Although the true incidence of COVID-19 acute respiratory distress syndrome (ARDS) is unknown, associated mortality is nearly 45%.
In appropriately selected patients with COVID-19 ARDS, venovenous extracorporeal membrane oxygenation (VV 1 ECMO) may offer a promising bridge to lung recovery, with reported survival of more than 50%. Cannulation for VV 2 ECMO, however, is invasive and previously reported a 15% risk for major vascular injury, bleeding or limb ischemia whereas neurologic complications alone during COVID-19 may be up to 25%. 3 , In addition, dual-site cannulation techniques for VV 4 ECMO (femoral-IJ, femoral-femoral) may not optimally drain deoxygenated blood flow in a high cardiac output state, such as that often encountered in COVID-19 sepsis. Cannulation via a RIJ DLC may provide improved delivery of oxygenated blood by using bicaval venous drainage without sacrificing extracorporeal flow rates, minimizing recirculation to 1–3%, and exhibiting improved survival. 5 , Avoidance of femoral venous cannulation may also lessen sedation requirements and facilitate early mobilization. Placement of RIJ DLC, however, requires guidance by either fluoroscopy or transesophageal echocardiography (TEE) to minimize the risk of cardiac perforation, vascular injury, or malposition. Transporting to the operating room or catheterization laboratory is not always feasible in rapidly deteriorating, unstable patients. Furthermore, fluoroscopy in the intensive care unit (ICU) room is not always available. Finally, care of COVID-19 patients must consider infection control measures as well as safety for medical personnel. 6
Here, we describe our center’s experience developing a bedside protocol for RIJ DLC cannulation under TEE guidance alone and show that it can be performed safely. We offer comprehensive guidance regarding COVID-19 patient selection for VV
ECMO, necessary equipment and personnel to be deployed at the bedside, RIJ DLC technique, periprocedural TEE monitoring and imaging, and infection control measures. Materials and Methods
Patient Selection and
We developed and disseminated guidelines for the consideration of patients with COVID-19 ARDS. Patients selected demonstrated poor gas exchange, as defined in Appendix 1. Ultimately, patients were referred to and selected for VV
ECMO by a multidisciplinary panel guided by available critical care resources. Patients under consideration for VV ECMO were preferably transported to our “COVID ECMO ICU,” a negative pressure unit in the cardiac/surgical ICU.
Activation of the VV
ECMO team for cannulation involved notification of the on-call perfusionist, cardiothoracic surgical team, cardiac anesthesiology team, intensivist, and ICU charge nurse. Materials mobilized to the bedside included an ECMO circuit, TEE machine, and “ ECMO Go-bag” (Appendix 2). Selection of cannula (Avalon Elite, Getinge or Crescent, MC3) and cannula size were determined by the attending surgeon and guided by the patient’s size, manufacturer recommendation, and desired flow rates. Cannula configuration was determined by surgeon preference. Procedural Considerations and Technique
Although cannulation of pronated patients has been reported in a limited series William Jakobleff commentary, we supinate patients for cannulation, which may lead to acute deterioration in gas exchange and the need for deepening of sedation or paralytic. In addition, patients with preexisting RIJ lines may require line replacement for hemodynamic support. Finally, TEE is performed in patients without contraindications to placement.
Team members within the ICU room include an attending cardiothoracic surgeon (surgical assistant), cardiothoracic surgery trainee (operator), cardiothoracic anesthesiology attending and trainee, perfusionist, ICU nurse, and respiratory therapist (
Figure 1). An assistant outside of the room further promotes safety and efficiency. Surgical field preparation is illustrated in Figure 2. Heparin 3000-5000 IU is administered intravenously. Under full barrier drapes and Trendelenburg positioning, RIJ venous cannulation is obtained by the operator via wire exchange of an existing RIJ catheter or new RIJ cannulation under ultrasound guidance to minimize the risk of vascular injury or pneumothorax. Transesophageal echocardiography is used to confirm wire passage into the infrahepatic inferior vena cava, excluding wire termination in hepatic venous branches. Guidewires are distance-marked, and typically insertion limited to <40 cm to avoid inadvertent wire looping and cardiac injury. Next, a skin incision is made, and a purse string suture is performed through the incision in anticipation of venous securement after eventual decannulation. Seldinger technique is used to pass a series of dilators, each repeated until ease (Video 1, Supplemental Digital Content, ), which ultimately facilitates smooth placement of the dual-lumen RIJ cannula. During cannulation, the surgical assistant maintains the guidewire near the 100 cm mark, although the operator secures the cannula obturator with one hand (Figure A, Supplemental Digital Content, https://links.lww.com/ASAIO/A843 ) and incrementally advances the cannula near the insertion site with the other hand (Video 2, Supplemental Digital Content, https://links.lww.com/ASAIO/A842 ). Transesophageal echocardiography is used for monitoring cardiac injury, wire displacement, bowing or bending of the wire (Figure B, Supplemental Digital Content, https://links.lww.com/ASAIO/A844 ), and successful deployment of the cannula into the hepatic IVC. Each lumen of the dual-lumen cannula is deaired and flushed with saline by bulb-syringe and connected to the corresponding preprimed perfusion circuit tubing https://links.lww.com/ASAIO/A842 via tubing connectors in an air-tight fashion. Tubing clamps are removed, VV ECMO flow is initiated while monitoring for circuit air entrainment and confirmatory outflow blood color change and the cannula position is assessed by TEE. Optimal placement should result in an “outflow” color jet directed toward the tricuspid valve ( Figure 3, Video 3, Supplemental Digital Content, ). The RIJ dual-lumen cannula is initially oriented such that the outflow lumen is anterior, but may require subtle manipulation if clinical and TEE monitoring suggests suboptimal placement. https://links.lww.com/ASAIO/A845 Figure 1.:
In-room choreography of bedside cannulation team, which includes a cardiothoracic surgeon, cardiothoracic surgery assistant, cardiac anesthesiologist, intensive care unit nurse, and perfusionist. All team members are wearing either approved personal protective equipment or sterile surgical gowns.
Surgical instrument and field preparation use equipment from our “
ECMO Go-bag” and in-room furniture. The patient is prepared with a sterile procedural drape, and instruments are laid out on a bedside table. Figure 3.:
Transesophageal echocardiography confirmation of cannula position is demonstrated by a high-velocity color flow jet, originating from the superior vena cava-atrial junction, and directed toward the tricuspid valve.
Initial pump settings are adjusted to a speed of 3,000–4,000 RPMs corresponding with 4–7 LPM blood flow, circuit FiO
2 of 100%, and sweep between 3 and 6 LPM based on the precannulation PCO 2. Pump calibration and initial adjustments are based on starting Hb and initial arterial blood gas. It may be necessary to proactively resuscitate patients with crystalloid or blood products caused by precannulation hypovolemia. Infection Control Measures
To protect providers from COVID-19 exposure, we recommend an
ECMO team consisting of the minimum number of necessary providers. All providers were donned in N95 masks with face shields or powered air purifying respirator, contact isolation gown, and two sets of sterile gloves. Surgical providers were donned in sterile surgical gowns. The TEE machine was donned in a clear cover and later doffed into a container sealed by two hazard bags. All reusable equipment was cleansed with an enzymatic cleaner before exiting the isolation area. After cannulation, we preferred to perform immediate tracheostomy to minimize infection risk by minimizing the number of procedures and removing the existing endotracheal tube. Results
We cannulated 47 patients from March 30, 2020 to November 21, 2021, for
ECMO. We noted three distinct cohorts, corresponding to the following dates: 1) March 30, 2020 to August 21, 2020, 2) October 27, 2020 to May 10, 2021, and 3) August 9, 2021 to November 21, 2021 (suspected delta variant). Cannulated patients demonstrated a mean age of 46.4 years, 28% female gender, and 70% African American or Latinx ethnicity. Mean precannulation PF ratio and number of ventilator days were 64 and 4.3, respectively. Mean Sequential Organ Failure Assessment score at the time of cannulation was 8. A total of 26 patients were cannulated for VV ECMO via RIJ DLC catheterization. All cannulations occurred at the bedside. Patient cannulation configurations and survival are shown in Table 1.
Table 1. -
Configurations and Survival of VV
Survival to hospital discharge was 45% for all configurations (50% when excluding the delta cohort) and 52% for patients supported by RIJ DLC technique (60% when excluding the delta cohort). The mean duration of VV
ECMO was 51.2 days for survivors. Periprocedural complications were primarily minor but did include one major complication of carotid artery injury. No providers contracted COVID-19 from exposure to our cohort of patients. Discussion
In all cases,
ECMO cannulation was considered salvage therapy, after the failure of rescue therapies including inverse-ratio ventilation, prone positioning, neuromuscular blockade, and inhaled nitric oxide administration in nearly all patients pre cannulation. The data reported here represent three distinct cohorts of our COVID experience, with the last cohort representing the delta-variant surge and consequently a dismal survival rate. We observed an overall declining survival trend throughout the progression of pandemic surges, which is most likely because of the severity of ARDS and consequent irreversible lung injury associated with latter variants. However, after subgroup analysis within each surge cohort, there is a tendency toward improved survival with the RIJ DLC configuration (52%) vs. dual-site configurations (32%), which may be coincidental, a translation of decreased site infection risk, or a reflection of an anticipated learning curve for pericannulation and ICU ECMO support, but may also demonstrate a benefit of DLC configurations as previously reported. 1
We observed one major complication of carotid artery puncture. This was attributable to multiple factors, including TEE malfunction at the time of attempted IJ cannulation, which could not be resolved immediately because of the patient’s near-arrest hemodynamic profile and precluded our ability to confirm wire placement in the right atrium before dilation. Later, we emphasized the need to cannulate earlier in the patient’s course, which we hope will prevent rushed cannulations and also allow adequate time to confirm functional TEE.
Although further study is needed to determine whether any cannulation strategy is superior, in our experience, critically ill patients with COVID-19 ARDS can be safely and successfully cannulated at the bedside
via RIJ dual-lumen VV ECMO technique under TEE guidance without significant risk of harm to the patient or medical staff or sacrifice in medical education. Acknowledgments
The authors would like to acknowledge all the support and work of their colleagues, including the Johns Hopkins CVSICU advanced practice provider team, intensivist team, respiratory therapists, perfusionists, cardiothoracic surgery fellows, cardiothoracic anesthesiologists, pharmacologists, nutritionists, and bedside nurses. Photos by Sophia Pan.
Appendix 1: Patient Selection for VV
Patients selected must demonstrate respiratory failure with worsening gas exchange despite conventional interventions that include lung-protective ventilation with the use of positive end expiratory pressure as per ARDSnet protocol, prone positioning, neuromuscular blockade, and if available, inhaled nitric oxide. In select cases, a trial of airway pressure release ventilation may be appropriate.
Despite these measures, if gas exchange continues to worsen as defined by:
PF ratio <80 for 6 hours or PF ratio <50 for 3 hours
pH <7.25 with PCO
2 >60 with plateau pressure >30
Age >60 years
Body mass index (BMI) >45
Poor neurologic function or unknown mental status before presentation
Mechanical ventilation ≥7 days
Bleeding diathesis (inability to be anticoagulated)
Immunocompromised state (
i.e., untreated human immunodeficiency virus or on immunosuppressive treatment) Malignancy with expected survival <5 years
Multisystem organ failure (MOF) or pending MOF with high pressor requirement and evidence of poor perfusion
Unwitnessed cardiac arrest
Chronic end-organ disease:
Chronic renal disease stage III or worse
Moderate to severe chronic obstructive pulmonary disease (COPD)
Ischemic or nonischemic cardiomyopathy with a history of congestive heart failure (CHF)
Severe peripheral vascular disease
Exclusion extensions for patients
<45 years of age
>14 days Appendix 2:
ECMO “Go-Bag” of Necessary Surgical Equipment to Be Deployed at the Bedside
Dual-lumen catheter cannula (1)
Heparin medication bag (1)
Sterile towel packs (8)
4 × 4 Sterile gauze pack (2)
Central line drape (2)
Chloraprep swab sticks (3)
Magnetic needle counter (1)
Ultrasound transducer sterile probe cover (1)
Micropuncture kit (1)
Central venous catheter (single lumen) kit (1)
20 ml Slip tip syringes (2)
Bulb irrigation tray (1)
Mayo scissors (1)
Dual-lumen kit Guidewire (2)
Needle holders (2)
Trauma shears (1)
Tubing clamps (6)
Zip ties (10)
Tegaderm dressings (2)
Cannula dressings (2)
0 Silk sutures (2)
Skin marker (1)
Enzyme spray (1)
Used Instrument tray (1)
Instrument tray cover (1)
1. Thiagarajan RR, Barbaro RP, Rycus PT, et al.; ELSO Member Centers: Extracorporeal Life Support Organization Registry International Report 2016. ASAIO J. 63: 60–67, 2017.
2. Badulak J, Antonini MV, Stead CM, et al.; ELSO COVID-19 Working Group Members: Extracorporeal membrane oxygenation for COVID-19: Updated 2021 guidelines from the Extracorporeal Life Support Organization. ASAIO J. 67: 485–495, 2021.
3. Kannapadi NV, Jami M, Premraj L, et al.: Neurological complications in COVID-19 patients with
support: A systematic review and meta-analysis. Heart Lung Circ. 31: 292–298, 2022.
4. von Segesser LK, Berdajs D, Abdel-Sayed S, et al.: New, optimized, dual-lumen cannula for veno-venous
. Perfusion. 33(1_suppl): 18–23, 2018.
5. Togo K, Takewa Y, Katagiri N, et al.: Impact of bypass flow rate and catheter position in veno-venous extracorporeal membrane oxygenation on gas exchange in vivo. J Artif Organs. 18: 128–135, 2015.
6. Jia D, Yang IX, Ling RR, et al.: Vascular complications of extracorporeal membrane oxygenation: A systematic review and meta-regression analysis. Crit Care Med. 48: e1269–e1277, 2020.