Special Articles Original Research

Transcription

1 Special Articles Original Research Morden et al. Displacement of CT-Injectable Peripherally Inserted Central Catheter Tip Special Articles Original Research Peter Morden 1 Farnoosh Sokhandon 1 Laura Miller 2 Michael Savin 1 Matthias Kirsch 1 Michael Farah 1 Richard Silbergleit 1 Morden P, Sokhandon F, Miller L, et al. Keywords: catheter tip displacement, CT powerinjectable peripherally inserted central catheter, CT protocol, saline flush DOI: /AJR Received January 18, 2013; accepted after revision May 13, Department of Radiology, Beaumont Health System, Oakland University, William Beaumont School of Medicine, 3601 W 13 Mile Rd, Royal Oak, MI Address correspondence to F. Sokhandon (Farnoosh.sokhandon@beaumont.edu). 2 AngioDynamics, Latham, NY. WEB This is a web exclusive article. AJR 2014; 202:W13 W X/14/2021 W13 American Roentgen Ray Society The Role of Saline Flush Injection Rate in Displacement of CT- Injectable Peripherally Inserted Central Catheter Tip During Power Injection of Contrast Material OBJECTIVE. The purpose of this article is to determine the rate and the cause of displacement of CT power-injectable peripherally inserted central catheters (CT-PICCs) during contrast material and saline flush injection and to modify CT-scanning protocols to decrease the frequency of displacement. MATERIALS AND METHODS. In the laboratory setting, in vitro modeling of CT- PICC displacement during power injection was examined while varying the initial rate of injection of the saline flush. In the clinical setting, the CT images of all patients at a large academic hospital for one calendar year who underwent power injection of CT contrast media were reviewed for CT-PICC displacement. A retrospective comparison of the rate of displacement during the 8 months before implementing a protocol with a lower initial rate of saline flush and the rate of displacement for the 4 months after the protocol change was performed. RESULTS. Laboratory modeling showed dramatic movement of the CT-PICC at higher rates of saline flush. This movement was attributed to differences in viscosity between contrast media and saline. The clinical arm of the study found that 8.2% of the 243 examinations performed before implementing the new protocol resulted in displacement, in comparison with 2.2% of the 138 examinations performed afterward. This difference was considered statistically significant (p = 0.023). CONCLUSION. Initiation of saline flush at high injection rates correlates with a higher rate of CT-PICC displacement. The use of a slower initial rate of saline flush injection significantly reduces the rate of displacement. C T-injectable peripherally inserted central catheters (CT-PICCs) are U.S. Food and Drug Administration approved for power injection of CT contrast material, with certain guidelines, depending on the model, including flow rate and average static burst pressure [1]. Since receiving Food and Drug Administration approval, the use of CT-PICCs for power injection has become increasingly popular, yet there are relatively few articles addressing potential complications with their use. A review of the literature finds that there are technical articles evaluating and comparing flow rates, static burst pressures, and other device thresholds [2] and in vitro catheter rupture [3]; however, there are not many studies looking into the root cause of in vivo displacement. Recently, four instances of CT-PICC tip displacement caused by power injection of contrast medium were reported, though the precise cause of displacement was not investigated [4]. The authors suggested reducing the contrast medium injection rate to no more than 2 ml/s, which may reduce CT image quality [4, 5]. Manufacturers have suggested catheter displacement as a complication (package insert Morpheus Power PICC, AngioDynamics). Secondary complications of catheter malposition, such as venous thrombosis, line occlusion, and cost and convenience issues, have also been reported [6 12]. To the best of our knowledge, there have been no published studies investigating the root cause of CT-PICC tip displacement after power injection of contrast material. A number of CT-PICC displacement events during power injection were diagnosed (Fig. 1) by body imagers at a large academic institution from October through December 2010, prompting further analysis. Materials and Methods This is a retrospective study of the rate of CT- PICC tip displacement after injection of contrast material when the saline flush is initiated at the AJR:202, January 2014 W13

2 Morden et al. same injection rate as contrast material (protocol A), as compared with when a slower initial injection rate of the saline flush is used (protocol B). The change to slow-increase injection of saline flush was implemented because of the results of communication with the catheter vendor (Angio- Dynamics) and a collaborative effort to understand the root cause of displacement, as a part of our department s quality improvement process. For the purpose of this article and to better convey our results, we are reporting this study as three separate phases. Phase 1 is an observational study to record the rate of catheter displacement during power injection to determine the need for protocol modification. Phase 2 is the in vitro testing performed by physicists at AngioDynamics. Phase 3 determines the rate of catheter tip displacement when a slowrate-increase saline flush protocol is implemented. Ultimately, this study compares the frequency of CT-PICC displacement before and after the injection protocol was modified. Phase 1 Patients All instances of PICC power-injected CT examinations at our 1100-bed hospital (both inpatient and outpatient) from January 1, 2011, through August 31, 2011, were logged. The research population was defined by the indication for the CT examination with IV contrast medium administration as ordered by the clinical team, patients who had a power-injectable CT-PICC in place and no peripheral line was obtainable, and CT-PICC administration of iodinated contrast material. There was no active recruitment of patients, and there was no communication to ordering physicians informing them of this analysis. Institutional review board approval was obtained. This study was initiated as part of the department s quality assurance efforts; thus, the need for patient consent was waived. The clinical arm of the study was not funded by any company or organization. CT-PICC injection protocol The radiology department s CT-PICC injection protocol requires that a topogram of the chest be acquired before and after all CT-PICC injections. Furthermore, the radiology technicians check each patient s CT-PICC position before examination and verify the maximum allowed rate and pressure. During phase 1, CT- PICC protocol A was used. Contrast agent injection rates varied according to the type of examination, ranging from 2 to 5 ml/s. A rate increase modification was performed, which involved starting the injection at 1 ml/s and progressively increasing the rate to the desired level. All examinations were performed using a power injector (Stellant, MedRad). The pressure was monitored and was not to exceed 300 psi (2,068,410 Pa). CT-PICC protocol A was in effect from January 1, 2011, through August 31, This protocol called for a power-injected 40- ml saline flush immediately after the administration of iopamidol (Isovue 370, Bracco) contrast material. The rate of saline flush was the same as the maximum rate of contrast material. The hospital used the AngioDynamics Morpheus CT-PICC exclusively during the duration of the study. Seven catheters were of an unknown brand because of placement at outside facilities and were excluded from the study. Data collection The primary outcome variable was catheter tip displacement. This was determined by retrospective analysis of the CT images (including the pre- and postscan topograms) by a fellowship-trained body imager with 12 years of experience. Catheter tip placement was verified to be in acceptable position (proximal superior vena cava [SVC] to cavoatrial junction) on the prescan topogram and was then reevaluated on the postscan topogram. Additional data abstracted from the electronic medical record and CT order forms included catheter placement site (left vs right upper extremity), catheter gauge, number of lumens, maximum pressure and rate during administration of contrast agent, and type of study. Phase 2A An AngioDynamics research team conceived of a model mimicking CT injection of contrast material through a CT-PICC line with its tip in the SVC, as follows: multiple trials of injection were completed through a 4-French single-lumen AngioDynamics CT-PICC line that had been trimmed to 45 cm and placed in a 100-mL graduated cylinder (simulating the SVC). The graduated cylinder was partially filled with saline, and the tip of the PICC was at the bottom of the cylinder. The entire apparatus was placed into a water bath at 37 C (Fig. 2A). A pressure transducer was connected to a three-way stopcock between the CT injector and catheter, near the catheter s Luer end. An aqueous glycerin solution of viscosity equal to standard contrast media was power-injected up to the PICC line s labeled rate, initially using the rate increase protocol but without the contrast-enhanced saline flush. The power injections were recorded with digital video. Initial results found no CT-PICC displacements, sparking further communication between our institution and the AngioDynamics team. This led to repeating the model, this time including the contrast-enhanced saline flush. Phase 2B During phase 2 of the project, the reduced displacement after power injection with elimination of subsequent saline flush prompted further inquiry. It was hypothesized that a drastic change in pressure within the CT-PICC secondary to the change in media viscosity between the contrast medium and saline was the root cause of the movement. The Poiseuille law of fluid dynamics states that Q = ΔpπR 4 / 8nl, where Q = flow, Δp = change in pressure from one end of a tube to the other, R = radius of the tube, n = viscosity of the fluid, and l = length of the tube. Flow, length, and radius remain constant when switching from injection of contrast material to saline. However, viscosity decreases by 21 fold (n for contrast medium = 20.9 cp [20.9 mpa s]; n for saline = 1 cp [1 mpa s]). Rearranging the equation above shows that Δp is proportional to n; thus, a 21-fold decrease in viscosity has a resulting 21- fold decrease in Δp. Phase 3 Phase 2 of the project found that, while testing catheter performance during product development, the common practice of using a saline flush of the CT-PICC after contrast material injection was not considered. Further testing indicated that a dramatic change in the pressure within the catheter secondary to the large difference in viscosity of CT contrast material and saline caused the catheter to undergo a whipping phenomenon. It was hypothesized that this whipping phenomenon was the cause of catheter displacement. The hospital s CT-PICC power-injection protocol was modified to minimize the abrupt change in catheter pressure. CT-PICC protocol B was put into effect from September 1, 2011, through December 31, 2011, and called for the saline flush to be injected using the rate increase technique. The saline flush was started at 2 ml/s and progressively increased to the rate of contrast material injection. The change in rate of saline flush was the only modification made from CT-PICC protocol A. There was no change in patient selection in comparison with phase 1. Data collection was limited to catheter displacement, catheter gauge, maximum pressure, and rate. Data were collected in the same manner described in phase 1. Statistical Analysis The rate of catheter displacement was calculated, and the Fisher exact test was applied to calculate a two-sided p value. Results Phase 1 During phase 1, of 243 CT scans performed after CT-PICC IV contrast material injection, 20 CT-PICCs were displaced. This corresponds to a displacement rate of approximately 8.2%. There was no correlation to catheter W14 AJR:202, January 2014

3 Displacement of CT-Injectable Peripherally Inserted Central Catheter Tip TABLE 1: CT Power-Injectable Peripherally Inserted Central Catheter (CT-PICC) Displacement Rates for Protocols A and B Protocol No. of Nondisplaced CT-PICCs No. (%) of Displaced CT-PICCs A (n = 243) (8.2) B (n = 138) (2.2) a Total (n = 381) (6.4) a Fisher exact test two-sided p = gauge, number of lumens, rate of contrast material injection, type of study, or maximum pressure. Among displaced catheters, 45% were 4-French and 55% were 5-French, compared with 40% 4-French, 57% 5-French, and 3% 6-French among the nondisplaced catheters. The lumen number was similar as well: 45% single-lumen and 55% doublelumen catheters were displaced, compared with 47% single-lumen, 52% double-lumen, and 1% triple-lumen nondisplaced catheters. The rate of contrast agent injection and maximum pressure averaged 4 ml/s and 281 psi (1,937,410.7 Pa), respectively, in the displaced group compared with 3.7 ml/s and 247 psi (1,702,990.9 Pa), respectively, in the nondisplaced group. In both groups, approximately 60% of CT studies were of the chest, abdomen, or pelvis, with pulmonary artery angiography and head imaging accounting for the bulk of the remaining studies. Phase 2A Fourteen trials of power injection without subsequent saline flush were performed and documented with video. This resulted in a slow curving movement of the catheter within the graduated cylinder with minimal displacement of the catheter tip (Fig. 2B). Trials with the subsequent saline flush yielded a dramatically different result: the CT-PICC, on initiation of the saline flush, whipped wildly and flipped out of the top of the graduated cylinder (Fig. 2C). These same results were obtained after repeating the experiment with a CT-PICC from a different vendor (PowerPICC, Bard Access Systems). Phase 2B A comparison of the pressure within the catheter with saline flush at the same rate as contrast agent injection (4 ml/s) versus with saline flush at half the rate of contrast agent injection (2 ml/s) found a marked decrease in the variance of pressure within the catheter (Fig. 3). This decrease in variance correlated with a decrease in observed movement of the catheter. The decrease in saline injection rate seems to modulate the effect of the decrease in viscosity. Furthermore, there is imaging that supports the hypothesis that the catheter displacement occurs at some point later than the initiation of iodinated contrast material injection (Fig. 4). Phase 3 A total of 397 CT-PICC power-injected CT examinations were performed during phases 1 and 3 of the study. Of these 397 examinations, 16 were excluded (11 from phase 1 and five from phase 3) because of undocumented PICC brand, size, or lumens (seven examinations); lack of postinjection topograms (five examinations); the PICC line being improperly positioned before injection (two examinations); CT protocol not being followed and maximum contrast agent injection rate exceeding the advised rate in the protocol (one examination); and one patient becoming agitated during image acquisition and forcibly pulling out her CT-PICC. Of the 243 examinations performed with protocol A during phase 1, 8.2% resulted in displacement, compared with 2.2% of the 138 examinations performed with protocol B during phase 3. This difference in frequency corresponds to a p value of 0.023, which is statistically significant at the α = 0.05 level (Fig. 5 and Table 1). Discussion Catheter displacement is a relevant complication for a variety of reasons. In some cases, a displaced CT-PICC may delay treatment, and patients often rely on vascular access for chemotherapy, blood products, or antibiotics. Furthermore, we think that many of these displacements are missed (particularly on studies not focused on the chest). There are potential complications of delivering chemotherapy or other agents in an inappropriate vein. Finally, catheter tip malpositioning increases rates of thrombotic complications and catheter occlusion [6 12]. In addition, consultation with the PICC team or interventional radiology for repositioning or replacement of the catheter is an added health care cost and carries with it the potential for additional complications. As a result of repositioning, the patient is exposed to additional fluoroscopy or radiation from a chest radiograph to confirm the PICC position. Patient satisfaction is another issue, because the patient must tolerate an additional procedure as well as a potentially lengthened hospital stay, if an inpatient, or an increased visit length, if an outpatient. Most important, PICC displacement affects quality of care. Power injection of the saline flush at high rates is a common practice supported by the literature [13 15]. An informal survey of seven AngioDynamics client hospitals found that five routinely power-inject the saline flush at the same rate as the previously injected contrast media. Although power injection of the saline flush increases the risk of catheter displacement, it does provide benefit to the patient: according to the literature, the use of a saline flush reduces the volume of the injected iodinated contrast agent while improving image quality [16, 17]. Laboratory testing during phase 2 of our study suggested that contrast-enhanced saline flush injection should be initiated at rates no greater than 2 ml/s. If a higher rate of saline flush is desirable, then testing suggests that at least 10 ml of saline flush at 2 ml/s should be administered before incrementally increasing the rate to the desired level. Admittedly, these suggestions are somewhat arbitrary because they are based partially on subjective observation of catheter movement in an in vitro model. However, the scientific reasoning is plausible. Furthermore, on clinical application during phase 3, we found beneficial results. There was a roughly fourfold decrease in the rate of CT-PICC tip displacement after changing the saline flush injection protocol. Conclusion Catheter tip displacement and malposition after CT-PICC IV contrast agent injection is likely underreported, particularly when power injection is done for imaging of body parts other than the chest. In addition, power injection of a saline flush initiated at high injection rates appears to be the primary cause of catheter tip displacement. We recommend that, when CT-PICCs are used for power injection, the rate of catheter tip displacement be assessed. Pre- and postscan chest topogram is an effective tool to accomplish this goal with minimal disruption to image acquisition efficiency and minimal danger to the patient. On the basis of our experience, we recommend using a slow-increase technique when powerinjecting a saline flush. AJR:202, January 2014 W15

4 Morden et al. References 1. U.S. Food and Drug Administration. Special 510(k) premarket notification: device modification Morpheus CT PICC and procedure kit. Food and Drug Administration website.www. accessdata.fda.gov/cdrh_docs/pdf7/k pdf. Published August 27, Accessed July 8, Coyle D, Bloomgarden D, Beres R, Patel S, Sane S, Hurst E. Power injection of contrast media via peripherally inserted central catheters for CT. J Vasc Interv Radiol 2004; 15: Salis AI, Eclavea A, Johnson MS, Patel NH, Wong DG, Tennery G. Maximal flow rates possible during power injection through currently available PICCs: an in-vitro study. J Vasc Interv Radiol 2004; 15: Lambeth L, Goyal A, Tadros A, et al. Peripherally inserted central catheter tip malposition caused by power contrast medium injection. J Vasc Interv Radiol 2012; 23: Plumb AA, Murphy G. The use of central venous catheters for intravenous contrast injection for CT examinations. Br J Radiol 2011; 84: Kearns PJ, Coleman S, Wehner JH. Complications of long arm-catheters: a randomized trial of A C central vs peripheral tip location. JPEN J Parenter Enteral Nutr 1996; 20: Schutz JC, Patel AA, Clark TW, et al. Relationship between chest port catheter tip position and port malfunction after interventional radiologic placement. J Vasc Interv Radiol 2004; 15: Puel V, Caudry M, Le Metayer P, et al. Superior vena cava thrombosis related to catheter malposition in cancer chemotherapy given through implanted ports. Cancer 1993; 72: Petersen J, Delaney JH, Brakstad MT, Rowbotham RK, Bagley CM Jr. Silicone venous access devices positioned with their tips high in the superior vena cava are more likely to malfunction. Am J Surg 1999; 178: Cohn DE, Mutch DG, Rader JS, Farrell M, Awantang R, Herzog TJ. Factors predicting subcutaneous implanted central venous port function: the relationship between catheter tip location and port failure in patients with gynecologic malignancies. Gynecol Oncol 2001; 83: Luciani A, Clement O, Halimi P, et al. Catheterrelated upper extremity deep venous thrombosis in cancer patients: a prospective study based on Doppler US. Radiology 2001; 220: B D 12. Racadio JM, Doellman DA, Johnson ND, Bean JA, Jacobs BR. Pediatric peripherally inserted central catheters: complication rates related to catheter tip location. Pediatrics 2001; 107:E Schoellnast H, Tillich M, Deutschmann HA, et al. Improvement of parenchymal and vascular enhancement using saline flush and power injection for multiple-detector-row abdominal CT. Eur Radiol 2004; 14: Bae KT. Intravenous contrast medium administration and scan timing at CT: considerations and approaches. Radiology 2010; 256: Behrendt FF, Bruners P, Keil S, et al. Effect of different saline chaser volumes and flow rates on intravascular contrast enhancement in CT using a circulation phantom. Eur J Radiol 2010; 73: Tillich M, Schoellnast H. Optimized imaging of pulmonary embolism. Eur Radiol 2005; 15(suppl 5):E66 E Haage P, Schmitz-Rode T, Hübner D, Piroth W, Günther RW. Reduction of contrast material dose and artifacts by a saline flush using a double power injector in helical CT of the thorax. AJR 2000; 174: Fig. 1 Examples of CT power-injectable peripherally inserted central catheter (CT-PICC) displacement during power injection in two different patients. A and B, 55-year-old woman. Unenhanced (A) and contrast-enhanced (B) coronal maximum-intensity projection (MIP) reformatted CT scans show interval displacement of tip of CT-PICC. C and D, 76-year-old man. Scout topogram (C) shows CT-PICC tip (arrowhead) projecting over right main bronchus within superior vena cava, whereas contrast-enhanced MIP CT image (D) shows distal CT-PICC coiled within right internal jugular vein. W16 AJR:202, January 2014

5 Displacement of CT-Injectable Peripherally Inserted Central Catheter Tip Fig. 2 Photographs of laboratory power injection model. A, Model is shown before injection of material. B, Catheter curving movement is seen with minimal tip displacement (arrowhead) during injection of iodinated contrast material. C, Catheter tip displacement (arrowhead) is shown immediately after initiation of 4 ml/s saline flush. Pressure (psi) A 4 ml/s Saline Flush Finish 1:00 1:10 Time (mins) 8 1:20 A Pressure (psi) B 2 ml/s Saline Flush :00 1:10 Time (mins) Fig. 3 Pressure variance within catheter with different rates of saline infusion after contrast agent injection at 4 ml/s. A and B, Immediately after injection of contrast agent at 4 ml/s, there is much greater variance in intraluminal pressure when injecting saline at 4 ml/s (A) than at 2 ml/s (B). 1:20 Finish C B AJR:202, January 2014 W17

6 Morden et al. Fig year-old man. Pulmonary embolism CT shows timing of catheter displacement. A, Scout topogram shows catheter tip (arrowhead) projecting over distal superior vena cava. Per pulmonary embolism CT protocol, 80 ml of iodinated contrast is injected at 4 ml/s and image acquisition is triggered once pulmonary artery measures 80 HU or greater. Image acquisition in this study began 10 seconds after initiating contrast injection (i.e., in midst of iodinated contrast agent injection). B, Maximum-intensity-projection CT image shows mild curving motion and slight catheter tip displacement (arrowhead), similar to that seen in laboratory model (Fig. 2B). C, Postimaging topogram acquired after saline flush shows about 8 cm of catheter looped within internal jugular vein (arrowhead) and catheter tip at junction of subclavian and jugular veins. 11 Excluded 223 Negative A B C 254 Protocol A 20 Positive 397 CT-PICC power injections reviewed 3 Positive 143 Protocol B 5 Excluded 135 Negative Fig. 5 Flowchart of results from phases 1 and 3. CT-PICC = CT power-injectable peripherally inserted central catheter. W18 AJR:202, January 2014