An Unusual Case of Pulmonary Artery Catheter Malfunction

Abstract

Pulmonary artery catheters (PACs), like all medical devices, may occasionally malfunction. We report a case in which an indwelling PAC was inadvertently cut in the operative field during a pulmonary thromboendarterectomy with closure of a patent foramen ovale. This resulted in loss of cardiac output (CO) measurement capability, possible electrical hazard, and increased risk of infection. Case Report A 51-yr-old woman with chronic thromboembolic pulmonary hypertension presented for pulmonary thromboendarterectomy. Preoperative evaluation included pulmonary angiography, which showed extensive bilateral thromboembolic pulmonary artery disease. Cardiac catheterization revealed a pulmonary artery pressure of 82/17 mm Hg, right atrial pressure of 12 mm Hg, and pulmonary vascular resistance of 1100 dynes · s · cm−5. On the morning of surgery, a 20-gauge radial arterial catheter and a 16-gauge peripheral IV catheter were inserted. After preoxygenation and application of standard and arterial catheter monitoring, general anesthesia was induced with 0.16 mg/kg midazolam, 10 μg/kg fentanyl, and 0.15 mg/kg pancuronium. After endotracheal intubation, the right internal jugular vein was cannulated with a 9F introducer sheath (B. Braun Medical Inc., Bethlehem, PA) by using a modified Seldinger technique. Placement of the PAC (Baxter Healthcare Corp., Irvine, CA) was moderately difficult and required three attempts because of coiling in the right atrium. Wedging the catheter was briefly attempted but was unsuccessful; the PAC was secured at 48 cm. The pulmonary artery pressure was 74/42 mm Hg, and the morphology of the wave form was normal. This depth was thought to be appropriate, considering the massively dilated right atrium and ventricle noted on transesophageal echocardiography. An agitated saline test was positive for the presence of a patent foramen ovale. Thermodilution CO measurements were successfully made before cardiopulmonary bypass (CPB) by use of a Hewlett-Packard (Palo Alto, CA) CO module (M1012A), part of the Component Monitoring System M1046-9100C. Upon right pulmonary arteriotomy, the surgeon noted that the PAC was coiled back on itself. The bilateral thromboendarterectomies, primary closure of the foramen ovale via right atriotomy, and separation from CPB were uneventful. After CPB the pulmonary artery and central venous pressure tracings were normal, and temperature measurement appeared functional. However, attempts at thermodilution CO determination were unsuccessful. The CO thermodilution curve was flattened and irregular. Also, blood was found to be leaking from the connection site between the thermodilution cable and the PAC (Fig. 1). A serious malfunction of the PAC was suspected, so it was immediately disconnected from the thermodilution cable. The connection site of the PAC was filled with blood and was capped to prevent spillage. The remainder of the surgery was uneventful, and upon completion of surgery, the PAC was easily removed and a new catheter was introduced to 45 cm.Figure 1: Accumulation of blood (arrow) at the electrical connection site between the pulmonary artery catheter and the thermodilution cable.Close inspection of the removed PAC revealed a thin 2-mm transverse defect in the area of the thermistor wire at 23 cm from the tip, with the appearance of a scalpel incision (Fig. 2). Blood tracking proximally and distally in the thermistor wire conduit was evident.Figure 2: Evidence of trauma to the pulmonary artery catheter. A, The scalpel incision; B, a dark linear shadow of blood tracking along the thermodilution conduit.Discussion PAC malfunction is rare. Presumably, manufacturing defects, if present, are usually discovered when the catheter is tested before insertion and are not deemed reportable by those involved. Most reported malfunctions involve balloon rupture or thrombus formation (1). Numerous cases of central venous catheter fracture have been reported. Catheters that are in place for prolonged periods (months) and subjected to severe stress points (e.g., subclavian catheters passing between the clavicle and the first rib) are particularly at risk (2–4). We do not believe that our catheter was fractured, because it was in place for a short period of time and does not appear to have been subjected to a particular stress point. Suture entrapment has also been reported (5,6). Kodavatiganti et al. (5) described two cases in which surgical sutures snared the catheter, causing malfunction of the thermistor and difficulty with removal. In one of the cases, blood was noted at the hub of the thermistor connection port. In both cases the problem was discovered after the operation, when difficulty in removing the catheter was encountered. Likewise, Vucins et al. (6) reported two cases of catheter entrapment that were diagnosed in the postoperative period, when catheter removal proved difficult. Our case was unique in that the catheter damage became apparent in the operating room, upon separation from CPB. Also, this is the first reported case of PAC malfunction secondary to scalpel damage. Trauma to an indwelling PAC is possible during operations in close proximity to the catheter. These include tricuspid valve surgery, atrial septal defect repair, mitral valve surgery via right atriotomy, pneumonectomy, and pulmonary artery surgery. In our case, the presence of thrombus and fibrinous material in the pulmonary arteries, as well as dilation of the right atrium and ventricle, led to the coiling of the catheter upon placement. This may have resulted in the catheter lying near the wall of the right atrium at the time of atriotomy. The observation that the defect was 23 cm from the tip, or 25 cm from the entrance site of the introducer sheath, also suggests that the damaged portion of the catheter was in the right atrium during surgery. We believe that the damage occurred either when the venous cannulae were placed or, more likely, when the right atrium was incised for repair of the atrial septum. The damage to the PAC created direct fluid continuity between the thermodilution CO module and the pulmonary artery. As soon as the problem was noticed, we disconnected the catheter from the thermodilution module. We were concerned that leakage current could travel through the fluid column in the thermistor conduit and result in microshock. This concern was probably unfounded, because the leakage current rating for this module is <10 μA, and the threshold for ventricular fibrillation from intracardiac catheters is reported to be 50 μA (7). However, electrical current (electrocautery) could travel retrograde through the fluid column to the thermodilution module, causing damage. Although clinically significant microshock or equipment damage is unlikely, it seems prudent to disconnect the PAC from the monitor unit if the electrical integrity of the system is in question. The continuity between the thermodilution cable and the pulmonary artery presents a risk of infection, both to the provider (blood borne) and the patient (antegrade through the aberrant fluid column). Providers can protect themselves by observing universal precautions, and patients can be protected by timely removal of the catheter when such an event occurs. In this case, PAC malfunction presented in a unique and unexpected way. Inability to perform thermodilution CO measurements was the cardinal feature, whereas the pressure wave forms were completely normal. Cardiothoracic procedures near a PAC pose a unique risk to the device. This case serves as a reminder that PAC malfunctions that cannot easily be explained and solved by simple adjustment require prompt attention. If the PAC cannot easily be removed, suture entrapment or knotting should be considered, and the catheter should be left in situ pending further evaluation. Otherwise, the catheter should be replaced as soon as practically feasible.