|Home | About | Journals | Submit | Contact Us | Français|
Initiation of hemofiltration in a patient in septic shock can cause hemodynamic compromise potentially leading to cardiac arrest. We propose that the standard '4Hs and 4Ts' approach to the differential diagnosis of a cardiac arrest should be supplemented in critically ill patients with anaphylaxis and human and technical errors involving drug administration (the 5th H and T). To illustrate the point, we report a case where norepinephrine infused through a central venous catheter (CVC) was being removed by the central venovenous hemofiltration (CVVH) catheter causing the hemodynamic instability. CVVH has this potential of interfering with the systemic availability of drugs infused via a closely located CVC.
Hemofiltration is a routine procedure in the management of critically ill patients. It is well known for its safety and hemodynamic stability as compared to hemodialysis.[1,2]Yet, commencement of central venovenous hemofiltration (CVVH) in septic shock patients is frequently associated with hypotension and cardiovascular collapse. This can at times proceed to cardiac arrest. Management of cardiac arrest in such a severely sick patient needs a systematic yet standard approach. Human and technical errors are well-known contributors to patient morbidity and mortality. Anaphylaxis, which can present anywhere, is another differential diagnosis that can easily tip an already compromised patient into a peri arrest situation or cardiac arrest. We suggest adding an extra 'H' and 'T' along with anaphylaxis to the standard '4Hs and 4T' approach to the differential diagnosis of a cardiac arrest. To illustrate the point we describe a case where a hemofiltration catheter is interfered with the delivery of essential vasopressor, leading to cardiac arrest and delayed initiation of hemofiltration.
A 69-year-old patient was referred to our intensive care unit from the emergency department with symptoms and signs of severe sepsis having suffered a Pulseless electrical activity (PEA) cardiac arrest as treatment was being initiated. Cardiopulmonary resuscitation (CPR) was immediately instituted according to advanced life support guidelines. There was return of spontaneous circulation following two cycles of CPR and a bolus of one milligram of epinephrine. At this stage he was still very hypotensive (~70/30 mmHg) and tachycardic (~140/min) and required fluid and metaraminol boluses to support his blood pressure. Tracheal intubation was performed to aid mechanical ventilation. Arterial blood gas analysis revealed a severe metabolic acidosis (pH 6.86, PaO29.1 kPa, PaCO2 6.9 kPa, BE −23, lactate 13.5 mmol/L).
Treatment was initiated as recommended by the surviving sepsis guidelines. In the Intensive Care Unit, a central venous catheter (CVC) was inserted via right internal jugular vein. A MAP of >70 mmHg was achieved with the help of fluid resuscitation and norepinephrine infusion. On grounds of sustained oliguria and systemic acidosis, it was decided to initiate CVVH. A double lumen hemofiltration catheter was therefore inserted via the left internal jugular vein. A chest radiograph showed CVC and hemofiltration catheter tips adjacent to each other approximately at the level of the carina [Figure 1].
The initiation of CVVH caused precipitous hypotension proceeding to PEA arrest in a matter of seconds. Spontaneous circulation got restored promptly following 1 mg epinephrine bolus and 15 s of CPR. When CVVH was recommenced half an hour later, it required further fluid and epinephrine boluses and massive increase in the norepinephrine infusion rate (>10 times the pre-CVVH rate) to achieve a systolic blood pressure of >90 mmHg. Over the next 15 min, there were massive fluctuations in blood pressure (60-250 mmHg systolic). At this point CVVH was stopped again. Subsequently norepinephrine requirement reduced to pre-CVVH levels and hemodynamic stability was restored.
It was decided at this juncture to separate the tips of the CVC and the hemofiltration catheter. The hypothesis was that juxtaposition of the two catheters was interfering with vasopressor delivery most likely by withdrawing blood (containing norepinephrine) for filtration. Leaving the CVC undisturbed, the hemofiltration catheter was withdrawn by 5-6 cm. A repeat chest radiograph a few minutes later showed the hemofiltration catheter tip in the brachiocephalic vein [Figure 2]. CVVH was resumed at this point with little hemodynamic instability and did not require a significant increase in vasopressor infusion rate.
CVVH in contrast to hemodialysis usually results in less hypotension and is therefore better tolerated in critically ill patients.[1,2] Yet, hypotension commonly occurs with initiation of CVVH and is generally attributed to volume depletion, solute disequilibrium and/or vasodilatation. Fluid bolus and commencement of inotrope infusion prior to starting CVVH is standard of care. Norepinephrine (molecular weight 169.18) though freely filtered by both CVVH (filters molecules up to about 30 000 Daltons) and dialysis filters (allows clearance of molecules up to about 500 Daltons) has a half-life of < 3 s in the circulation and therefore requires no dose modification for that reason during RRT. Generally in the event of a cardiac arrest, it is recommended to clamp off the vascular access catheter lumens and stop CVVH, unless it is caused by hypovolemia (filter blood be washed back to the patient) or hyperkalemia (filter be kept running). If a patient on CVVH develops a shockable rhythm, defibrillation/cardioversion has been shown to cause no interference in the CVVH machine settings.
A systemic approach to causes of cardiovascular collapse/cardiac arrest worth considering in such a patient would conform, with a couple of additions, with the ′4Hs and 4Ts′ formula suggested in the ALS guidelines of the Resuscitation Council, UK:
The pattern of hemodynamic instability in our patient in response to hemofiltration led us to narrow down to the following possibilities:
Our decision to withdraw and thus separate the tips of the two intravascular devices was based on the logical assumption that if reasons a) or c) were causing the hemodynamic instability then this would resolve the problem. Also, this would not compromise the patient in any way if reduction in preload (b) was the root cause of the hypotension.
The above case report illustrates how a human and technical error (closely positioned vascular access catheter tips) led to an iatrogenic cardiovascular collapse and PEA cardiac arrest. This complication could have been easily avoided by positioning the two catheters tips on either side of the diaphragm or at least in separate veins. It is difficult to know how much the two catheter tips should be separated in the same vein in order to avoid interference with systemic drug delivery. This would depend on a lot of factors including the CVVH pump speed, intravascular volume status of the patient and the drug infusion rate. The clinicians should be aware of this possibility.
Although extensive literature search using Medline, Embase and Google revealed no earlier published report of a similar incident whereby one intravascular device had interfered with the delivery or systemic availability of drugs administered through another, but there are umpteen number of reports regarding mortalities and near misses due to human and technical errors.[7,9]Critical incident reporting, morbidity and mortality meetings, two person checking of drugs and blood products, labeling of syringes, use of prefilled syringes and bar codes, introduction and implementation of guidelines and protocols and proper use of alarm settings are vital tools in avoiding human and technical errors.
In conclusion, our experience suggests that:
Source of Support: Nil
Conflict of Interest: None declared