Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Is the Allen's test adequate to safely confirm that a radial artery may be harvested for coronary arterial bypass grafting?

Three Part Question

In [patients undergoing CABG surgery using radial artery grafts] is the [Allen's test, plethysmography, Doppler ultra-sound or MRI imaging] the best method of assessing [ulnar artery or collateral flow]

Clinical Scenario

You are at a clinical research meeting when you hear presentations comparing the use of magnetic resonance imaging (MRI), plethysmography and Doppler ultrasound techniques to assess adequacy of ulnar collateral flow in patients scheduled for radial artery graft conduit harvesting for CABG surgery. You decide to review the literature to identify just how good these techniques are and to find out whether they offer any advantage in identifying satisfactory collateral flow in the forearm over the Allen's test which you currently use in your own practice.

Search Strategy

Medline 1966 to March 2005 using OVID interface
EMBASE 1980 to March 2005
[ OR exp Thoracic Surgery/OR Coronary art$ OR Cardiopulmonary OR exp Cardiovascular Surgical Procedures/OR exp Thoracic Surgical Procedures/OR exp Coronary Artery Bypass] AND [exp Radial Artery/OR radial artery] AND [ OR exp Plethysmography/OR Magnetic resonance OR exp Magnetic Resonance Imaging/OR exp Magnetic Resonance Angiography/OR Angiography/OR Allens OR exp Ulnar Artery/OR Doppler OR exp Ultrasonography, Doppler] AND [collateral OR exp Collateral Circulation/OR exp Regional Blood Flow/OR exp Ulnar Artery/OR exp Hand/OR ulnar blood]

Search Outcome

A total of 176 papers were identified: 58 on Medline, 111 on Embase and 7 by hand searching of reference lists. Fifteen papers representing the best evidence on the subject are summarised below.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Glavin and Jones,
150 forearms in 75 adults Assessment of collateral forearm blood flow comparing Allen's test, (blushing of hand in <6 s following release of UA compression satisfactory), pulse oximetry and pulse monitor to 'gold standard' of Doppler assessmentCohort study (level 2b)Doppler flowPresent in 95% of ulnar arteriesAssumes Doppler as gold standard No information on blinding Study performed on anaesthetised patients, therefore may not be relevant to awake patients with coronary artery disease GA not standardised
Allen's test vs Doppler125/150 true positives; 3/150 false positives; 18/150 false negatives; 4/150 true negatives (sensitivity 0.87; specificity 0.57; +ve predictive value 0.98; -ve predictive value 0.18)
Pulse monitor vs Doppler138/150 true positives; 7/150 false positives; 5/150 false negatives; 0/150 false negatives (sensitivity 0.97; specificity 0; +ve predictive value 0).

Oximetry vs. Doppler – 143/150 true positives; 7/150 false positives; 0/150 false negatives; 0/150 true negatives (sensitivity 1.00; specificity 0; +ve predictive value 0.95; –ve predictive value 0)
Johnson et al,
452 radial arteries in 401 patients Modified Allen's test (MAT) using pulse oximetry to identify return of perfusion to control levelsCohort Study (level 2b)Return of SpO2 to control level within 12 s21/401 (5.2%) diagnosed as being 'RA dominant' on basis of return of control SpO2 exceeding 12 s cut-off pointNo explanation as to why 12 s cut-off for MAT No comparison to 'non-Allen's' technique
Jarvis et al,
93 hands in 47 patients scheduled for CABG surgery Comparison of modified Allen's test (MAT) and Doppler ultrasound assessment of collateral UA flow by analysing signal from princes pollicis artery (PPA) of thumb using receiver operating characteristics (ROC) of signal during release of UA compressionCohort Study (level 2b)PPA flow in response to RA compression

Identification of UA flow adequacy at different time points using ROC of Doppler system to identify PPA flow waveform
RA compression led to damping of Doppler signal in 33/93(35.5%) hands in 23/47(49%) of patients suggesting reduced ulnar collateral flowNo explanation as to why 12 s cut-off for MAT No comparison to 'non-Allen's' technique Small numbers Needs identification of Doppler as 'gold standard'
Reliability of Allen's test as indicator of UA flowAllen's test (6 s cut off) +ve in 23/93 (24.7%) hands (18 true positive; 5 false positive) and -ve in 70/93 (15 false negative and 55 true negative) - sensitivity 54.5%; specificity 91.7% and diagnostic accuracy 78.5%

ROC analysis revealed maximal diagnostic accuracy (79.6%) at 5 s Allen's test cut-off sensitivity of 75.8% and specificity of 81.7% with +ve Allen's test in 36/93(38.7%) hands (25 true positive; 11 false positive) and -ve in 57/93(61.3%) hands (49 true negative; 8 false negative)

Allen's test sensitivity of 100% at 3 s cut-off but this would increase +ve Allen's test rate to 77/93(83%) hands (33 true positive; 44 false positive) and -ve in 16/93(17%) hands (16 true negative; 0 false negative) with specificity 27% and diagnostic accuracy 52%
Sajja et al,
241 CABG patients requiring RA conduits Preoperative Allen's test and pulse oximetry with intraoperative presence of distal RA pulse during proximal RA occlusion used to assess efficacy of collateral flowCohort study (level 3b)Allen's test <6 s

Reappearance of pulse waveform on oximeter in <6 s
No case of hand/digital ischaemia when combinatoin of 3 tests usedNo information on numbers who did not achieve these 3 parameters
Intraoperative return of distal RA pulse with proximal RA occlusion within 6 sFalse -ve rate for Allen's test 0.4% based on lack of return of distal RA with proximal occlusion
Ruengsakulrach et al,
71 patients undergoin CABG surgery Non-dominant hand circulation assessed using modified Allen's test (MAT) and peak systolic flow (PSV) in superficial palmar branch of RA (SPA), ulnar artery at wrist (UA) and dorsal digital thumb artery (TA) with and without RA compression measured with Doppler ultrasonographyCohort study (level 2b)Allen's test >10 s defined as abnormal4/71(6%) had abnormal MAT (>10 s)59 men; 12 women Validation of MAT Lack of definition of 'abnormal' Doppler result
Changes in Doppler PSV in UA, SPA and TA with RA compression3/71(4%) UA's had no Doppler flow with RAC (2/3 had abnormal MAT)
Flow described as 'no flow', 'decreased flow', 'increased flow' or 'reversed flow'7/66(10.6%) SPA's had no Doppler flow with RAC (2/3 had abnormal MAT)

2/71 TA's (3%) had no Doppler flow with RAC (2/2 had abnormal MAT)

Patients with 'no flow' in either UA, SPA or TA had longer Allen's test recovery times

48/71(67.7%) RA's harvested for surgery

No ischaemic sequelae
Agrifoglio et al,
150 patients undergoing CABG with RA graft Assessment of non-dominant forearm using colour Doppler echo (ECD), Allen's test (AT), snuffbox test (SBT) and palmar arch test (PAT) ECD - assessment of vessel wall morphology and basal RA flow Measurement of UA PSV before and after RA compression Identification of retrograde flow in snuff-box wit RA compression Identification of backward flow in palmar arch during RA compressionCohort study (level 2b)Criteria for RA harvest;

basal RA PSV >0.2m/s

RA diameter >2 mm

increase in UA PSV with RAC

backward flow in snuff-box with RAC

backward flow in palmar arch with RAC
Clinical AT normal in all patients

8/150(5.3%) had preoperative ECD AT, SBT and PAT which contraindicated RA harvest and RA avoided in this group

Remaining 142/150 RA's harvested without evidence of postoperative forearm or hand ischaemia

97/150 patients followed up long-term for 24 months. 17/97(17.5%) complaining of hand paraesthesia only
No time cut-off given for clinical AT 36/150(24%) diabetic They raised the importance of Doppler assessment from a medico-legal perspective in patients with marginal or inadequate collateral flow
Starnes et al,
129 consecutive pre-CABG patients Modified Allen's test (MAT) with Doppler ultrasound used to assess blood flow in the superficial palmar arch (SPA) during RA compression and compared to 1st and 2nd digit blood pressures measured before and after RA compressionCohort study (level 2b)Result of MAT using decreased Doppler signal in SPA with RA compression

1st and 2nd digit pressures before and after RA compression

Decrease in digit pressure (DeltaP) >40 mm Hg with RA compression +ve

ROC curve analysis using plots of sensitivity against specificity to determine when MAT most accurate at predicting outcome of DeltaP
257 extremeties in 129 patients

14/115(12.2%) dominant and 16/112(14.3%) non-dominant arms had +ve MAT

7/14(50%) dominant and 8/16(50%) non-dominant limbs with +ve MAT had DeltaP <40 mmHg with RA compression(false positive)

MAT most accurate in non-dominant arm with DeltaP 40 mmHg - 50% sensitivity; 96.4% specificity; 90.6% accuracy

MAT most accurate in dominant arm with DeltaP 36-37 mmHg - 57.1% sensitivity; 85.2% specificity; 82.2% accuracy

RA harvested in 52/129(40%) patients. No symptoms/clinical signs of hand ischaemia in 50/52 patients followed up
107 male; 22 female Need to assume that DeltaP represents a 'gold standard' value of 40 mmHg an empirical value Identification of DeltaP value determinded by ROC of system and predicted by result of MAT - may be confounding factor
Abu-Omar et al,
287 consecutive patients undergoing total arterial revascularisation Assessment of blood flow using Allen's test and Duplex ultrasonographycohort study (level 2b)Normal left Allen's test, normal = capillary refill <5 s

Left Duplex ultrasonography measurements (calibre, flow, structure) in radial (RA), ulnar (UA) and brachial (BA) arteries if Allen's test abnormal

Right Duplex measurements if left abnormal

Occurrence of hand ischaemia
244/287 had normal left Allen's test and proceeded directly to RA harvest

43/287(15%) had abnormal Allen's test - 38/43 had normal RA, UA and BA calibre, flow and structure and were harvested - 5/43 had abnormal left RA duplex scans - 3/5 had normal RA's harvested; 2/5 did not have RA harvest

No significant differences between diameter of RA and UA's between groups

No ischaemic sequelae
Limitations of doppler monitoring in atherosclerosis
Meharwal and Trehan,
4172 radial artery grafts in 3977 cases undergoing CABG surgery Modified Allen's test (MAT) in ward Intraoperative pulse oximetry (PO) - time to recovery of trace/saturation measured at index finger during RA compression (10 s cut-off)Cohort study (level 2b)Functional outcome of hand post RA harvest/graft

Evidence of vascular or neurological complications in the arm
94 patients followed up

Early problems:0/3977 had acute ischaemic hand complications. 1113/3977(28%) had numbness/parathesia 477/3977(12%) had limitation of hand movement
No details of preoperative MAT cut-off time
Problems at discharge968/3977 (24.5%) had numbness/paraesthesia

80/3977 (2%) had limitation of hand ischaemia
Long term follow up194/3977 (5.2%) had weakness beyond 4 weeks. 15/3977 (0.4%) had weakness beyond 3 months

598/3977 (16%) had numbness beyond 4 weeks. 242/3977 (6.5%) had numbness beyond 3 months

314/3977 (8.4%) had paraesthesia beyond 4 weeks. 112/3977 (3%) had paraesthesia beyond 3 months

46/3977 (1.22%) had paraesthesia/numbness beyond 6 months
Kupinski et al,
146 preoperative CABG patients Imaging of forearm vessels with Duplex ultrasound and digital pulse volume recording (PVR) at rest and during RA compressionCohort study (level 2b)Vessel diameter and velocity volume flow data

Peak systolic velocities (PSV)

Anatomical imaging of vessels with B-mode ultrasound

Pulse volume at wrist

PVR measurement at rest and with RA compression
238 limbs in 146 patients

No statistical difference in Duplex data between right and left vessels

Velocity, diameter and volume flow greater proximally

RA larger than UA proximally and distally

Female PSV's > male PSV's for UA and RA both proximally and distally

Female vessel diameter significantly < male vessel diameters in both RA and UA proximally and distally

Male RA > female RA volume flow proximally and distally

UA volume flow not statistically different but male flow female flow distally

29/146 (20%) patients had abnormalities on Duplex scanning (12/29) bilateral)

10 limbs had UA or RA<1.8mm diameter

11 vessels with flow <5ml/min

17 RA's or UA's calcified

3 significantly stenosed

Resting PVR normal or mildly abnormal in 224/238 (94%) of limbs

PVR inadequate with RA compression in 41/238 (17%) limbs = incomplete palmar arch 10 patients had bilaterally abnormal digital PVR's with RA compression

PVR and Duplex abnormal in 17 limbs of 13 patients

RA's harvested in 83 patients 3/83 had abnormal preop PVR's and subsequently had some postop symptoms of ischaemia

60/83 followed up. 6/60 displayed a 'moderate to severe perfusion defect' but no clinical symptoms of digital ischaemia
115 male; 31 female Wide age range Confusion between patients and limbs Confusion between follow-up groups
Pola et al,
188 consecutive patients Patency of upper limb arteries/adequacy of UA in non-dominant arm assessed by static and dynamic Doppler evaluation (DDT)Cohort study (level 2b)Flow in each artery (PSV) and end-diastolic velocity (EDV)

Resistance index RI=(PSV-EDV)/PSV

Assessment of flow at ulnar artery (UA) at wrist, superficial palmar artery (SPA), main artery of thumb (I-ray), 2nd common palmar digital artery (II-ray), 3rd common palmar digital artery (III-ray)

Lack of UA flow increase associated with disappearance of SPA flow during RA compression = unsuitable for RA harvest
3/188 (1.6%) excluded from study on basis of decreased basal PSV

185/188 (99.4%) had 'normal' baseline flows

Divided on basis of response to RA compression

Group A - 174/185 (94.05%) considered adequate for RA harvest - significant increase in UA PSV (P<0.0001) and decrease at I-ray and II-ray arteries (P<0.001) - no change at III-ray - SPA retrograde flow - EDV significantly increase at UA (P<0.001); slightly decreased at I-ray artery (P<0.05); no change at II-ray and III-ray arteries - RI slightly down at UA (P<0.05); slight increase at I-ray artery; no change at II-ray and III-ray arteries

Group B - 11/185 (5.9%) had no UA PSV increase - no increase PSV at UA; significant decrease at II-ray and III-ray arteries P<0.001); slight decrease at III-ray artery (P<0.05); – flow disappearance at SPA– EDV – no change at UA; decrease at I-ray and II-ray arteries (P<0.05); more evident decrease at III-ray artery (P<0.001)– RI no change at UA and I-ray artery and slight increase at II-ray and III-ray arteries

100/185 (54%) RA's harvested (74 declined for various 'surgical reasons'

Early (10 day) and late (1 year) follow up confirmed significant increase UA PSV and EDV with decreased RI and SPA flow reversal when compared to control with patterns similar to that seen during preop DDT with RAC

No ischaemic sequelae
152 male; 36 female Lack of definition of 'normal' flows
Rodriquez et al,
346 arms in 187 CABG patients Doppler ultrasound to determine RA suitability for harvesting Plethysmography pressure measurements for 1st and 5th digit with and without RA compression in patients with normal Doppler assessmentsCohort study (level 2b)RA diameter <2 mm or diffuse calcification or congenital abnormalities not harvested

Perfusion defects during RA occlusion (>40% decrease in digital pressure, non-reversal of RA flow or <20% increase in UA velocity)= not harvested
94/346 (27.1%) RA's excluded from harvest on basis of Doppler measurements

44/346 (12.7%) excluded due to anatomical abnormalities (1.5% diameter <2mm; 8.7 diffusely calcified; 2.3% congenital abnormalities; 0.3% with RA occlusion)

50/346 (14.5%) excluded due to circulatory abnormalities (7.2% non-reversal of flow; 5.5% abnormal digital pressures; 1.7% inappropriate UA velocity)

266/346 assessed by plethysmography

19/266 demonstrated digital plethysmography BP fall >405 with RA compression (16/19 in 1st finger; 3/19 in 5th finger)

7/19 showed complete blunting of pressure waveform during RA compression

116/346 RA's harvested (110 non-dominant only; 3 bilateral)

No evidence hand ischaemia post-harvesting
80 limbs not assessed by plethysmography due to 'technical difficulties'
Winkler et al,
122/182 patients undergoing CABG surgery in 6 month period 137 extremities in 122 patients Blood flow in limbs assessed using 3-part radial artery mapping (RAM) consisting of Doppler measurement of upper extremity arterial system, segmental arterial and individual finger doppler pressures; Allen's testing of all 10 digits with photoplethysmography (PPG); and arterial Duplex scanning of UA, RA and SPA during UA/RA compression (UAC/RAC)Cohort study (level 2b)Obliteration of PPG waveform of all 5 digits with UAC=UA dominance

Loss of Doppler signal to SPA with UAC=UA dominance

Obliteration of PPG waveform of all 5 digits with RAC=RA dominance

Loss of Doppler signal to SPA with RAC=RA dominance

No complete loss of Doppler SPA signal with RAC or UAC=Mixed dominance SPA's - proceed to 'RA removal simulation' procedures

RA dominance absolute contraindication to RA harvest
137 extremities studied

8/137 (5.8%) RA dominant (0/8 harvested)

9/137 (6.6%) UA dominant (8/9 harvested). 64/137 (46.7%) mixed adequate (46/64 harvested). 56/137 (40.9%) mixed inadequate - mixed dominance but flow/circulation deemed inadequate with 'RA removal simulation' assessment (26/56 harvested)

RA not harvested if all 3 parts of RAM suggested inadequate flow

If only 2 tests suggested inadequate flow then combination of abnormal Duplex and segmental Doppler most likely to be associated with RA harvest

RA used most often in presence of abnormal Duplex; least often in presence of abnormal Allen's test

10/31 patients followed up reported 'minor' but resolving hand symptoms. No evidence of ischaemia
No reason given for selection of subgroup from patient group 93 male; 26 female
Barbeau et al,
1010 consecutive patients referred to cath lab modified Allen's test (MAT), Plethysmography (PL), Pulse oximetery (OX)Cohort study (level 2b)MAT cut-off 9 s

PL outcomes:

A no damping

B slight damping

C loss of trace with recovery

D no recovery of trace within 2 min

PO outcomes: positive or negative
MAT male mean recovery time 4.7 s (left and right); female 4.1 (left) and 4.0 (right)

MAT < 9 s in: 86.5% right arm, 87.8% left arm, 80.8% both arms, 93.6% either arm, 6.4% patients excluded from transradial catheter on basis of MAT (8.4% male; 2.2% female)

PL/OX type A/type B 90.8% right arm 89.5% left arm, 83.9% both arms 96.3% either arm 3.6% excluded (4.8% male; 0.9% female)

PL/OX type A/type B/type C 96% right arm 95% left arm 92.3% both arms 98.5% either arm 1.5% excluded (2% male; 0.3% female)

PL and OX more sensitive than AT in evaluating hand collaterals
32% female 19% diabetic 7% previous CABG Need further evaluation of PL group C to determine exact cut-off points although no sequelae reported in this study
Winterer et al,
21 patients presenting for CABG surgery Comparison of Doppler assessment of blood flow velocity in ulnar artery at wrist and main arteries of 1st-5th fingers with Gadolinium-enhanced magnetic resonance angiographyCohort study (level 3b)Persistance of finger digital pulse and increase in UA flow during RA compression = -ve Doppler

No UA flow acceleration or-flow decrease in one or more digital arteries with RA compression = +ve Doppler

Anatomy of RA, UA and completeness of palmar arch vessels assessed by MRI angiography

Patients with +ve Doppler rejected from RA harvest
21/21 imaged with Doppler - 18/21 (86%) -ve, 3/21 (14%) +ve

20/21 imaged with MRI angiography - 17 -ve Doppler; 3 +ve Doppler patients

All -ve Doppler patients had patent branches between UA and RA

All +ve Doppler patients had evidence of aberrant vessels or lack of collaterals between UA and RA
20 male; 1 female Small numbers New 'gold standard' How practical?


Definitions of hand/forearm ischaemia are subjective and variable but are low in patients with 'normal' Allen's tests. Agrifoglio reported no postoperative forearm or hand ischaemia, but 5.3% of their patients did not have radial artery harvest on the basis of Doppler assessment of hand flow during RA compression. However, papers by Sajja, Ruenaskulrach, Starnes and Abu-Omar report no ischaemic sequelae when using an Allen's test to guide suitability for RA harvest. The paper by Meharwal is perhaps the most reassuring in that it reported no acute ischaemic hand symptoms in a series of 4172 harvests in 3977 patients and whilst 5.2% complained of some hand weakness at 4 weeks, this had fallen to 0.4% beyond 3 months. Their incidence of numbness and paraesthesia although as high as 25% in the early postoperative period had fallen to 1.22% at 6 months. The decision to harvest in this series was based on preoperative Allen's test and intraoperative pulse oximetry studies. Of note cut-off points from 3 to 12 s are quoted as the time limit for return of palmar flush in these studies. Several studies have compared the Allen's test to a 'gold standard'. Glavin demonstrated that the test was associated with both false positives and false negatives using a 6-s cut-off when compared to Doppler and that pulse monitor and pulse oximetry were also unable to identify the absence of ulnar flow demonstrated with Doppler. Johnson used pulse oximetry to identify return of perfusion to control levels during an Allen's test in 452-forearms using a 12-s cut-off time. Jarvis compared a modified Allen's test (MAT) to Doppler ultrasound assessment of collateral ulnar flow in CABG patients. By examining the receiver operating characteristics of the Doppler signal from the princeps pollicis artery of the thumb during release of ulnar artery compression they concluded that whilst a 3-s cut-off would give the highest sensitivity, the false positive rate would also be high with a diagnostic accuracy rate of only 52%. Maximum diagnostic accuracy was achieved with a cut-off point of 5 s although the sensitivity would fall. In response to this paper, Sajja published a case series of 241 patients in which the Allen's test with pulse oximetry (6-s cut-off) was used in combination with intraoperative assessment of distal RA pulse during proximal RA occlusion to assess efficacy of collateral blood flow. They quoted a false negative rate of 0.4%. In a smaller study, Ruengsakulrach studied the non-dominant arm of 71 patients undergoing CABG surgery with a modified Allen's test (10 s cut-off). They identified abnormal Doppler flow in patients with abnormal Allen's tests and concluded that their study confirmed the validity of the Allen's test and that absence of flow in the dorsal digital thumb artery was an absolute contraindication to RA harvest. Some studies support the use of the RA even in the presence of an abnormal Allen's test. Starnes compared the modified Allen's test with Doppler ultrasound assessment of the superficial palmar arch blood flow during RA compression, to digital blood pressures before and after RA compression (P). They identified a false positive rate of 50% in their series, and defined 'maximum sensitivities' for P for both arms. RA harvest was performed in 52/129 patients with no ischaemic sequelae in 50/52 followed up. Abu-Omar used Duplex ultrasonography in the presence of an abnormal Allen's test in a subgroup of 43 out of 287 patients undergoing total arterial revascularisation who had an abnormal preoperative Allen's test (5 s cut-off). Duplex scanning was normal in 38 patients in this group and these together with a further 3 'Duplex abnormal' RA's were subsequently harvested without ischaemic sequelae. In some patients with equivocal tests a 'non-Allen's test' may be required. These are principally Doppler-based and there are a number of papers reporting various techniques to identify adequate collateral circulation. Kupinski imaged forearm vessels with Duplex ultrasound and digital pulse volume recording at rest and during RA compression. They identified 'some postoperative symptoms of ischaemia...' in a small subgroup that had abnormal preoperative digital pulse volume recordings and evidence of 'perfusion defects' without symptoms of ischaemia in 10% of a follow-up group. Finally, Winkler used a 3-part radial artery mapping technique together with a 'RA removal simulation procedure' to assess circulation preoperatively. They reported that surgeons were least keen to proceed to RA harvest if the Allen's test was abnormal and most if only the Duplex scan was positive. They also followed up a small subgroup and whilst 32% of patients reported minor but resolving hand symptoms, they felt that there was no evidence of ischaemia. Adding a second technique might significantly increase the number of arteries which could be harvested without adverse outcome. When Barbeau studied 1010 consecutive patients presenting to the Catheter Lab for transradial catheterisation they noted that 80% of the abnormal Allen's test group (52 out of 65 patients) had satisfactory plethysmography and pulse oximetry responses to RA compression. This decreased the potential 'RA rejection' rate from 6.4% (9-s Allen's test) to as little as 1.5%. Starnes reported a false positive rate as high as 50% in their series when they compared the Allen's test to a digital pressure change test. Finally in 2001, Winterer compared Doppler assessment of blood flow velocity in the ulnar artery at the wrist to Gadolinium-enhanced MRI angiography in a series of 21 patients presenting for surgery. Three patients had a positive Doppler with no UA acceleration or flow decrease in one or more digital arteries during RA compression. Twenty patients (including all with positive Doppler's) proceeded to MRI which confirmed patent vessels between UA and RA in all the Doppler negative, and aberrant vessels/lack of collaterals in the Doppler positive patients. Whilst this study involved small numbers, it may represent a new ultimate 'gold standard' assessment modality.

Editor Comment

Full report data anticipated by 06/10/05. To read the complete report see*&displaysectionid=Best+Evidence+Topic&journalcode=icvts

Clinical Bottom Line

A negative Allen's test safely selects patients for radial artery harvest, although the cut-off point is controversial. However, if the test is positive, then a 2nd test such as dynamic Doppler ultrasound or measurement of digital pressure changes with radial artery occlusion may allow safe harvest.


  1. Glavin RJ, Jones HM. Assessing collateral circulation in the hand – four methods compared. Anaesthesia 1989;44:594–595.
  2. Johnson WH 3rd, Cromartie RS 3rd, Arrants JE, Wuamett JD, Holt JB. Simplified method for candidate selection for radial artery harvesting. Ann Thorac Surg 1998;65:1167.
  3. Jarvis MA, Jarvis CL, Jones PR, Spyt TJ. Reliability of Allen's test in selection of patients for radial artery harvest. Ann Thorac Surg 2000;70:1362–1365.
  4. Sajja LR, Mannam G, Sompalli S. Is Allen's test not reliable in the selection of patients for radial artery harvest (letter)? Ann Thorac Surg 2002;74:296.
  5. Ruengsakulrach P, Brooks M, Hare DL, Gordon I, Buxton BF. Preoperative assessment of hand circulation by means of Doppler ultrasonography and the modified Allen test. J Thorac & Cardiovasc Surg 2001;121:526–531.
  6. Agrifoglio M, Dainese L, Pasotti S, Galanti A, Cannata A, Roberto M, Parolari A, Biglioli P. Preoperative assessment of the radial artery for coronary artery bypass grafting: Is the clinical Allen test adequate? Ann Thorac Surg 2005;79:570–572.
  7. Starnes SL, Wolk SW, Lampman RM, Shanley CJ, Prager RL, Kong BK, Fowler JJ, Page JM, Babcock SL, Lange LA, Erlandson EE, Whitehouse WM Jr. Noninvasive evaluation of hand circulation before radial artery harvest for coronary artery bypass grafting. J Thorac & Cardiovasc Surg 1999;117:261–266.
  8. Abu-Omar Y, Mussa S, Anastasiadis K, Steel S, Hands L, Taggart DP. Duplex ultrasonography predicts safety of radial artery harvest in the presence of an abnormal Allen test. Ann Thorac Surg 2000;77:116–119.
  9. Meharwal ZS, Trehan N. Functional status of the hand after radial artery harvesting: Results in 3977 cases. Ann Thorac Surg 2001;72:1557–1561.
  10. Kupinski AM, Huang J, Khan AM, Zorn TJ, Mathus LH, Mick JA, Hoskins MS, Shah DM. Noninvasive upper extremity arterial assessment in patients undergoing radial artery harvest. J Vasc Technol 1998;22:181-197.
  11. Pola P, Serricchio M, Flore R, Manasse E, Favuzzi A, Possati GF. Safe removal of the radial artery for myocardial revascularization: a Doppler study to prevent ischemic complications to the hand. J Thorac & Cardiovasc Surg 1996;112:737–744.
  12. Rodriguez E, Ormont ML, Lambert EH, Needleman L, Halpern EJ, Diehl JT, Edie RN, Mannion JD. The role of preoperative radial artery ultrasound and digital plethysmography prior to coronary artery bypass grafting Eur J Cardiothorac Surg 2001;19:135–139.
  13. Winkler J, Lohr J, Bukhari RH, Hearn A, Goller R, Parlato D, Schmeltzer M, Van Wagenen T, Smith JM. Evaluation of the radial artery for use in coronary artery bypass grafting. J Vasc Technol 1998;22:23–29.
  14. Barbeau GR, Arsenault GF, Dugas L, Simard S, Lariviere MM. Evaluation of the ulnopalmar arterial arches with pulse oximetry and plethysmography: comaprison with the Allen's test in 1010 patients. Am Heart J 2004;147:489–493.
  15. Winterer JT, Ennker J, Scheffler K, Rosendahl U, Schafer O, Wanner M, Laubenberger J, Langer M. Gadolinium-enhanced elliptically reordered three-dimensional MR angiography in the assessment of hand vascularization before radial artery harvest for coronary artery bypass grafting: First experience Investigat Radiol 2001;36:501–508.