Journal of Neonatal Nursing (2014) 20, 31e36

www.elsevier.com/jneo

Procedural pain in neonates: Do nursesfollow national guidelines? A survey toSwedish neonatal units

Elna Scherman a, Linda Johansson a, Maria Gradin b,Johannes van den Berg c,*, Mats Eriksson b,d,e

aDept. of Nursing, Umea University, SwedenbDept. of Paediatrics, Orebro University Hospital, SwedencDept. of Clinical Science e Paediatrics, Umea University, Umea, SwedendCentre for Health Care Sciences, Orebro University Hospital, Swedene School of Health and Medical Sciences, Orebro University, Sweden

Available online 13 May 2013

KEYWORDSDocumentation;National guidelines;Neonatology;Newborn;Procedural pain

* Corresponding author. Barn4/NICUfax: þ46 907852898.

E-mail addresses: Johannes.berg@

1355-1841/$ - see front matter ª 201http://dx.doi.org/10.1016/j.jnn.2013

Abstract Purpose: To investigate if nurses at neonatal units in Sweden haveadopted national guidelines when neonates are exposed to intravenous catheter,capillary heel prick, venepuncture and injections, to identify the frequency ofdocumentation of pharmacological and behavioural treatments and to comparethe answers from the nurses with results from an earlier national survey completedby the chief neonatologists at the same units.Design and sample: Four nurses at a total of 44 neonatal units in Sweden,received questionnaires. A total number of 116 surveys were analysed (responserate 66%).Main outcome and results: All units had written guidelines for prevention andtreatment of pain. Behavioural treatments were used in every painful procedurein the study, but only1/5 used EMLA� often or always. There was a higher ten-dency to document the use of drugs than behavioural treatments. The chiefneonatologist reported higher use of glucose than did nurses.Conclusions: Swedish national guidelines are not used consistently in someneonatal units. There is a considerably larger cohort of nurses who use behaviour-al treatments, rather than using drugs when painful procedures are performed. It

, Department of Pediatrics, Umea University, SE-90185 Umea, Sweden. Tel.: þ46 907850300;

pediatri.umu.se (J. van den Berg), mats.h.eriksson@oru.se (M. Eriksson).

3 Neonatal Nurses Association. Published by Elsevier Ltd. All rights reserved..04.010

32 E. Scherman et al.

was also evident that it was more common to document the use of drugs than be-havioural treatments.ª 2013 Neonatal Nurses Association. Published by Elsevier Ltd. All rights reserved.

Introduction

Invasive procedures are the most common reasonsfor pain in hospitalized infants. Carbajal et al.(2008) demonstrated that newborns on averagewere subjected to 12 painful procedures per dayduring the first two weeks of their NeonatalIntensive Care Unit (NICU) stay, mainly withoutanalgesia. The pain caused by these proceduresadds to the burden of stress in the neonate,causing a risk for short- and long term conse-quences (Abdulkader et al., 2008; Bouza, 2009).

The increasing insight that pain alleviation is anecessity in newborn care has lead to the creation ofnational and international guidelines for the man-agement of newborn pain. Following the interna-tional document (Anand, 2001), early examples ofnational guidelines can be found in USA and Canada(American Academy of Pediatrics, 2006), Sweden(Larssonetal., 2002)andAustralia (RoyalAustralasianCollege of Physicians, 2006). Typically, these guide-lines have sections on pain assessment and non-pharmacological and pharmacological interventions.The non-pharmacological interventions aim at mini-mizing stress- and painful events, offering supportlike non-nutritive sucking and skin-to-skin contactandgiving sweet solutionsorally prior tominorpainfulprocedures.

National guidelines should be transformed intolocal written guidelines at every unit providing carefor newborn infants. Gharavi et al. (2007) showedthat units with local guidelines will provide a higherfrequency of pain treatment and documentation ofpain. In Sweden88%of theneonatal units hadwrittenpain guidelines 2008 (Eriksson and Gradin, 2008)compared to for example 15% in Australia (Harrisonet al., 2006) and 44% in Austria, Switzerland andGermany (Gharavi et al., 2007).

A previous survey to the chief neonatologist atall Swedish NICUs revealed that behavioural (non-pharmacological) interventions were given beforesome skin breaking procedures at 53e95% of theunits. Only one (3%) of the units reported usingpharmacological interventions (EMLA cream),before subcutaneous injection. Oral glucose solu-tion for the same procedures were given at 66e91%of the units (Eriksson and Gradin, 2008).

The purpose of this study was to investigate ifnurses on neonatal units in Sweden followed the

national guidelines for some painful procedures(peripheral venous catheter placement, capillaryheel stick, venepuncture and s.c/i.m. injections),and secondly to investigate documentation ofpharmacological and non-pharmacological pain-alleviating interventions. The results werecompared with those of a preceding chief neona-tologist survey (Eriksson and Gradin, 2008).

Methods

Design and setting

The study design was a semi-structured survey sentto a sample of nurses at every NICU in Sweden. Theunits were divided into four categories, from levelA: university hospital with full neonatal intensivecare services, to level B: county hospital with aneonatal intensive care unit, level C: county andgeneral hospital with partial and short-timeneonatal intensive care, to level D: hospitalwithout neonatal intensive care, following theofficial Swedish ranking (National Board of Healthand Welfare, 1997). Every head nurse received aninformation letter, four survey-forms and a pre-paid return envelope. The head nurses wereasked to distribute the surveys to a representativesample of nurses that had been working at least sixmonths at the unit. The surveys were providedwith code-numbers to enable reminders and drop-out analysis. After two and four weeks reminderswere sent to units that had not returned the forms.Totally 180 surveys were distributed, to all 45 unitslisted as providers of neonatal care in Sweden.According to Swedish legislation at the time of thesurvey, no research-ethical application wasneeded for a survey sent from the university to thestaff at the unit. The informants were howeverinstructed that participation was voluntary andthat analysis and presentation of results would bedone on a group-level where neither individualnurses nor units could be identified.

The survey that had nine open and 17 closedquestions plus nine Likert-type questions (see Ap-pendix 1) was developed from the earlier chiefneonatologist-survey (Eriksson and Gradin, 2008).The survey was pilot-tested by two paediatricnurses before the final distribution.

Neonatal pain guidelines 33

Analysis

The data from the returned questionnaires wereentered into a MS Excel (Microsoft Corp., Rich-mond, WA, USA) worksheet. Statistical calculationswere performed in Excel, SPSS version 15 (IBMCorp., Armonk, NY, USA) and MedCalc version 9.6(MedCalc software, Mariakerke, Belgium). Foranalysis of non parametric data the KruskaleWallistest or chi-square test were used. For differences inparametric data Student’s t-test was used. Skeweddata are presented as median values. A p-value lessthan 0.05 was considered statistically significant.

Table 1 Fraction of nurses reporting that they usepharmacological and behavioural supportive in-terventions for painful procedures. Number in squarebrackets indicates how many nurses reported theprocedure being done at their unit, n ¼ 116.

Procedure Pharmaco-logicaltreatment

Behaviouralsupportiveinterventions

Both

% (n) % (n) % (n)

Venepuncture[116]

64 (74) 96 (11) 62 (72)

Capillary heelstick [57]

44 (25) 98 (56) 39 (22)

Peripheralvenouscatheter[114]

68 (77) 96 (109) 63 (72)

Subcutaneousand/orintramuscularinjection[101]

44 (44) 87 (88) 39 (39)

Results

One-hundred and twenty surveys (response-rate67%) were returned from a total of 36 units One ofthe 45 units that had received the survey reportedthat they did not provide care for newborn infants,so the response rate for units was 82%. Four of thereturned surveys from one unit reported policiesand practices concerning older children and werethus excluded, leaving 116 surveys for analysis, from34 units. Twenty of the reporting units returned allfour surveys, ten returned three surveys and twounits each returned two respectively one survey. Noresponding units were categorized as level D; hos-pital without neonatal intensive care.

The majority (76%; n ¼ 88) of the respondingnurses had a specialist nurse education in paedi-atric, intensive or midwifery care nursing. Therespondents were also asked to specify whatfraction of all the nurses on their own unit had anyof the above mentioned specialist nursing educa-tions, and the mean percentage for the answersfrom each unit was calculated. The reportedfraction of specialist education among all nurses atany of the three levels of neonatal units did notdiffer from that of the respondents (no significantdifference, chi-square test), and the respondents’level of education can therefore be seen asrepresentative in this aspect.

The nurses who worked in either level A or levelB hospitals had a median neonatal experience of 8years and on level C 11 years. No significant dif-ferences could be seen (Kruskal Wallis test).

Ninety per-cent of the respondents (n ¼ 104)from all (100%) of the participating units reportedthat the unit they worked in had written guidelinesabout neonatal pain, compared to 88% of the unitsin the previous survey (Eriksson and Gradin, 2008).In level C units, three of the four respondingnursed stated that the unit did not have guidelinesand in one level B unit two nurses answered that

the unit had guidelines and two answered that noguidelines were available.

The respondents were asked if they would usepharmacological or behavioural supportive in-terventions or both types for four specified painfulevents: venepuncture, capillary heel-stick, place-ment of peripheral venous catheter, and subcu-taneous/intramuscular injection. The results thatare specified in Table 1 shows that almost all nurseswould use some kind of behavioural supportivetreatment for all the listed procedures. The mostcommon intervention was to give 30% glucose solu-tion orally, prior to the procedure. Non-nutritivesucking and swaddling/holding were alsofrequently used. On the specific question if thenurses would give glucose prior to capillary bloodglucose checks, 22% (n ¼ 25) answered that theywould not. The reasons given were fear of false testresults or that they used new devices that were notpainful. In a secondary analysis the results werecompared to the answers from the chief neonatol-ogist of the same units in a previous study (Erikssonand Gradin, 2008), which revealed a lower use ofglucose administration for the listed interventionsthan stated in the first survey (Fig. 1).

Twenty-one per cent of the nurses (n ¼ 23) re-ported that they often or always would use EMLA�-cream at venepuncture. For peripheral venouscatheter, 22% (n ¼ 24) would use EMLA� and forsubcutaneous/intramuscular injection 17%(n ¼ 16), often or always. No one stated that theywould use it often or always for capillary heel stick

0

10

20

30

40

50

60

70

80

90

100

Venepuncture (p=0.013) Capillary heel stick (n.s.) Peripher venous catheter(p=0.017)

Sucutaneous/ intramuscularinjection (p=0.033)

Chief neonatologistNurses

Fig. 1 Reported use of oral glucose for specified interventions, in a previous survey to chief neonatologists (Erikssonand Gradin, 2008) and nurses at the same units (n.s. ¼ not significant).

34 E. Scherman et al.

(Table 2). One hundred nurses gave their reasonsfor not using EMLA� in general or in specific situ-ations. The most frequent reason was low age and/or weight of the infants (35%), that the same effectcould be achieved with glucose (25%) and lack oftime to wait for the effect of EMLA� (24%).

Sixty percent of the respondents (n ¼ 68) ans-wered that they could take capillary blood samplesfrom the toes, rather than the heel and other sitesused for blood sampling were arms, legs, fingers orears, as some blood glucose devices only require asmall sample of blood.

The nurses were also asked about documenta-tion of pain-relieving interventions. Behaviouralsupportive interventions were much less docu-mented than pharmacological treatment (EMLA�).Orally given glucose, non-nutritive sucking andswaddling were the most documented behaviouralinterventions, often in the written shift-report,whereas EMLA� was documented on the medica-tions list.

Table 2 Reported use of EMLA�-cream at different proc

Venepuncture(n ¼ 111)

Capillaryheel stic(n ¼ 57)

% (n) % (n)

Never 33 (37) 92 (53)Seldom 35 (39) 4 (2)Sometimes 11 (12) 4 (2)Often 14 (15) 0 (0)Always 7 (8) 0 (0)

Discussion

This study supports the idea that most nurses inSwedish neonatal units provide support, accordingto the national guidelines, for the infants duringcommonly performed painful procedures, and iscoherent with a recent study showing that Sweden,Denmark and France are most likely to employ evi-dence based pain-alleviation recommendations(Losacco et al., 2011). This indicates an awarenessamong the nurses, that providing treatment,behavioural support or both tominimize the infant’spain is important. This realization is the first step torecognizing and in a systematicway evaluate pain ininfants and thereby increase the likelihood to treatpain (Allegaert et al., 2003). Even if this part of theresults seems promising, other findings however,indicates that there is still work to be done to findand break down psychological, structural or otherbarriers, implement research findings and keepreaching for full clinical implementation of the

edures.

kPeripheralvenouscatheter(n ¼ 110)

Subcutaneousand/orintramuscularinjection (n ¼ 95)

% (n) % (n)

43 (48) 64 (61)24 (26) 8 (8)11 (12) 12 (10)8 (9) 8 (8)

14 (15) 8 (8)

Neonatal pain guidelines 35

Swedish national guidelines. For example; capillaryheel stick is considered to be more painful thanvenepuncture (Shah and Ohlsson, 2011). In thissurveymore than a third of the nurses reported thatthey provide treatment or support to the infantduring capillary heel stick. In other words: bloodsampling by heel stick is still practiced in neonatalunits in Sweden and yet not fully replaced by otherless painful techniques such as venepuncture (Shahand Ohlsson, 2011) or transcutaneous measurement(Carceller-Blanchard et al., 2009). This on-going useof heel-sticks is also shown in a large Europeansurvey (Losacco et al., 2011).

Another issue of concern is that close to onefourth of the nurses avoided giving 30% glucoseorally to infants prior to blood sampling for bloodglucose checks, despite evidence that it is possibleto give glucose orally prior to capillary blood glu-cose checks (Harrison, 2008).

Despite the scientific support for the use of lido-caineeprilocaine cream (EMLA�) (Weise and Nahata,2005) almost 70% of the nurses seldom or never usedEMLA� prior to venepuncture or insertion of a pe-ripheral venous catheter. This high proportion in-dicates that the use of topical anaesthesia needs tobe promoted, especially in the light of the reportedarguments against using EMLA� (low gestational age,lack of time to wait for the effect of EMLA� or thatsweet solution is sufficient). There are studiesreporting that it is safe to use EMLA� for topicalanaesthesia in infants having a gestational age of 30weeks or higher (Weise and Nahata, 2005) which isthe case of the waste majority of the infantsadmitted to neonatal units in Sweden.

When a blood sample or insertion of a peripheralvenous catheter are needed in an emergency situa-tion it is reasonable to exclude EMLA�. Whenordering blood sampling it should be consideredwhether the test could be delayed for at least anhour to get the opportunity to use EMLA� cream. Thethird argument, that 30% glucose orally is sufficient,does not hold for a closer evaluation. Recentresearch indicates that orally given glucoseandEMLAcan work synergistic, increasing the chance for lesspain for the infants (Biran et al., 2011). Recently thepain relieving effects of orally given glucose havebeen questioned (Slater et al., 2010) which isanother reason for finding combination of pain-relieviingmethods that can have a synergistic effect.Such combinations of course need further studies toevaluate their effect and safety.

The nurses reported that pharmacological treat-ment (EMLA�) was documented to a higher degreethan non-pharmacological. This raises questionsabout how nursing interventions is viewed at inrelation to traditional medical interventions.

Written guidelines were reported to be adaptedon all units participating in the survey. When com-pared to the results of a previous study (Eriksson andGradin, 2008), the units that did not have writtenguidelines at the time of the survey now have painmanagement guidelines in place. Not every res-ponding nurse was aware of those guidelines, whichindicates the need for a more thorough imple-mentation process of the local guidelines. Theobvious question to this and other studies reportingon the existence of guidelines is whether the guide-lines actually have impact on the daily clinical work.And secondly, how it affects the outcome of the pa-tients. In other words: do infants benefit from beingtreated at a unit that has adopted pain guidelines?After a review of the literature on paediatric painassessment Franck and Bruce raised the question“Putting pain assessment into practice: why is it sopainful?”. They concluded that there is a lack ofstraightforward evidence that the implementationof routines for pain assessment in itself will enhancethe pain management at the unit (Franck and Bruce,2009). It is tempting to believe that successful painmanagement is a chain where every link is asimportant as the others; evidence based guidelinesthat are accepted and adopted by all staff, acontinuous educational program for education aboutpain issues, regular pain assessment and documen-tation of pain and providing of pain alleviation inaccordance to the guidelines. Recent research alsopinpoints the role of parents in neonatal pain man-agement (Axelin et al., 2010; Franck et al., 2011).

Strengths and limitations

This study investigated pain management prac-tices among nurses at 82% of Swedish neonatalintensive care units. The self-reported frequencyof glucose use was lower than previously reportedfrom the chief neonatologists at the same units butthis difference can reflect either the time lag be-tween the two surveys, different awareness ofwhat is actually done, or both.

Conclusion

A vast majority of responding NICU nurses providedbehavioural and/or pharmacological support atskin breaking procedures, thereby following theSwedish national neonatal pain guidelines. Phar-macological interventions were documented to ahigher extent than behavioural support. All unitshad written guidelines about neonatal pain man-agement. Chief neonatologists reported higher useof glucose, which is recommended in the Swedish

36 E. Scherman et al.

national guidelines, than did nurses and specialisttrained nurses.

Acknowledgements

This project was supported by the European Com-munity’s Seventh Framework Programme undergrant agreement no. 223767. We also wish to thankthe staff at all units that replied to the survey.

Appendix A. Supplementary data

Supplementary data associated with this articlecan be found in the online version, at http://dx.doi.org/10.1016/j.jnn.2013.04.010.

References

Abdulkader, H.M., Freer, Y., Garry, E.M., Fleetwood-Walker, S.M.,McIntosh, N., 2008. Prematurity and neonatal noxious eventsexert lasting effects on infant pain behaviour. Early Hum. Dev.84 (6), 351e355.

Allegaert, K., Tibboel, D., Naulaers, G., Tison, D., De Jonge, A.,Van Dijk, M., Devlieger, H., 2003. Systematic evaluation ofpain in neonates: effect on the number of intravenous an-algesics prescribed. Eur. J. Clin. Pharmacol. 59 (2), 87e90.

American Academy of Pediatrics, Committee on Fetus andNewborn and Section of Surgery, Canadian Paediatric Soci-ety and Fetus and Newborn Committee, 2006. Preventionand management of pain in the neonate: an update. Pedi-atrics 118 (5), 2231e2241.

Anand, K.J., International Evidence-Based Group for NeonatalPain, 2001. Consensus statement for the prevention andmanagement of pain in the newborn. Arch. Pediatr. Adolesc.Med. 155 (2), 173e180.

Axelin,A., Lehtonen,L., Pelander,T., Salantera, S., 2010.Mothers’different styles of involvement in preterm infant pain care.J. Obstet. Gynecol. Neonatal. Nurs. 39 (4), 415e424.

Bouza, H., 2009. The impact of pain in the immature brain.J. Mat. Fetal Neonatal. Med. 22 (9), 722e732.

Biran, V., Gourrier, E., Cimerman, P., Walter-Nicolet, E.,Mitanchez, D., Carbajal, R., 2011. Analgesic Effects of EMLA

Cream and Oral Sucrose During Venipuncture in PretermInfants. Pediatrics 128 (1), e63ee70.

Carbajal, R., Rousset, A., Danan, C., Coquery, S., Nolent, P.,Ducrocq, S., Breart, G., 2008. Epidemiology and treatmentof painful procedures in neonates in intensive care units. J.Am. Med. Assoc. 300 (1), 60e70.

Carceller-Blanchard, A., Cousineau, J., Delvin, E.E., 2009. Pointof care testing: transcutaneous bilirubinometry in neonates.Clin. Biochem. 42 (3), 143e149.

Eriksson,M.,Gradin,M., 2008.Painmanagement inSwedishneonatalunits e a national survey. Acta Paediatr. 97 (7), 870e874.

Franck, L.S., Oulton, K., Nderitu, S., Lim, M., Fang, S.,Kaiser, A., 2011. Parent involvement in pain management forNICU infants: a randomized controlled trial. Pediatrics 128(3), 510e518.

Franck, L.S., Bruce, E., 2009. Putting pain assessment intopractice: why is it so painful? Pain Res. Manag. 14 (1), 13e20.

Gharavi, B., Schott, C., Nelle, M., Reiter, G., Linderkamp, O.,2007. Pain management and the effect of guidelines inneonatal units in Austria, Germany and Switzerland. Pediatr.Int. 49, 652e658.

Harrison, D., Loughnan, P., Johnston, L., 2006. Pain assessmentand procedural pain management practices in neonatal unitsin Australia. J. Paediatr. Child Health 42 (1e2), 6e9.

Harrison, D.M., 2008. Oral sucrose for painmanagement in infants:myths and misconceptions. J. Neonat. Nurs. 14 (2), 39e46.

Larsson, B.A., Gradin, M., Lind, V., Selander, B., 2002. Swedishguidelines for prevention and treatment of pain in thenewborn infant. Lakartidningen 99 (17), 1946e1949.

Losacco, V., Cuttini, M., Greisen, G., Haumont, D., Pallas-Alonso, C.R., Pierrat, V., Sizun, J., 2011. Heel blood sam-pling in European neonatal intensive care units: compliancewith pain management guidelines. Arch. Dis. Child. FetalNeonatal. Ed. 96 (1), F65eF68.

National Board of Health and Welfare, 1997. Intensive Care ofNew-born Infants. Stockholm, p. 67.

Royal Australasian College of Physicians, Peadiatrics & ChildHealth Division, 2006. Guideline statement: management ofprocedure-related pain sin neonates. J. Paediatr. ChildHealth 42 (Suppl. 1), S31eS39.

Shah, V.S., Ohlsson, A., 2011. Venepuncture versus heel lance forblood sampling in term neonates. Cochrane Database Syst. Rev.10, CD001452. http://dx.doi.org/10.1002/14651858.CD001452.

Slater, R., Cornelissen, L., Fabrizi, L., Patten, D., Yoxen, J.,Worley, A., Fitzgerald, M., 2010. Oral sucrose as an analgesicdrug for procedural pain in newborn infants: a randomisedcontrolled trial. Lancet 376 (9748).

Weise, K.L., Nahata, M.C., 2005. EMLA for painful procedures ininfants. J. Pediatr. Health Care 19 (1), 42e47.

Available online at www.sciencedirect.com