Cardiovasc Intervent Radiol https://doi.org/10.1007/s00270-017-1857-0
CLINICAL INVESTIGATION
Chest Radiograph Measurement Technique Facilitates Accurate Bedside Peripherally Inserted Central Catheter Placement in Children Aishu Ramamurthi1,2 • Jeffrey Forris Beecham Chick1 • Rajiv N. Srinivasa1 • Anthony N. Hage1 • Jason J. Grove1 • Joseph J. Gemmete1 • Timothy D. Johnson3 Ravi N. Srinivasa1
•
Received: 19 October 2017 / Accepted: 5 December 2017 Springer Science+Business Media, LLC, part of Springer Nature and the Cardiovascular and Interventional Radiological Society of Europe (CIRSE) 2017
Abstract Purpose To report the chest radiograph measurement technique for placing bedside peripherally inserted central catheters (PICCs). Materials and Methods Two hundred and thirty-two consecutive pediatric patients, mean age of 56.3 months (range: 0–203 months), underwent PICC placement from January 2015 to May 2017 (28 months) with a total of 232 PICCs placed. Measurements were taken from the cavoatrial junction to clavicle, clavicle to medial margin of midhumeral head, and medial margin of mid-humeral head to mid-humerus. The difference between total radiographic measured length and actual PICC length was then calculated, and the percent difference (from actual cut length) was recorded. An equivalence test was performed using the two, one-sided test method. Results Mean ± standard deviation cavoatrial junction to clavicle length was 5.29 ± 2.20 cm (range: 2.1–12.6 cm). Mean clavicle to shoulder length was 8.20 ± 3.59 cm (range: 3.23–19.06 cm). Mean shoulder to mid-humerus length was 7.88 ± 3.87 cm (range: 2.01–16.8 cm). Mean total radiographic measured length was 21.37 ± 9.19 cm
(range: 7.42–43.6 cm). Mean actual cut PICC length was 20.64 ± 8.72 cm (range: 8.5–44 cm). The mean difference between predicted, or total radiographic measured length, and actual cut PICC length was 0.73 ± 2.51 (range: - 5.42–8.60 cm). The mean percent difference was 4.07 ± 12.65% (range: - 23.84–47.80%). An equivalence test rejected the null hypothesis of the true percent difference greater/less than ± 6.67% with a p value of 0.002. Conclusion The chest radiograph measurement technique is an accurate method to determine catheter length for PICC placement at bedside in the pediatric population. Keywords Bedside peripherally inserted central catheters PICC placement Central venous catheters Bedside Pediatric Children Interventional radiology Abbreviations PICC Peripherally inserted central catheter
Introduction & Ravi N. Srinivasa
[email protected] 1
Department of Radiology, Division of Vascular and Interventional Radiology, University of Michigan Medical Center, 1500 East Medical Center Drive, Ann Arbor, MI 48109, USA
2
Department of Biology, Duke University, Durham, NC 27705, USA
3
Department of Biostatistics, School of Public Health, University of Michigan, 1415 Washington Heights, Ann Arobr, MI 48109, USA
Peripherally inserted central catheters (PICCs) are a standard, cost-effective, alternative to centrally inserted tunneled catheters for long-term venous access [1–6]. PICC placement is the most commonly performed venous access procedure in the pediatric population [7]. The growth in PICC placements may be attributed to safer insertion, increased convenience, and reducing the risks associated with tunneled central venous catheters [5, 6, 8].
123
A. Ramamurthi et al: Chest Radiograph Measurement Technique Facilitates Accurate Bedside…
Despite these advantages, complications may result from incorrect catheter tip positioning, including cardiac arrhythmias, venous thrombosis, thrombophlebitis, and cardiac perforations [1, 5, 9]. Image-guided placement facilitates optimal positioning of the catheter tip at the junction of the superior vena cava and right atrium (cavoatrial junction) [1, 9]. Non-image-guided PICC placement at the bedside; however, is often necessary due to medical comorbidities or the absence of image guidance [1, 5, 10]. In such cases, various methods have been used to determine proper catheter length including age, height, weight, electrocardiographic, and surface measurementbased techniques [1]. The literature suggests that the rates of successful PICC placement in these bedside cases may vary widely from 37 to 99% across all age-groups [6, 10–13]. In the pediatric population, one study reported that 85.8% of 843 pediatric PICCs placed without image guidance resulted in a non-central tip position [14]. A simple and accurate technique for estimating appropriate PICC length would be beneficial in facilitating bedside PICC placement. While there are studies that have developed formulae to predict PICC length from patient height, there are no studies reporting the use of chest radiography to aid in proper central tip placement [1]. This report describes the chest radiograph measurement technique to determine PICC length for bedside placement and the accuracy of this technique for central tip placement in the pediatric population.
or without a recorded weight or height within 2 weeks prior to PICC placement were excluded from the study. Patients who had midline or non-central catheter placements were also excluded. Measured and Defined Variables Patient age, weight, height, vein insertion site, catheter size, number of catheter lumens, predicted PICC length (as described below), and actual cut PICC length were recorded (Fig. 1). The cavoatrial junction was defined as being approximately two vertebral bodies below the carina, as previously reported [15]. Proper PICC placement was defined as a catheter tip within 1 cm of the cavoatrial junction on chest radiography. For each patient, the total radiographic measured length was calculated by a single operator. Age was recorded in months, weight in kilograms, and height in centimeters. Total radiographic measured length (predicted PICC length) was determined by a single author and verified by a board-certified radiologist (RNS with 6 years of experience) and reported in centimeters (Fig. 1). The actual cut PICC length was recorded in centimeters; this was verified in two locations, including the annotated length on the PICC placement image as well as the radiology report. The difference between total radiographic measured length (predicted PICC length) and actual cut
Materials and Methods Patient Selection This study was conducted with institutional review board approval and complied with the Health Insurance Portability and Accountability Act. Informed consent was not required for this retrospective study. Patients were identified via retrospective review of the electronic medical record (EPIC; Epic Systems Corporation; Verona, WI) in conjunction with the department’s prospectively maintained database (Microsoft Access 2016; Seattle, WA). Pediatric patients, \ 18 years, who underwent PICC placement in the upper extremity from January 2015 to May 2017 (n = 232) were identified. Inclusion/Exclusion Criteria All patients \ 18 years of age who underwent upper extremity PICC placement between January 2015 and May 2017 were considered for inclusion in this study. Patients [ 18 years of age, without a chest radiograph 1 month prior to PICC placement, a malpositioned PICC,
123
Fig. 1 Diagram illustrating the technique for estimating PICC length on a right-sided PICC. Measurements are taken from a straight line drawn from the cavoatrial junction to sternal end of clavicle (Line A; ClCa), a straight line drawn from the sternal end of clavicle to the medial margin of the middle humeral head (Line B; S-Cl), and a straight line drawn parallel to the long axis of the humerus from the medial margin of middle humeral head to the mid-humeral shaft (Line C; Mh-S). The sum of these three measurements determines the PICC length
A. Ramamurthi et al: Chest Radiograph Measurement Technique Facilitates Accurate Bedside…
PICC length was calculated, and the value was reported in centimeters. Patient Demographics Two hundred and thirty-two consecutive pediatric patients, mean age of 56.3 months (range: 0–203 months), underwent PICC placement with a total of 232 PICCs placed. Mean weight and height were 19.29 kg (range: 0.51–91.90 kg) and 92.2 cm (range: 30.6–182.9 cm), respectively (Table 1). Chest Radiograph Measurement Technique (Figs. 1, 2, 3) For each PICC placement with proper tip positioning, a supine chest radiograph within one month of placement was identified. Images were identified and accessed via the hospital Picture Archiving and Communication System (PACS) (McKesson, San Francisco, CA). The line measurement tool on the PACS software was utilized for all subsequent distance measurements. In order not to bias measurements, a radiograph that did not contain the PICC was utilized. The side that the PICC was placed on (left or right) was the only variable known to the operator prior to performing measurements. The cavoatrial junction was first identified. Measurements recorded included lengths of a straight line drawn from the cavoatrial junction to the sternal end of the clavicle (Line A; C-Ca), a straight line drawn from the sternal end of clavicle to the medial margin of the middle humeral head (Line B; S-Cl), and a straight line drawn parallel to the long axis of the humerus from the medial margin of middle humeral head to the mid-humeral diaphysis (Line C; Mh-S). Total radiographic measured length (predicted PICC length) was computed by summing these three measurements. After these values were reported, the image following PICC placement was reviewed to ensure actual PICCs were positioned within 1 cm of the cavoatrial junction (two vertebral bodies below the carina).
Table 1 Patient information Age (months)
Weight (kg)
Height (cm)
Mean
56.32
19.3
92.2
Standard deviation
64.9
19.7
41.2
Median
24
10.9
82.6
Range
0–203
0.51–91.90
30.6–182.9
Fig. 2 Example of the chest radiograph measurement technique for estimating PICC length on a right-sided PICC. Measurements are taken from a straight line drawn from the cavoatrial junction to sternal end of clavicle (Line A; C-Ca), a straight line drawn from the sternal end of clavicle to the medial margin of the middle humeral head (Line B; S-Cl), and a straight line drawn parallel to the long axis of the humerus from the medial margin of middle humeral head to the midhumeral shaft (Line C; Mh-S). The sum of these three measurements determines the PICC length
Fig. 3 Example of the chest radiograph measurement technique for estimating PICC length on a left-sided PICC. Measurements are taken from a straight line drawn from the cavoatrial junction to sternal end of clavicle (Line A; C-Ca), a straight line drawn from the sternal end of clavicle to the medial margin of the middle humeral head (Line B; S-Cl), and a straight line drawn parallel to the long axis of the humerus from the medial margin of middle humeral head to the midhumeral shaft (Line C; Mh-S). The sum of these three measurements determines the PICC length
123
A. Ramamurthi et al: Chest Radiograph Measurement Technique Facilitates Accurate Bedside…
Statistical Analyses Calculations of mean, range, standard deviation, and percentages were performed using the R statistical software package [16]. Paired differences were calculated between total radiographic measured length (predicted PICC length) and actual cut PICC length. An equivalence test was conducted using the percent difference with true cut length as the normalizing variable. For the lower and upper equivalence bounds, the authors assumed a percent change of ± 6.67% (2/30 cm) as a tolerable percent change. The authors determined this percentage by computing that for an average PICC line of 30 cm they were willing to accept a ± 2 cm error. The dataTOSTone function in the R TOSTER package [17] was used to conduct the two onesided tests (TOST) on the percent difference. The null hypothesis was that the true percent difference was \ - 6.67 or [ 6.67%. Pearson correlation coefficients were also calculated comparing patient height with actual cut length and patient height with predicted cut length.
Results PICCs were placed in 232 pediatric patients from January 2015 to May 2017. One hundred and five PICCs (45.3%) were placed in the right basilic vein, 45 PICCs (19.4%) in the left basilic vein, 38 PICCs (16.4%) in the right brachial vein, 16 PICCs (6.9%) in the right cephalic vein, 16 PICCs (6.9%) in the left brachial vein, 9 PICCs (3.9%) in the left cephalic vein, and 1 PICC (0.4%) each in the right subclavian, right antecubital, and left antecubital veins. A total of 161 (69.3%) right-sided PICCs and 71 (30.7%) left-sided PICCs were placed. One hundred and twenty-five (53.9%) 3-French, 68 (29.3%) 4-French, 15 (6.5%) 2.6-French, 15 (6.5%) 1.9French, and 9 (3.8%) 5-French PICCs were placed. Of the 232 PICCs inserted, 154 (66.3%) were single lumen and 78 (33.7%) were double lumen, as listed in Table 2. Mean total radiographic measured length (predicted PICC length) was 21.37 ± 9.19 cm (range: 7.42–43.6 cm). Mean actual cut PICC length was 20.64 ± 8.72 cm (range: 8.5–44 cm). The mean difference between total radiographic measured length (predicted PICC length) and actual cut PICC length was 0.73 ± 2.51 (range: - 5.42–8.60 cm). The mean percent difference was 4.07 ± 12.65% (range: - 23.84–47.80%), as listed in Table 3. The dataTOSTone function in the R TOSTER package [17] was used to conduct the two one-sided tests (TOST) on the percent difference. For the upper TOST limit, the t value was 13.2 with p \ 0.001. For the lower TOST limit, the t value was - 2.31 with p = 0.002. The 95% confidence interval was (2.90, 5.63%), well within the tolerated equivalence bounds of ± 6.67%. The authors rejected the
123
Table 2 PICC type/insertion N
Percent
Vessel accessed Right basilic
105
45.3
Left basilic
45
19.4
Right cephalic
16
6.9
Left cephalic
9
3.9
Right brachial
38
16.4
Left brachial Right subclavian
16 1
6.9 0.4
Left subclavian
0
0
Right antecubital
1
0.4
Left antecubital
1
0.4
232
100
161
69.3
Left
71
30.7
Total
232
100
1.9
15
6.5
2.6
15
6.5
Total Sidedness Right
Catheter size (French)
3
125
53.9
4
68
29.3
5
9
3.8
232
100
Total Number of lumens Single
154
66.3
Double
78
33.7
232
100
Total
null hypothesis that the true percent difference was \ - 6.67 or [ 6.67% (p \ 0.001 and p = 0.002). Thus, the authors conclude that there is no statistically significant difference between the predicted and actual PICC cut lengths. Pearson correlation coefficient between patient height and actual cut length was 0.954 (R2 = 0.927) and was 0.963 (R2 = 0.927) between patient height and predicted cut length. Therefore, PICC lengths correlated well across all patient heights.
Discussion This report shows that simple measurements taken from a chest radiograph may accurately determine appropriate PICC length for placement of the tip at the cavoatrial junction during bedside placement.
A. Ramamurthi et al: Chest Radiograph Measurement Technique Facilitates Accurate Bedside… Table 3 Chest radiograph measurements Cavoatrial to clavicle length (cm)
Clavicle to shoulder length (cm)
Shoulder to mid-humerus length (cm)
Total predicted length (cm)
Actual catheter length (cm)
Mean
5.29
8.2
7.88
21.37
20.64
Standard deviation
2.20
3.59
3.87
9.19
8.72
Median Range
4.45 2.10–12.61
6.96 3.23–19.06
6.58 2.01–16.80
18.46 7.42–43.6
18.25 8.50–44.0
Malposition of a PICC after placement at the bedside is encountered commonly in a hospital setting [18]. Accepted guidelines advise PICC tip placement in the lower third of the superior vena cava at the cavoatrial junction with adequate tip positioning best facilitated using image guidance [1, 9, 19, 20]. Bedside PICC insertion; however, is often necessary in critically ill children. In such cases, catheter length has traditionally been determined using a tape ruler and other operator-dependent methods [1]. Given the variability in patient height, weight, and arm length, these methods are often ineffective, with reported rates of successful catheter tip placement varying from 37 to 99% [6, 10–12, 21, 22]. In the pediatric population, the rate of PICC malposition with blind insertion was reported as high as 85.8% in a large series [14]. Electrocardiographic-guided placement has also been reported in neonates [23]; however, such systems may not be universally available. Severe complications may occur due to malpositioning of a PICC tip [1, 5, 9]. Racadio et al. [24] reported a complication rate of 28.8% for patients with non-centrally placed PICC tips compared to 3.8% for correctly placed PICCs. Venous thrombosis was reported in 38% [25]. Elsharkawy et al. [9] documented a case of arrhythmia resulting from incorrect PICC positioning. These studies show a high complication rate associated with a malpositioned PICC placed at the bedside and highlight the need to develop an effective technique for determining appropriate PICC length for bedside placement. Such a method would reduce the guesswork involved in bedside PICC placement and may reduce malposition associated complications. There are a few methods reported in the literature to predict optimal PICC length during bedside placement. Lum et al. [26] offered a measurement guide to estimate central venous catheter length from patient height. This technique correlated the length of the upper extremity bones and veins in adults, but is unsubstantiated in the pediatric population given a majority of these patients are still growing. Cho et al. [1] detailed a formula for predicting right-sided PICC length from patient height; however, this technique has not been applied to the pediatric population and is limited to right-sided
placement. Another consideration is that PICCs may change in position over time and with movement. Gnannt et al. [27] reported that pediatric PICCs inserted into the cephalic vein and that were made of silicone were less likely to move compared with PICCs inserted into basilic or brachial veins and made of polyurethane. Nurses may be helpful in assessing appropriate PICC tip position relative to the carina on chest radiography [28]. Prematurity is also a factor in where a PICC tip should be placed. PICCs should be placed 0.5–1 cm from the right atrium in premature infants and 1–2 cm from the right atrium in larger infants [29]. Considering the limitations in the above reports, this study technique using chest radiography to determine PICC length provides a patient-specific reference that is applicable to right- and left-sided bedside PICC placement. This study reported PICCs in 232 pediatric patients. Measurements were taken from a straight line drawn from cavoatrial junction to the sternal end of clavicle, a straight line drawn from the sternal end of clavicle to the medial margin of the middle humeral head, and a straight line drawn parallel to the long axis of the humerus from the medial margin of middle humeral head to the mid-humeral shaft. With a mean difference of 0.73 cm between predicted PICC length and actual PICC length, this study suggests that chest radiograph measurements accurately predict optimal PICC length at the bedside without the need for image guidance. Caution needs to be exercised in interpretation of the results given the possible confounding variables and errors inherent in a retrospective study. Although the sample size was large, patients were only considered from a single institution. Furthermore, there are many factors that influence PICC positioning, but this study limited its focus to PICC length. Limitations also include the fact that the arm may have been accessed at sites higher or lower than the mid-arm; compensation for this may be made with simple arithmetic while still using this technique. Additionally, variability in positioning of the arm on the chest radiograph either prior to or following placement of the PICC may have accounted for some differences in predicted length or
123
A. Ramamurthi et al: Chest Radiograph Measurement Technique Facilitates Accurate Bedside…
ultimate positioning. Further, the authors relied on reported actual cut length from the dictated report during retrospective review of the PICC lengths; incorrect transcription may have occurred in some patients yielding an inaccurate number. These factors may account for some of the outliers in the study. Future studies should consider other relevant factors including vein selection and insertion technique. Other appropriate factors such as body habitus and point of insertion should also be considered.
Conclusion The chest radiograph measurement technique is an accurate method to determine PICC length for central tip placement at bedside in the pediatric population. This technique may be useful in the absence of image guidance. Compliance with Ethical Standards Conflict of interest All the authors declare that they have no conflict of interest.
References 1. Cho HH, Jeon E, Lee HJ, et al. A new formula to estimate the length of right upper extremity vein from elbow crease to carina calculated by peripherally inserted central catheter insertion through right basilic vein puncture. J Korean Soc Radiol. 2012;66:229–33. 2. Ng PK, Ault MJ, Ellrodt AG, Maldonado L. Peripherally inserted central catheters in general medicine. Mayo Clin Proc. 1997;72:225–33. 3. Goltz JP, Scholl A, Ritter CO, Wittenberg G, Hahn D, Kickuth R. Peripherally placed totally implantable venous-access port systems of the forearm: clinical experience in 763 consecutive patients. Cardiovasc Intervent Radiol. 2010;33:1159–67. 4. Amerasekera SS, Jones CM, Patel R, Cleasby MJ. Imaging of the complications of peripherally inserted central venous catheters. Clin Radiol. 2009;64:832–40. 5. Chopra V, Flanders SA, Saint S. The problem with peripherally inserted central catheters. JAMA. 2012;308:1527–8. 6. Venkatesan T, Sen N, Korula PJ, et al. Blind placements of peripherally inserted antecubital central catheters: initial catheter tip position in relation to carina. Br J Anaesth. 2007;98:83–8. 7. Braswell LE. Peripherally inserted central catheter placement in infants and children. Tech Vasc Interv Radiol. 2011;14:204–11. 8. Loughran SC, Borzatta M. Peripherally inserted central catheters: a report of 2506 catheter days. JPEN J Parenter Enteral Nutr. 1995;19:133–6. 9. Elsharkawy H, Lewis BS, Steiger E, Farag E. Post placement positional atrial fibrillation and peripherally inserted central catheters. Minerva Anestesiol. 2009;75:471–4. 10. Neuman ML, Murphy BD, Rosen MP. Bedside placement of peripherally inserted central catheters: a cost-effectiveness analysis. Radiology. 1998;206:423–8. 11. Davis J, Kokotis K. A new perspective for PICC line insertions: cost effectiveness and outcomes associated with an independent PICC service. J Assoc Vasc Access. 2004;9:93–8.
123
12. Glauser F, Breault S, Rigamonti F, Sotiriadis C, Jouannic AM, Qanadli SD. Tip malposition of peripherally inserted central catheters: a prospective randomized controlled trial to compare bedside insertion to fluoroscopically guided placement. Eur Radiol. 2017;27:2843–9. 13. Linda J, Bledsoe L, Hadaway LC. A retrospective look at tip location and complications of peripherally inserted catheter lines. J Infus Nurs. 1993;16:104–9. 14. Fricke BL, Racadio JM, Duckworth T, Donnelly LF, Tamer RM, Johnson ND. Placement of peripherally inserted central catheters without fluoroscopy in children: initial catheter tip position. Radiology. 2005;234:887–92. 15. Song YG, Byun JH, Hwang SY, Kim CW, Shim SG. Use of vertebral body units to locate the cavoatrial junction for optimum central venous catheter tip positioning. Br J Anaesth. 2015;115(2):252–7. 16. R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2016. URL https://www.R-project.org/. 17. Lakens D. Equivalence Tests: a Practical Primer for t Tests, Correlations, and Meta-Analyses. Soc Psychol Personal Sci. 2017;8(4):355–62. 18. Ng PK, Ault MJ, Maldonado LS. Peripherally inserted central catheters in the intensive care unit. J Intensive Care Med. 1996;11:49–54. 19. Crowley JJ, Pereira JK, Harris LS, Becker CJ. Peripherally inserted central catheters: experience in 523 children. Radiology. 1997;204:617–21. 20. Chait PG, Ingram J, Phillips-Gordon C, Farrell H, Kuhn C. Peripherally inserted central catheters in children. Radiology. 1995;197:775–8. 21. Trerotola SO, Thompson S, Chittams J, Vierregger KS. Analysis of tip malposition and correction in peripherally inserted central catheters placed at bedside by a dedicated nursing team. J Vasc Interv Radiol. 2007;18:513–8. 22. James L, Bledsoe L, Hadaway LC. A retrospective look at tip location and complications of peripherally inserted central catheter lines. J Intraven Nurs. 1993;16:104–9. 23. Zhou L, Xu H, Liang J, Xu M, Yu J. Effectiveness of Intracavitary Electrocardiogram Guidance in Peripherally Inserted Central Catheter Tip Placement in Neonates. J Perinat Neonatal Nurs. 2017;31(4):326–31. 24. 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:E28. 25. Allen AW, Megargell JL, Brown DB, et al. Venous thrombosis associated with the placement of peripherally inserted central catheters. J Vasc Interv Radiol. 2000;11:1309–14. 26. Lum P. A new formula-based measurement guide for optimal positioning of central venous catheters. J Assoc Vasc Access. 2004;9:80–5. 27. Gnannt R, Connolly BL, Parra DA, Amaral J, Moineddin R, Thakor AS. Variables decreasing tip movement of peripherally inserted central catheters in pediatric patients. Pediatr Radiol. 2016;46(11):1532–8. 28. Zhang X, Jia D, Ke N, Liu C, Fu L, Hu X. Excellent interobserver agreement between radiologist and nurse: tracheal carina-based identification of peripherally inserted central catheter tip position. J Vasc Access. 2017;6:0. 29. Sneath N. Are supine chest and abdominal radiographs the best way to confirm PICC placement in neonates? Neonatal Netw. 2010;29(1):23–35.