Data in this article were presented at the American Academy of Pediatrics Section of Emergency Medicine, Atlanta GA, October 2006.
The Anatomic Relationship of Femoral Vein to Femoral Artery in Euvolemic Pediatric Patients by Ultrasonography: Implications for Pediatric Femoral Central Venous Access
Article first published online: 29 MAR 2008
© 2008 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 15, Issue 5, pages 426–430, May 2008
How to Cite
Warkentine, F. H., Pierce, M. C., Lorenz, D. and Kim, I. K. (2008), The Anatomic Relationship of Femoral Vein to Femoral Artery in Euvolemic Pediatric Patients by Ultrasonography: Implications for Pediatric Femoral Central Venous Access. Academic Emergency Medicine, 15: 426–430. doi: 10.1111/j.1553-2712.2008.00087.x
- Issue published online: 29 MAR 2008
- Article first published online: 29 MAR 2008
- Received June 28, 2007; revision received November 22, 2007; accepted November 26, 2007.
- emergency ultrasound;
- femoral vein size;
- pediatric emergency medicine;
- patient safety
Background: Knowledge of the femoral vein (FV) anatomy in pediatric patients is important in the selection of appropriate size central line catheters as well as the approach to central venous access. This knowledge may avoid potential complications during central line access.
Objectives: To describe the relationship of the FV to the femoral artery (FA). To measure FV diameter and FV depth using ultrasonography (US) in newborns, infants, and children up to 9 years of age.
Methods: This study was a prospective descriptive study at a tertiary care children’s hospital. A convenience sample of euvolemic children was enrolled aged 0–9 years presenting to an urban pediatric emergency department. All patients underwent a standardized US evaluation using a Sonosite Titan bedside machine by a single emergency physician. The FA and FV were identified by four criteria: relative positions, FV compressibility, FV enlargement by Valsalva maneuver, and absence of FV pulsatility. The position of the FV relative to the FA was described as being completely overlapped by the FA, having partial (<50%) overlap by the FA, and having no overlap by the FA. The FV depth was measured from the skin to the superior border of the vein using the US machine’s caliper function.
Results: A total of 84 patients were studied. The FV was found to be completely overlapped by the FA in 8% of subjects and partially overlapped by the FA in 4% of subjects. The mean FV diameter ranged from 4.5 mm in young infants to 10.8 mm in patients 9 years of age. The mean FV depth ranged from 6.5 mm in neonates to 11.2 mm in patients 9 years of age.
Conclusions: External landmarks were not always predictive of internal anatomy. The FV was completely or partially overlapped by the FA in 12% of cases. Thus, visualization of femoral vessels should be recommended prior to attempting pediatric femoral central venous access.
Central venous lines are frequently placed during resuscitation of critically ill children. Traditionally, central lines have been placed in the femoral vein (FV), internal jugular vein (IJV), and subclavian vein (SCV). Physicians conducting a resuscitation may prefer FV central line placement because it does not interfere with chest compressions or airway instrumentation. In addition, placement of IJV or SCV central lines may be less desirable since these sites have been associated with a significant risk of arterial puncture in young children.1
External landmarks are traditionally used for placement. Differences between external landmarks and internal anatomy may lead to complications such as arterial puncture. Femoral artery (FA) puncture has associated with other complications such as pseudoaneurysm, arteriovenous fistula, and embolic events.2 FV overlap by the FA is known to occur in adults.3 This relationship may lead to inadvertent FA puncture when attempting to access the FV. Increased knowledge of the internal anatomy may help practitioners place femoral central lines in children more successfully and with fewer complications.
When using the FV for pediatric central venous access, FV diameter determines the selection of appropriate size catheters. In addition, knowledge of FV anatomy may avoid potential complications during central line placement such as arterial puncture, hematoma, and failure to place the central line. The proportion of children with FA overlapping the FV is unknown, but overlap is known to occur in adults.3 Additionally, there are only three studies in the literature that have assessed FV diameter in pediatric patients using computed tomography, angiography, and ultrasound (US).4–6 The single previous study using US was limited because it was carried out only in neonates. There are little data describing the relationship of the FV to the FA in children, FV diameter, and FV depth in pediatric patients.
This study was prospective and descriptive. The institutional review board approved the study. Informed consent was obtained from the parent or guardian. Written assent was obtained from all patients ≥7 years old.
Study Setting Population
This study was conducted between February 1 and March 15, 2006, in the emergency department of an urban, tertiary care children’s hospital. A total of 105 patients were invited to participate; 84 subjects were enrolled. Fourteen subjects were enrolled in each of the following age categories: young infant, 1 year of age, 3 years of age, 5 years of age, 7 years of age, and 9 years of age. Subjects up to 2 months of age were enrolled in the young infant age group. Subjects were enrolled in the remaining age categories if their date of birth fell within 6 months of the age category. The sample size of 14 subjects per group was selected so as to guarantee reasonable precision for the interval estimates of internal FV size. In particular, using a prior estimate of 2.2 mm for the standard deviation of FV size, it was calculated that a sample size of 14 subjects per group would produce a 95% confidence interval (CI) half-width of 1.2 mm, a desirable degree of precision.5
Inclusion criteria included a history of euvolemia. Patients were determined to be euvolemic by history (absence of vomiting, diarrhea, or other medical history suggesting hypovolemia and history of adequate oral intake). Exclusion criteria included: history of vascular attempts in the femoral region, severe hip contractures that obscured femoral vessel anatomy, and allergy to US gel.
All US procedures and measurements were performed by a single, senior pediatric emergency medicine fellow utilizing a single Sonosite Titan (SonoSite, Inc., Bothell, WA) machine to minimize interobserver variability. The investigator performing the US measurements had 40 hours of didactic training in US and 3 years of clinical experience using US.
Ten milliliters of US transmission gel, Aquasonic 100 (R.P. Kincheloe, Dallas, TX), was applied to the skin overlying the femoral vessels. A transducer was placed 1 cm below the inguinal ligament, over the pulsation of the FA. The FA and FV were identified by four criteria: their relative positions, FV compressibility, FV enlargement by Valsalva maneuver, and absence of FV pulsatility.
The FV and FA images were captured on the US machine and saved to a compact flash card. Compression of the FV was avoided by applying gentle pressure when the image was captured.
This standardized technique was repeated in both supine position and in the position of 30° of reverse Trendelenburg. Subjects were examined with their hips slightly externally rotated. Both left and right FVs diameter and depth were measured in both positions for all individuals studied. The relationship of the FV to the FA was also described.
The US image was displayed on an 8.4-inch square screen. The Sonosite Titan is equipped with two sets of internal calipers. For this study, two different transducers were used, based on age recommendations from the manufacturer. For patients 0 to 7 years of age, a 25-mm broadband (10-5 MHz) linear array hockey-stick transducer was used. For patients greater than 7 years of age, an 11-mm broadband (7-4 MHz) tightly curved array transducer was used. The precision of caliper measurement is ±2% of the distance measured plus 1% of the depth of the image.
The internal diameter of the vein was measured from the lateral wall to the medial wall using the internal calipers. The depth of the vein was also measured using the US machine’s caliper function. We defined the depth of the FV as the distance from the skin to the superior border of the vein. The relative anatomic position of the FV was described as either completely medial to the FA, posterior to the FV (the FA overlapped the FV by more than 50%), or posterior-medial to the FA (the FA could not overlap the FV by more than 50%; Figure 1).
FV diameter, in both the supine and the reverse Trendelenburg positions, and FV depth in both the right and left legs were estimated with 95% CIs in each of the age groups. The associations between FV diameter and age, height, weight, body surface area (BSA), and body mass index, (BMI) were tested with separate, simple linear regressions. Separate linear regressions were applied for age, height, weight, BSA, and BMI due to the high degree of correlation among these five parameters and the inherent difficulty of fitting regression models with highly correlated predictors. Analyses were conducted in SPSS 14.0.2 (SPSS Inc., Chicago, IL, 2005).
Eighty-nine patients were assessed for eligibility: only 5 declined study participation. Therefore, a convenience sample of 84 patients was studied. Fourteen subjects were enrolled in each of the following age categories: young infant, 1 year of age, 3 years of age, 5 years of age, 7 years of age, and 9 years of age. The demographic characteristics of the sample are presented by age group in Table 1. No subjects were excluded for hypovolemia.
|Age Categories||Age (yr)||Weight (kg)||Height (cm)||BSA (m2)||BMI (kg/m2)||Gender (M/F)||Race (White/African American/Other|
|Young Infant (n = 14)||0.08 (0.03)||4.2 (0.8)||52.4 (2.7)||0.25 (0.03)||15.4 (2.77)||7/7||10/4/0|
|1 yr (n = 14)||0.85 (0.27)||9.2 (1.5)||70.6 (4.6)||0.42 (0.05)||18.34 (2.08)||9/5||5/9/0|
|3 yr (n = 14)||3.27 (0.33)||15.5 (2.0)||95.6 (6.2)||0.64 (0.06)||17.00 (2.11)||6/8||6/7/1|
|5 yr (n = 14)||4.99 (0.40)||18.5 (2.1)||109.1 (6.1)||0.75 (0.05)||15.65 (2.02)||6/8||8/6/0|
|7 yr (n = 14)||7.16 (0.32)||26.8 (5.9)||124.6 (7.1)||0.96 (0.11)||17.27 (3.48)||8/6||7/7/0|
|9 yr (n = 14)||9.08 (0.30)||35.0 (7.1)||136.7 (7.8)||1.15 (0.15)||18.56 (2.13)||9/5||7/7/0|
Mean diameters of the right and left FVs for each of the age categories are listed with ranges and with 95% CIs in Tables 2 and 3. Pooling the data for all 84 subjects showed that the supine and reverse Trendelenburg positioning measurements were 6.9 mm versus 7.8 mm for left FV diameter and 7.2 mm versus 8.0 mm for right FV diameter.
|Age Category||LFV||RFV||FV Catheter Size (Fr)|
|Mean (mm)||95% CI||Range||Mean (mm)||95% CI||Range|
|Young Infant||4.1||3.7, 4.5||3–5.1||4.9||4.4, 5.4||3.6–6.8||3.0 (1 mm)|
|1 yr||5.2||4.6, 5.8||3–7.2||5.6||5.2,6.1||4–7.7||3.0–4.0 (1–1.3 mm)|
|3 yr||6.7||6.0, 7.4||4.1–9.2||6.4||5.7, 7.1||4–8.8||4.0–5.0 (1.4–1.6 mm)|
|5 yr||6.6||6.0, 7.2||4.7–8.6||7.4||6.7, 8.1||4.7–10.0||4.0–5.0 (1.4–1.6 mm)|
|7 yr||8.1||7.3, 8.8||6–10.6||8.2||6.9, 9.5||5.1–15.8||4.0–5.0 (1.4–1.6 mm)|
|9 yr||10.9||10.0, 11.8||8.1–13.9||10.6||9.0, 12.2||6.2–16.5||5.0–8.0 (1.6–2.6 mm)|
|Age Category||LFV-T (mm)||RFV-T (mm)||LFV Depth (mm)||RFV Depth (mm)|
|Neonates||4.5 (3.9, 4.9)||4.4 (4.1, 4.75)||6.6 (5.8, 7.5)||6.5 (5.7, 7.2)|
|1 yr||6.2 (5.6, 6.8)||5.5 (4.9, 6.1)||5.5 (4.3, 6.8)||6.3 (5.6, 7.0)|
|3 yr||7.0 (6.0, 8.0)||7.5 (6.3, 8.7)||6.2 (5.5, 6.9)||6.3 (5.4, 7.1)|
|5 yr||8.0 (7.3, 8.7)||8.0 (7.5, 8.6)||7.7 (5.8, 9.7)||8.6 (7.2, 10.1)|
|7 yr||9.3 (8.3, 10.2)||9.8 (8.6, 11.0)||10.3 (6.5, 14.1)||9.2 (5.7, 12.8)|
|9 yr||11.9 (10.8, 13.0)||12.4 (10.8, 14.0)||11.2 (8.6, 13.8)||10.2 (6.8, 13.6)|
The relationships between FV diameter and age, height, weight, BSA, and BMI were tested with separate simple linear regressions. Separate regression models were performed rather than multiple regression models due to the high degree of correlation among age, height, weight, and BSA. This high degree of correlation indicates that these predictors can be considered as proxies for each other. Similar regression analyses were conducted for FV depth onto age, height, weight, BSA, and BMI.
Table 4 provides the R2 measures of goodness-of-fit for each of the regression models tested. Age, height, weight, BSA, and BMI were strongly significantly associated with FV diameter and depth, with a few exceptions. Examining the results for the left FV diameter, the regression coefficient indicates that a 1-year increase in age is associated with a 0.65-mm increase in left FV diameter. Similarly, our results showed that each additional 1 kg of weight is associated with a 0.18-mm increase in FV diameter. Our results also showed that for each 1 cm of increase in height the FV increases by 0.07 mm.
|LFV diameter (mm)||RFV diameter (mm)||LFV depth (mm)||RFV depth (mm)|
In 8% percent of our patients, the FV lies completely inferior to the FA, just below the inguinal ligament. In addition, in 4% percent of our patients, the FV is partially overlapped by the FA (Figure 1). FV overlap by FA occurred across all age groups (Table 5).
|Age Category||No Overlap||Partial (<50%) Overlap||Complete Overlap|
External landmarks are traditionally used for central venous line placement. The discordance between external landmarks and internal anatomy may lead to complications such as arterial puncture. Increased knowledge of the internal anatomy may help practitioners place femoral central lines in children more successfully and with fewer complications.
In our study, FV diameter was strongly associated with weight in newborns and children up to 9.5 years of age. These results are similar to Akingbola et al.,5 who showed a statistical correlation between weight and FV diameter. To capture images, we used an internal capture system and calipers that were more precise than the instant camera used by Akingbola et al.
Our results showed that age and BSA were highly correlated with FV diameter. These findings were consistent with previous studies that found age, weight, and BSA were highly correlated with FV diameter.1,6 Furthermore, our findings that height is correlated with FV diameter may be helpful for those practitioners who use the Broselow tape system for selection of venous catheters. In our study, mean FV diameter was weakly correlated with BMI. Estimating FV diameter by BMI may not be accurate in pediatric patients.
At present, recommendations for venous catheters for pediatric patients include: 3.0 French (1 mm) for neonates, 3.0–4.0 French (1–1.3 mm) for infants less than 12 months of age, 4.0–5.0 French (1.3–1.6 mm) for patients ages 1–8 years of age, and 5.0–8.0 French, (1.6–2.6 mm) for pediatric patients older the 8 years of age.7 These recommendations propose catheter sizes that are smaller than the measured venous diameter of the patients in their respective age categories. This suggests that failure to cannulate the FV for central venous catheter placement may be related to factors other than vein size.
Thirty-degree reverse Trendelenburg position increased FV diameter on average by 12%. The effect of increased FV diameter in reverse Trendelenburg was observed in patients over 3 years of age. Previous findings suggested that positioning of patients increased FV diameter by decreasing pooling of blood in the lower extremities.5,6 Reverse Trendelenburg positioning may increase success rates of central line placement using an external landmark technique since it may be easier to cannulate larger vessels.
Our study also examined FV depth from the skin. Age correlates in determining FV depth as the subject reaches >5 years of age. Increases in weight and BSA with age are likely the strongest contributors to the increase in correlation of age to FV depth. It is of interest that FV depth was weakly correlated with weight.
Our study differed from previous studies by examining the relationship of the FV to the FA in children. The FV was positioned very close to the FA, just below the inguinal ligament. Previous studies in adults have shown that up to 8% of the time the FV is partially overlapped by 25% of the FA.3 We found that 4% of pediatric patients have a FV that is partially overlapped by the FA. More importantly, we found that 8% of pediatric patients have a FV that courses completely under the FA. This is a finding that was previously not reported in studies of adult anatomy.3 This finding may have significant clinical implications, because it defines that 12% of our patient population was at high risk for FA puncture during FV central line placement.
Our data support the fact that the FA can overlap the FV and thus clinicians should look with US on both sides before attempting blindly. We also believe that our study suggests that US may provide an idea of the FV size to help select an appropriate catheter.
There was little variance of weight and BMI within our age categories. A cohort of patients at the extremes of weight or BMI for age may yield different results.
Patients in the neonate and 1 year of age categories tended to cry more throughout the first part of the exam while the subjects were in the supine position. As the US exam progressed and the subjects were moved into a reverse Trendelenburg position, the subjects would often stop crying as they became familiar with the ultrasonographer. This observation may account for some of the differences in measurements between the supine and reverse Trendelenburg positions seen across age categories. A study with a larger sample size may better evaluate this finding.
Ultrasound is known to be an operator-dependent technology. Some variance in measurements may have occurred secondary to caliper placement. We minimized interobserver variability for all measurements by using a single ultrasonographer with previous US training and experience.
The FV in children is very small and may be difficult to cannulate, which may lead to higher failure rates in central line placement. External landmarks are not always predictive of internal anatomy. The FV may lie beneath the FA, which may complicate FV cannulation during central line placement. These data may guide clinicians in the selection of appropriate sized catheters and support the use of US prior to attempting pediatric FV access.
- 7ChameidesL, HazinskiMF (Eds). Pediatric Advanced Life Support. 3rd ed. Dallas, TX: American Heart Association, 1997.