The data showed that pain was higher for our standardized method of sampling

Objective

To investigate the analgesic effect (measured with NIPS) of breastfeeding in addition to skin-to-skin contact versus other methods of non-pharmacological analgesia during blood sampling through heel lance in healthy term neonates. The influences of non-pharmacological methods on crying time, percentage of crying while sampling, heart rate, number of attempts and duration of sampling were studied.

Methods

Randomised controlled trial performed on 136 healthy term newborns in the maternity ward of a tertiary hospital. The inclusion criteria were: healthy term neonates, wish to breastfeed and absence of feeding during the previous 60 minutes. Neonates were randomly assigned to four groups: group 1, breastfed with skin-to-skin contact ; group 2, oral sucrose with skin-to-skin contact ; group 3, skin-to-skin contact ; or group 4, receiving oral sucrose Data for the primary objective was analysed per intention to treat. This study was approved by local ethical committee.

Healthy term newborns commonly undergo painful procedures even if they do not need intensive care. Heel lance for neonatal screening is the most frequent standardized painful procedure performed in these newborns. Until recently, analgesics were rarely administered to neonates because it was believed that they were less sensitive to pain than more mature infants but current studies suggest that newborns are able to perceive pain. Regardless of the available evidence, appropriate systematic analgesia in common painful procedures in neonates is far from adequate.

Painful stimuli in neonates may have short-term physiologic (increase in intracranial pressure, increase in heart rate or decrease in oxygen saturation) and behavioural consequences (cry, eye squeeze); as well as long-term consequences (altered of pain response in later infancy). Different non-pharmacological methods of analgesia, such as sucrose, skin-to-skin contact (SSC), breastfeeding(BF) or music10, have been used to reduce pain in neonates undergoing venipuncture or heel lancing, and different physiologic pathways to explain the underlying mechanism have been proposed.

A variety of valid and reliable pain assessment instruments have been developed over the past decades. Yet, behavioural pain assessment remains challenging and controversial due to the lack of a gold standard for neonatal pain expression. The Neonatal Infant Pain Scale (NIPS) and the Premature Infant Pain Profile (PIPP) were selected by an international consensus neonatal pain group for acute procedural pain in neonates.

Our main objective was to investigate the analgesic effect (measured with NIPS) of breastfeeding in addition to SSC versus other methods of non-pharmacological analgesia during blood sampling through heel lance in healthy term neonates. Furthermore, the influences of non-pharmacological methods on crying time, percentage of crying while sampling, heart rate, number of attempts and time of sampling were studied.

METHODS

Protocol

This randomised controlled trial was performed on 136 healthy term newborns in the maternity ward of a tertiary hospital. The inclusion criteria were as follows: healthy term neonates (37-416 weeks of gestation) confirmed through a routine physical exam during the first 24h of life, wish to breastfeed and absence of feeding during the previous 60 minutes. Exclusion criteria were as follows: maternal use of opioids, birth in general anesthesia, artificial feeding, previous capillar or venous sampling, and previous admission to the neonatal unit.

Written informed consent was asked to parents during consultation. Study protocol and informed-consent forms were approved by the local ethics committee.

Intervention

Participating neonates were randomly assigned to four groups: group 1, breastfed with SSC (n=35); group 2, receiving oral sucrose with SSC (n=35); group 3, being held in SSC (n=33); or group 4, receiving oral sucrose (n=33). Randomisation was by closed envelopes and nurses and parents were masked to the randomization group but not blinded to the treatment assignment. In group 1 (BF+SSC), neonates dressed with a diaper were held in prone, in SSC with the mother; BF was started at least 5 minutes before heel lance and maintained during sampling. In group 2 (sucrose + SSC), neonates were held in prone between the mothers' breast at least 5 minutes before sampling and 2 ml 24% sucrose was given with a sterile syringe in the mouth 2 minutes before heel lance. In group 3 (SSC), neonates were held between the mother's breast as in group 2, but no sucrose was given. In group 4 (Sucrose), 2 ml 24% sucrose was administered through a sterile syringe in the mouth 2 minutes before heel lance to neonates laid on supine on a cot; the procedure was done in the presence of the mother. Mothers were allowed to speak or touch their babies in all the groups. The standardized procedure of heel prick in our hospital is such as described in group 4.

Throughout the duration of the test, babies were continuously recorded with a video camera at least 2 minutes before sampling and 2 minutes after the procedure. The heel was warmed up by a glove with lukewarm water at least 2 minutes before the sample. Heel lance was made with an automated piercing device for routine neonatal screening for congenital disorders at 48 hours of life. Heart rate was monitored by a pulse oximeter (Radical MasimoSet Datascope, Masimo Corporation, Irvine, CA) set on the infant's hand or foot. Hear rate was measured continuously but special attention was given to three time points: t0 (2 minutes before sampling); t1 (the highest value of the first 10 seconds after heel prick); and t2 (2 minutes after the procedure).

Blood sampling was performed through a standardized procedure by five experienced nurses, who obtained 5 dried spots of blood collected on a filter paper card. If the sample was not enough to complete all of the dried spots on the filter card, a new heel lance was practiced in few seconds. In this case, neonates were assessed for NIPS measure only after the first heel lance; secondary outcomes were evaluated during the whole sampling. Crying time was defined as the duration of crying while sampling. Percentage of crying was defined as the ratio between crying time while sampling and time of the procedure.

Pain scale

The NIPS scale is a validated 6-indicator scale for the assessment of acute pain in neonates11,13. It measure movement of arms and legs, breathing patterns, cry, facial expression and state of arousal. Score ranges from 0 (no pain) to 7 (severe pain). NIPS score <4 means no pain-mild pain. NIPS score ≥4 means moderate-severe pain. NIPS score was measured at three time points: NIPS0 (2 minutes before sampling); NIPS1 (the highest value of the scale in the first 10 seconds after heel prick); and NIPS2 (2 minutes after the procedure). NIPS2 was not recorded if sampling had not finished at this time point. This took place in one newborn (2,8%) of group 1, five (14,2%), six (18.1%), and five (15.1%) of group 2, 3 and 4 respectively. NIPS was measured by three researchers who watched the videos: one expert neonatologist (Observer 1) and two young paediatricians (Observers 2 and 3). Coincidence was 90.8% (95% CI 87.8-93.8) between observers 1 and 2; 75% (70.5-79.5) between observers 1 and 3; and 78.7% (74.7-82.9) between observers 2 and 3. Consequently, only data of the expert neonatologist is shown.

Statistical Analysis

A sample size of 67 infants in each group was calculated to achieve a power of 80% with an α of 0.05 to detect a 0.5-point difference in the NIPS score (assuming SD=1). We decided to make a mid-point analysis when half of the neonates were recruited in order to detect if any of the analgesic methods was better than the standard procedure, and thus modify our clinical practice.

Results were expressed as mean ± standard deviation (SD). Categorical variables were analysed with the χ2 test and the Fisher test. We compared mean between groups with the t-test, ANOVA test and Bonferroni test. We used Mann-Whitney U test, non-parametric ANOVA and Kruskal-Wallis to compare median between groups. The SPSS v.14.0 software package was used to perform all statistical analysis (SPSS Inc. IL, USA). A p-value of 0.05 was considered a statistically significant level of difference. The analysis of the main objective was conducted on an intention-to-treat. Secondary objectives were analysed according to protocol.

Inclusion Criteria:

• Healthy term neonates (37-41+6 weeks of gestation) confirmed through a routine physical exam during the first 24h of life
• Wish to breastfeed
• Absence of feeding during the previous 60 minutes.

Exclusion Criteria:

• Maternal use of opioids.
• Birth in general anesthesia.
• Artificial feeding.
• Previous capillar or venous sampling.
• Previous admission to the neonatal unit.