST segment depression or elevation, and some also have VT or VF (Gastroenterol Clin North Am 2013;42:639-57; Gastrointest Endosc 2012;76:707-18).

Use of propofol itself has been associated with a higher incidence of cardiac arrest and death – 6.07/10,000 in patients sedated with propofol versus 4.28/10,000 with nonpropofol sedation (Saudi J Gastroenterol 2015;21:400-11). Propofol has a narrow therapeutic window and is the most commonly used sedative in anesthesia-assisted endoscopies (JAMA Intern Med 2013;173:1659-60). For GI endoscopy, difficulty in titrating sedation with propofol leads to respiratory complications such as aspiration and inadequate ventilation and oxygenation, which lead to cardiac events. Airway difficulties resulting in inadequate ventilation and oxygenation remain the most common cause of cardiac arrest in patients undergoing GI endoscopy under propofol sedation (JAMA Intern Med 2013;173:551-6).

Systems factors associated with this case are numerous. First, lack of EtCO₂ monitoring during sedation and especially after intubation of the airway was pivotal to this case. The 2018 Practice Guidelines for Moderate Procedural Sedation and Analgesia specifically recommend continual monitoring of ventilatory function with capnography to supplement standard monitoring by observation and pulse oximetry (Anesthesiology 2018;128:437-79; asamonitor.pub/41AAsxq; asamonitor.pub/3IFSqpI). Health care facilities must invest in, and practitioners must utilize, appropriate monitoring. The purpose of these guidelines is to allow clinicians to optimize the benefits of moderate procedural sedation regardless of site of service, to guide practitioners in appropriate patient selection, and to decrease the risk of adverse patient outcomes (e.g., apnea, airway obstruction, respiratory arrest, cardiac arrest, death). Second, a nonoperating room area (NORA) without access to standard OR facilities and equipment is known to be already a challenging environment for optimal anesthesia care. Additionally, a darkened room for endoscopic procedures makes determination of respirations even more challenging until a drop in SaO₂ occurs. Thorough documentation of the events would aid in determination of standard of care, which was also lacking in this case. Advanced cardiac life support (ACLS) and pediatric advanced life support (PALS) training and emergency protocols may have also been an issue at the facility. Qualified personnel and equipment, including EtCO₂, should be on hand to manage emergencies. Personnel with training in age-appropriate advanced resuscitative techniques (e.g., ACLS, PALS) should be immediately available. Policies and procedures need to be established in guiding response to emergencies and unanticipated patient transfer to an acute care facility. Ideally, scenario training should be done yearly with the facility staff (Gastroenterol Clin North Am 2013;42:639-57; Saudi J Gastroenterol 2015;21:400-11).

Closed claim reviews of claims related to airway concerns revealed overwhelmingly that permanent injury resulting in death occurred in 75% of the cases, with grave or major permanent injury in 21% of cases. The cause of injury was most commonly related to inadequate ventilation/oxygenation (65%), followed by difficult intubation (30%).

Of the cases resulting in death, 48% were general anesthesia cases and 40% were MAC cases. Twenty seven percent of all the cases resulting in death were endoscopy (upper, lower, or ERCP) being performed under MAC. Lack of recognition of and monitoring of EtCO₂ for adequate ventilation or oxygenation are the most common causes of this catastrophic event. Inadequate ventilation or oxygenation during procedural sedation (MAC) using propofol is still occurring.

Defense experts may recommend a quick settlement in a case such as this where the anesthesiologist did not adequately ventilate the patient when they first developed respiratory obstruction. The anesthetic record and notes suggest that the patient was inadequately ventilated shortly after the propofol was administered to the time he was intubated (18 minutes). End-tidal capnography and an anesthesia machine were not used. There is also a question of whether the endotracheal tube was inserted in the correct position, which was raised by the EMTs. ACLS protocols were not appropriately followed, and code records were incomplete.