SciELO - Scientific Electronic Library Online

vol.50 issue3Keeping advanced seating services appointments in a Western Cape setting: A qualitative exploration of the experiences of carers of persons with cerebral palsyOccupational Therapy Association of South Africa position paper on vocational rehabilitation author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



Related links

  • On index processCited by Google
  • On index processSimilars in Google


South African Journal of Occupational Therapy

On-line version ISSN 2310-3833
Print version ISSN 0038-2337

S. Afr. j. occup. ther. vol.50 n.3 Pretoria Dec. 2020 



Therapists' consensus on an infant massage programme for high-risk infants from resource constrained contexts: a delphi study



Lauren Michelle PerksI; Gina RenckenII; Pragashnie GovenderIII

IB.OT (UKZN), M.OT (UKZN).; Occupational Therapist, Private Practice
IIB.OT (UP), M.OT (UFS).; Lecturer, Department of Occupational Therapy, School of Health Sciences, University of KwaZulu Natal
IIIB.OT (UDW), M.OT (UKZN), PhD (UKZN).; Associate Professor, Discipline of Occupational Therapy, Academic Leader: Research, School of Health Sciences, University of KwaZulu Natal





BACKGROUND: There are various infant massage programmes designed for infants born without complications, but there are limited options available for caregivers with high-risk infants from a resource-constrained context
AIM: This study aimed to develop a caregiver-initiated infant massage programme specific for high-risk infants in a resource-constrained context to facilitate the bonding and attachment process between this dyad
METHOD: A three-round Delphi study was conducted with purposively selected physiotherapists and occupational therapists with experience in the field of high-risk babies in a resource-constrained context. The design of the round one survey was informed by the available literature, and the subsequent two rounds of surveys were designed based on the prior rounds. Iteration of results was provided prior to the subsequent rounds
RESULTS: Attrition between rounds occurred from an initial 14 participants in round one to 8 participants in round three. Consensus on 19 massage strokes were achieved in addition to considerations for implementation of the programme with this population
CONCLUSION: The implication of attrition requires further study. Notwithstanding this, an infant massage programme for high-risk infants was developed with the inclusion of considerations for this dyad's social and medical context

Key words: Infant massage, Infant massage programme, High-risk infants, Delphi technique, Bonding




The birth of a high-risk infant is frequently the result of birth trauma1. It may result from a variety of factors such as delay in seeking medical attention during labour, non-initiation of antenatal care, late admission in pregnancy, as well as facility and management factors including inappropriate response to ruptured membranes, inadequate facilities and lack of transport and ventilator and resuscitation equipment2. These infants then subsequently spend a significant amount of time in hospital3.

High-risk infants often do not respond appropriately to caregiver's cues which may lead to the primary caregiver feeling out of touch with their infant as well as in the infant feeling overwhelmed4. The future anticipated special needs that the infant may require, as well as the possibility of losing the medically fragile infant, results in negative emotions in the caregiver3. In addition to the underlying causes of the infant being born high-risk, there are also a variety of cultural and environmental factors which inhibit this bonding process such as the death of the infant's parent/s and resource-constrained living conditions5,6. Alongside these social factors is that breastfeeding, which is a source of bonding between the dyad7, is often a challenge with infants who are born high risk. This is due to several risk factors such as the medical condition of the infant, the length of hospitalisation, the distance of the infant from the maternal residence and the type of delivery8. These medical and social factors impact on the bonding and attachment process between the dyad. Positive maternal or caregiver touch is essential as this contributes to the infant's feeling of safety, as well as its cognitive and social development9,10.

The purpose of this study was to establish consensus on infant massage strokes deemed appropriate for high-risk infants, that can be included in a programme initiated by primary caregivers in their context, towards the aim of bonding and attachment in the mother-infant dyad.



A "high-risk infant" is an infant who requires a greater degree of monitoring and care compared to a healthy infant who was born full term11. The St Apollinaris high-risk baby policy has been included as a reference of high-risk infants12. This policy was developed due to rural district therapists experiencing limitations in their rehabilitation care owing to a lack of resources and policies13, and so the rehabilitation team at this facility designed a guideline for the management of these clients. Early intervention rehabilitation teams are generally inclusive of occupational therapists (OT), physiotherapists (PT), speech therapists (SLT) and audiologists (Au), with OT and PT focusing on sensory-motor development, SLT concentrating on feeding and early communication and Au focusing on hearing in this context according to the focus of their undergraduate training. High-risk infants present as having one or more of the following910; born with a low birth weight (i.e. <1800g), preterm birth (i.e. <37 weeks), neonatal encephalopathy resulting in low APGAR scores of 7/10, infants requiring mechanical ventilation for more than 24 hours, infants presenting with seizures post-delivery, infants presenting with infections (e.g. meningitis, syphilis, CMV, toxoplasmosis, herpes simplex), infants presenting with major morbidities (i.e. chronic lung disease, intraventricular haemorrhage, peri-ventricular leucomalacia) hyperbilirubinemia where the infant requires a blood transfusion, major malformations and abnormal neurological examination at discharge, cardiopulmonary distress, drug withdrawal syndrome requiring monitoring, suspected or proven neonatal sepsis and metabolic abnormalities ie. anaemia, hypoglycemia, hypocalcemia.

In their first year, many infants who are born prematurely or with a low birth weight present with more significant deficits in mental function as well as in neuromusculoskeletal and movement-related dysfunctions compared to children born at an average birth-weight or full term14. Approximately 50-70% of children born prematurely present with dysfunctions such as cognitive deficits, learning disabilities, attention problems, behavioral problems, and neuropsychologi-cal deficits when they reach school age15. These high-risk infants have a lower threshold for sensory input than other infants and in turn experience difficulty with tolerating physical handling and interaction16. The infant who has been in the neonatal unit has also been separated from her/his mother and so may have increased levels of dysregulation due to the clinical environment, the loss of opportunity to co-regulate in close physical contact with his/her mother and the limited experience of regular positive touch and nurturing physical contact and care, which will require extra sensitive handling and if this infant has a special need then this will need to be coupled with emotional support17.

Primary caregivers are also faced with various extraneous factors such as a shortage of financial resources, limited transport and under-resourced hospitals18. These social risk factors affect the infants' cognitive and behavioural development19 and also make it difficult for the sustainability of therapy programmes and medical intervention13. Unfortunately, these caregivers are further faced with long queues at hospitals, inefficient administration procedures where patients have to wait lengthy periods for their files and have a perception of disrespectful and "uncaring" attitudes from the staff at these hospitals18. Clients often have to be put on long waiting lists for therapy appointments and have poor access to equipment and assistive devices due to a lack of district funding18.

These social factors (alongside the medical concerns of the infant, difficulty with breastfeeding and the emotional experience of the primary caregiver) hinder the bonding and attachment process. Infant massage therapy, in conjunction with other early intervention programmes, provides the supportive environment needed for attachment and bonding and appropriate stimulation for the infant and is used to optimise the infant's sensory experience which, in turn, improves development and functional outcomes20. The initial primary goal of intervention with preterm infants (after ensuring survival) is to prevent further exposure to dangerous environmental stimulation21. Additional goals of the intervention are to encourage and create a supportive environment which provides appropriate stimulation for the infant and which mirrors that provided in utero22. Other goals include providing support and strategies to the families of the infant so that they can become competent and confident in optimising their child's development and to empower them to be an active participant in their child's recovery programme23. It is therefore crucial that to increase caregiver competence and sustainability of the programme that they are taught the stimulation techniques to use with their children23.

Bonding and attachment are essential developmental and neu-robehavioural processes occurring between the mother and infant, starting in pregnancy and increasing in intensity of action in the early weeks of life. The infant develops security by having its psychological needs met through nurturing touch and proximity to the primary caregiver24,25,26.

Infant massage consists of gentle, slow stroking of each part of the body through tactile stimulation by human hands27. This nurturing touch is beneficial in that it contributes to the infant's optimal development both psychologically and physically through stimulation, relaxation, relief and interaction28 through enhancing the parent-child relationship and strengthening the emotional attachment29. Further benefits include improvements in social interaction, emotional functioning and soothing temperaments and sleep organisation in full-term infants30.

A further benefit in the attachment process is that infant massage creates a purposeful and satisfying interaction between the dyad through a constructive and positive activity31. This dedicated infant massage time is also a protected time between the mother and the infant31 which is relaxing for both amongst the many other demands which she is required to do during the infant's waking times such as nappy changing, bottle or breastfeeding, trying to decipher the different cries, alongside a lack of sleep as well as recovering from the physical birthing process if she is the mother28. Participation in infant massage has been shown to decrease depression, anxiety and stress experience in mothers, and improve their perceptions of self-efficiency in caregiving29,30,32,33.

Early identification and intervention is recommended for all infants and toddlers who are presenting with difficulties as there is much evidence that this intervention has a positive effect on the development of the child throughout her/his life34. This is evident in neuroplasticity where young animals' and humans' exposure to sensory information and experiences from their environment results in growth and development as a result of specialisation and maturation of the nervous system34.



Study design: A three-round Delphi technique (Figure 1 p74) was used in this study with the aim of programme development35,36 which was based on the available time and funding as well as the consideration of participation fatigue37,38. The strength in this method lies in the stability of the group consensus rather than on individuals' opinions39. Participants were required to respond on a survey comprising Likert scales40 via a group facilitation process38 to provide a holistic overview41.



The study included the following phases:

Literature review (Phase 1): The initial phase of the study focused on a comprehensive literature review on the high-risk infant, the contextual considerations for this dyad from the resource-constrained context as well as on infant massage to form a basis upon which the initial online survey could be created38.

Delphi Process Round One: Item Generation (Phase 2): Purposive sampling was used for this Delphi study42 where specific participants were considered expert based on specific selection criteria35,43. Following the recruitment of the sample, the initial survey was uploaded online onto The format and content of this initial survey were based upon the first authors' experience, her training as an Infant Massage Instructor through the International Association of Infant Massage (IAIM), and a comprehensive literature review38,44. Items for inclusion were based on consideration of the infant massage strokes as taught by the IAIM and the potential effect on the central nervous system and arousal levels of the high risk baby16,34. The fast paced, tickly and alerting massage strokes could be disorganising to the arousal levels of the high-risk infant, and as such were excluded. The slower, deeper-pressured, calming and organizing strokes were favoured for inclusion16,34,45.

Delphi Process Round 2: Consensus on item generation (Phase 3): The list of infant massage techniques formulated from the initial survey and from the additions from the expert participants in round one was pared down in the round two survey46. The three-point Likert scales from survey one were collapsed into a two-point scale where "Important" and "Nice to have" were reduced to "Important"; and "Not important" remained as is. Items which were infrequently rated by participants in round one were still included in this survey to gain clarity37. Iteration was not provided at this stage as round one and round two aimed to establish consensus on similar items.

Delphi Process Round 3: Consensus (Phase 4): The third round included iteration, where an organised summary of the previous round one and round two results was distributed to the participants so that they were aware of the range of opinions and positioning of the programme based on these opinions35. Following this feedback, the participants were provided with the round three survey based on the findings of round two36. The two-point Likert scale from round two was maintained with a change in terminology from "Important" to "Agree" and "Not important" to "Disagree". The list of items in survey three was reduced by including only those items that achieved a priori consensus threshold of >70%47. This threshold is concurrent with previous studies which was noted as an acceptable threshold48,49. Development of the Programme (Phase 5): This phase comprised the final collation and development of the infant massage programme and was based on the consensus obtained in Round Three.

Sampling and sampling techniques:

Purposive sampling was employed in the selection of experts for this study based on specific selection criteria36,43. This approach was used to ensure that a homogenous sample was maintained for this study which would yield satisfactory results even though the sample size is smaller40. The following selection criteria were used in this sampling approach which was based on the qualities of the expert panel rather than on the sample size37. The selection criteria based on Akins and colleagues42 and Bruce and colleagues43 included, (i) having a bachelor degree in occupational therapy or physiotherapy, (ii) knowledge and practical engagement in intervention with high-risk infants for a minimum of one year, (iii) knowledge of and practical engagement in community-based clinics or rural district hospitals for a minimum of one year, (iv) capacity and willingness to participate in and contribute to the development of an infant massage programme for high-risk infants from resource-constrained contexts, (v) good written communication skills in English and (vi) adequate computer literacy and computer skills38. Participants who did not meet the selection criteria were excluded from the study. To maintain the rigour of the results from the Delphi technique, this study aimed for a response rate of 70% throughout the study38. Therefore, for this study to obtain, what was considered to be a small sample size of n=l5 respondents, the study should have begun with n=23 participants43. This selection of a homogenous small sample size was based on the theory that there are marginal benefits of a larger sample size above a certain threshold as a large sample size may increase the reduction in group error, but the management of the data analysis and Delphi technique may become cumbersome40.

Demographic Profile of Participants

A total of 29 participants began the survey and completed the demographic section (Table I p74). Of these 14 participants completed the round one survey, 13 completed round two, and 8 participants completed round three.



Table 2


Table 3


Validity and Reliability: The authors were aware of the concepts of validity and reliability during the data collection and data analysis processes. Validity was maintained through the appraisal and piloting of the online surveys prior to implementation of the surveys by two expert professionals who were not participants in the study37,38 as well as the inclusion of selection criteria which meant the participants were representative of the field of expertise50 and their consensus contributed towards inter-rater reliability. Concurrent validity was increased through the inclusion of three successive rounds31. Reliability was maintained as the data from the comments in each survey guided the subsequent surveys, and so the tool was based on the responses of the participants and not on that of the authors.

Data analysis: Descriptive statistics assisted in determining the level of consensus by a summary of the data via central tendencies and standard deviations38. The mean assisted in establishing the group opinion and the standard deviation assisted in assessing the level of disagreement on each item51. The qualitative responses in the open-ended 'comments' section were analysed using content analysis by grouping similar items together which had several different terms but which appeared to be the same issue to provide one universal description38. The content analysis and the statistical summaries formed the quantitative basis for developing the online survey for subsequent rounds52. Ethical considerations: Ethical clearance was obtained from the Humanities and Social Sciences Research Ethics Committee at the University of KwaZulu-Natal (ethical clearance number HSS/0533/017M). Participants volunteered participation via an informed consent document and were free to withdraw without prejudice. Confidentiality was ensured by the participants being given an option to include their email address for further correspondence if they chose to. Responses without this contact information were included and remained anonymous.



Delphi Round One Survey (n=14): Consensus was obtained on all massage strokes, which were included in the initial survey and obtained from the literature review conducted by the authors in phase one of the study38. Participants who commented noted that specific strokes should be included in the programme. One participant indicated the following when commenting on the lower limbs "hug and glide provides much more proprioceptive input" and which was confirmed by another participant who also noted "hug and glide (squeeze and twist)". This observation was also noted by a third participant who recommended the inclusion of "deep compression along long bones - from hip to knee and knee to heel. Calming, well tolerated in supine and good prep for therapeutic activities" which has similar characteristics to the hug and glide massage stroke. Another participant suggested the inclusion of "swedish milking". A further recommendation was "touching each toe separately". When commenting on the stomach, participants recommended "circular sweeping motion across the stomach in a clocking motion to mimic bowel movements" and "clockwise circular sweep" as well as "along large colon- up from right hip, across from right to left and then down to aid digestion and soothe cramps". It was recommended by one participant that "butterfly" be included as a massage stroke for the chest which is a technique which stimulates and deepens breathing in the infant28. For the upper limbs, two participants indicated that "hug and glide" should be included, one participant recommended the inclusion of "swedish milking" and two participants recommended "finger rolling" or "rolling each finger separately". One participant suggested "small circles around the jaw with fingertips" as well as a further "maybe 'relax the jaw' or jaw clenching".

It was noted in this survey that the average recommended duration for an infant massage routine is 10 minutes as well as the repetition of each stroke three times. There was consensus obtained on excluding certain oils/creams as recommendations for this population and these included "cold pressed sunflower oil", "sunflower oil" and "olive oil".

Delphi Round Two Survey (n=13): All strokes which had obtained consensus in round one were confirmed with a subsequent consensus in round two. There was also a further consensus on six additional strokes added in round two which were based on the comments gathered in round one. However, there was a non-consensus noted for the added sub-items "hug and glide" and "swedish milking" for the lower limbs as well as "butterfly" for the chest and "hug and glide" for the upper limbs.

One of the participants recommended "a cuddle" as a conclusion consideration in survey one and this was confirmed in survey two by obtaining consensus. Participants noted in survey one that further beneficial sensory input could include "encourage skin to skin". The recommendation was also to include "facial expressions" which were confirmed with a result of consensus in survey two. Further additions included that participants recommended the inclusion of "vaseline" and "coconut oil" for use during the massage routine. The use of "coconut oil" did not obtain consensus and "vaseline" achieved consensus of not being important.

It was also recommended by participants in survey one that this routine be implemented "after bath time when cream should be applied" and "I would suggest after bath time", "when baby is happy and calm" and "so long as baby is calm-alert state" as well as "when the baby is in the quiet alert state". These were confirmed by consensus in survey two.

Delphi Round Three Survey (n=8): This survey confirmed the consensus strokes from the round one and round two surveys. However, there was a further non-consensus on the four sub-items discussed above in round two on "hug and glide" and "swedish milking" for the lower limbs as well as "butterfly" for the chest and "hug and glide" for the upper limbs. It was confirmed in this survey that this programme would be able to be completed in the recommended limit of 10 minutes. Participants suggested that should the primary caregiver be unable to complete the massage routines on the whole body in 10 minutes then they should, "complete one area, and end in the above-mentioned routine (conclusion considerations). The next session start at a missed area or only do resting hands on a previously completed area", "can complete routine later", "at the next massage, the caregiver could start where they left off' and confirmed with "start the next time with the body parts she did not do". The recommendation of completing the routine three times a day if possible did not obtain a final consensus in survey three with participants stating; "build-up to 3x per day" and "this will give mothers a number to aim for, which may get in the way, rather than promoting bonding in this dyad where separation post-birth has been necessary due to medical reasons". A further non-consensus on the recommendation of the use of "coconut oil" was noted in survey three, as in survey two. This is confirmed with comments from the participants of "new research showing coconut oil is not as good as previously thought", "I don't think this is easily accessible in this context", "may not be easily accessible" and "may be costly for population group". "Vaseline", a commercially produced mineral oil, was excluded in survey three based on its consensus of not being relevant in survey two. However, one participant did note in survey two that "although Vaseline may not be the best choice- it is most commonly used by caregivers in resource constrained settings". The inclusion of "aqueous cream" as a recommendation obtained consensus. Non-consensus was also noted in the final positioning sub-item.



Massage strokes for high-risk infants

High-risk infants have a lower threshold for sensory input. They may experience difficulty tolerating tactile input from handling and interaction16 which is an imbalance between discriminative interpretation and the need for a defense response53. The use of deep touch pressure and the resultant activation of the dorsal column medial lemniscal system (DCML) blocks the protective response to touch and so, in turn, results in decreased levels of distractibility53 and pain relief54. Therefore, the massage stroke must have pressure to be able to gain these benefits of DCML activation as light touch, activation of the anterolateral system, can be aversive for the infant and result in protective responses and strong emotional reactions53.

Resting hands is a simple holding technique which activates the DCML system and is especially beneficial with a medically fragile, premature or colicky baby4. This is done by warming of the hands, placing them on the infant and then letting the hands go heavy and warm4. Due to the nature of this technique, it helps the dyad to become more conscious of the experience of touch4 and improve the infant's body awareness of these areas through the activation of the DCML system53. This technique will be implemented in a massage stroke with the lower limbs, upper limbs, chest and the back. Deep pressure will further be added to the ankles by circulating the ankle with the caregiver's thumb as this will further stimulate growth and development and joint flexibility28. This deep pressure will also be added to the infant's toes by touching each toe separately. However, caution must be given when massaging the feet as commented by a participant that "These babies often have quite tender feet from all the prodding and poking in NICU". The infant's fingers may also not have been mobilised very much in the unit and so the individual will be required to try and open the infant's hand gently with their thumb and press into the palm as well as roll each of the fingers. These strokes are beneficial as they are "very useful to prevent contractures and increase input into palm".

These high risk infants who have been in the neonatal unit or in kangaroo mother care have also not been mobilised very often or picked up unless it was medically necessary. Indian milking of the upper and lower limbs is included in the programme as it relaxes the limb and encourages blood flow to the infant's feet or hands28 and so encourages normal movement patterns.

The high-risk infant may also have experienced gastro-intestinal concerns due to the nature of their medical condition and so may be presenting with constipation. Through massage of the abdomen area of the infant, the gas and intestinal matter is moved toward the bowel as these strokes assist in mimicking the movement of the bowel and this will assist with alleviating gas and discomfort4. Three strokes will be done on the stomach area which include a sweeping motion with deep pressure in a clockwise direction, circular sweeping motion from hip to hip and deep pressure movement across the bowels.

High-risk infants may have been overstimulated on their face during their stay in the unit in the anterolateral system, as commented; "babies who have had NGTs hate their faces being touched or heads held firmly - beware of fear response" and "babies in the NICU faces are usually overstimulated due to be oscillated, ventilated, tube feeding through the nose. Also the regular suction and wiping of the face". This may also be as a result of the high receptor density on the face which is very sensitive to touch and so can result in overstimulation54. However, it is still recommended that these two strokes are conducted on the face where deep pressure is applied around the perimeter of the infant's face as well as applying small circles around the infant's jaw. These strokes are beneficial in that they alleviate tension in the jaw, support chewing and speech28.

Touch of the infant's back activates a larger receptor field where the receptor density is low and so is less sensitive to touch54 and thus has a calming effect for the infant. The massage stroke to be done on the back will require deep pressure combing down the lateral sides of the spine from superior to posterior. The added benefit of this stroke is that it will also stimulate body awareness in the infants as their backs are not an area that is frequently touched during their daily routine28.

The study highlighted a non-consensus on the inclusion of "hug and glide" for the upper and lower limbs where the stroke involves. This massage stroke motion involves a twisting motion of the limb which would activate the anterolateral system and so result in the dysregulating effects of activating this system with this sensitive infant54. Therefore, these strokes will be excluded from the programme to avoid this aversive response.

"Swedish milking" of the lower limbs was also a stroke that did not obtain consensus in this study. The purpose of this stroke is to aid circulation from the feet and back towards the heart4 and it is for this purpose that it will be included in the programme.

The final non-consensus was regarding the "butterfly" technique on the chest. This technique is beneficial for the infant in that it stimulates and deepens breathing28 which is beneficial for this high-risk infant who often has a history of breathing difficulties due to preterm birth or low birth weight and so it will be included in the programme.

Inclusion of other sensory stimulation during massage routine

Consensus from this study highlighted the inclusion of contingent tactile stimulation, which is massage conducted alone or in conjunction with visual and/or auditory stimulation55. The participants agreed on the inclusion of eye contact between the dyad, skin to skin contact, as well as the primary caregiver singing softly and quietly and talking to the infant whilst showing facial expressions27. The benefits of these are that they assist in increasing vocalisations, smiles, eye contact and approach behaviours in the infant who is possibly experiencing difficulty with responding appropriately to the mother due to the nature of her/his medical condition56. A further benefit is that by the primary caregiver responding to different facial expressions and vocalisations from the infant it will further assist the infant in becoming familiar with the caregiver's tone of voice, physical prompts, and cues, thus supporting bonding and integration of a variety of social-interactive experience23. The infant also uses the caregiver's emotional expressions to help make sense of their environment23. The inclusion of skin to skin contact is beneficial in helping the infant to remain calm and relaxed, which is in addition to the calming effect of the massage routine47.

Duration of massage routine

A consensus was not obtained in this study about the recommendation of repeating the massage routine three times a day. However, with the support from Field's protocol57 as well from a participant's comment of build up to 3x per day it is recommended that this routine be conducted up to three times a day if the dyads are comfortable with this. This study also obtained consensus on the inclusion of infant massage before the infant's bed time or after bath time as this will help with obtaining the benefits of massage on sleep58.

However, whilst keeping these recommendations as a priority it is further considered that primary caregivers initiate that infant massage routine when both they and the infant are in a state which is accepting and ready for the process as it is remembered that this dyad is also faced with many other responsibilities and contextual factors in these early days.

The massage routine should be concluded by integrating the infant's body by using resting hands on each part of the body that has been massaged and naming the areas. This should be in addition to saying thank you to the infant for accepting the massage and dressing the infant. Finally, the dyad should share a cuddle as this is a technique which will also aid in the bonding and attachment process through nurturing touch59 and is a means of expressing emotions without words45.

Use of oils/creams

It is recommended that the massage routine is initiated with the use of oils/creams; such as non-perfumed baby cream or aqueous cream. However, it is essential to note that the substance used is non-occlusive and does not block the skin pores but allows the skin to breathe60. The substance should also be safe and suit the baby's delicate skin, and the ingredients of the formula should have been tested for their potential to cause contact sensitivity61. The use of coconut oil with this population did not obtain consensus on its inclusion or exclusion. However, research has shown the positive effects of preterm weight gain with the use of coconut oil over mineral oils61 due to the transcutaneous absorption of the vegetable oil through the thinner and more vascular skin of the infant62. Although there are clear benefits of using this oil it "may not be easily accessible" and "may be costly" for this population in their resource-constrained context. It must also be noted that certain cosmetic products and baby creams such as Vaseline and Johnsons baby products are also sensitising agents as they contain formaldehyde and may contribute to allergic contact dermatitis (ACD) and so should be avoided when conducting infant massage63.



There was participant attrition between subsequent Delphi rounds in this study. The potential reasons for this attrition from the initial group of n=29 participants; could be attributed to, (i) therapists not presenting with an interest in this area of practice once they started the survey36, (ii) having insufficient current knowledge of the topic36, and (iii) insufficient time available to them to be part of the research process36. The decrease observed in the response rate over the three rounds may have been due to the participants being busy and experiencing difficulty to participate fully40. Previous studies have noted that as the number of rounds increase and the effort required by the participants, that a drop in the response rate is often observed38,39,40,42,48,48-50. There is limited evidence of reliability when using the Delphi technique38 and the primary author was unable to test and retest the outcomes of the surveys with a different sample due to time constraints.



A three-round Delphi technique completed by physiotherapists and occupational therapists guided the development and consensus on an infant massage programme for high-risk infants from resource-constrained contexts. It is recommended that a user-friendly format is designed which can be provided to caregivers from therapists during their intervention session as a home programme. This programme and its relevant considerations are available and accessible for therapists to use to train these primary caregivers from this context, to implement these massage techniques with their high-risk infants.

A collaborative goal of the intervention of physiotherapists and occupational therapists working with high-risk infants is to encourage and create a supportive environment which provides appropriate stimulation for the infant and which mirrors that provided in utero22. As therapists, the focus needs to be on including strategies in the early intervention rehabilitation programme that optimises the infant's sensory experience and thus improve developmental and functional outcomes20 but one that is still appropriate and accessible for this population. The aim of this research study was achieved in providing an infant massage guideline that can be utilised and accessed in this setting to achieve that collaborative goal.



1. Hynan MT. The Perinatal Posttraumatic Stress Disorder (PTSD) Questionnaire. In: Evaluating stress: A Book of Resources. CP Zalaquett & RJ Wood, RJ, Editors. London, United Kingdom:The Scarecrow Press Incorporated; 1997: 199-200.         [ Links ]

2. Lloyd LG, De Witt W. Neonatal mortality in South Africa: How are we doing and can we do better? South African Medical Journal [Internet]. South African Medical Association NPC; 2013 Jul 2;103(8): 518.        [ Links ]

3. Minde K. Prematurity and Serious Medical Illness in Infancy: Implications for Development and Intervention. In: Handbook of Infant Mental Health. C Zeanah Jr, Editor. New York: Guilford Press; 1995:87-105. New York: Guilford Press; 1995.<240::AID-IMHJ2280160310>3.0.CO;2-Z        [ Links ]

4. McClure V. Infant Massage: A Handbook for Loving Parents. Fourth Edition. Croydon, United Kingdom:Bantam; 2017.         [ Links ]

5. Anderson B, Phillips H. Trends in the percentage of children who are orphaned in South Africa: 1995-2005. Statistics South Africa; 2006.         [ Links ]

6. UNICEF Statistical data on the status of children aged 0-4 in South Africa. 2007.        [ Links ]

7. Else-Quest NM, Hyde JS, Clark R. Breastfeeding, Bonding, and the Mother-Infant Relationship. Merrill-Palmer Quarterly [Internet]. Project Muse. 2003; 49(4): 495-517.        [ Links ]

8. Dall'Oglio I, Salvatori G, Bonci E, Nantini B, D'Agostino G, Dotta A. Breastfeeding promotion in neonatal intensive care unit: impact of a new program toward a BFHI for high-risk infants. Acta Paediatrica [Internet]. Wiley. 2007 Nov; 96(11): 1626-31.        [ Links ]

9. Weiss SJ, Wilson P Hertenstein MJ, Campos R. The tactile context of a mother's caregiving: implications for attachment of low birth weight infants. Infant Behavior and Development [Internet]. Elsevier BV 2000 Jan; 23(1): 91-111.        [ Links ]

10. O'Brien M, Lynch H. Exploring the role of touch in the first year of life: Mothers' perspectives of tactile interactions with their infants. British Journal of Occupational Therapy. 2011 Mar; 74(3): 129-36.        [ Links ]

11. Elzouki AY Harfi HA, Nazer HM, Stapleton FB, Oh W, Whitley RJ, editors. Textbook of Clinical Pediatrics. Springer Berlin Heidelberg; 2012.        [ Links ]

12. Boyd J. High risk baby programme: Rehab 01. KwaZulu-Natal Department of Health. 2015. (unpublished)        [ Links ]

13. van Stormbroek K, Buchanan H. Community Service Occupational Therapists: thriving or just surviving? South African Journal of Occupational Therapy [Internet]. Academy of Science of South Africa. 2016; 46(3).        [ Links ]

14. Beck S, Wojdyla D, Say L, Pilar Bertran A, Meraldi M, Harris Requejo J, et al. The worldwide incidence of preterm birth: a systematic review of maternal mortality and morbidity. Bulletin of the World Health Organization [Internet]. WHO Press. 2010 Jan 1; 88(1): 31-8.        [ Links ]

15. Sullivan MC, Hawes K. A decade comparison of preterm motor performance at age 4. Research in Nursing & Health [Internet]. Wiley. 2007; 30(6): 641-54.        [ Links ]

16. Miller MQ, Quinn-Hurst M. Neurobehavioral Assessment of High-Risk Infants in the Neonatal Intensive Care Unit. American Journal of Occupational Therapy [Internet]. AOTA Press; 1994 Jun 1; 48(6): 506-13.        [ Links ]

17. Pick R. Understanding Childhood. London: Child Psychotherapy Trust; 1999.         [ Links ]

18. Saloojee GM, Rosenbaum PL, Stewart AV Using caregivers' perceptions of rehabilitation services for children with Cerebral Palsy at public sector hospitals to identify the components of an appropriate service. South African Journal of Physiotherapy [Internet]. 2011; 19;67(3): 35-40.        [ Links ]

19. Lee VE, Brooks-Gunn J, Schnur E, Liaw F-R. Are Head Start Effects Sustained? A Longitudinal Follow-Up Comparison of Disadvantaged Children Attending Head Start, No Preschool, and Other Preschool Programs. Child Development [Internet]. 1990 Apr; 61(2): 495.        [ Links ]

20. Symington AJ, Pinelli J. Developmental care for promoting development and preventing morbidity in preterm infants. Cochrane Database of Systematic Reviews [Internet]. Wiley; April 2006; 19.        [ Links ]

21. Anand KJS, Scalzo FM. Can Adverse Neonatal Experiences Alter Brain Development and Subsequent Behavior? Biolo Neonate. 2000; 77(2): 69-82.        [ Links ]

22. Westrup B, Kleberg A, von Eichwald K, Stjernqvist K, Lagercrantz H. A Randomized, Controlled Trial to Evaluate the Effects of the Newborn Individualized Developmental Care and Assessment Program in a Swedish Setting. Pediatrics [Internet]. American Academy of Pediatrics. 2000 Jan 1; 105(1): 66-72.        [ Links ]

23. Colyvas JL, Sawyer LB, Campbell PH. Identifying Strategies Early Intervention Occupational Therapists Use to Teach Caregivers. American Journal of Occupational Therapy [Internet]. 2010; 64(5): 776-85.        [ Links ]

24. Brazelton TB & Nugent JK. The Neonatal Behavioral Assessment Scale: Fourth Edition. United Kingdom: Mac Keith Press; 2011.         [ Links ]

25. Rutter M. Clinical Implications of Attachment Concepts: Retrospect and Prospect. Journal of Child Psychology and Psychiatry [Internet]. 1995 May; 36(4): 549-71.        [ Links ]

26. Atzil S, Hendler T, Feldman R. Specifying the Neurobiological Basis of Human Attachment: Brain, Hormones, and Behavior in Synchronous and Intrusive Mothers. Neuropsychopharmacology. 2011 Aug 31; 36(13): 2603-15.:        [ Links ]

27. Vickers A, Ohlsson A, Lacy J, Horsley A. Massage for promoting growth and development of preterm and/or low birth-weight infants. Cochrane Database of Systematic Reviews [Internet]. 2004;19.        [ Links ]

28. Simpson R. Baby massage classes and the work of the International Association of Infant Massage. Complementary Therapies in Nursing and Midwifery [Internet]. 2001 Feb; 7(1): 25-33.        [ Links ]

29. Shipwright S, Dryden T. Paediatric massage: an overview of the evidence. Focus on Alternative and Complementary Therapies [Internet]. 2012 Mar 30; 17(2): 103-10.        [ Links ]

30. Field TM. Massage therapy effects. American Psychologist [Internet]. American Psychological Association; 1998; 53(12): 1270-81.        [ Links ]

31. Clarke CL, Gibb C, Hart J, Davidson A. Infant massage: developing an evidence base for health visiting practice. Clinical Effectiveness in Nursing [Internet]. 2002 Sep; 6(3-4): 121-8.        [ Links ]

32. Karimi F Abolhassani M, Ghasempour Z, Gholami A, Rabiee N. Comparing the effect of kangaroo mother care and massage on preterm infant pain score, stress, anxiety, depression, and stress coping strategies of their mothers. International Journal of Pediatrics. 2020.        [ Links ]

33. Mousavi S, Rostami S, Haghani H, Borimnejad L. Effect of mother's infant massage on neonatal weight and perceived self-efficacy of mothers with low birth weight infants. Koomesh. 2020; 22(2): 237-43. Available from:        [ Links ]

34. Smith-Roley S. Ayres Sensory Integration for Infants and Toddlers. The South African Institute for Sensory Integration. 2016: 1-14. (unpublished)        [ Links ]

35. Hsu C, Sandford BA. The Delphi technique: making sense of consensus. Practical Assessment, Research, and Evaluation. 2007; 12(1): 10.        [ Links ]

36. du Plessis E, Human SP The art of the Delphi technique: highlighting its scientific merit. Health SA Gesondheid [Internet]. 2007 Nov 17; 12(4).        [ Links ]

37. Powell C. The Delphi technique: myths and realities. Journal of Advanced Nursing [Internet]. 2003 Feb; 41(4): 376-82.        [ Links ]

38. Hasson F Keeney S, McKenna H. Research guidelines for the Delphi survey technique. Journal of Advanced Nursing [Internet]. 2000 Oct; 32(4): 1008-15.        [ Links ]

39. Adler M & Ziglio E. Gazing into the Oracle: The Delphi Method and its Application to Social Policy and Public Health. Bristol: PA: Jessica Kingsley Publications. 1996.         [ Links ]

40. Skulmoski GJ, Hartman FT Krahn J. The Delphi Method for Graduate Research. Journal of Information Technology Education: Research. Informing Science Institute. 2007; 6: 001-21.        [ Links ]

41. Mason j. Qualitative Researching. Thousand Oaks, USA: Sage Publications. 1996.         [ Links ]

42. Akins RB, Tolson H, Cole BR. Stability of response characteristics of a Delphi panel: application of bootstrap data expansion. BMC Medical Research Methodology [Internet]. 2005 Dec; 5(1).        [ Links ]

43. Bruce J. The use of experts and their judgments in nursing research: an overview. Curationis [Internet]. 2008 Sep 28; 31(4).        [ Links ]

44. Kerlinger F. Foundations of Behavioral Research. New York: Holt, Rinehart, and Winston. Inc. 1973.         [ Links ]

45. Simpson R. Baby massage classes and the work of the International Association of Infant Massage. Complementary Therapies in Nursing and Midwifery [Internet]. 2001 Feb; 7(1): 25-33.        [ Links ]

46. Schmidt RC. Managing Delphi Surveys Using Nonparametric Statistical Techniques. Decision Sciences [Internet]. 1997 jul; 28(3): 763-74.        [ Links ]

47. Cremers D, Rosenhahn B, Yuille AL, Schmidt FR, editors. Statistical and Geometrical Approaches to Visual Motion Analysis. Lecture Notes in Computer Science [Internet]. Springer Berlin Heidelberg; 2009.        [ Links ]

48. Graham B. Delphi as a method to establish consensus for diagnostic criteria. Journal of Clinical Epidemiology [Internet]. 2003 Dec; 56(12): 1150-6.        [ Links ]

49. Naidoo P Joubert RWE. Consensus on Hypotonia via Delphi Process. The Indian journal of Pediatrics [Internet]. 2013 May 17; 80(8): 641-50.        [ Links ]

50. Goodman CM. The Delphi technique: a critique. Journal of Advanced Nursing [Internet]. 1987 Nov; 12(6) :729-34.        [ Links ]

51. Greatorex j, Dexter T. An accessible analytical approach for investigating what happens between the rounds of a Delphi study. Journal of Advanced Nursing [Internet]. 2000 Oct ;32(4): 1016-24.        [ Links ]

52. McKenna HP The Delphi technique: a worthwhile research approach for nursing? journal of Advanced Nursing [Internet]. 1994 jun; 19(6): 1221-5.        [ Links ]

53. Ayres AJ. Sensory Integration and Learning Disorders. Los Angeles: Western Psychological Services; 1972.         [ Links ]

54. Bundy AC, Lane SJ, Murray EA. Sensory Integration Theory and Practice: Second Edition. Philadelphia: F.A. Davis Company; 2002.         [ Links ]

55. Peláez-Nogueras M, Field T, Gewirtz jL, Cigales M, Gonzalez A, Sanchez A, et al. The effects of systematic stroking versus tickling and poking on infant behavior. Journal of Applied Developmental Psychology [Internet]. 1997 Apr; 18(2): 169-78.        [ Links ]

56. Bystrova K, Widström A-M, Matthiesen A-S, Ransjö-Arvidson A-B, Welles-Nyström B, Wassberg C, et al. Skin-to-skin contact may reduce negative consequences of "the stress of being born": a study on temperature in newborn infants, subjected to different ward routines in St. Petersburg. Acta Paediatrica [Internet]. 2007 Jan 2; 92(3): 320-6.        [ Links ]

57. Field TM, Schanberg SM, Scafidi F Bauer CR, Vega-Lahr N, Garcia R, Nystrom J, Kuhn CM. Tactile/kinesthetic stimulation effects on preterm neonates. Pediatrics. 1986 May 1; 77(5): 654-8.         [ Links ]

58. Agarwal KN, Gupta A, Pushkarna R Bhargava SK, Faridi MM, Prabhu MK. Effects of massage & use of oil on growth, blood flow & Sleep. Indian j Med Res. 2000 Dec; 112: 212-7.        [ Links ]

59. Schneider EF. The power of touch: Massage for infants. Infants & Young Children [Internet]. Infants anf young Chgildren. 1996 jan; 8(3): 40-55.        [ Links ]

60. Dhar S. Newborn skin care revisited. Indian Journal of Dermatology [Internet]. 2007;52(1): 1.        [ Links ]

61. Sankaranarayanan K, Mondkar jA, Chauhan MM, Mascarenhas BM, Mainkar AR, Salvi RY Oil massage in neonates: an open randomized controlled study of coconut versus mineral oil. Indian pediatrics. 2005 Sep 1; 42(9): 877.        [ Links ]

62. Nachman RL, Esterly NB. Increased skin permeability in preterm infants. The journal of Pediatrics [Internet]. 1971 Oct; 79(4): 628-32.        [ Links ]

63. Moennich JN, Hanna DM, Jacob SE. Formaldehyde-Releasing Preservative in Baby and Cosmetic Products. Journal of the Dermatology Nurses' Association [Internet]. 2009 May; 1(3): 211-4.        [ Links ]

64. Avery A, Savelyich B, Sheikh A, Cantrill J, Morris C, Fernando B, et al. Identifying and establishing consensus on the most important safety features of GP computer systems: e-Delphi study. Journal of Innovation in Health Informatics [Internet]. 2005 Mar 1;1 3(1): 3-11.        [ Links ]



Gina Rencken



Lauren Michelle Perks completed this study as part of the requirements towards a Masters degree and was responsible for the conceptualisation, data collection, analysis and drafting of the first version of the manuscript. Gina Rencken & Pragashnie Govender were supervisors of the study and assisted in conceptualisation of the study, guidance on the choice of methodology as well as for critical inputs throughout the drafting of the manuscript. GR & PG were responsible for the revisions through the review process.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License