Article Text

Download PDFPDF

Sweeten, soother and swaddle for retinopathy of prematurity screening: a randomised placebo controlled trial
  1. A O'Sullivan1,
  2. M O'Connor1,
  3. D Brosnahan1,2,3,
  4. K McCreery1,2,
  5. E M Dempsey4,5
  1. 1Department of Paediatrics and Newborn Medicine, Coombe Women and Infants University Hospital, Dublin, Ireland
  2. 2Our Lady's Hospital for Children, Crumlin, Dublin, Ireland
  3. 3Royal Victoria Eye and Ear Hospital, Dublin, Ireland
  4. 4Department of Neonatology, Cork University Maternity Hospital, Cork, Ireland
  5. 5Department of Paediatrics and Child Health, University College Cork, Cork, Ireland
  1. Correspondence to Dr Dempsey, Department of Neonatology, Cork University Maternity Hospital, Cork, Ireland; gene.dempsey{at}


Objective To assess the efficacy of oral sucrose combined with swaddling and non-nutritive suck (NNS) as a method for reducing pain associated with retinopathy of prematurity (ROP) screening.

Design Randomised placebo controlled study.

Setting Tertiary level neonatal intensive care unit.

Sample 40 infants undergoing primary eye examination for ROP screening.

Intervention The control group were swaddled, and received 0.2 ml of sterile water given by mouth using a syringe and a soother. The intervention group were swaddled, and received 0.2 ml of sucrose 24% given by mouth using a syringe and a soother.

Results 40 infants were included in the study. There was no difference in mean gestational age at birth, mean birth weight or corrected gestational age at first examination between both groups. The sucrose group had a significantly lower median Neonatal Pain, Agitation and Sedation Scale (N-PASS) score during ROP screening, initially following insertion of the speculum (6.5 vs 5, p=0.02) and subsequently during scleral indentation (9.5 vs 7.5, p=0.03). Fewer infants experienced episodes of desaturations or bradycardia in the intervention group (1 vs 4, p=0.18).

Conclusion ROP screening is a necessary but recognised painful procedure. Sucrose combined with NNS and swaddling reduced the behavioural and physiological pain responses. However, pain scores remained consistently high and appropriate pain relief for ROP screening remains a challenge.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Retinopathy of prematurity (ROP) can occur when the normal retinal vascular development is interrupted by premature birth.1 ROP is one of the causes of childhood visual disability and is largely preventable with appropriate screening.2 The incidence of complete or partial blindness decreased to 3% in 2000 from 5–8% prior to 1990,3 despite an overall increased survival of high risk neonates at lower gestational ages. The contribution of screening guidelines has been central to the reduction in this adverse outcome and the Royal College of Ophthalmologists and the British Association of Perinatal Medicine currently recommend screening for all infants at risk, defined as those infants of less than 32 weeks gestation and less than 1501 g.4

ROP screening is one of many potentially painful diagnostic and therapeutic procedures performed routinely on preterm infants in the neonatal intensive care unit.5 Autonomic pathways for pain processing are in place from mid to late gestation and suboptimal inhibitory mechanisms contribute to increased sensitivity to pain.6 7 Numerous studies demonstrate that repeated pain can have direct and long-term consequences on the neurological and behavioural development of infants.8,,10 Therefore strategies for stress reduction and pain management are essential to promote growth and development and minimise long-term sequelae. Procedural analgesia should include concepts of developmental care, non-nutritive suck (NNS), and pharmacological and non-pharmacological agents.5 Sucrose is thought to stimulate the body's activation of lingual sweet taste receptors and release of endogenous opioids,11 12 and when combined with NNS, non-opioid mechanisms are also activated.13 The administration of sucrose or the combination of sucrose and NNS is one of the most frequently studied non-pharmacological interventions for relief of pain in neonates14 and oral sucrose has been shown to be an effective and safe therapy for common neonatal procedures such as heal lance, bloodletting and venepuncture.15 16 There is conflicting evidence on the benefit of sucrose in ROP screening.17,,21 Therefore, the purpose of this study was to determine the efficacy of sucrose combined with swaddling and NNS as a potential regime for reduction of pain associated with ROP screening.

What is already known on this topic

Retinopathy of prematurity screening is a painful procedure frequently performed in the neonatal intensive care unit.

What this study adds

  • Sucrose combined with non-nutritive suck and swaddling reduced the behavioural and physiological pain scores recorded during screening for retinopathy of prematurity.

  • Retinopathy of prematurity screening is a painful procedure for the preterm infant and current pain reduction techniques are inadequate.


A prospective randomised placebo controlled study was carried out. Infants who met the criteria outlined by the Royal College of Ophthalmologists and the British Association of Perinatal Medicine were considered. Exclusion criteria included infants requiring mechanical ventilation (excluding continuous positive airway pressure), infants receiving sedation, infants with congenital malformations and infants where consent to participate was not obtained. The Clinical Research Ethics Committee of the National Maternity Hospital, Dublin reviewed and subsequently granted ethics approval for this study. Parents of eligible infants were given verbal and written information about the study, and written consent was obtained.

The neonatal pharmacist undertook a computer based randomisation process. Sequentially numbered, opaque sealed envelopes were utilised. Infants were randomised to solution A or B. Both solutions were clear and colourless. Only the pharmacist was aware of the identity of these solutions. Efficacy was measured by assessing the infants' pain measured by a validated composite criterion, the Neonatal Pain, Agitation and Sedation Scale (N-PASS). This pain assessment tool is multidimensional in nature and current guidelines suggest that the multifaceted nature of pain makes the use of a multidimensional assessment tool preferable to a one-dimensional option.22 N-PASS, which controls for differences in gestational age, is designed to assess acute prolonged pain and chronic pain. Five indicators graded 0, 1 and 2 are included in the N-PASS. The indicators chosen for their established validity, clinical applicability and ease of assessment include crying/irritability, behaviour/state, facial expression, extremities/tone and vital signs. Scores for sedation graded 0, −1 and −2 were not used as sedated infants were excluded from this study.23 24 Infants in this study were swaddled, however swaddling was limited to the upper body so lower limbs could be assessed and this approach was consistent in both groups. The ophthalmologist, the nurse recording data, parents and personnel scoring the N-PASS were unaware of the group assignment. The study was carried out on the first screening examination for each infant; this was to overcome potential confounding effects that might occur if a baby was to anticipate a procedure following previous exposure. Eye examinations were carried out by one of two experienced ophthalmologists using indirect ophthalmoscopy. The same type of speculum and indentors were used in an effort to maintain consistency.


Preparation for screening included the instillation of mydriatic eye drops Cyclomydril (cyclopentolate 0.2% and phenylephrine 1%) 60 and 30 min prior to the examination. Every neonate received local anaesthetic eye drops (tetracaine hydrochloride 1%) 30 s prior to the examination. Infants were randomised to receive one of the following two interventions 2 min before the start of the eye examinations. The control group were swaddled and received 0.2 ml of sterile water given by mouth using a syringe and a soother. The intervention group were swaddled and received 0.2 ml of sucrose 24% given by mouth using a syringe and a soother. Each infant was video recorded during the examination and for 5 min after completion of the examination. Only examinations of the first eye were used to record the N-PASS score. Patient demographics and baseline physiological data were recorded before the start of the examination and continuously throughout the examination using a Dash 4000 monitor (GE Healthcare, Little Chalfont, UK) which relayed data to a laptop computer with a data recording program (Bedmaster; Excel Medical Electronics, Jupiter, Florida, USA). The researcher, assistant researcher and consultant neonatologist blinded to the group status analysed the N-PASS score at a later date away from the clinical setting. Hummel et al demonstrated high inter-rater reliability, internal consistency and test–retest reliability of the N-PASS.23 24 In this study, training and in-depth knowledge of the N-PASS promoted interobserver consistency, scoring was done independently by the research team initially and where discrepancies arose, infants were re-scored and consensus reached. Participants were monitored routinely as per unit standards. This study had no impact on the care provided. All patients had ROP screening as per unit protocol. Adverse events to be recorded included choking episodes or vomiting following administration of the sucrose as well as changes in physiological parameters including tachycardia >180 bpm, bradycardia <80 bpmor oxygen desaturations <80% for >10 s.

On the basis of the results from observational data in a similar group of neonates in our intensive care unit, a power calculation indicated that a sample size of 20 in each group would be required to allow 80% probability of detecting a three-point difference in N-PASS score with significance at the 5% level. Data were analysed with a PC based statistics package (StatsDirect version 2.6.1) using the Student t test and Mann–Whitney U test. p<0.05 Was accepted as significant for our study.


A total of 66 infants met the eligibility criteria for ROP screening during the duration of the study. Of these, 40 infants were included and randomised. The primary reason for non-enrolment was failure to approach for consent (11 cases); the other reasons for exclusion are presented in figure 1. The demographic details of the 40 infants included in the study are presented in table 1. There was no difference in mean gestational age at birth, mean birth weight or corrected gestational age at first examination between both groups.

Figure 1

Flow diagram showing flow of patients through each step of the trial.

Table 1

Demographic details

Baseline physiological parameters, and behavioural and physiological pain responses during ROP screening are outlined in table 2. The sucrose group had a significantly lower N-PASS score during ROP screening of the first eye. This was both for insertion of the speculum where there was a 1.5 point reduction in pain scores recorded (6.5 vs 5, p=0.002) and during scleral indentation where the intervention resulted in a 2 point reduction in pain scores (9.5 vs 7.5, p=0.003). Fewer patients had episodes of desaturations (3 vs 1, p=0.32), bradycardia (3 vs 1, p=0.32) and overall adverse outcome (4 vs 1, p=0.18) in the intervention group; however, this did not reach statistical significance.

Table 2

Pain profile and adverse events in the control and intervention groups


ROP screening in preterm infants evoked a measurable pain response in both groups of patients as determined by the N-PASS. These findings are consistent with the inclusion of ROP screening as a recognised cause of pain in the neonatal intensive care unit and are also consistent with other studies addressing ROP screening.19 25 We looked at two phases of the screening process, initial insertion of the speculum followed by scleral indentation, as we felt both of these interventions were likely to cause pain. The score was high following initial insertion of the speculum and even higher following scleral indentation. In both phases the combined use of sucrose, swaddling and NNS was associated with a reduction in the N-PASS score compared to the use of swaddling and soother alone.

Gal et al and Mitchell et al also report a positive effect of sucrose on pain scores during ROP screening. Gal17 studied 23 infants of whom 12 received 2 ml of sucrose 24% and noted that sucrose reduced pain during the insertion of the eyelid speculum; however, the analgesic effect was not sustained. Mitchell18 included NNS in their study; they examined the analgesic effect of NNS and oral sucrose during eye examinations for ROP. While the mechanism involved in NNS is speculative, it is suggested that a non-opioid mechanism is involved,13 which helps the infant to achieve organisation and self-regulation and promotes calming.26 The investigators found a statistically significant higher pain score in the placebo group during the eye examination, with no significant differences found in Premature Infant Pain Profile (PIPP) scores between the two groups immediately after the examination, again suggesting a beneficial short-term effect of sucrose and NNS.

Boyle et al noted reduced distress responses in neonates with NNS, but a synergistic effect of sucrose was not demonstrated.19 Infants were randomised to one of four interventions: 1 ml of sterile water with a pacifier, 1 ml of sucrose 33% with a pacifier, 1 ml of sterile water without a pacifier or 1 ml of sucrose 33% without a pacifier. Significantly lower PIPP scores were seen in infants who received a pacifier compared to those who did not, while there was no significant difference between infants who received sucrose alone compared to those who received water alone. There was, however, a trend towards lower scores in the group who received both sucrose and a pacifier, although the extent was not significant. The authors speculate that a greater effect may have been seen if repeated doses of sucrose were used. Swaddling was also incorporated into the study undertaken by Rush et al20 who randomised 14 infants to comfort care, sucrose and NNS versus 16 infants to the control group who received no intervention. While unable to demonstrate a difference between the two groups, they suggest that lack of effect may be related to the length of the eye examination compared with other shorter procedures.20 Grabska et al found no benefit to the intervention (swaddling, sucrose and NNS) and they suggest that discomfort caused by lid retraction and scleral depression cannot be alleviated by sucrose, a relatively mild analgesic despite using weight adjusted doses from 0.12 to 0.48 g.21 Notwithstanding our reporting a positive effect from sucrose, swaddling and NNS, the reduction in pain scores was modest, and a 1.5–2 point reduction is unlikely to translate into clinically significant effects. We therefore concur with the findings of Grabska et al that future studies are required if optimal pain management is to be achieved.

In a recent review on the efficacy of sucrose doses, as little as 0.012–0.12 g (0.05–0.5 ml of 24% sucrose) was effective in reducing PIPP scores in term and preterm infants undergoing heel lance or venepuncture.15 While higher doses of sucrose up to 0.48 g (2 ml of 24% sucrose) have been more efficacious in studies of term infants, the optimal dose of sucrose remains to be determined and may be procedure specific. It is important to remember the sensitive nature of the eye and caution is advised when drawing conclusions from studies undertaken on procedures involving other areas of the body. In the consensus statement for the prevention and management of pain in the newborn, the recommended dose of sucrose for preterm infants is 0.1–0.4 ml of 12–24% sucrose and 2 ml for a term infant. On the basis of these recommendations, we decided on a dose of 0.2 ml of sucrose 24%. The consensus statement does encourage the combination of behavioural and pharmacological interventions such as comfort measures and a pacifier to give an additive or synergistic effect to sucrose.5

We chose to evaluate pain with the N-PASS for its multidimensional qualities. Hummel et al demonstrated the reliability and validity of this tool to assess both prolonged and acute pain and concluded that their research provides the beginning of evidencethat the N-PASS is a valid and reliable tool for assessing pain,23 24 and reported a strong correlation between the PIPP and the N-PASS, particularly for high scores.23 Interestingly, the pain scores reported by the Hummel group for acute pain (heelstick procedure) are much lower (mean (SD) 3.93 (2.30)) compared to the pain scores we observed during the screening examination for ROP for scleral indentation. While we recognise that there are differences in the procedures, it is concerning that the pain scores during our study remained high in the sucrose group. It is also worth noting that N-PASS scores increased over the duration of each individual eye examination from insertion of the speculum to use of the indentors, suggesting a more effective pain management strategy is required. Other strategies may include an increased dose or repeated doses of sucrose, an alternative local anaesthesia eye drop or a combination of both of these agents. These questions remain to be answered and require further research. Different methods of assessment, including avoidance of indentation or the use of the RetCam may be a less painful alternative for ROP screening.

There are some limitations to the study. These include failure to obtain a pre-examination N-PASS score as preparation of the infant for eye examination may influence the infant's response to the examination nor did we document sleep state and proximity of feeding time which could have an impact on pain and agitation scores. However, a standard approach was adopted to prepare infants for the ROP examination. Other limitations pertain to the procedure, where there were some differences noted in the holding and swaddling techniques for the screening procedure. Seven infants lost their soothers before the procedure was complete, limiting the potentially beneficial effects of NNS (three in the treatment group and four in the control group). There were also differences in the types of soothers used, and some were observed to be too small for the corrected gestational age of the infants in question. Poor light on four of the video recordings made accurate scoring difficult. The results of this study need to be interpreted in light of these limitations. However, the results do indicate that ROP screening is a painful procedure for the preterm infant and current pain reduction techniques are inadequate.


In conclusion, ROP screening is a painful procedure and pain responses increased during the screening examination for ROP. Sucrose combined with NNS and swaddling resulted in a modest reduction in behavioural and physiological pain scores recorded with N-PASS and is an easy strategy to adopt. However, alleviating pain in ROP screening remains a challenge.


The nursing staff of the Neonatal Department of the Coombe Womens and Infants University Hospital are acknowledged and thanked for their assistance in the eye examinations and administration of treatments according to the study protocol. Peter Duddy, the neonatal pharmacist, is thanked for his role in randomisation and in the preparation and dispensing of study solutions A and B. Maricar Publico RN, and Siobhan O Sullivan, medical student, are thanked for their roles as research assistants responsible for videotaping the eye examinations and Rose Akmad RN for IT support.



  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the Clinical Research Ethics Committee, National Maternity Hospital, Dublin.

  • Provenance and peer review Not commissioned; externally peer reviewed.