Background
Pain in the newborn is associated with acute behavioral and physiological changes. Cumulative pain is linked to morbidities, including adverse neurodevelopmental outcomes. Studies have shown reduced changes in physiological parameters and pain score measurements following preemptive administration of analgesics in neonates experiencing pain or stress.
Non-pharmacological measures (such as holding the child, swaddling, and breastfeeding) and pharmacological measures (such as paracetamol, sucrose, and opioids) have been administered for analgesia. This is an update of a review first published in 2006 and updated in 2012.
Goals
The primary objective was to evaluate the effectiveness of breastfeeding or supplemental breast milk in reducing procedural pain in neonates.
The secondary objective was to perform subgroup analyzes according to the type of control intervention, gestational age, and the amount of supplemental breast milk administered.
Search methods
We searched CENTRAL, MEDLINE, Embase, CINAHL and trials registries (ICTRP, ISRCTN and clinicaltrials.gov) in August 2022; searches were limited from 2011 onwards. We also checked the reference lists of included studies and relevant systematic reviews.
Selection criteria
We included randomized controlled trials (RCTs) or quasi-randomised trials of breastfeeding or supplementary breast milk versus no treatment/other measures in neonates.
Both full-term (≥ 37 completed weeks of postmenstrual age) and premature (< 37 completed weeks of postmenstrual age) neonates were included up to a maximum of 44 weeks of postmenstrual age. The study must have reported physiological markers of pain or validated pain scores.
Data collection and analysis
The methodological quality of the trials was assessed using information from the studies and personal communication with the authors. Data on relevant outcomes were extracted, effect sizes estimated and reported as mean difference (MD). The GRADE method was used to assess the certainty of the evidence.
Main results
Of the 66 included studies, 36 evaluated breastfeeding, 29 evaluated supplemental breast milk, and one study compared them with each other.
The procedures performed in the studies were: heel prick (39), venipuncture (11), intramuscular vaccination (nine), eye examination to detect retinopathy of prematurity (four), suction (four) and removal of tape as a procedure (one ).
Marked heterogeneity was observed in control interventions and pain assessment measures across studies. Since many studies included multiple groups with breastfeeding/supplementary breast milk as the primary comparator, not all interventions could be synthesized together.
Individual interventions are compared to breastfeeding/supplementary breast milk and reported. The numbers of studies/participants presented with the results are not from pooled analyzes (as is common in Cochrane reviews), but are the overall totals for each comparison.
Overall, the included studies were at low risk of bias, except for intervention blinding and outcome assessment, where almost a third of the studies were at high risk of bias.
Breastfeeding versus control
Breastfeeding may reduce the increase in heart rate compared to maternal holding, skin-to-skin contact, bottle-feeding, moderate sucrose/glucose concentration (20% to 33%) and skin-to-skin contact . skin (low-certainty evidence, eight studies, 784 participants).
Breastfeeding is likely to reduce the duration of crying compared to no intervention, lying on a stretcher, rocking, warming heels, being held by the mother, skin-to-skin contact, bottle-feeding breast milk and moderate glucose concentration (moderate-certainty evidence, 16 studies, 1866 participants).
Breastfeeding may reduce the percentage of crying time compared with maternal holding, skin-to-skin contact, bottle-feeding, moderate-concentration sucrose, and moderate-concentration sucrose with skin-to-skin contact (evidence low certainty, four studies, 359 participants).
Breastfeeding is likely to reduce Neonatal Infant Pain Scale (NIPS) score compared to no intervention, maternal holding, heel warming, music, EMLA cream, moderate glucose concentration , being swaddled, being swaddled and hugged (moderate-certainty evidence, 12 studies, 1432 participants).
Breastfeeding could reduce the Neonatal Facial Coding System (NFCS) score compared to no intervention, restraint, pacifier and moderate glucose concentration (low-certainty evidence, two studies, 235 participants).
Breastfeeding may reduce the Douleur Aigue Nouveau‐né (DAN) score compared to position, restraint or placebo (low‐certainty evidence, four studies, 709 participants).
In most other comparisons there was little or no difference between the breastfeeding group and the control group on outcome measures.
Supplemental breast milk versus control
Supplementary breast milk may reduce the increase in heart rate compared to water or no intervention (low-certainty evidence, five studies, 336 participants).
Supplementary breast milk probably reduces the duration of crying compared to positioning, massage or placebo (moderate-certainty evidence, 11 studies, 1283 participants).
Supplemental breast milk makes little or no difference in percentage of crying time compared to placebo or glycine (low-certainty evidence, one study, 70 participants).
Supplemental breast milk makes little or no difference in NIPS score compared with no intervention, pacifier, moderate concentration of sucrose, eye drops, gentle touch and verbal comfort, and breast milk smell and verbal comfort (low-certainty evidence , three studies, 291 participants).
Supplementary breast milk may reduce the NFCS score compared to glycine (overall low-certainty evidence, 1 study, 40 participants). DAN scores were lower compared to massage and water; there were no differences compared to no intervention, EMLA and moderate sucrose concentration; and were superior compared to rocking or pacifying (low-certainty evidence, two studies, 224 participants).
Due to the high number of comparison interventions, other pain measures were evaluated in a very small number of studies in both comparisons, making the evidence of low certainty.
Most studies reported no adverse events, considering the benign nature of the intervention. Those reporting adverse events did not identify any in participants.
Subgroup analyzes were not performed due to the small number of studies.
Authors’ conclusions
|
