Lessons of the month 1: Learning from Harvey; improving blood-taking by pointing the needle in the right direction

The taking of blood for diagnostic purposes is a frequent cause of difficulty for physicians. In patients with intact visible or palpable large veins, such as those often seen in the antecubital fossa, a needle or cannula entering from any direction will usually be rewarded with any quantity of blood. In smaller veins in less convenient locations, such as in the hand, the direction of the needle becomes much more important. Failure to take blood is very commonly because of failure to appreciate the direction of flow of venous blood up the arm, and the ubiquitous presence of valves in the veins, both aspects of the circulation clearly described by William Harvey nearly 4 centuries ago. This paper encourages more frequent success with phlebotomy by remembering Harvey’s work and pointing the needle in the right direction; this is not always towards the heart.


Venous infusions and drainage
It follows from these observations that, if the anatomy follows the pattern depicted in Fig 1 a, delivery of substances into the blood via a venous puncture is best served by a cannula facing up the arm in the direction of flow and that drainage of blood via a cannula must permit flow from the periphery. For delivery of laudanum into the hind-limb vein of a dog in 1656, Boyle provided a detailed account of the surgical access achieved by Christopher Wren in Oxford using a 'slender pipe of a syringe' directed 'towards the heart'. 3,4 Conversely, for drainage of blood, if a tube occupies the whole lumen of the vein, it must clearly be pointing towards the periphery from which the blood is flowing. If it is pointing in the opposite direction, blood will need to be able to creep past the cannula before doing an about-turn to flow back out. Early advocates of blood transfusion from animals and humans to patients were clear about the appropriate orientation of drainage and infusion cannulae. Thus, Elsholtz in 1667 (Fig 1 b) showed his curved silver phlebotomy drainage cannula facing the hand and the infusion line facing the heart. 5 A very detailed explanation is given emphasising that 'the curved ends of the tubes, when they are inserted, must be properly directed to pour out and receive blood'. 6 Aveling in 1873, in the same manner, showed infusion of blood into a woman suffering post-partum haemorrhage from her carriage driver, the apparently slightly uphill passage of blood being driven by a small manual pump connecting the silver drainage and infusion pipes (Fig 1 c). 7,8 Strategies to prevent clotting The clotting of blood in this pre-anticoagulant era was a constant problem. An attempt to overcome this used speed. By Lessons of the month 1919, Lindeman reported several hundred transfusions using a cycle of sterilized 20 mL syringes with a team of operators to take blood from a donor and inject it into the recipient over very short periods of time, each syringe holding blood for only 6-10 seconds. [9][10][11] Large (14G) cannulae were inserted over introducers; interestingly, Lindeman nowhere states the direction of insertion of his special cannula in either donor or recipient, but the speed with which large volumes of blood (often over 1 L) were transferred suggests unquestioned use of the then traditional orientations. The introduction of citrate anticoagulation permitted a more leisurely approach to harvesting blood, as shown by Lewisohn in Fig 1 d, summarising practice over the period 1914-1924. 12 Drainage was again in the direction of venous flow.

Blood tests and standardisation
What is the explanation for the transition to the current total adherence to the practice of phlebotomy conducted with needles pointing centrally? Three factors seem to be relevant. First, the arrival of routine blood tests; second, the technological development of relatively fine needles that could be used to draw modest quantities of blood for laboratory tests from a vein without necessarily occluding the flow from the periphery; and third, the dominance of the use of the median antecubital vein that is usefully supplied by neighbouring superficial veins as well as, in part, by penetrating veins from deep in the arm.
The first widely used screening blood test dates back to Wassermann et al in 1906. This screening for syphilis progressed between the World Wars in the USA, in part because of compulsory pre-marital tests on both husbands-and wives-to-be and accelerated with the arrival of penicillin. 13 Other tests early in the century included agglutinin reactions prior to transfusion and blood glucose assays on large populations. 14 An image of a sample being taken for a Wassermann test in Florida in 1941 (Fig 2 a) serves to illustrate the now almost universal practice of performing phlebotomy with a syringe and needle pointing headwards in an antecubital vein.

Antecubital and forearm superfi cial veins
The median vein of the antecubital fossa is a generous gift of nature to healthcare staff. In many patients it offers a large target generously supplied with blood from more than one Post-partum haemorrhage is shown being treated by forearm drainage from a donor silver tube facing the hand, via an 'indiarubber tube to form an anastomosis between the emittent and recipient veins and a little bulb in the middle to act as an auxiliary heart'. 25 The operator needed intermittently manually to occlude donor and infusion tubes as the bulb was respectively squeezed and released. Reproduced with permission from Aveling J. Immediate transfusion in England. Obstetrics Journal 1873;1:303. Credit: Wellcome Collection, CC BY. d) Harvesting of blood from the forearm of a donor into a beaker containing citrate to prevent coagulation, as depicted by Lewisohn summarising practice in the period 1914-1924. 12 Here a metal cannula is shown positioned in an antecubital vein pointing towards the hand; blood fl ow is vigorous. Reproduced with permission from Lewisohn R. The citrate method of blood transfusion after ten years:  15 It seems unfortunate that 90% of the valves are positioned such a manner that a headwards-pointing phlebotomy needle occluding the vein close to the site of entry would receive no replenishment of flow from a perforating vein from deep in the arm into the cannula, and would apparently empty at most a mere 5 cm, or so, of the superficial vein of blood.
The median basilic vein was the site recommended for blood donation by Keynes in his influential detailed text Blood transfusion published in 1922; this is the segment of the median vein closest to the basilic vein. 17 Partly on grounds of the suitability of veins, Keynes recommended donors being 'young men between the ages of 18 and 25', adding that 'the use of women as blood donors is to be avoided' . Fig 2 b shows the recommended route for donation using citrate as an anticoagulant, it being qualified that 'If the vein be of a good size, it makes no difference whether the needle be inserted towards the heart or away from it.' 17 An earlier depiction of blood being taken in the same direction via a suction bottle containing glass beads for the method of defibrination can be found dated to 1913 in the USA. 18 This route and direction have now become the well-established default for phlebotomy. 19

Drawing blood from the direction of flow
Benefi ts of a phlebotomy device facing the periphery A needle or cannula pointing towards the hand in a patent vein will receive a flow of blood unrestricted by closure of a valve. 20 Fig 4 a shows drainage in this manner using a 'butterfly' needle in a small vein on the anterior aspect of the wrist ; Fig 4 b shows the use of a cannula to draw blood from a vein on the dorsum of the hand. Insertion of a needle or cannula facing the periphery carries, of course, all the challenges of identifying and entering the vein associated with all phlebotomy, but once in the vein the flow cannot be impeded by the closure of a valve. We recommend this approach to taking blood in patients in whom the antecubital fossa fails to provide a favourable site. A cannula sited in this way can be useful for repeated sampling, for example during exercise testing requiring serial lactate assay. 21 Potential complications of facing the periphery Harvey noted: I have often noticed in dissecting veins, that no matter how much care I take, it is impossible to pass a probe from the main venous trunks very far into the smaller branches on account of the valvular obstructions. On the contrary it is very easy to push it in the opposite direction, from the branches towards the larger trunks. 2 Correspondingly, a cannula inserted pointing towards the periphery is at risk of being advanced into the closed cusp of a valve. This will obstruct drainage of blood and has the potential  Lessons of the month to puncture the vein or damage the valve itself. It can therefore prove useful to identify the location of any nearby valve by using Harvey's method. This method is to empty a segment of vein by stroking along it with a finger, moving away from the heart, and noting the point beyond which a valve prevents the retrograde filling of the vein from the proximal end of the segment, the distal segment still being occluded by the finger. 2,20 A site of insertion can then be chosen well away from the valve. There is the possibility that a cannula placed facing the periphery may be used for drug or fluid administration. Although this appears unlikely to result in harm it may best be avoided by removing the phlebotomy cannula immediately after use or labelling it as for drawing of blood only.

Conclusion
The restrictions proposed by Keynes on blood donors back in 1922 imply that blood taking can be difficult and the vast majority of clinicians will recognise such difficulties from their own practice. The internet is awash with accounts from patients of difficult and failed phlebotomies, and suggestions from healthcare staff on how to improve technique. A large study of the success rate of outpatient phlebotomy mainly in the USA suggested that fewer than 1% of phlebotomies failed because of inability to obtain a sample, yet most laboratories took account of the need to have more than one attempt at drawing blood, many up to four times. 22 Inpatient blood draws by phlebotomists, also mainly in the USA, met with a success rate of 93%, but failure was not always because of technical difficulty. 23 A recent study of 1,200 hospital patients from India found that only 27% of phlebotomies were achieved with just one needle prick, most requiring two or three attempts. 24 Medical staff, particularly anaesthetists and physicians, are familiar with being asked to help with blood taking from patients in whom other staff have failed. While there are clearly multiple different factors contributing to the success or failure of phlebotomy, we believe that there needs to be a focus on the 'elephant in the room' missing from recent studies on, and courses in, phlebotomy: the significance of needle or cannula direction. By highlighting what one of the Royal College of Physicians' best-known alumna demonstrated in the 17th century, we hope a few more patients and clinicians may be spared the discomfort and frustration of failed phlebotomy.