Skills development neonatal capillary sampling
Skills development: Neonatal Capillary Sampling
Capillary blood sampling of the neonate via a heel prick is a common procedure carried out by midwives, neonatal nurses and paediatricians. It involves making a small shallow incision in the heel and collecting a drop or drops of blood on a card or stick. It is indicated routinely in national screening of the neonate for phenylketonuria (PKU) and congenital hypothyroidism (the ‘Guthrie Test’). The heel prick is also used to obtain capillary blood for blood glucose estimation where hypoglycaemia is suspected or known, and for serum bilirubin estimation in certain cases of jaundice. Effective capillary sampling is important because early diagnosis of PKU and hypothyroidism has been shown to increase the benefit of subsequent treatment to affected babies (Corbett et al 1998), and where a neonate may be compromised, fast and accurate diagnosis is key to establishing and maintaining effective treatment. A study carried out in the North Thames region found that 1.5% of samples taken were inadequate or unsuitable, requiring resampling and a delay in diagnosis and commencement of treatment (Corbett et al 1998). It is therefore important that midwives are confident and competent in the skill. Student midwives are taught the skill relatively early in their training, and in my case this consisted of a lecture in the classroom, followed by observation in a community setting, and finally carrying out the skill myself under the supervision of a midwife. I have performed this skill many times and have confidence in my practice, considering myself a competent practitioner.
This essay will look at my personal experience of capillary sampling and compare it with the available evidence. Little has been written about the practical application of this skill (Warren 1998) and in order to ensure sufficient depth of discussion I will be looking in particular at choosing the correct site for heel prick, preparation of the heel prior to the procedure and ensuring sufficient blood flow during the procedure, and minimising the pain felt by the neonate and by extension minimising distress to the parents. I have chosen these issues in particular because I initially lacked confidence in choosing the site for heel prick and was anxious about causing harm to the infant, and because I have seen a variety of different practices from different midwives regarding preparation of the heel, ensuring sufficient blood flow and minimising pain, and I was keen to examine the evidence.
Opinion regarding the most appropriate site for capillary sampling has long been based on Blumenfeld et al’s 1979 study which looked at the heels of 40 children post mortem weighing between 561 and 13150 grams, 35 of which were neonates. The distance between the skin surface and the perichondrium of the calcaneus was measured, the rationale being that serious complications of heel prick in the neonate such as necrotising chondritis or calcaneal osteomyelitis can occur after penetration of the calcaneus by the lancet. In the smallest infant, weighing 561 grams, skin-perichondrium distance was 2.38mm. The anatomy of the foot is such that in a neonate the calcaneus rarely reaches the lateral and medial edges of the plantar beyond imaginary lines drawn from the middle of the big toe and from between the fourth and fifth toes. Blumenfeld et al used this knowledge along with their findings to recommend a maximum lancet depth of 2.4 mm used on the lateral or medial edges of the plantar, thereby ensuring that even on the smallest infant the risk of damage to the calcaneus would be minimised. Interestingly, their results show that where weight is between 2001 and 4000 grams, the skin-perichondrium distance ranged from 5.81mm to 6.8mm on the left and right plantar. The majority of capillary samples performed by midwives, and particularly student midwives, are on term babies. Low and very low birthweight babies needing capillary sampling are likely to be cared for in a Special Care Baby Unit or Neonatal Intensive Care Unit and cared for by neonatal nurses. Blumenfeld et al’s recommendations were made at a time when manual lancets or scalpels were used for this procedure, increasing the risk of human error leading to complications. It is unlikely therefore that in the course of my own practice I would encounter a situation where damage to the calcaneus was likely.
Blumenfeld et al remains the primary source for guidance on capillary sampling (see for example Baird & Witt 1996, Baston 2002, Johnston & Taylor 2000, Meehan 1998, Warren 1998) and their evidence is the basis for the training I received, both in the classroom and from midwives in practice. Jain & Rutter (1999) have attempted to update Blumenfeld’s work by studying a group of 80 live neonates. Using ultrasound, they measured the distance between skin and perichondrium at the point at which the perichondrium was closest to the skin and found that in all but 2 of the infants, the shortest skin-perichondrium distance was 4mm. The two infants who showed a 3mm skin-perichondrium distance were 610g and 1560g in weight. No infants showed a skin-perichondrium distance smaller than 3mm at any point on the plantar surface. Jain & Rutter argue that confining heel pricks to a small area as defined by Blumenfeld leads to hypersensitivity, making each subsequent heel prick more painful, and that their research has demonstrated that using an automated lancet which penetrates to a depth of 2.4mm would be safe on any part of the plantar surface, avoiding only the posterior aspect of the heel. Where infants are only subject to one heel prick, for example the Guthrie Test, this is of small importance. However, where infants are under investigation, for example for bilirubin or glucose estimation, they may have multiple pricks.
My current practice is to puncture the heel on the medial or lateral sides as prescribed by Blumenfeld et al and by local protocol. Although Jain & Rutter’s arguments are interesting, further research is needed in order to support their findings. In my practice, where heel pricks are usually only undertaken once per infant, and where local policies support the status quo, it would be inappropriate for me to consider changing the site of the heel pricks I undertake, particularly as I am confident that my current practice is causing no harm.
Preparation of the heel before heel prick varies between practitioners. It is common practice for midwives to suggest that parents apply an extra layer of clothing to the baby’s foot before a heel prick is undertaken (Johnson & Taylor 2000, Baston 2002) in order to achieve good perfusion of the area. Other methods include warming the heel by holding it in the hand (Gerds 1997) placing the heel in warm water (Baston 2002) or rubbing the heel (Warren 1998). Commercial heel warmers are available (Baird & Will 1996) but are perhaps an unnecessary expense. My practice has been to gently rub the heel with my hand prior to heel prick. Barker et al (1996) conducted a randomised controlled trial of 81 procedures in 57 infants and found that warming the heel had no effect on the outcome of the sampling in terms of time taken to collect a sufficient sample, the number of repeat samples needed or the infant’s behavioural response. They concluded that time spent warming the heel was time wasted. In light of this, warming the heel should be viewed as unnecessarily prolonging the process and therefore the distress to the infant and parents. There is no local policy to which I have to adhere, and I will therefore in future consider changing my practice in order to avoid unnecessary distress to the infant.
Heel cleansing prior to heel prick is advocated by some. Blumenfeld et al (1979) advocate thorough cleansing of the site, citing as evidence the case of an infant who had a damaged calcaneus from heel prick, with an infected puncture track, dermal cellulites with abscess formation and necrotising calcaneal chondritis. They concede that the chondritis would probably not have occurred had the calcaneus not been damaged, and that the risk of any infection at a puncture site is increased where the puncture is made through an older infected puncture site. Meehan (1998) recommends cleaning with an alcoholic wipe and drying with sterile gauze. Gerds (1997) agrees and suggests allowing the heel to dry naturally after wiping, or using sterile gauze if desired. Johnston & Taylor (2000) specify that if a swab is used, the heel should be cleaned for 30 seconds and allowed to dry before heel prick. They stipulate however that alcohol impregnated wipes should not be used for Guthrie Tests or for glucose readings as the alcohol can affect the accuracy of results. As the heel pricks I undertake in practice are almost exclusively for Guthrie testing and blood glucose estimation I have not used alcoholic wipes in practice, and I have found no evidence to suggest this is incorrect. I will not therefore be changing my practice.
Squeezing the heel in order to encourage blood flow is sometimes necessary but is recognised as causing or increasing pain in the neonate (Jain et al 2001). It can also cause haemolysis and soft tissue damage (Gerds 1997). The heel should not therefore be routinely squeezed, and if it is, then gentle squeezing and releasing should be used (Johnston & Taylor 2000). Ensuring the infant is in a position where the heel is lower than the body may facilitate blood flow (Warren 1998, Gerds 1997). In practice, I have been squeezing the heel where the blood is not flowing well. In future, in order to reduce the risk of soft tissue damage and haemolysis, and to reduce distress to the infant, I will use the position of the infant to help blood flow before resorting to gently squeezing where necessary. The first drop of blood should be wiped away to avoid contamination (Johnson & Taylor 2000, Gerds 1997, Warren 1998). This is something which I have not seen in practice, and although I was aware of the theory, I have also been using the first drop of blood. This has been partly because it is often not obvious that there will be a sufficient sample, and therefore discarding the first drop would be a ‘waste’. In future, I am confident that better positioning of the baby will facilitate better blood flow and remove this obstacle. It is particularly important to wipe away the first drop of blood where an alcohol impregnated wipe has been used. Although I do not use wipes, I will in future ensure I wipe away the first drop of blood to reduce the risk of a contaminated sample leading to an inaccurate test result or the need for resampling.
There is no doubt that heel prick is painful for the neonate (Oksanen & Salantera 2002) and that health professionals often overlook this issue due to the infant’s limited ability to communicate pain and particularly its intensity (Fletcher 1993). Venepuncture has been found to cause less pain to the infant and be more efficient than heel prick (Larsson et al 1998, Shah & Ohlsson 2001), which has led to a question mark being placed over the ethics of continuing to use heel prick as the standard format for obtaining blood from neonates (Kvist et al 2002). In reality, the high level of skill required to carry out venepuncture on a neonate renders this option practically impossible. The equipment used in the trust in which I am training is an automated device with a spring-loaded blade. An automated lancet has been shown to cause less bruising of the heel and foot and facilitate faster healing of the wound than a manual lancet and is therefore the tool of choice for neonatal capillary sampling (Vertanen et al 2001). The policy employed by the trust in which I work is therefore in line with the evidence as far as is practical.
My practice has been to undertake the procedure as quickly as possible, ensuring a sufficient sample is obtained in order to reduce the risk of having to repeat the procedure, and then passing the baby to a parent, preferably its mother, for comforting. Various studies have been undertaken to investigate different methods of reducing pain during heel prick. Gray et al (2000) found that 10-15 minutes skin to skin contact between a mother and baby reduces the infant’s pain responses during heel prick. This may often not be practical as it necessitates undressing the infant and requesting that the mother do the same, which may not be appropriate. That it is time consuming must also be taken into consideration. Carbajal et al (2003) studied 44 term neonates who were breastfed throughout venepuncture and found that 35 scored 3 or less out of 10 on the Douleur Aiguë Nouveau-né pain scale, indicating that they felt minimal or no pain. They found that simply being held by their mothers did not have any significant effect on the pain they felt. Although heel prick has been seen to be more painful for the neonate than venepuncture (Larsson et al 1998, Shah & Ohlsson 2001), it can be assumed that breastfeeding throughout the procedure would have some effect on the level of pain felt by the infant.
The use of sucrose as an analgesia has been investigated. Bellini et al (2002) found that sucking 1ml of oral glucose 33% was an effective analgesic, which was made more effective when combined with sensorial saturation (massage, voice, eye contact and perfume smelling). Bilgen et al (2001) studied 139 infants during heel prick and found that 30% sucrose solution was more effective than breastfeeding as an analgesic when pain was measured in terms of crying time and heart rate. This evidence is counteracted by Oksanen & Salantera (2002) who found that a 30% glucose solution made no difference to neonates experience of pain during 129 heel pricks. Jain et al (2001) investigated the use of topical amethocaine gel to reduce the pain of heel prick, but found it to be ineffective. Similar results were found for topical anaesthetic (Barker & Rutter 1995). Once again it is important to consider practicality. While glucose has been shown to be an effective analgesic by some studies, it may not be something parents wish their child to have. The person carrying out the procedure would also need to obtain and maintain supplies of the solution, with the added expense that entails. Furthermore, this is not supported by local policy in the trust in which I work. I would not therefore consider suggesting this method to parents. Encouraging mothers who have chosen to breastfeed to do so during the heel prick procedure may help reduce the pain felt by the infant. This would also have the added advantage of allowing the heel to dangle below the rest of the body, facilitating better blood flow and reducing the need for squeezing of the heel as discussed above.
Current practice regarding the site of capillary sampling is based on Blumenfeld et al’s 1979 research. This has since been challenged in certain respects by Jain & Rutter but further research is needed before changes in practice can be considered. I am therefore happy that in respect of choosing the appropriate site for heel prick I am practicing in line with evidence. I have been warming the heel by rubbing it prior to capillary sampling, but what little research there has been into this practice has shown that it is ineffective. In order to reduce distress to the infant, unnecessary handling and prolonging of the process I will therefore consider not attempting to warm the heel prior to heel prick. Cleansing of the heel using an alcohol-impregnated swab is advocated by some, but as it can affect the results of the Guthrie Test and glucose estimation I am happy that my practice of not using wipes is sound. I am aware that in future I should wipe away the first drop of blood which is formed in order to avoid contamination of the sample. Squeezing of the heel has been shown to be necessary sometimes, but positioning of the baby with its heel lower than the rest of its body may be just as effective and is something I intend to try in the future. Heel prick is a procedure which causes pain to infants, and various ways of reducing this have been studies. Although glucose solution has been shown to be effective this is not a practical option for my practice. Breastfeeding throughout the procedure has been shown to be very effective at reducing pain in neonates undergoing painful procedures, and where appropriate I will in future be exploring this option with breastfeeding women.
Capillary sampling of the neonate is a procedure routinely carried out by midwives and student midwives, and as such is at risk of becoming guided by habit rather than evidence. I have learned that while my basic practice is sound, there are issues I should consider in future in order to ensure I am providing the best care I can.
Baird PB, Witt CL (1996) Neonatal glucose screening Neonatal Network 15 (7) 63-66
Barker DP and Rutter N (1995) Lignocaine ointment and local anaesthesia in preterm infants Archives of Disease in Childhood Fetal and Neonatal Edition 72 F203-F204
Barker DP, Willetts B, Cappendijk VC (1996) Capillary Blood Sampling: should the heel be warmed? Archives of Disease in Childhood Fetal and Neonatal Edition 74 (2) F139-F140
Baston H (2002) The Guthrie Test The Practicing Midwife 5 (3) 32-35
Bellini CV, Bagnoli F, Perrone S, Nenci A Cordelli DM, Fusi M, Ceccarelli S, Buonocore G (2002) Effect of multisensory stimulation on analgesia in term neonates: a randomised controlled trial Pediatric Research 51 (4) 460-463
Bilgen H, Ozek E, Cebeci D, Ors R (2001) Comparison of sucrose, expressed breast mik, and breast-feeding on the neonatal response to heel prick Journal of Pain 2 (5) 301-305
Blumenfeld TA, Turi GK, Blanc WA (1979) Recommended site and depth of newborn heel skin punctures based on anatomical measurements and histopathology The Lancet 181 (10) 230-233
Carbajal R, Veerapen S, Couderc S, Jugie M, Ville Y (2003) Analgesic effect of breast feeding in term neonates: randomised controlled trial BMJ http://bmj.com/cgi/reprint/326/7379/13.pdf accessed 09/09/2003
Corbett V, Bedford H, Grant J, Addison M, Streetly A (1998) Heelprick testing of neonates: the role of the midwife British Journal of Midwifery 6 (6) 390-394
Fletcher V (1993) Pain and the Neonate in Alexander J, Levy V, Roch S Midwifery in Practice: A research-based approach Basingstoke, Macmillan
Gerds RJ (1997) Nursing care and sensory enrichment of the normal newborn in Reeder J, Martin LL, Koniak-Griffin D Maternity Nursing: family newborn and women’s healthcare Philadelphia, Lippinant-Raven
Gray L, Watt L, Blass EM (2000) Skin-to-Skin contact is analgesic in healthy newborns Pediatrics 105 (1) 460-463
Jain A, Rutter N (1999) Ultrasound study of heel to calcaneum depth in neonates Archives of Disease in Childhood 80 F243-F245
Jain A, Rutter N, Ratnayaka M (2001) Topical amethocaine gel for pain relief of heel prick blood sampling: a randomised double blind controlled trial Archives of Disease in Childhood Fetal and Neonatal Edition 84 F56-F59
Johnson R, Taylor W (2000) Skills for Midwifery Practice London Churchill Livingstone
Kvist LJ, Jonsson K, Tornestrand BM, Mansson ME (2002) Can venepuncture reduce the pain of neonatal PKU sampling? A randomised study Vard I Norden Nursing Science and Research in the Nordic Countries 22 (4) 27-30
Larsson BA, Tannfeldt G, Lagercrantz H (1998) Venepuncture is more effective and less painful than heel lancing for blood tests in neonates Pediatrics 101 (5) 882-886
Meehan RM (1998) Heelsticks in neonates for capillary blood sampling Neonatal Network 17 (1) 17-24
Oksanen I and Salantera S (2002) Infant pain relief with glucose solution during heel lancing Hoitiede 14 (5): 233-242
Shah V, Ohlsson A (2003) Venepuncture versus heel lance for blood sampling in term neonates (Cochrane Review) in The Cochrane Library 3 Oxford, Update Software
Vertanen H, Fellman V, Brommels M (2001) An automatic incision device for obtaining blood samples from the heels of preterm infants causes less damage than a conventional manual lancet Archives of Disease in Childhood Fetal and Neonatal Edition 84 (1) F53-F55
Warren C (1998) Blood without tears: taking capillary blood from newborns The Practicing Midwife 1 (10) 10-11
If you are the original writer of this essay and no longer wish to have the essay published on our website then please click on the link below to request removal:
comments powered by Disqus