Abstract
Objective: The Allen test is a widely used screening method of hand circulation. Our aim was to study whether the Allen test alone gives sufficient information for harvesting the radial artery in coronary artery bypass grafting. Methods: One hundred and forty-five patients scheduled for coronary artery bypass grafting underwent the Allen test, upper arm Doppler ultrasonography and digital plethysmography. In ultrasonography both anatomical and circulatory measurements were performed. The Allen test was then compared with more objective tests and sensitivity; specificity and diagnostic accuracy were calculated. Results: Most of the patients had a negative Allen test, but 23% were positive (abnormal). Ultrasound scanning revealed anatomical anomalies in 10 patients and circulatory deficits in 17 patients. Thirteen patients had both circulatory and anatomical abnormalities. Sensitivity of the Allen test was 73.2% and specificity 97.1% based on our findings. There were no abnormalities in the recovery of the arms with harvested radial grafts. Conclusions: The Allen test is a good and valid screening test for the circulation of the hand. If the Allen test is negative it is safe to harvest the radial artery. If it is positive further examinations are needed to ensure safe harvesting of the radial artery.
1 Introduction
Radial artery as a conduit for coronary artery bypass grafting (CABG) was introduced by Carpentier et al. [1]. It was abandoned relatively soon after because of problems in early patency [2]. A new interest in the radial artery in coronary surgery came with the concept of total arterial revascularisation. The radial artery is the second major artery to the hand and in order to avoid ischaemic complications it is important to reliably assess the circulation before harvesting it. Traditionally this has been done with the Allen test which was described by Edgar V. Allen in 1929 [3]. The Allen test, albeit being cheap and simple, has been criticised for being too unreliable and subjective, even to a degree that it should be abandoned in favour of more objective tests [4].
The circulation of the hand is supported by two main arteries, ulnar artery and radial artery. Some individuals have even a third, or median, artery supporting the hand. These arteries form several collaterals in the hand which makes the harvest of the radial artery possible without compromising the circulation [5]. Sufficient circulation to the hand can be endangered in situations with lack of palmar collaterals, absent or insufficient ulnar artery or sclerotic changes within the supra-aortic, brachial or antebrachial vessels if the radial artery is sacrificed. For reasons mentioned above 7.5–27.1% of the radial arteries are not suitable for harvesting [6,7].
Our aim was to study whether the Allen test alone is a reliable screening method before harvest of the radial artery regarding sensitivity, specificity and diagnostic accuracy.
2 Material and methods
Between October 2000 and April 2005, 145 patients were enrolled in this study. Subjects were considered for CABG with a possible use of radial artery as a graft. Emergency cases and patients older than 60 years were excluded. Mean age was 52.0 years (SD 5.7). Male gender was predominant (130 vs 15) and most of the patients were right handed (136 vs 9). Patients underwent the Allen test, Doppler ultrasonography and digital plethysmography. All tests were performed by the same examining physician in the laboratory of clinical physiology. The non-dominant arm was studied and if there were contraindications for the harvest of radial artery, the dominant arm was also tested. Only non-dominant arms were included in this study. Patients were advised not to smoke or consume caffeine 3 h prior to the examination. Dipyridamole was prohibited on the day of examination.
The Allen test was performed by asking the subject to clench his fist for 1 min while both radial and ulnar arteries were compressed with the examiner’s fingers. The wrist was held at the level of the heart and extension was avoided. The ulnar artery was then released and the time that elapsed between the release and the recovery of normal pallor of the thumb and thenar area was recorded. Cut-off point was determined at 6 s. A positive test was reported when there was abnormality in the capillary filling of the fingers in 6 s.
Biplane ultrasonography was performed to assess calcification, sclerosis of the media and anomalies. The inner diameters of the radial and ulnar arteries were measured both distally and proximally. A transducer with emission frequency between 5 and 10 MHz was used, based on best visibility (Aloka, Pro Sound 5500). The harvest of the radial artery was contraindicated when diffuse intimal or medial calcification was present, when inner diameter of the radial artery was less than 2 mm or an anomaly, such as high bifurcation of the brachial artery or hypoplasia of the radial or ulnar artery, was seen. Circulatory measurements with Doppler ultrasonography included peak systolic velocity (PSV) and end diastolic velocity (EDV) of both radial and ulnar artery. A resistance index was calculated (PSV-EDV/PSV). These measurements were repeated on the ulnar artery while the radial artery was compressed. During radial compression the distal part of the radial artery was controlled for reverse flow. An absence of reverse flow in the radial artery and an increase of ulnar PSV less than 20% were interpreted as abnormal findings and were considered as contraindications for radial artery harvest [6,7].
Digital plethysmography (Finapress) was performed on every digit both at rest and during radial compression. A 40% decrease in systolic pressure in any digit was considered abnormal and was a contraindication for harvest [6].
Statistical analysis was performed by using SPSS 15.0 software.
3 Results
Seventy-seven percent of patients had a negative Allen test (111/144), while 23% were positive (i.e. abnormal). On one patient the Allen test was not performed, but other measurements were carried out. Proximal inner diameter of the radial artery was 3.06 ± 0.63 mm (range 1.2–5.3) and distal inner diameter was 2.6 ± 0.46 mm (0.9–3.06). Ulnar artery was proximally 3.25 ± 0.72 mm (1.3–5.8) and distally 2.39 ± 0.49 mm (1.0–3.5), respectively. In six patients (4.1%) the inner diameter of the radial artery was less than 2 mm. There were more ulnar arteries with small calibres as 27 patients (18.6%) had smaller vessel than 2 mm. The brachial bifurcation level, e.g. within the proximal third of the antebrachium, was normal in 96.6% of the patients (140/145).
Changes of flow velocity and resistance index in ulnar artery between rest and during radial compression are summarised in Table 1 . All changes in flow velocities were significant. In eight patients the increase of ulnar PSV was less than 20%. Reverse flow on distal radial artery during radial compression was present in 122 patients. Intimal calcification in radial artery was found in 12 patients and on ulnar artery in 13 patients. Sclerosis of the media was seen on seven patients’ radial artery and on six ulnar arteries.
Changes of flow velocities in ulnar artery between rest and radial compression (RAC)
Ultrasonography Doppler studies showed that 10 patients had anatomical anomalies, 17 patients had circulatory deficits and 13 patients had both anatomical and circulatory abnormal findings. Altogether 40 patients had ultrasonography scanning, suggesting that the radial artery should be left untouched.
Digital plethysmography showed significant changes in digital pressures between rest and radial compression. These are summarised in Table 2 . Thirty-three patients had a 40% decrease on digital pressure in thumb, 19 of these (13.1%) had zero pressure. Corresponding figures for the fifth finger are 25 and 15 (10.3%), respectively.
Changes in digital pressures between rest and radial compression (RAC)
Based on the findings of the Allen test, Doppler ultrasonography and plethysmography we calculated that the Allen test has a sensitivity of 73.2% and a specificity of 97.1%. Positive Allen tests were compared against abnormal findings in Doppler ultrasonography (absence of reverse flow, lack of sufficient increase in ulnar PSV, intimal or medial calcification, anomaly) and plethysmography (decrease of digital pressure more than 40%). Positive predictive value is 90.9% and negative predictive value 90.1%, and therefore a diagnostic accuracy of 90.3%. When the Allen test is compared to zero digital pressure of the thumb we received positive predictive value of 57.6% and negative predictive value of 98.2%. This comparison also gives the highest sensitivity, 89.2%, among the different single-parameter analysis. We had no ischaemic complications in the arms of the patients with harvested radial artery during hospital stay.
4 Discussion
The use of the radial artery as a conduit for coronary bypass grafting is a routine procedure today. Ischaemic complications associated with the harvest of radial artery are rare but they have been reported [8]. On a single patient such a complication is disastrous and should be avoided. There has been a debate whether the Allen test is a sufficient screening method to exclude patients with possibly compromised hand circulation.
There are several reasons for insufficient ulnar hand circulation. The collaterals between the radial artery and the ulnar artery are several. Altogether there are four arches, two in the carpal area and two in the palmar area. The palmar arches, superficial and deep, are especially important. They have been studied in detail in several cadaver studies [5,9]. Although there is considerable variation between the reported results, it is important to notice that the superficial and deep arch support each other in majority of cases, and thus, create a safety marginal.
The Allen test is a clinical test and, as such, a subjective measure, which contains several possible biases. It tells us nothing about the vascular anatomy of the hand, only the functional circulatory status which is interpreted by the examiner, whose interpretation is based on experience that may vary considerably. A false-negative result may be due to inadequate compression of the radial artery and false-positive result may be due to extension of the wrist. Furthermore there is no consensus regarding the cut-off point which may vary between 3 and 10 s. We chose to set the cut-off point to 6 s as a compromise between absolute sensitivity and specificity.
Jarvis et al. [4] studied 93 hands in 47 patients with Doppler ultrasonography and the Allen test to examine the overall reliability of the Allen test and determine the optimal cut-off point. They found that diagnostic accuracy was maximal at 5 s cut-off point, being 79.6%. Sensitivity and specificity were 75.8% and 81.7%, respectively. Sensitivity was at its peak level at 3 s (100%), but diagnostic accuracy was only 52% and specificity decreased to 27%. The authors concluded that the Allen test is unreliable and should be substituted with more objective tests.
Ruengsakulrach et al. [10] used a 10 s cut-off point and studied 71 patients with the Allen test, Doppler ultrasonography dynamic test and biplane for evaluation of the anatomy. Diagnostic accuracy was 97.2% when compared to ultrasonography regarding the flow in thumb artery. The absence of flow in thumb artery was considered as a contraindication for harvest of radial artery. Sensitivity and specificity were 100% and 97.1%, respectively. Other parameters, namely flow of the superficial palmar arch and flow of the ulnar artery, were not as reliable as the flow of thumb artery. The authors concluded that the Allen test is a valid screening method and that the use of ultrasound in conjunction with the Allen test permits safe harvest of the radial artery.
Earlier ultrasonography studies have shown that collateral circulation is insufficient in 10–23% of cases where radial artery is considered for harvesting [6]. Our own study showed that 27.7% (40/145) of patients had anatomical variations, pathologic changes or abnormal circulatory findings that would have made the harvest of radial artery questionable. Their radial arteries were left untouched.
There are multiple anatomical variations that unable the use of radial artery. A small diameter may carry problems in harvesting and in anastomosis technique. Also medial or intimal calcification can cause problems which may lead to a poorer patency or graft failure. Hypoplasia of ulnar artery is rare but does exist. In our material one patient had a hypoplastic and one had occluded ulnar artery. In 6.8% of cases there was anatomical contraindication for radial artery harvest.
Anatomy of the radial or ulnar artery per se is seldom a reason why radial artery is not suitable for harvesting. More often there is a circulatory deficit which can be measured as insufficient increase of ulnar flow, absence of reverse flow in radial stump or as a loss of blood pressure in digits. In our study 11.7% patients had a circulatory contraindication for radial artery harvest. Anatomical and circulatory changes often correlate which each other and 9% of our patients had both anatomical and circulatory contraindications.
Circulation to the hand is secured if there is adequate collateral circulation between radial and ulnar arteries and the ulnar artery is able to respond to the increased demand. If these two prerequisites are met there should be sufficient circulation to the hand to ensure safe harvest of the radial artery. Adequate collaterals are difficult to point out except with angiography, which is impractical. Ulnar response can be measured with Doppler scanning as we have done. The functional circulatory status after radial artery harvest can be measured as blood pressure in digits or as flow digital arteries. We chose to measure blood pressure as this is easily done with plethysmography. During radial compression pressure of the thumb decreased to zero in 13.1% of the patients, which suggests radial dominance. Earlier on, this has been reported in 26–28% of patients [11].
We compared the Allen test with Doppler ultrasonography and plethysmography and found that the negative Allen test correlates well with sufficient circulation in the hand after radial artery harvest. Specificity rate was 97.1% at 6 s cut-off point. Sensitivity was 73.1%; therefore it is possible that the radial artery is harvestable even if the Allen test is positive. If the Allen test is positive or there is suspicion of pathologic changes in arterial wall of radial artery, as it can be in diabetes or in arteriosclerotic disease we recommend plethysmography and Doppler ultrasound scanning preoperatively.
Contraindications as outlined in the section 2 are summarised in Table 3 . We regard them as valid contraindications; however, none of them is absolute. In our study we found that zero pressure in thumb during radial artery compression is a marker of significantly reduced circulation of the hand and as such an absolute contraindication for radial artery harvest.
Contraindications for radial artery harvest
If the Allen test is positive we recommend preoperative plethysmography and Doppler ultrasonography scanning. Plethysmography, especially of the thumb, should be performed primarily. Correspondingly we recommend these tests if we suspect pathologic changes such as those mentioned above in the arterial wall of radial artery to avoid unnecessary explorations of the radial artery.
We conclude that the Allen test is a good and valid screening test for the circulation of the hand. If the Allen test is negative it is safe to harvest the radial artery. If it is positive further examinations are needed to ensure safe harvesting of the radial artery.
