OBJECTIVETo describe and analyze the complications of subcutaneous venous access port for patients with gastrointestinal malignancy.

METHODSData of 1 912 patients with gastrointestinal malignancy who accepted chemotherapy in our department via subcutaneous venous access ports, including 127 cases in upper arm, 865 cases in subclavicular vein and 920 cases in internal jugular vein, from June 2007 to April 2016 were analyzed retrospectively. Associated complications and risk factors were emphatically investigated.

RESULTSPostoperative complications were confirmed in 233 patients(12.2%), and complication morbidity was 37.0%(47/127), 15.5%(134/865), 6.7%(62/920) in upper arm group, subclavicular vein group, internal jugular vein group respectively, whose difference was statistically significant (χ=71.060, P=0.000). Sixty-one(3.2%) patients developed early complications (in the day of insertion, including catheter dislocation, pneumothorax, arterial damage). Early complication morbidity of upper arm group (14.2%, 18/127) was higher as compared to subclavicular vein group (3.4%, 29/865) and internal jugular vein group(1.5%, 14/920) with significant difference (χ=57.867, P=0.000). Postoperative long-term complications (catheter dislocation, thrombosis, pinch-off syndrome, infusion base exposure, catheter detachment) were found in 182(9.5%) patients. Morbidity of long-term complication was 5.2%(48/920) in internal jugular vein group, which was significantly lower than 22.8% (29/127) in upper arm group and 12.1% (105/865) in subclavicular vein group with statistically significant difference (χ=50.828, P=0.000). Multivariate analysis indicated that subclavicular vein intubation (OR=0.536, 95%CI: 0.341 to 0.843; P=0.007 OR=0.156, 95%CI: 0.096 to 0.253, P=0.000), internal jugular vein intubation (OR=0.156, 95%CI: 0.096 to 0.253, P=0.000), operation time <40 minutes (OR=0.458, 95%CI: 0.342 to 0.613, P=0.000) and standardized training (OR=0.233,95%CI: 0.171 to 0.318, P=0.000) were protective factors of postoperative complication; besides, subclavicular vein intubation (OR=0.458, 95%CI: 0.342 to 0.613, P=0.000), internal jugular vein intubation (OR=0.233, 95%CI: 0.171 to 0.318, P=0.000) and standardized training (OR=0.313, 95%CI: 0.173 to 0.568, P=0.000) were protective factors of thrombosis.

CONCLUSIONSSubcutaneous venous access port implantation is a preferable access to central vein. Appropriate intubation approach and standardized training may reduce postoperative complications effectively. Internal jugular vein approach is safer and more reliable than upper arm vein and subclavian vein approach.