Article Text
Abstract
Objective To describe the use of heel blood sampling and non-pharmacological analgesia in a large representative sample of neonatal intensive care units (NICUs) in eight European countries, and compare their self-reported practices with evidence-based recommendations.
Methods Information on use of heel blood sampling and associated procedures (oral sweet solutions, non-nutritive sucking, swaddling or positioning, topical anaesthetics and heel warming) were collected through a structured mail questionnaire. 284 NICUs (78% response rate) participated, but only 175 with ≥50 very low birth weight admissions per year were included in this analysis.
Results Use of heel blood sampling appeared widespread. Most units in the Netherlands, UK, Denmark, Sweden and France predominantly adopted mechanical devices, while manual lance was still in use in the other countries. The two Scandinavian countries and France were the most likely, and Belgium and Spain the least likely to employ recommended combinations of evidence-based pain management measures.
Conclusions Heel puncture is a common procedure in preterm neonates, but pain appears inadequately treated in many units and countries. Better compliance with published guidelines is needed for clinical and ethical reasons.
Statistics from Altmetric.com
Introduction
Heel puncture for neonatal blood sampling is a controversial procedure, because of the pain and the risk of tissue injury.1 In experienced hands, venipuncture is considered less painful and more efficient.1 Yet, in very preterm infants heelstick may sometimes be the only viable alternative. Use of mechanical rather than manual lance and non-pharmacological analgesic procedures are recommended, while topical anaesthetics and heel warming are not.1 2 Non-pharmacological analgesic measures are better applied in combination, as their effect appears to be synergic.2 In the face of well-established recommendations, data regarding compliance are scarce.
This paper aims at describing the self-reported practices of heel blood sampling in a large representative sample of European neonatal intensive care units (NICUs), and compare them with criteria set forth in published guidelines.1 2
What is already known on this topic
▶ Neonates have the capability of feeling pain, and neonatal pain can have adverse short- and long-term health consequences.
▶ Heel blood sampling is a frequently used painful procedure in neonatal intensive care units.
▶ Evidence-based guidelines for pain control in case of heel blood sampling are available.
What this study adds
▶ Mechanical lance and non-pharmacological analgesia are still underused in many European units.
▶ There are international differences, with Sweden, Denmark and France being the most likely, and Belgium and Spain the least likely to comply with guidelines.
Methods
The study is part of a larger project on NICU practices of developmental care and family visiting in eight countries.3 All NICUs in Italy, Denmark, Sweden and the Netherlands, together with third-level units in France and Belgium, units with at least five intensive care cots in the UK, and those with ≥50 yearly very low birth weight (VLBW) admissions in Spain were invited to participate.3 Overall 284 accepted (response rate 78%, ranging from 100% in Sweden, Denmark and the Netherlands to 70% in the UK). To ensure comparability, only the 175 NICUs with at least 50 VLBW admissions per year were considered for this analysis.
Data on NICU policies and practices were collected through a structured mail questionnaire. Questions regarding heelstick included the type of device (mechanical or manual) and related usage frequency, as well as the use of the following non-pharmacological procedures for pain control: oral sweet solutions (sucrose or glucose), non-nutritive sucking (ie, pacifier), swaddling or positioning, heel warming and topical anaesthetic (eutectic mixture of local anaesthetics (EMLA, eutectic mixture of local anaesthetics) cream). Frequency was measured on a four-point scale ranging from ‘rarely or never’ (defined as less than 10% of events) to ‘most of the times’ (ie, over 90% of events).
For this analysis, the predominant mode of blood sampling (ie, by mechanical or manual device) was identified by assigning to each NICU the mode reported with highest frequency. Equal frequencies were defined as ‘no clear policy’, while units reporting only occasional (<10% of cases) use of both manual and mechanical lance were considered not to use heel blood sampling. Regarding analgesic procedures, only the ‘most of the times’ use was considered to satisfy the guideline recommendations.
Data are reported as number and proportion of units per country. Among units using heel blood sampling, cumulative frequencies were computed to describe the predominant use of mechanical lance in combination with one or more of the recommended analgesic procedures, and without the non-recommended ones, that is heel warming and EMLA cream (see summary of published recommendations in the online Appendix). Data analysis was carried out through the STATA statistical package, V.10 (StataCorp 2007, Stata Statistical Software: Release 10, College Station, Texas, USA).
Results
Most participating units were third level except in Denmark, and most belonged to a teaching hospital except in Italy (table 1). The median number of annual VLBW admissions ranged from 60 in Sweden to 150 in the Netherlands. Parental visiting was allowed over the 24 h in most countries except Spain and Italy. A written protocol on pain management was available in most units in Sweden, the Netherlands and France, but in less than half of those in the UK.
The use of heelstick for blood sampling was widespread; only 13 units (six in France, three in Sweden and two in Belgium and Spain) appeared not to use it unless occasionally (<10% of the events). Most units in the Netherlands, UK, Denmark and France predominantly adopted mechanical devices, while manual lance was still used in Italy, Belgium and Spain. Routine (ie, over 90% of events) use of pacifier or sweet solutions during heelstick was common in every country except Spain and the UK. Swaddling or positioning was used especially in the Scandinavian countries and, to a lesser extent, the Netherlands and Belgium. With regards interventions considered ineffective, heel warming was still routinely applied in many units, especially in Denmark and Spain, while only six NICUs (five French and one Italian) used topical EMLA cream.
Figure 1 shows the cumulative proportions of units complying with one or more recommendations for pain control, and avoiding useless procedures (ie, heel warming and EMLA <10% of events) among those practising heel blood sampling. Predominant use of mechanical lance was routinely combined with administration of sweet oral solutions in the two Scandinavian countries and, to a lesser extent, France and Italy. Further association with pacifier, swaddling/positioning and no EMLA was practised by a substantial proportion of units only in Denmark (seven out of nine units using heelstick) and Sweden (three out of six). Simultaneous routine compliance with all the criteria explored could be found in two Swedish units and one French unit, but not in Denmark because of the relatively high use of the non-recommended heel warming.
Discussion
This study showed that heel blood sampling is still a very common procedure in European NICUs, with only 13 out of 175 surveyed units using it less than 10% of the time. Manual lance was still predominantly used in some countries, despite evidence that it is more traumatic and leads more often than mechanical device to repeated punctures to obtain a sufficient amount of blood.4 Non-pharmacological analgesia appeared to be underused, confirming findings from other studies.5 There were, however, important international differences, with Scandinavian countries and France being the most likely, and Belgium and Spain the least likely to comply with recommendations. UK and the Netherlands scored positively because of the adoption of mechanical devices, but their use of non-pharmacological analgesia was poor.
Although considered a minor procedure, heel puncture in neonates has been rated as very painful by physicians and nurses.6 Term newborns exposed to repeated heel lancing learn to anticipate pain, and subsequently exhibit more intense pain responses during venipuncture.7 In preterm infants, heel puncture and squeezing have been associated with several indicators of pain, including accelerated heart rate and decreased oxygen saturation, and may increase the likelihood of long-term adverse effects.8 Yet Carbajal et al,5 in a multicentric study carried out in French NICUs, found that heelstick was the most frequently employed procedure after nasal and tracheal aspiration; in about 40% of cases it was carried out without any analgesia. Clearly, willingness to prevent and treat neonatal pain has lagged behind the ability to recognise and measure it.
This study has limitations. The different levels of evidence associated with the surveyed procedures were not taken into account in the analysis, which weighted each procedure equally. Oral administration of sweet solutions, particularly sucrose, at a 12–24% concentration has been granted the highest evidence score (A).1 Its action appears to be both immediate, due to taste, and sustained because it stimulates the release of endorphins. The use of pacifier has a lower level of evidence (B), but sucking seems to act synergically with sweet taste in increasing the calming effect.1 Other effective procedures, such as breastfeeding, holding and skin-to-skin care were not explored in our questionnaire. The limitations inherent to self-reporting of practices should also be acknowledged.
However, this is the only study comparing practices and policies in a large, representative sample of NICUs in eight European countries. The high response rate is reassuring in terms of external validity, and the selections of units with at least 50 VLBW annual admissions increases comparability of results. Our findings have implications for neonatal intensive care practice, quality assurance and professional education programmes.
Recently, worries about the possible toxic effects of anaesthetics and sedatives on the developing brain raised by animal studies have been contrasted with the well-known harms of repetitive pain.9 Whether or not these worries will be confirmed, no conflict whatsoever exists for the use of non-pharmacological evidence-based interventions such as those explored in this paper.8 Appropriately administered, they are safe and effective, responding both to good clinical practice criteria and to ethical requirements in neonatal intensive care.
Acknowledgments
The authors are very grateful to all the 284 units that participated in the survey.
Footnotes
-
ESF Network C Amiel-Tison, France; JV Browne, USA; M Cuttini, Italy; G Greisen, Denmark; D Haumont, Belgium; PS Hűppi, Switzerland; H Lagercrantz, Sweden; K Stjernqvist, Sweden; Warren, UK. ESF Officers: C Moquin-Pattey, M Minkowsky and B Schaller.
-
Funding This study was funded by the European Science Foundation Network on Research on Early Developmental Care for Extremely Premature Babies in neonatal intensive care units (http://www.esf.org/activities/networks/all-current-networks.html) and by the Ospedale Pediatrico Bambino Gesù of Roma, Italy (Grant no. OPBG 200580X001247).
-
Competing interests None.
-
Ethics approval As this study was carried out on units' policies, and not on human subjects, requirements for informed consent and Ethics Committee approval do not apply.
-
Provenance and peer review Not commissioned; externally peer reviewed.