Sucking is one of the first coordinated muscular activities in infants, already observable in fetuses as young as 13 weeks.1,2 Similar to other rhythmic actions (eg, locomotion), sucking is generally considered to be controlled by an innate neural network known as central pattern generator (CPG).3 In human infants, there are 2 main types of sucking.4 Nonnutritive sucking (NNS) is an ororhythmic behavior already occurring in utero. It is organized in bursts consisting of 6 to 12 suck cycles that occur at approximately 2 Hz, followed by pause periods.5,6 Development of nutritive sucking (NS) begins later and finalizes at fluid intake. It comprises expression or the combination of suction and expression for fluid intake.7,8 Suction is the negative intraoral pressure generated by lowering the tongue and jaw and closure of the nasopharynx to draw milk out, while expression is the stripping/compression of the nipple between the tongue and the hard palate to eject milk.9,10
To be effective, NS generally requires the infant to have mature and functional neural networks and coordinated swallowing and breathing.11 For the successful development and function of NS, the organization of the suck, swallow, and respiration must occur at 2 levels. First, each of the 3 components must be well established for their synchronous function. Second, this cumulative mechanism function must be still effective with the introduction of the bolus to support transfer to the stomach.12
Because of the immaturity and significant medical needs of infants hospitalized in the neonatal intensive care units, these skills are not well established, especially in preterm infants, defined as neonate born prior to 37 weeks of gestation (GA), and can represent an important hurdle in their early development.13 Chronic lung disease, infections, neurologic disorders, infections, and necrotizing enterocolitis are associated with feeding dysfunctions and neurodevelopmental outcomes in later childhood.14 Infants in all categories of prematurity, including extremely preterm (<28 weeks of GA), very preterm (28 to <32 weeks of GA), moderate (32 to<34 weeks of GA), and late preterm (34 to < 37 weeks of GA), may require support for the attainment of coordinated sucking for functional and effective breastfeeding or bottle-feeding.13-15,16
Recent evidence suggests that GA age at birth and birth weight are the strongest predictors of the time for the transition from tube feeding to the first oral feeding and subsequent oral feeding independence. And while the achievement of oral feeding is typically one of the primary requirements for the discharge of a neonate, the initial neonatal intensive care unit discharge to home may take place while infant remains feeding through the feeding tube. This increases the risk of an increased rate of emergency department visits and rehospitalisation17 and puts the feeding tube-dependent infants at an increased risk of significant neurodevelopmental challenges.18
Infants who were dependent on an orogastric tube upon discharge were recently reported to have significantly lower cognitive, communication, and motor scores on a standard developmental assessment at 2 years, independent of infant's brain injury and overall illness.18 Therefore, the dependence on orogastric technology upon discharge presents long-term difficulties for the infant's neurodevelopment and the family overall.17 Up to 40% of infants referred to feeding clinics in early months of life are reported to have been born preterm.14 This is consistent with preterm birth often being accompanied by a number of medical and neurodevelopmental needs that have long-term developmental consequences, ranging from mild delays in general developmental that are overcome with supportive rehabilitative services to permanent disorders and disabilities.
Therefore, numerous studies have focused on intervention strategies to facilitate oral feeding of preterm infants while still hospitalized in the neonatal intensive care unit.19 In particular, NNS via pacifier has been consistently proposed as an intervention to improve sucking in preterms and was found to decrease length of hospital stay and to facilitate the transition to full oral-/bottle-feeds.20-22 Similarly, strategies based on sensorimotor inputs (such as cheek/chin support, oral, tactile, kinesthetic, auditory, and vestibular, and/or visual stimulation) were utilized in other studies showing beneficial effects on qualitative aspects of oral feeding performance in preterm infants.23-27
Several studies have demonstrated that early intervention strategies have beneficial effects on the oral feeding performance of preterm infants, and the majority of them used qualitative assessments as the primary outcome measures. A recent systematic review of NNS interventions and the effect on the physiologic stability of preterm infants showed significant decrease in the length of hospital stay and facilitation of the transition from gavage or partial oral feeds to full oral feeds, with limited information on the basic mechanisms underlying the observed improvements.21 Similar results of improvements of some of the feeding/swallowing variables were reported in another review article exploring the effects of a wider range of interventions, namely, NNS and oral/perioral stimulation.28 Yet, another Cochrane review reported benefits of NNS outweighing nonoral interventions and standard of care in regard to transition to oral feeding and duration of the initial hospitalization.20
A number of previous systematic reviews and meta-analyses have been published with valuable information about the types and outcomes of developmental interventions provided to preterm infants with oromotor or feeding difficulties, the use of pacifiers, and specific intervention protocols. In general, their conclusions show that interventions are promising for enhancing feeding performance in preterm newborns, although methodological heterogeneity and variations in results across studies hamper the application and the selection of the most appropriate intervention across preterm populations.
However, no review has focused on the quantitative categories of suck performance. This information may be useful in the implementation of early treatment and more individualized strategies of interventions. Therefore, the current review was designed to review preterm infant intervention studies with quantitative outcomes of sucking performance to summarize the evidence of the effect of interventions on specific components of sucking including efficiency, morphology, frequency, and timing parameters of preterm neonate feeding. These parameters are directly related to the infant nutrition capacity and therefore have a significant clinical relevance. Furthermore, to fill the gap in knowledge of the quantitative parameters of sucking maturation, this review included only randomized controlled trial designs of NS and NNS interventions with preterm infants with quantitative outcomes of sucking performance.
What This Study Adds
* Further studies on sucking interventions in preterm infants should include both direct and indirect quantitative and qualitative outcome parameters to maximize the evidence and increase the impact on clinical practice. Collecting data on the efficiency, morphology, and frequency parameters may be used to objectively determinate preterm infant's sucking skills.
METHODS
Literature Search
Studies were identified by searching PubMed, CINAHL, MEDLINE, EMBASE, and PSYCOLIST databases. The searches were limited to peer-reviewed publications in English and included articles published through August 2017. References were exported into a bibliographic management database and duplicates were removed. The search strategy was performed using the following Medical Subject Headings (MeSH) terms: ("preterm infants" or "preterms" or "prematurity newborns") and ("sucking" or "suck" or "oral feeding skills") and ("intervention" or "training").
Study Selection
Criteria for inclusion in the study were established prior to the literature search. Inclusion was limited to randomized controlled trial designs. Studies were selected if they fulfilled the following criteria: (i) involved infants born preterm (gestational age at birth below 37 weeks), (ii) tested experimental interventions to improve sucking or oral feeding skills, and (iii) included quantitative outcome measure based on objective parameters of sucking performance.
Studies were excluded if they were protocols, not in English, did not include preterm infants, or were any study design other than randomized.
Validity Assessment
The methodological quality of the studies was assessed using the Cochrane Collaboration's tool29 for assessing risk of bias and the following criteria were considered: sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, and completeness of outcome data. Each criterion was judged as "yes," "no," or "unclear." Two review authors separately assessed each study, and disagreement was resolved by discussion together with the senior author.
RESULTS
Description of Studies
The PRISMA flow diagram of the review process is reported in Figure 1. The search, completed in September 2017, yielded 780 articles, and then 59 duplicates were removed. Two authors independently reviewed 721 articles on the basis of the titles and abstracts, and selected 47 articles. Full texts of the selected articles were analyzed by 2 reviewers and the eligibility of the study inclusion was assessed independently; in case of mismatched opinion, consensus was reached after discussion. Twenty-eight of the 47 articles were excluded on the basis of failure to meet the inclusion criteria.
In particular, we excluded these studies because 6 were not randomized controlled trials, 2 were protocols, 9 did not include quantitative outcome measures, 1 was a poster, 8 were systematic reviews, and 2 were meta-analyses. The remaining 19 studies were included in this review: 15 randomized, 1 quasi-randomized, and 3 crossover randomized controlled trials.5,7,30-46
The methodological quality of the selected studies was evaluated as reported in Table 1. All studies had risk of bias in at least 1 of the 4 items.
Most of the studies involved low-risk preterm infants, and 2 studies included preterm infants with either specific respiratory pathology or with low birth weight.5,41-43 Sample sizes ranged from 11 to 230 preterm infants. The types of interventions were grouped into 6 categories on the basis of their features: (i) NNS, (ii) NNS with auditory reinforcement, (iii) sensorimotor stimulation, (iv) oral support (OS), (v) combined training, and (vi) NS. Interventions occurred before, during, or after feeding (either by orogastric tube or bottle-feeding), or outside of the feeding times. Figure 2 summarizes the results of the review by category (efficiency, frequency, morphology, and duration).
Very preterm infants were included in 19/19, extremely preterm were included in 11/19, moderately preterm were included in 11/19, and late preterm were included in 2/19 included studies. Since all of the studies the majority of the included infants were born at the very preterm GA age category, we did not identify any patterns of the type of training and its effectiveness as it relates to the GA age at birth. Table 2 summarizes the population, intervention/s, and outcome measures of the included studies.
Outcome Parameters
In accordance with the specific aims of our review, we selected all outcome measures consisting of quantitative parameters. Based on their characteristics, sucking parameters were subdivided into 4 groups: (i) efficiency, (ii) frequency, (iii) morphology, and (iv) duration. They are described both in NS and the NNS.
Efficiency parameters were defined as those strictly related to the infant nutrition capacity, including milk volume measures and parameters referred to the coordination of sucking and swallowing. Frequency parameters were defined as those exploring the occurrence of a periodic event per unit time (burst, pause, suck, expression, and suction). Some of these parameters describe the whole sucking pattern (the alternation of sucking burst and pause) or the single suck cycle (alternation of expression and suction phases). Morphology parameters were defined as those describing the shape, size, and phase distribution of the sucking curves, including either suction/expression amplitude or pressure. Finally, duration parameters were defined as those exploring the time length of all events related to sucking (Table 3).
Effects of the Different Types of Intervention
Nonnutritive Sucking
Nonnutritive sucking alone was evaluated in 6 studies, generally consisting of a brief intervention before infant's feedings. Training modalities were heterogeneous and their total duration ranged from 5 days to 1 week, performed for 2, 5, or 15 minutes before feedings. Overall, NNS training studies showed inconsistent effects on quantitative outcome measures.
Four efficiency parameters were identified in 5 studies. The parameter reported to have been significantly affected by the training was the formula taken at first 5 minutes (1 study). The parameters unaffected by the training were overall transfer (4 studies) and proficiency (2 studies). One parameter, namely, rate of transfer, was affected in 1 study and unaffected in another 2. Three frequency parameters were explored in 2 studies. The parameter significantly affected by the training was frequency of pauses in NS (1 study). Two parameters, namely, frequency of bursts in NS and frequency of sucks per burst in NS, were affected in 1 study and unaffected in the other. Only 1 morphology parameter, suction amplitude, was explored (1 study) and was found to be affected by the training. Four duration parameters were explored in 3 studies. The parameters significantly affected by the training were pause duration (1 study). The only parameter unaffected by the training was pause duration in NS (1 study). Three parameters, namely, first burst duration in NS, burst duration in NS, and feeding duration, were affected in one study and unaffected in another.
NNS With Auditory Reinforcement
Nonnutritive sucking with auditory reinforcement was explored in 2 studies. They used a pacifier adapted so that a suck of predetermined strength activated an audio player with lullabies36 or mother's voice.37 Infants were trained for 15 to 20 min/d, in a range of 15 to 45 min before feedings, for 5 days.
Nonnutritive sucking with auditory reinforcement trainings showed significant effects on all quantitative outcome measures explored. In particular, they were found to be effective on rate of transfer and oral volume intake (efficiency parameters), number of feeds per day (frequency parameter), mean pressure in NNS (morphology parameter), and feeding duration (duration parameter).
Sensorimotor Stimulations
Sensorimotor stimulation programs were explored in 11 studies, some of which assessed more than 1 training type. Sensorimotor stimuli consisted of oral/intraoral (O/IO), perioral (PO), or extraoral inputs. The O/IO was generally based on gum and tongue stimulation that could be delivered by therapist's hands or via a pacifier. Barlow et al5 applied an intraoral stimulus through a specific pacifier, a 'pulsating nipple' programmed to mimic the temporal features of a well-formed NNS burst. The treatment was administered for 3-minute epochs for 3 to 4 times per day for 10 days. Perioral stimulation programs, instead, were variable and composed of stroking and stimulating infants' cheeks and lips. In 3 studies,7,35,38 the perioral stimulation was associated with an oral one. Treatment duration ranged between 10 and 14 days for 12 to 30 minutes per session. Finally, an extraoral stimulation program implemented by Fucile and colleagues44 included sensorimotor tactile input to the infants' head, neck, trunk, and limbs, even in combination with IO/PO stimuli. The treatment duration was 12 minutes, twice daily, for 10 days.
Studies in this group showed significant effects of sensorimotor training on the majority of quantitative outcome measures explored.
Seven efficiency parameters were explored in 5 studies and were all found to be significantly affected by the training, except for the overall transfer and proficiency that were unaffected in 1 study.35 Seven frequency parameters were explored in 5 studies and were all found to be significantly affected by the training. Four morphology parameters were explored in 4 studies and were all found to be significantly affected by the various sensorimotor trainings with the exception of mean NNS pressure, which was unaffected in 1 study.41 A single duration parameter, mean sucking activity, was explored in one study38 and was found to be improved by the training.
Oral Support
Oral support was assessed in 2 studies and consisted of cheeks and chin support during feeding sessions. Studies used different treatment duration and time distribution.
The efficiency parameters explored, formula taken at first 5 minutes in one and oral volume intake in the other, were significantly affected by the training. Frequency (4 parameters), morphology (1 parameter), and duration (4 parameters) were explored in both studies and were always found to be significantly affected by the training.
Combined Training
Seven studies combined sensorimotor stimulation with either NNS (5 studies) or OS (1 study) or with other stimuli (1 study).
Nine efficiency parameters were explored in 5 studies and were all found to be significantly affected by the training. The exceptions were overall transfer, unaffected in 2 studies, and mean volume per suck and rate of transfer, unaffected in 1 study. Three frequency parameters were explored in 4 studies. Frequency of sucks per burst was affected in 1 study.46 Frequency of sucks per minute was unaffected (1 study) and frequency of feeds per day was unaffected in one and unaffected in the other study. Three morphology parameters were explored in 3 studies and were all found to be significantly affected by the different types of trainings. Two duration parameters were explored in 2 studies. Mean sucking activity was affected (1 study) and feeding duration was unaffected (1 study).
Nutritive Sucking
Two studies applied an NS intervention.30,44-46 The swallowing program consisted of placing a milk bolus on the tongue to facilitate swallow mechanisms for 15 minutes for 5 days. The intervention proposed by Fucile et al44 was based on a particular device, a controlled flow vacuum-free bottle system, capable to maintain milk continually at the level of the infant's mouth. The experimental group received feeding session with this device for 20 minutes.
Both of the studies demonstrated a good effect on efficiency outcome parameters, in particular, in overall transfer and rate of transfer. The swallow program also demonstrated an effect on proficiency outcomes. Fucile and colleagues44 also reported significant differences between experimental and control groups in frequency outcomes (suction frequency and expression frequency) and in feeding duration. However, no differences were found in suction amplitude and NS burst duration.
DISCUSSION AND CONCLUSIONS
The main aim of this systematic review was to determine the effects of early intervention on quantitative parameters of sucking in preterm infants as reported by randomized controlled designs studies. In general, the results demonstrate that most types of interventions aimed at improving sucking in preterm infants yield significant effects on quantitative parameters of sucking, with the exception of trainings based exclusively on NNS. The studies using NNS-only training report inconsistent results. Of the 12 different outcome parameters explored in those studies, 3 were positively affected by training, 3 were not affected by training, and the remaining 6 showed inconsistent results among studies.
It is of interest that the types of parameters most frequently found to be unaffected by the NNS-only training were the efficiency parameters, which are considered to be most directly related to the infant nutrition capacity and therefore may have the highest clinical relevance. Our findings might appear in conflict with other reviews, which found significant reduction of premature infants' length of hospital stay and improved bottle-feeding performance and transition from tube to bottle as potential positive clinical outcomes of NNS training.21-28 However, these outcome measures are not comparable with the objective sucking parameters explored in our review, as most of them represent general clinical variables. Moreover, clinical quantitative variables, such as weight gain, energy intake, oxygen saturation, or intestinal transit time, were not found to be affected by the training.
One of the proposed reasons for the limited effect of NNS-only trainings on sucking parameters is habituation to the teat.47 In this view, the positive effect of NNS on sucking in the trained infant gradually fades out and a novel stimulus is necessary to revitalize the sucking behavior. This interpretation aligns with many studies in which NNS was coupled with sensorimotor reinforcement (ie, sensorimotor stimulation, OS) and showed significant effects of training on sucking parameters. In those studies, the reinforcement provided by the sensory stimulation, which was human-mediated in the great majority of the cases, is likely to be responsible for the persistence of the positive effects on sucking patterns. This is also confirmed by the results obtained with interventions based on NNS with auditory reinforcement. Music or mother's voice was used as a contingent reward to promote NNS during the intervention sessions.
Following NNS, which was the most common intervention in our review (13/19 studies), sensorimotor stimulation was the second most common intervention, applied in 11 out of 19 studies. Most sensorimotor stimulation programs showed to positively affect sucking performance. This applied to all 4 domains explored and, in particular, to morphology and frequency. This finding is consistent with another recent review of oral motor interventions with preterm-born infants that highlighted quantitative parameters of oral feeding success.48 The sensorimotor stimulation seemed to be effective irrespective of the stimulated region, that is, intra-, peri-, or extraoral. This might suggest that benefits could be related to a broader effect of the intervention on the infant's state regulation rather than a direct effect on the coordination of sucking. Also, a critical role is played by the adult-infant relationship, as the sensorimotor stimulation involved a human caregiver. It is noteworthy, however, that the training was also effective in the 3 studies adopting a mechanical device to provide the intraoral stimulation.5,41,45
Two of the 11 studies using sensorimotor stimulation, either alone or in combination, showed predominantly nonsignificant results. Hwang et al42 found no effects of combined training (PO/IO + NNS) in efficiency, frequency, and time but reported an increase of formula intake in the first 5 minutes of a feeding as a result of the training intervention. This apparently conflicting finding may be due to the small sample size (n = 16) and heterogeneous infant clinical characteristics. Nonsignificant results were also reported by Rocha and colleagues43 although their study explored only 1 quantitative parameter (the number of feeds per day), and the quality of the study was low.43
Both studies31-38 assessing the effects of OS showed improvements in all areas of sucking performance, both when the training was used in isolation and when it was combined with oral stimulation. Oral support seems to provide the necessary stability for the jaw and to assist the infants in maintaining a more organized sucking pattern. Also, NS studies generally report positive effects on sucking performance, although the small number of studies and the different types of interventions do not allow for definitive conclusions.
Several limitations need to be considered in the interpretation of the results of this review, principally related to the heterogeneity of the reviewed studies. For example, sample size ranged from 10 to 160 preterm infants and subject characteristics were not homogeneous, particularly in terms of gestational age at birth, birth weight, and respiratory support. Quantitative outcome measures were also very heterogeneous and were assessed differently, preventing the possibility to perform a meta-analysis. Lastly, this review does not include information about the practices of breastfeeding versus bottle-feeding methods, as this was outside of the scope of this work. Most institutions caring for hospitalized infants provide or have information about the community source of supportive and rehabilitative services, including lactation specialists, speech-language pathologists, occupational therapists, and other professionals. These services assist mothers and primary caregivers in informed decision making and provide information and support to ensure successful feeding methods for their infants.
In conclusion, the evidence indicates that a variety of interventions, based on different principles and methodologies, are effective in enhancing quantitative aspects of sucking. Efficiency parameters are the most frequently explored, as they are directly related to the infant nutrition capacity and therefore present the highest clinical relevance. They are positively influenced by most types of intervention, though are less affected by NNS training only. Frequency, morphology, and efficiency parameters outline some of the mechanisms of functional sucking but are infrequently investigated. Future studies may benefit from the concurrent exploration of the different types of outcome measures. This could serve as a support for a better tailored therapeutic approach consisting of the selection of more personalized interventions based on a pretraining profiling of the specific sucking components that are compromised or weakened.
References