Central venous placement using ultrasound has significantly reduced the complications associated with blind puncture. The central venous catheter can still get misplaced if it follows an anomalous route after appropriate puncture of desired vessel. We report a case of misplaced dialysis catheter into the accessory hemiazygos vein which resulted in a large hemothorax, and we recommend the routine use of a fluoroscope for placement of dialysis catheters so as to avoid serious complications.

A 63-year-old woman, known case of chronic renal failure was admitted to the hospital for urgent dialysis. Her serum creatinine was 1444 μmol/L, and her blood urea nitrogen was 48.7 mmol/L. Her serum potassium was 6.2 mmol/L. She was mentally retarded with aggressive behavior. She had had dialysis catheter placed many times which she had pulled out during her bursts of aggressive behavior. Arteriovenous fistula was done one time which also failed. This time she presented without a dialysis catheter. We decided to place a permacath in the operation theater. She was given light general anesthesia. We chose the internal jugular as the vein of choice for the procedure. With the ultrasound, we assessed her internal jugulars. Her previous tunneled catheter which she had pulled out before was in the right internal jugular. This time the right internal jugular could not be visualized at all with ultrasound. It was presumed that either it was thrombosed or hidden in the fibrosis of the surrounding soft tissues. The left internal jugular was visible clearly. Under sterile precautions, it was punctured under ultrasound guidance, and the guide wire was passed. There was a slight difficulty in passing the guide wire but after some manipulation it passed. There were no ectopics in the electrocardiogram when the guide wire was passed. Ultrasound was used and guide wire was identified as a bright speck inside the internal jugular vein (IJV) in short axis. Fluoroscope was not used as it was engaged in some other case. Having clearly confirmed the guide wire presence in the IJV, we decided to proceed further. The dialysis catheter was tunneled from a suitable site in her chest. A peel away introducer was passed over the guide wire, and the wire removed. Typical venous blood was observed exiting from the introducer and the permacath was passed as the introducer was peeled off. Typical venous blood was aspirated from both permacath lumens and lumens were flushed and closed with heparin lock. The procedure seemed uneventful, and the patient was made awake. A routine portable X-ray was ordered before shifting the patient to the dialysis unit. When we saw the X-ray, we were shocked to find the catheter entirely in the left side of the heart as though it had passed through the aorta [Figure 1]. We took the patient back to occupational therapy and connected a transducer to the catheter port which showed a venous waveform. A blood gas analysis of the aspirated blood also showed venous values. Anyhow it was decided to leave the catheter untouched at that time as the patient was unstable. We urgently placed a regular dialysis catheter in her right femoral vein, and she was sent for dialysis. After 3 days of dialysis, the patient became stable and computed tomography (CT) chest was done. The preliminary X-ray done during CT clarified the position of the catheter in the lateral edge of aortic knuckle [Figure 2], which is the location of left superior intercostal vein. The reconstructed coronal and sagittal CT images [Figures 3 and 4] showed the catheter tending to go into the accessory hemiazygos vein which communicates with the left superior intercostal vein. The X-ray and CT films also showed large hemothorax on the left side which suggested that the veins were partially damaged by the large catheter. As the patient was very stable, we just removed the catheter and applied pressure dressing at the entry site. The patient tolerated it without any problems. The hemothorax was kept under observation as the patient did not have any respiratory distress, and the relatives did not give consent for chest tube insertion. The hemothorax was persistent but not increased in size. After another 3 days, once consent was obtained, we inserted an intercostal drainage (ICD) tube on the left side and drained about 1500 mL of altered blood. The lung expanded completely, and the ICD was removed. The patient was safely discharged subsequently.