Author + information
- Received February 5, 2005
- Revision received June 29, 2005
- Accepted July 11, 2005
- Published online December 6, 2005.
- Michael J. Greenwood, MBBS, FRACP⁎ (, )
- Anthony J. Della-Siega, MD,
- Eric B. Fretz, MD,
- David Kinloch, MD,
- Peter Klinke, MD,
- Richard Mildenberger, MD,
- Malcolm B. Williams, MD and
- David Hilton, MD
- ↵⁎Reprint requests and correspondence:
Dr. Michael J. Greenwood, Gold Coast Heart Center, Spendelove Street, Southport 4218, Australia
Objectives The purpose of this study was to assess the accuracy of the Allen’s test (AT) in predicting hand ischemia in patients undergoing transradial coronary angiography.
Background Patients with poor vascular communications between the radial artery (RA) and ulnar artery (UA), as indicated by an abnormal AT, are usually excluded from transradial coronary angiography to avoid ischemic hand complications.
Methods Over a four-month period, patients undergoing coronary angiography were screened for AT time. Circulation in the RA, UA, principal artery of the thumb (PAT), and thumb capillary lactate were measured before and after 30 min of RA occlusion.
Results Fifty-five patients were studied (20 normal, 15 intermediate, 20 abnormal). Three patients with an abnormal AT were excluded, owing to absence of detectible flow in the distal UA. Patients with an abnormal AT were all men, had a larger RA (3.4 vs. 2.8 mm; p <0.001), and smaller UA (1.9 vs. 2.5 mm; p <0.001), compared with patients with a normal AT. After 30 min of RA occlusion in patients with abnormal AT, blood flow to the PAT improved (3.2 to 7.7 cm/s; p <0.001) yet remained reduced relative to patients with normal AT (7.7 vs. 21.4 cm/s; p <0.001. Thumb capillary lactate was elevated in patients with an abnormal AT (2.0 vs. 1.5 mmol/l; p = 0.019).
Conclusions After 30 min of RA occlusion, patients with an abnormal AT showed significantly reduced blood flow to the thumb and increased thumb capillary lactate (compared with patients with a normal AT) suggestive of ischemia. Transradial cardiac catheterization should not be performed in patients with an abnormal AT.
Radial artery (RA) cannulation carries a risk of RA occlusion with an incidence of 4.8% to 19% (1–7). This is usually of no consequence, because the hand receives blood from both the radial and ulnar arteries with extensive collateral channels; however, some patients have incomplete palmar arches and might not have adequate communications between the ulnar and radial arteries (8–10). In these patients, there is a potential risk of hand ischemia in the event of RA occlusion.
A simple bedside test to check for communications between the ulnar and radial arteries is the modified Allen’s test (AT). Patients with an abnormal test will usually have their cardiac catheterization performed via the femoral artery, thus denying them the potential advantages of transradial cardiac catheterization. In patients undergoing coronary angiography, the incidence of an abnormal AT ranges from 6.4% to 27% (11,12).
Whether the AT can predict ischemic complications after RA cannulation is controversial, and some centers no longer exclude patients with an abnormal AT.
Over a period of four months, patients undergoing cardiac catheterization at The Royal Jubilee Hospital were screened with the AT. On the basis of the AT, patients were categorized as normal (0 to 5 s, group A), intermediate (6 to 10 s, group B), or abnormal (>10 s, group C). Patients were enrolled consecutively until we reached a pre-specified number of subjects in each group.
Patients were excluded if they had symptomatic peripheral vascular disease, history of Raynaud’s phenomenon, severe aortic stenosis, atrial fibrillation, bleeding disorder, or were not taking antiplatelet therapy. All patients gave informed consent.
Perfusion of the hand was assessed with:
1. Doppler ultrasound (SonoSite [Bothell, Washington] with 10-MHz hockey stick vascular probe). Blood flow was recorded at: 1) RA, at level of radial head; 2) distal radial artery, from dorsum of hand at base of first metacarpal; 3) ulnar artery (UA), and 4) principle artery of the thumb (PAT), from palmar surface of hand at most distal point before branching.
2. Pulse oximetry of thumb (Good = signal strength >50%; weak = signal strength <50%; absent = no waveform or saturation reading).
3. Thumb capillary lactate concentration (Accutrend Lactate Analyzer, Boehringer Mannheim, Mannheim, Germany), a single drop of blood from the distal thumb was collected for analysis.
The above measurements were compared at baseline, immediately after RA occlusion (except capillary lactate), and after 30 min of RA occlusion with an RA compression device (Fig. 1).This device placed focal pressure over the RA at the wrist without affecting UA flow. Obstruction of RA flow was confirmed with Doppler. The hand was kept warm before and between measurements. Heparin 70 IU/kg (to a maximum of 5,000 IU) was given before RA compression if the patient was not already heparinized.
Patients were studied either before angiography (majority) or after their procedure if they were heparinized and the RA had not been cannulated.
Several methods of statistical analysis were used to analyze the data. Differences between Allen’s groups for baseline characteristics were analyzed with analysis of variance (ANOVA) for continuous variables and chi-square tests for categorical variables. The normality assumptions for ANOVA were assessed with the following tests available in Proc Univariate of SAS 9.1 (SAS Institute, Cary, North Carolina): Shapiro-Wilk, Kolmogorov-Smirnov, Anderson-Darling, and Cramer-von Mises. If the normality assumption was violated, the Kruskal-Wallis non-parametric test was used instead. Paired ttests were used to compare blood flow at different times for a given Allen’s group. Correlations in the data were investigated with the Pearson product-moment correlation. All analyses were done with SAS 9.1 (SAS Institute).
Over a four-month period, 55 patients were studied—20 normal, 20 abnormal, and 15 intermediate AT. Three patients with an abnormal AT were excluded because of absent ulnar flow at baseline. The baseline characteristics are summarized in Table 1.
Results are summarized in Table 2.The diameter of the RA became larger (group A = 2.8 mm, group B = 3.2 mm, group C = 3.4 mm; p = 0.0016) and the UA smaller (group A = 2.5 mm, group B = 2.2 mm, group C = 1.9 mm; p = 0.0002) as the AT time increased.
Blood flow to the PAT was significantly reduced immediately after RA occlusion in all groups, but more marked in patients with an abnormal AT (group A: 29.7 to 16.1 cm/s; group B: 26.3 to 7.8 cm/s; group C: 29.4 to 3.2 cm/s). After 30 min of RA occlusion, flow improved significantly (group A: 16.1 to 21.4 cm/s; group B: 7.8 to 14.6 cm/s; group C: 3.2 to 7.7 cm/s).
Capillary blood lactate levels after 30 min of RA occlusion increased as AT time increased (group A = 1.46 mmol/l; group B = 1.87 mmol/l; group C = 2.1 mmol/l; p = 0.007). The degree of lactate elevation correlated with blood flow in the PAT (r = −0.4; p = 0.004).
All patients had a strong pulse oximetry signal at baseline. Immediately after RA occlusion, 100%, 67%, and 0% of patients had a strong signal in groups A, B, and C, respectively. After 30 min of RA occlusion, 100%, 93%, and 64% of patients, respectively, had a strong signal.
One patient, with an abnormal AT and no recordable PAT flow at 30 min, developed numbness of his thumb. Another patient with an abnormal AT complained of mild paresthesia of the thumb. No other symptoms were reported.
The vascular anatomy of the hand is complex and highly variable (8). The UA continues into the hand and usually anastomoses with the RA to complete the superficial palmar arch. The RA continues into the hand and usually anastomoses with the UA completing the deep palmar arch. The prevalence of incomplete superficial or deep palmar arches ranges from 3.6% to 34% and 3% to 33.3%, respectively, in cadaver dissections (8,10,13,14); however, a study of 50 hands showed that all specimens had either a complete superficial or deep palmer arch (10). This would suggest that all patients have some anastomosis and that occlusion of the RA should not result in ischemic complications.
The AT is a simple test to assess for adequacy of ulnar collateral circulation before RA cannulation. False normal rates range from 3% to 45.5%, and false abnormal results might be as high as 73%, depending on technique (15–17). Some authors have suggested that the AT should be replaced by more objective and reliable tests such as Doppler ultrasound and plethysmography before the RA is cannulated or harvested for coronary bypass surgery (11,17–20). Conversely, other authors have suggested that assessment of the ulnar collateral circulation is unnecessary, owing to the absence of complications in their series (which included patients with abnormal AT) and the fact that complications often occurred in patients with a normal AT (21–23).
There are case reports of digital ischemia in patients undergoing RA cannulation in the surgical, intensive care, and perioperative setting, but none after transradial cardiac catheterization (24). These patients don’t routinely receive heparin, and the catheters are often left in place for prolonged periods. Other factors such as distal embolization, vasospasm, and distal thrombosis have been proposed (21). Of interest, the majority of these patients were reported to have a normal AT.
Previous studies have used blood flow in the hand immediately after RA occlusion as an end point. This ignores the fact that the circulation in the hand is dynamic and it might take some time for collateral channels to reach their full potential. We occluded the RA for at least 30 min and found a significant improvement in the PAT flow (Fig. 2,Table 3).This improvement appeared to be most marked in the group with an abnormal AT, where flow increased by an average of 240%, compared with 135% in patients with a normal AT. No detectible PAT flow was seen in 35% of patients with an abnormal AT, immediately after RA occlusion. After 30 min, there was detectible flow to the thumb in all but one patient. No patient with an abnormal AT had a good oximetry signal immediately after RA occlusion, compared with 64% after 30 min of RA occlusion.
In an attempt to demonstrate thumb ischemia, we measured capillary blood lactate levels in the thumb before and after 30 min of RA occlusion. If blood flow to the thumb was inadequate, we would expect to see a gradual rise in lactate concentration as the tissues became dependent on anaerobic metabolism. Patients with an abnormal AT had a significantly higher capillary lactate level after 30 min of RA occlusion. This would suggest inadequate thumb perfusion, which could lead to ischemic complications with prolonged occlusion. The degree of lactate elevation correlated with the measured blood flow in the thumb; however, there was a large amount of scatter (Fig. 3).
Although collateral circulation of the hand is dynamic, with significant improvement of blood flow to the thumb over a 30 min period of RA occlusion (particularly in patients with an abnormal AT), there is still potential for hand ischemia after prolonged RA occlusion. We recommend that, in the presence of an abnormal AT, the RA should not be used for cardiac catheterization unless the risk of using the femoral approach is excessive (e.g., severe peripheral vascular disease, morbid obesity, large abdominal aortic aneurysm).
The authors would like to thank Chanelle Edwards, BSc, for her assistance with patient testing and data collection.
This work was supported by an anonymous grant to the Victoria Heart Institute Foundation.
- Abbreviations and Acronyms
- analysis of variance
- Allen’s test
- principle artery of the thumb
- radial artery
- ulnar artery
- Received February 5, 2005.
- Revision received June 29, 2005.
- Accepted July 11, 2005.
- American College of Cardiology Foundation
- Kleinert J.M.,
- Fleming S.G.,
- Abel C.S.,
- Firrell J.
- Benit E.,
- Vranckx P.,
- Jaspers L.,
- Jackmaert R.,
- Poelmans C.,
- Coninx R.
- Rodriguez E.,
- Ormont M.L.,
- Lambert E.H.,
- et al.
- Davis F.M.,
- Stewart J.M.
- Lee K.L.,
- Miller J.G.,
- Laitung G.