| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Medical Imaging, Children's Memorial Hospital, 2300 Children's
Plaza, Box 9, Chicago, IL 60614.
2 Department of Radiology, Northwestern University, Feinberg School of Medicine,
Chicago, IL 60611.
3 Department of Pediatrics and Preventive Medicine, Children's Memorial
Hospital, Chicago, IL 60614.
4 Mary Ann and J. Milburn Smith Child Health Research Program, Children's
Memorial Hospital, Chicago, IL 60614.
Received January 19, 2006;
accepted after revision June 28, 2006.
Address correspondence to C. K. Rigsby
(crigsby{at}childrensmemorial.org).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. All patients with a central line who were referred for body CT examinations requiring an IV contrast agent were prospectively evaluated. The power injector was pressure limited to 25 psi (172 kPa). A standard dose of 2 mL/kg of iodinated contrast medium was power-injected through the central line. Two pediatric radiologists scored all examinations on a scale of 1 (poor) to 5 (superior) for adequacy of contrast enhancement. Regression and receiver operating characteristic analyses were performed.
RESULTS. The subjects were 63 patients 0.3-22 years old. Nineteen of these patients had tunneled lines, 18 had ports, and 26 had peripherally inserted central catheters. There were no complications related to power injection. Regression analysis showed a significant association between patient weight and contrast enhancement adequacy score (p < 0.001), higher patient weights yielding lower contrast enhancement adequacy scores. Receiver operating characteristic analysis showed a weight cutoff of 30 kg as a reasonable predictor of adequacy of contrast enhancement. For patients weighing 30 kg or more, the average contrast enhancement score was 2.4 (suboptimal to adequate). For patients weighing less than 30 kg, the average contrast enhancement score was 3.4 (adequate to good).
CONCLUSION. Pressure-limited power injection through central lines in children is safe. The contrast enhancement obtained with 25 psi (172 kPa) pressure-limited injection is acceptable only for patients who weigh less than 30 kg.
Keywords: catheters • contrast media • CT technique • pediatrics • safety
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Power injection of iodinated contrast medium traditionally has been accomplished with power injectors, and injection rate has been the limiting factor for power injection with a constant maximum pressure rate. To achieve consistent contrast enhancement at our institution, we have power-injected contrast medium through central lines and limited the contrast injection rate to 0.5 mL/s at 300 psi (2,068 kPa) for all types of central lines. With this method of power injection, there have been infrequent line ruptures, but the specific rate of rupture has not been recorded.
Pressure-limiting power injectors (EnVision, Medrad) have been marketed that allow the pressure to be the limiting factor. For the injection, a pressure limit is set along with a maximum injection rate. With the injector, contrast medium is administered at up to the maximum rate entered unless the pressure reaches the maximum pressure allowed. If the pressure reaches the selected limit, the injector reduces the flow rate to prevent an overpressure condition. In short, the injector is flow-rate controlled with pressure-limit monitoring (injector manual, EnVision, Medrad). No guidelines have been set by power injector manufacturers on the pressure limit to use for injections through central lines. Among the literature on the tunneled lines, PICC lines, and ports inserted at our institution at the time of this study, Cook Incorporated brochures specifically warn against power injection through its PICC lines. Hickman, Leonard, and Broviac (all Bard Access Systems) brochures recommend a maximum infusion pressure of 25 psi (172 kPa) but do not specifically mention power injection of contrast material. The Arrow International brochures reviewed give no specific guidelines on power injection through catheters made by Arrow.
There have been several reports [3-6, 9-11] on the safety and efficacy of power injection through central lines, but to our knowledge, no study has been conducted to evaluate a pressure-limiting system in vivo. The goal of this study was to evaluate the safety and efficacy of pressure-limited power injection through central lines in children undergoing body CT examinations.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
An EnVision (Medrad) CT injector was used for all injections. The injector was set to a pressure limit of 25 psi (172 kPa), which is the minimum pressure limit allowable for this injector. The accuracy of the EnVision injector for limiting the pressure when set at 25 psi (172 kPa) is ± 10% (Mertz DR, personal communication, 2004). The maximum injection rate was set at 2 mL/s. A standard contrast dose of 2 mL/kg (maximum, 125 mL) of 68% ioversol (320 mg I/mL, Optiray 320, Mallinckrodt) warmed to 37°C was injected for each study. At 37°C, 68% ioversol has a viscosity of 5.8 cP. All ports were accessed with 19-gauge noncoring needles. All catheters were inspected for integrity, were flushed with saline solution, and were guaranteed by a nurse to have a blood return to ensure line patency before power injection. If a double-lumen catheter was in place, the larger of the two lumens was used for injection.
Scans were performed on a 4-MDCT scanner (LightSpeed Plus, GE Healthcare) with standard weight-based scanning parameters as a guide [2]. Images of the chest, abdomen, and pelvis were obtained with a slice thickness of 3.75 mm for patients weighing less than 13 kg and 5 mm for patients weighing more than 13 kg. Images of the neck with a slice thickness of 2.5 mm were obtained for patients weighing less than 13 kg, 3.75 mm for patients weighing 13-40 kg, and 5 mm for patients weighing more than 40 kg. For the chest and neck, the standard scanning initiation delay was 20 seconds after initiation of contrast injection. For the abdomen and pelvis, the scanning was initiated with a minimum delay of 20 seconds after completion of contrast injection with a maximum scanning initiation delay of 50 seconds after initiation of contrast injection
Patient age and weight, type of central catheter (PICC line, port, or tunneled line), number of lumens, and catheter brand and size (if determined) were recorded. Patient charts were reviewed to determine the catheter brand and size if this information was not available at the examination. The amount of contrast medium injected and injection duration were recorded from the injector console after injection. The amount of contrast medium injected would be less than the 2 mL/kg prescribed if pressure limitation significantly limited the injection rate so that the scan was completed before completion of the contrast injection. The technologists were instructed to terminate power injection at the end of the scan if the contrast injection was not yet complete. The catheters were inspected for integrity, flushed with saline solution, and guaranteed to have blood return by the same nurse who performed this test before power injection to ensure line integrity after power injection. The findings at this catheter inspection after power injection were evidence of presence or absence of catheter rupture. Any complication related to power injection was recorded.
At imaging, a body CT radiologist determined whether contrast enhancement was adequate. For images considered inadequate for interpretation, we planned to repeat the examination with injection through a peripheral catheter. Adequacy for interpretation was a subjective determination based on the indication for the study and the level of comfort of the body CT radiologist interpreting the images with the amount of contrast enhancement present. All images obtained with central line injection were reviewed by two pediatric radiologists for subjective adequacy of contrast enhancement. The reviewers scored each set of images on a scale of 1-5 (1, poor; 2, suboptimal; 3, adequate; 4, good; 5, superior). For examinations that included more than one body part (e.g., chest and abdomen), the individual body parts were scored separately.
Data were summarized with mean and SD for continuous variables and frequencies for proportions. Safety was evaluated by recording any complications related to power injection. Efficacy was determined with a contrast enhancement adequacy score, in which the two raters separately qualitatively ranked the examinations from 1 to 5 (poor to superior). For patients with more than one clinically indicated visit for CT during the study, only the examinations performed on the first visit were considered for statistical analysis. Only one visit was included in the analysis because the number of patients with multiple examinations was too small for repeated measures analysis. Because each predictor—weight, central line type, and contrast injection rate—had different relevance for each examination type, analyses were done separately for the thoracic, abdominal, and pelvic studies. No separate analyses were conducted for the cervical studies because there were only seven such studies. Cervical studies, however, were included in analyses that combined all examination types.
Mixed effects models for categoric data with the assumption of a multinomial model and a cumulative link function were used to determine the effect of weight, age, central line type, and contrast injection rate on the adequacy of each examination [12]. Using this procedure, we modeled the probability of obtaining a better rating. Because two independent raters scored the examinations, the rater effect was included as a random effect in the model. For analyses in which all examination types were combined, a random effect for the patient was included. Results are presented as odds of obtaining a better rating and corresponding 95% CI.
Receiver operating characteristic (ROC) curve analyses were conducted to determine whether numeric cutoffs for the covariates existed that were better predictors of examination adequacy. For these analyses, the adequacy score was redefined as low (1 or 2) or high (3-5). The scores were dichotomized into adequate or inadequate with a cutoff value of 3. Most of the time the two raters agreed about whether an examination was adequate or inadequate. When the raters did not agree (one rater determined an examination adequate, and the other did not), the examination was scored inadequate. This procedure resulted in only one adequacy score per examination. Area under the curve and best combinations of sensitivity and specificity were determined. SAS software version 9.1 (SAS Institute) was used, and all conclusions were made at a 0.05 level of significance.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Nineteen (30%) of 63 patients had tunneled lines, including 14 double-lumen catheters and five single-lumen catheters. Eighteen (29%) of 63 patients had single-lumen ports. Twenty-six (41%) of 63 patients had PICC lines, including 14 double-lumen PICC lines and 12 single-lumen PICC lines. The specific brand name of the catheter was unavailable in three (5%) of the cases (one port, one single-lumen central line, and one double-lumen central line). Our institution is a tertiary care facility with many physicians placing different types of central lines, and many patients are referred to us for studies having undergone placement of a central line elsewhere. Therefore, specific operative information on central line brands and sizes was not available in these three cases despite careful chart review. A complete listing of the types of central lines, catheter brand names, number of patients with each catheter type, and the internal lumens of each of the catheters in this study is presented in Table 1. All central lines were trimmed to appropriate length according to the size of the child. The specific length of the catheters in the children in the study was not determined because this information is not recorded at our institution.
|
Safety
There were no complications related to power injection. No lines were
damaged, and no reactions to contrast material occurred. Because all of the
studies were deemed adequate for interpretation, no patients underwent
placement of a peripheral catheter for repetition of a study. After completion
of data collection for the study, we applied the 25-psi (172-kPa) rate limit
to our routine CT studies. As of this writing, we have used the 25-psi
(172-kPa) pressure limit for injections through central lines in more than 250
CT studies, and no cases of line rupture have occurred.
Efficacy
Sixty-three patients who underwent 124 studies yielding 248 interpretations
were included in the statistical analysis. Regression analysis of the
cervical, thoracic, abdominal, and pelvic studies combined showed a
significant association between contrast injection rate and contrast
enhancement adequacy score (p < 0.001), higher contrast injection
rates yielding higher contrast enhancement adequacy scores. Weight also was
significantly associated with contrast enhancement adequacy score (p
< 0.001), higher weights yielding lower contrast enhancement adequacy
scores (Figs. 1A,
1B,
1C and
2A,
2B,
2C). Figures
1A and
2A have little if any
appreciable contrast enhancement. At the completion of the scan, the patient
in Figure 1A had a total of 37
mL of the expected 125 mL injected, and the patient in
Figure 2A had a total of 12 mL
of the expected 90 mL injected. With these limited contrast injections, both
of the studies appear as though there is no IV contrast material present.
|
|
|
|
|
|
Regression analysis on individual types of studies (thoracic, abdominal, and pelvic) showed a significant effect of contrast injection rate on contrast enhancement adequacy scores of thoracic (p = 0.03), abdominal (p = 0.007), and pelvic (p = 0.056) studies, increasing contrast injection rate being associated with higher contrast enhancement adequacy scores. A significant effect of weight on the contrast enhancement adequacy score of abdominal (p = 0.01) and pelvic (p = 0.008) studies was found, higher weight being associated with a poorer contrast enhancement adequacy score. There was no significant association between weight and contrast enhancement adequacy score for the thoracic examinations. For the individual thoracic, abdominal, and pelvic examinations, type of central line had no significant effect on contrast enhancement adequacy score.
ROC analysis yielded sensitivities for weight, age, and contrast injection rate cutoffs for prediction of contrast enhancement adequacy score in the range of 74-100% and specificities in the range of 63-76% (Table 2). In cases in which ROC analysis yielded more than one choice of cutoff, the point with higher sensitivity was chosen. Although only one adequacy score per examination was included in these analyses, for the dichotomized score, disagreements between the two raters were 7%, 9%, and 10% for the chest, abdomen, and pelvis, respectively.
|
Age and weight were highly correlated in our patient population. Patient weight was always available at CT examinations and was used for determining the contrast dose, so clinical use of this parameter for predicting study adequacy was convenient. Use of contrast injection rate as a predictor of contrast enhancement adequacy would not be convenient because a test bolus would have to be obtained before the examination. The ROC curves show a cutoff of 28 kg for an adequacy score of 3 or greater with a sensitivity of 74% and a specificity of 63% for the chest, a cutoff of 30 kg for an adequacy score of 3 or greater with a sensitivity of 76% and a specificity of 76% for the abdomen, and a cutoff of 30 kg for an adequacy score of 3 or greater with a sensitivity of 78% and a specificity of 75% for the pelvis (Fig. 3A, 3B, 3C). For example, for abdominal studies, 76% of patients with adequacy scores of 3 or greater weighed 30 kg or less, and 76% of patients with adequacy scores less than 3 weighed more than 30 kg.
|
|
|
With a weight cutoff of 30 kg for the thoracic, abdominal, and pelvic studies, Table 3 shows the mean contrast enhancement adequacy scores for patients weighing less than 30 kg were 3.4 versus 2.4 for patients weighing 30 kg or more. Table 3 also shows that the mean and median contrast enhancement adequacy scores in patients who received injections through central lines were higher than in those who received injections through PICC lines. Table 4 shows that the mean contrast injection rate for patients weighing less than 30 kg was 30% greater than the injection rate for patients weighing 30 kg or more. Six of seven patients who had a contrast injection rate of 2 mL/s weighed less than 30 kg. Table 4 also shows that the mean injection rate for central lines was significantly higher than for ports or PICC lines. In 56 of the 63 central catheters through which injections were administered, the contrast injection rate was limited by pressure, so only seven catheters had an injection rate of 2 mL/s.
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
An in vitro study [6] of in-line pressure generated in small-bore pediatric central venous catheters ranging in size from 3.0-French single lumen to 7.0-French double lumen with the catheters at full and half lengths showed that the pressure generated in these catheters increased linearly with increasing flow rate. The maximum allowable flow rate for injection also was determined. The pressure used for injection to determine the maximum flow rate followed the manufacturers' guidelines. These pressure limits were 25 psi (172 kPa) for 6.6-French catheters (Bard Access Systems) and 35 psi (241 kPa) for 3.0-French, 50 psi (345 kPa) for 4.0-French, and 50 psi (345 kPa) for 7.0-French double-lumen PICC lines (Cook Incorporated).
The maximum allowable flow rates for these pressures determined in that study were 0.7 (full length catheter) to 1.5 (half length) mL/s for the 6.6-French catheter (Bard) and 0.2 (full length) to 0.5 (half length) mL/s for the 3.0-French, 0.8 to 1.4 mL/s for the 4.0-French, 1.0 to 2.0 mL/s for the 7.0-French double-lumen small-lumen, and 2.2 to 3.5 mL/s for the largelumen PICC lines (Cook). These highest flow rates were achieved with 43% meglumine iothalamate (Conray 43, Mallinckrodt Medical) because it was the least viscous of the contrast agents tested. Flow rates decreased with increased viscosity of the contrast medium.
When our study was performed, manufacturers' guidelines for power injection of central lines were not available, as they were to Ruess et al. [6]. Our starting point of 25 psi (172 kPa) as an initial power injector pressure limit, however, was in accordance with the guidelines available to Ruess et al. because any increase in this pressure would exceed the recommendations of the manufacturers of some of the lines. The range of contrast injection rates through the various catheters injected in our in vivo study did not exceed 2 mL/s (the maximum rate set on the injector for the studies), with a mean injection rate of 0.8 mL/s. Our mean injection rate was significantly lower than most rates in the in vitro study by Ruess et al.
Many factors account for differences between in vitro and in vivo studies, two of which may be most significant. First, the contrast agent routinely used in our department is warmed ioversol (Optiray 320), which is more viscous than the agents used in our in vivo study and would decrease flow rate. Second, the overall decrease in flow rate might have been secondary to formation of a fibrin cap or partial fibrin plug in the catheter lumen in patients with catheters in place for longer times.
Herts et al. [5] compared the images of 174 adults who received IV contrast power injections through central lines, 51 of whom received the power injection through a peripheral catheter. The indwelling lines in that study included ports, tunneled lines, and nontunneled lines ranging in size from 7-French triple lumen to 9.6-French double lumen. Contrast medium was injected at rates of 1.5-2 mL/s at 100 psi (689 kPa) for the central line group and 2.5-3 mL/s at 300 psi (2,068 kPa) for the peripheral catheter group. No significant catheter complications occurred as a result of power injection. The authors found that vascular enhancement was less in the central line group for the pulmonary artery and thoracic aorta but not less in the abdominal aorta. Liver enhancement was less in the central line group, the degree of enhancement being deemed possibly acceptable. The authors did not specifically assess the adequacy of the examinations for diagnosis. Our study showed that contrast enhancement of images is generally adequate for diagnosis if the patient weighs less than 30 kg. Patients receiving injections through tunneled lines had higher contrast injection rates and higher contrast enhancement adequacy scores than those receiving injection through ports or PICC lines.
The internal diameters of the tunneled central lines through which injections were administered in our study were greater than the internal diameters of any of the PICC lines. This difference in luminal diameter probably accounts for the higher injection rates and higher adequacy scores among the patients with tunneled lines. The internal luminal diameters of the central lines in this study were equal to the luminal diameter of the ports in most of the patients. We are not certain of the exact mechanism for the decrease in contrast injection rate and adequacy scores for the ports, but it may be related to the addition of the noncoring needle and the port hub as additional barriers to contrast flow in these patients.
In a study of in vitro power injection through several brands of PICCs, Rivitz and Drucker [10] found that polyurethane PICC lines handled higher pressures than all but one of their silicone counterparts. Those authors also found that the tolerated flow rates and pressures, except in the smallest silicone and polyurethane catheters, generally exceeded the manufacturers' limits. A Cook 3.0-French PICC line tolerated a maximum flow rate of 0.9 mL/s of meglumine iothalamate (Conray 60, Mallinckrodt Medical) at a pressure of 108 psi (745 kPa). This catheter burst at an injection rate of 1.2 mL/s at a pressure of 126 psi (869 kPa). Our initial pressure limit was set at not greater than 25 psi (172 kPa) because of the serious consequence of rupture of any central line in vivo. In addition, the differences between the cathetertolerated and burst pressures in the study by Rivitz and Drucker are not large.
One limitation to our study was lack of knowledge of the length of the central catheters. At our institution, catheter length is not recorded when catheters are inserted, and the catheters of almost all pediatric patients are trimmed to fit the size of the patient. The only general statement that can be made is that most small pediatric patients have shorter catheters than do larger pediatric patients. This statement is based on the assumption that as patients grow, their vascular trees lengthen proportionally. Catheter length affects flow rate, but the importance of this effect was not assessed in our study. Patients who weighed more than 30 kg had lower mean and median contrast injection rates than patients weighing less than 30 kg, probably because patients who weigh more are larger and have longer catheters, which are associated with increased resistance to flow and decreased flow rates. This finding does not take into effect the diameter of the catheter, which may be greater in larger patients. On the basis of the results of this study, we determined that 30 kg is a reasonable cutoff weight. At weights greater than 30 kg, contrast-enhanced CT studies in which contrast medium is injected through any type of central line have poor or suboptimal contrast enhancement. Below 30 kg, the mean contrast enhancement score is 3.4, and above this weight, the mean contrast enhancement score is 2.4. The ROC curves show that the 30-kg cutoff is not exact because this weight does not yield a specific ROC curve cutoff, and this study can be criticized for this deficiency. Thirty kilograms was chosen from the ROC curves to minimize the possibility of examination with poor contrast enhancement, but the cutoff does not eliminate this possibility. For clinical use, however, a weight cutoff is the most convenient measurement readily available for pediatric patients undergoing CT, and 30 kg serves as a reasonable cutoff for adequacy on the basis of ROC curves.
Since the completion of this study, we have performed power injections through more than 250 central lines for body CT studies and used the 25-psi (172-kPa) pressure limit without a single line rupture. Therefore, in our current everyday practice, we almost never administer power injections through central lines in patients weighing more than 30 kg. For patients weighing less than 30 kg, the attending body CT radiologist assesses individual cases before central line power injection, and most of these patients receive power injections through a central line. Some body CT attending physicians, however, are reluctant to administer power injections through central lines in patients whose weight approaches 30 kg if good or superior (4 or 5 on our scale) contrast enhancement is crucial. These radiologists consider these cases to have the most variable contrast enhancement results. Further studies with increasing pressure limits may prove useful for fine-tuning the maximum safe pressure limit so that the contrast injection rate can be safely maximized.
We determined that pressure-limited power injection through all types of central lines (larger than 3-French) in children is safe and that 30 kg can be used as a cutoff weight for power injection to maximize the potential of performing adequate nonangiographic body CT examinations at our institution. These guidelines may help others develop protocols for safe power injection through central lines in children.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. F. Donnelly, J. Dickerson, and J. M. Racadio Is Hand Injection of Central Venous Catheters for Contrast-Enhanced CT Safe in Children? Am. J. Roentgenol., December 1, 2007; 189(6): 1530 - 1532. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
