A retrospective study was conducted in the nephrology department of a tertiary care teaching hospital situated in southern India. Approval was taken from the Institutional Ethics Committee for collation and analysis of the data. The data for 243 CVCs placed by a single operator from March 2014 till September 2017 was extracted and analysed retrospectively. These included patients with end stage kidney disease, acute kidney injury requiring haemodialysis and few patients on Continuous Ambulatory Peritoneal Dialysis (CAPD). Though all efforts were made to place the CVC under guidance, the use of ultrasound and fluoroscopy was limited by the availability of the equipment. Due to variable access to equipments, most of the CVC placed in the evening and night hours were landmark guided. Three cohorts of patients were identified, landmark guided cannulation arm (n=99), ultrasound guided arm (n=108) and ultrasound plus fluoroscopy guided arm (n=36). The data analysed included, the demographic profile, co-morbidities, indication for CVC placement, history of prior IJV cannulation, number of passes (attempts) for successful cannulation and the complications encountered. A complete blood count, blood grouping, and coagulation parameters were obtained before the procedure.

All consenting patients without a functional vascular access requiring renal replacement therapy were included in the study. Haemodialysis being an obligatory lifesaving therapy, none of the patients were excluded.

The choice of the catheter: The choice between cuffed and uncuffed catheters were dependent on the availability of the hardware and the clinical circumstances. The guiding principle was the anticipated duration of haemodialysis, with cuffed CVC being used when the duration was likely to be more than two weeks. CVC was always used as a bridge to AVF except in occasional patients who had exhausted all vascular access. The uncuffed CVC used was a 12 F, 13.5 centimeters, pre-curved double lumen dialysis catheter (elite-Harsoria, Elite Biomedicals, New Delhi). The cuffed CVC used was 14.5 F, 24 centimeters (tip to cuff), double lumen silicon catheter with split-tip (HaemoSplit, Long Term Haemodialysis Catheter, Bard, Utah, USA).

The choice of the vein: The preferred site for cannulation was the right IJV. In landmark guided cannulation, the procedure was abandoned in the event of either an arterial puncture or failure to cannulate the IJV despite two passes. In such cases, the further course was decided based on the clinical urgency and the availability of an ultrasound machine. In the absence of an ultrasound machine, when the clinical circumstances dictated urgent vascular access, the left IJV was cannulated using a landmark guided technique.

Catheterisation technique: Standard part preparation, cleaning, draping, and aseptic precautions were practiced for all the cases. For the landmark based cannulation, the apex of the Selidotts triangle, formed between the two heads of sternocleidomastoid muscle was the point of insertion. The direction of the needle was towards the ipsilateral nipple. The initial cannulation of the IJV was done using a 21 G x 7-centimeter-micro-puncture needle (Merit Mak mini access kit, Merit Medical System, Utah, USA). The needle tip was deemed to be in the IJV based on the color and the nature of the blood column. Once the position of the micro-puncture needle was confirmed, a 0.018-inchx40 centimeters stainless steel guidewire was introduced, followed by 4Fx10 centimeters co-axial introducer sheath. The guidewire was exchanged over the sheath with a 0.035-inch guidewire followed by removal of the sheath. The CVC was then introduced over the wire and secured.

For the ultrasound guided cannulation, a 7.5 MHz linear probe was placed perpendicular to the long axis of the vein. The IJV and carotid artery were visualised and the compressibility of the vein and pulsatility of the artery was checked. The course of the puncture needle was followed in real-time. When additionally fluoroscopy was available, the course of the guidewire, dilators, and the catheter were followed real time.

In cases of cuffed CVC, to allow a wide-angle tunnel towards the exit site, a low jugular venous entry point was chosen. After the venous puncture and insertion of the guidewire, the exit site was planned using fluoroscopy. The cuffed CVC was kept externally over the body and manipulated so that the tip overlies the right atria and the course mimics the actual final position. A small incision was made at the planned exit-site, and the catheter was negotiated through the tunnel using the tunneling trochar. The soft tissue around the guidewire was serially dilated, followed by insertion of the peel-away sheath and the catheter. A post-procedure chest X-ray was taken in all the cases and complications if any, were recorded.

Outcome assessment: The surrogate markers to assess the difficulty level associated with various cannulation modalities included the number of passes before successful cannulation, the use of alternative venous sites and technique failure. The procedure related safety was assessed using complications encountered. These CVC related complications were classified using the American Society of Diagnostic and Interventional Nephrology (ASDIN) classification [7].

For the purpose of statistical analysis, the patients were divided into two cohorts based on initial cannulation technique i.e., landmark and ultrasound guidance cohort. The numeric values were expressed as mean and standard deviation. The statistical analysis was carried out by manually creating a 2x2 contingency table and using online resources (www.graphpad.com/quickcalcs/contingency1?/). The test of significance applied was two tailed Fisher’s-exact test. A p-value of <0.05 was accepted as statistically significant.

A total of 243 patients, who required haemodialysis vascular access during the period were included in the study [Table/Fig-1]. These included 174 males (71.6%) and 69 females (28.4%). The cannulation techniques used included landmark guided, ultrasound guided, and ultrasound plus fluoroscopy guided in 99 (40.8%), 108 (44.4%) and 36 (14.8%) patients respectively. The mean age of these three arms was 48±8.8, 50±9.1, and 56±9.8 years respectively. The most common indication for CVC placement was a presentation with advanced symptomatic azotemia without vascular access. A total of 163 patients (67%) presented in this fashion, highlighting poor AVF awareness. Fistula failure- both primary and secondary, constituted the second most common indication. Thirty-three patients (13.5%), had acute kidney injury requiring renal replacement therapy and six patients (2.4%) had refractory CAPD peritonitis, requiring CAPD catheter removal. A total of 180 (74.1%) un-cuffed catheters and 63 (25.9%) cuffed CVC were placed during the study period. Cuffed CVC was always placed under the guidance, with ultrasound being obligatory and fluoroscopy desirable.

The rate of successful first-pass cannulation in the landmark guided arm was 75.7%, which improved to 92.4% by the second pass [Table/Fig-2]. The corresponding first-pass success for the ultrasound and ultrasound plus fluoroscopy guidance arm was 97.2% and 100%, respectively. The procedure was unsuccessful in eight patients (8.1%) in the landmark guided arm. Of these eight patients, CVC could be placed in six, using the landmark guidance in the left IJV, and two eventually required ultrasound assistance.

The most common complication encountered was ASDIN type I i.e., local haematoma formation [Table/Fig-3]. The incidence was 10.1%, 3.7% and 5.5% in the landmark arm, ultrasound arm, and ultrasound plus fluoroscopy arm respectively. Inadvertent carotid artery puncture (ASDIN type III complication) was noticed in 5 patients; all in the landmark guided arm. Non-sustained ventricular tachycardia (ASDIN type IX complication) was observed in one patient each, of landmark guided and ultrasound with fluoroscopic guided arm. One patient in the ultrasound guided arm had a central vein perforation (ASDIN type II complication) which required endovascular intervention and placement of stent graft [8]. None of the patients developed a pneumothorax, haemothorax, brachial plexus injury or pericardial tamponade.

Based on the modality of initial IJV cannulation, for statistical purposes, the study population was divided into two cohorts: landmark guidance and ultrasound guidance [Table/Fig-4]. The use of ultrasonography significantly improves the first-pass successful cannulation to over 97% (p-value <0.0001).

Haemodialysis is the commonest renal replacement modality worldwide. It requires vascular access capable of providing rapid extra-corporeal blood flow. The gold standard for vascular access remains an AVF. Despite the ‘Fistula First’ initiate, CVC continues to be used in a significant number of patients on haemodialysis. A recent study on haemodialysis vascular access in India showed that less than one-quarter of the patients started dialysis with an AVF [9]. Whether we love it or hate it, CVC will continue to play an important part in the therapeutic armamentarium of a nephrologist. Strategies like the use of ultrasound and fluoroscopy are often employed in a bid to reduce the CVC related acute complications. The existing studies evaluating the safety profile of these aids are confounded by the presence of multiple operators with varying skill levels making the comparison inappropriate. Peculiar local circumstances in the present study made it possible to eliminate this skill level bias.

The initial IJV cannulation is the most critical step in the performance of the procedure. An important finding of the present study is successful landmark guided cannulation of the right IJV in three-fourth of the patient in the first pass and in over 90% by the second pass. Only 8% of patients required the use of either an alternative site, i.e., the left IJV or assistance of ultrasound guidance. The study thus validates the efficacy of the landmark guided technique in a resource-poor environment. The use of ultrasound guidance significantly improves the first pass successful cannulation to over 97%. This improved success with ultrasound is a result of direct visualisation of the blood vessel, mitigating the anatomical variability of the IJV. Similar findings were noted by Farrell J and Gellens M, [10]. In their study, they noted a successful cannulation rate of 96.67% with the use of ultrasound guidance and 82% with use of landmark guidance. The use of ultrasound resulted in a significant reduction in the requirement for the second pass and use of the alternative sites for CVC placement [10].

The optimal CVC tip placement continues to be debated with conflicting recommendations and guidelines. Further confounding the issue is the different CVC tip designs and the role of the side holes. The general guiding principle is to avoid placing it too deep or too high so as to avoid trauma to the floor of the right atrium and the wall of the superior vena cava respectively. Fluoroscopy helps in the appropriate placement of the CVC tip and in the early detection of catheter kink. Though not a part of the present study, the catheter tip placement was more appropriate and the blood flow achieved was superior when additional fluoroscopy was available.

ASDIN defines procedure related complications as an unanticipated adverse event that requires therapy [7]. The most common complication noted in the present study was the formation of a local haematoma. The incidence of local haematoma formation,10.1%, and 4.2%, in the landmark and ultrasound guidance arm, is not statistically significant (p-value 0.1115). The higher-incidence of haematoma, in the landmark guided arm, despite the use of low profile micropuncture needle, was likely a consequence of higher rate of arterial puncture and increased requirement for the second pass. These haematomas were minor, and the only therapeutic intervention required was local compression. The incidence of inadvertent carotid artery puncture was significantly lower with the use of ultrasound guidance; 5% of patients in the landmark guided cohort and none in the ultrasound guided cohort had this complication (p-value-0.0106). Similar findings of higher carotid artery puncture with landmark guidance were observed by others [11-13]. One major difference between the present study and the existing studies is the complete absence of inadvertent carotid artery puncture in the ultrasound guided arm. This is likely due to the improvement in the resolution of the ultrasound over the years and the experience of the operator. Only one major life threatening complication in the form of iatrogenic central venous perforation was noted during the study period. In summary, of the 243 procedures performed, complications were few and required nominal or minor therapy; only one patient required major endovascular intervention.

The small sample size, due to the novel design of the study, aimed at eliminating the skill level bias, by compiling data of a single operator.

In resource poor environment, landmark guided approach when performed by an experienced operator continues to be a safe alternative for placement of dialysis catheters. The use of aids like ultrasound and fluoroscopy when available further improves the efficacy and the safety profile of the procedure.