Adult patients who underwent ultrasound-guided PICC performed by an intensive care medical trainee or an intensivist during their ICU stay were included in the study. Of these patients, patients younger than 18 years of age, those with insufficient medical records, and those discharged before 14 days after PICC placement were excluded. A total of 209 patients with PICC placement were analyzed in this study (Fig. 1). To assess the safety and feasibility of PICC performed by subspecialty trainees in critical care medicine, postprocedural outcomes were compared between trainee-led and intensivist-led procedures.

We retrospectively reviewed all placements of PICCs in the ICU during the study period. The illness severity on ICU admission was estimated by the Simplified Acute Physiology Score 3 (SAPS 3) and Sequential Organ Failure Assessment (SOFA) systems. Central line-associated bloodstream infections (CLABSIs) were defined as an infection resulting from the same bacteria as confirmed by line and blood cultures [4,11,13,14]. CLABSIs were also identified in collaboration with the Infection Prevention and Control Team of Samsung Medical Center who monitors for CLABSIs [11]. Compression ultrasonography and duplex Doppler were performed only when the patient had clinical signs and symptoms suggesting venous occlusion such as a swelled arm, unexplained local pain, or malfunctioned PICC [1,11].

In this study, the primary outcome was PICC-induced complications. The secondary outcome was initial success of central line placement.

An intensivist determined the PICC placement in critically ill patients. The ultrasound-guided PICC was preferred at the bedside for patients with hemodynamic instability, on mechanical ventilation, or those who were critically ill. In this study, all PICCs were inserted as an elective procedure. Indications for placement of PICC included the need for a central line for parenteral nutrition, infusion of drugs requiring a central line, need for frequent blood sampling, or difficult venous access [1,11]. Contraindications to PICC placement were small deep veins of the arm (diameter of target vein <3 mm), local contraindications due to specific arm conditions such as skin infection, burns, or an arteriovenous fistula for renal failure [11]. Obesity or severe edema were not considered contraindications for PICC placement. We used 5 Fr single-lumen silastic catheters or 5 Fr dual-lumen Turbo-Ject Power-Injectable PICCs (Cook, Bloomington, MN, USA), 5 Fr triple-lumen PowerPICC Catheters (Bard Access Systems, Salt Lake City, UT, USA), or 6 Fr dual-lumen Vaxel PICCs with PASV Valves (Navilyst, Marlborough, MA, USA) [11]. PICC placement was performed by an intensivists or intensive care medical trainee. After trainees had seen the procedure three times, intensivists supervised their procedures two or three times before they performed the PICC placement independently. In the supine position, the patient’s arm was abducted and externally rotated. Usually, the target vein is either the basilic, brachial, or cephalic vein, and a tourniquet is sometimes applied to help find the target veins. The PICC placement was performed by an ultrasound-guided puncture of the deep veins in the upper midarm using the microintroducer technique. A standard 5 to 10 MHz linear ultrasound probe was used. The optimal length was determined for placement of the catheter tip in the cavoatrial junction, which is commonly done by measuring the distance from the site of insertion through the humeral head to the sternal notch, and down to the 3rd intercostal space. Maximal barrier precautions were consistently used during the procedure [11]. Ultrasound was used to anesthetize the skin and reidentify veins. After the ultrasound-guided puncture of the deep vein, the syringe was removed and the guide wire was advanced through the needle. The needle was removed and the wire placement in the vein was confirmed using ultrasound. Using a scalpel, a small nick at the insertion site was created to accommodate the dilator. The dilator and introducer were inserted over the guide wire. After removal of the guide wire and dilator, only the introducer was left in place. Finally, the catheter was inserted through the introducer and advanced to the predetermined length before the introducer was removed. The target veins and procedure are shown in Fig. 2. The correct position of the tip of the catheter was verified by chest radiographs. Malposition was defined as a catheter tip that was not located at the cavoatrial junction in chest radiographs. Dressing changes occurred every 7 days, or if soiled in all placements [11].

All data are presented as mean±standard deviation for continuous variables and numbers (percentages) for categorical variables. We compared data using Student’s t test for continuous variables and chi-square test or Fisher’s exact test for categorical variables. All tests were two-sided and P values <0.05 were considered statistically significant. Data were analyzed using IBM SPSS statistics ver. 20 (IBM, Armonk, NY, USA).

A total of 209 patients underwent PICC placements during their ICU stay. The mean age of the patients with PICC was 59.3±15.9 years. Of 209 patients, 116 (55.5%) were males. Hypertension (54.5%) and malignancies (53.6%) were the most common comorbidities among patients who underwent PICC placement. Respiratory failure (27.3%) and sepsis (23.4%) were the most common causes of ICU admission. There were no significant differences in age, gender, body mass index, comorbidities, causes of ICU admission, or severity scores of illness on ICU admission between the two groups (Table 1).

Difficult venous access (42.6%) and requirement for central line infusion (39.2%) were the most common reasons for PICC placement. Almost all patients (92.8%) used a mechanical ventilator and 70 patients (33.5%) had hemodynamic instability. Renal replacement therapy was more frequently used in the trainee group compared to the intensivist group. The basilic vein (62.2%) and the brachial vein (20.6%) were the most common target veins among patients who underwent PICC placement. Activated partial thromboplastin time was prolonged in the trainee-led PICC group compared to the intensivist-led PICC group. There were no significant differences in reasons for PICC placement, use of anticoagulant and antiplatelet agent, insertional veins, platelet count, and international normalized ratio between the two groups (Table 2). The procedure times of the intensive medical trainees and intensivists according to procedure number are shown Fig. 3. Although the procedure times of intensive medical trainees decreased after three or four procedures (Fig. 3A), those of intensivists according to procedure number were similar (Fig. 3B).

There were no significant differences in complication between the two groups (P=0.473). There were five cases of CLABSI (2.4%) and three cases of symptomatic PICC-related venous thrombosis (1.4%) in all patients who underwent PICC placement. However, there were no insertional injuries and moderate or severe bleeding in patients with PICC (Table 3). In addition, there were no significant differences in the rate of initial success and procedure time between the two groups (P=0.108 and P=0.076, respectively).