Best Practices for Oral Motor Stimulation to Improve Oral Feeding in Preterm Infants: A Systematic Review

Research Question

In healthy pre-term infants with oral feeding difficulties, do non-nutritive oral motor stimulation and oral support methods improve infants’ oral feeding skills as demonstrated by time to full oral feedings, volume intake, weight gain, and/or length of hospital stay?

Search Methods

The author conducted a systematic search of the literature published from 1995 to 2015. The search included studies on pre-term infants born at least 28 weeks gestational age (GA) with no contraindications for oral stimulation or oral feeding. The databases searched included CINAHL Complete, CINAHL Plus with Full Text, E- Journals, Health Source: Nursing/Academic Edition, MEDLINE, OVID, and PubMed Clinical Queries. Search terms included oral motor OR oral stimulation, AND preterm infants OR prematurity, AND feeding. The focus was on articles reporting the effectiveness of peri-oral and intra- oral stimulation, NNS, and/or oral support on the outcomes of volume of oral intake, feeding efficiency/proficiency, weight gain, and length of hospital stay. Articles were excluded if they included infants with conditions in which oral stimulation or oral feeding are contraindicated or infants who were not medically stable to tolerate oral stimulation or oral feeding. After literature search results were collected and exact duplicates were removed, the abstracts of the remaining articles were reviewed by the author. Figure 1 depicts the flow of abstracts and articles through the process. One of the articles by Lessen [12] was retained although it included pre- term infants younger than 28 weeks GA since the information was relevant and also included infants through 29 weeks GA. Articles are summarized in Table 1.

Studies included in this review provide Level I and III evidence. Level IV and V evidence was excluded. The American Occupational Therapy Association [13] recognizes the following levels of evidence adapted from [14]. Level I - Systematic reviews, meta-analyses, randomized controlled trials Level II - Two groups, nonrandomized studies (e.g., cohort, case-control) Level III - One group, nonrandomized (e.g., before and after, pretest and posttest) Level IV - Descriptive studies that include analysis of outcomes (single-subject design, case series) Level V - Case reports and expert opinion that include narrative literature reviews and consensus statements

Quality Review

Cincinnati Children’s Hospital Medical Center’s (CCHMC) evidence evaluation tools & resources were used with permission [15]. CCHMC utilizes Let Evidence Guide Every New Decision system to guide the evaluation of evidence, develop best evidence statements and evidence-based care guidelines, and guide decision-making to “achieve the best, safest care for children” [16]. The author used the CCHMC’s LEGEND appraisal forms to appraise each of the articles. For the purpose of this review, the Intervention Systematic Review / Meta-Analysis, Intervention Randomized Controlled Trial or Controlled Clinical Trial, and Intervention Cohort Study forms were used [17]. LEGEND resources were also used to grade the body of evidence and judge the strength of each recommendation.

According to LEGEND’s Grading the Body of Evidence (BOE) system, a high BOE indicates that there is a sufficient number of high quality studies with consistent results on the topic; a moderate BOE indicates that the studies included a single well-done trial, multiple lesser quality trials, or multiple large, high-quality observational studies on the topic; a low BOE indicates that studies included were of lesser quality or with some uncertainty on the topic; a very low BOE indicates that the studies included were of insufficient quality including descriptive studies, case series, general reviews, insufficient design or execution, there were too few studies, and/or inconsistent results; and grade not assignable indicates local consensus only [18]. LEGEND’s dimensions for “judging the strength of a recommendation” include the components of safety/harm, benefits, burden to adhere to recommendations, cost-effectiveness, directness of the evidence, impact on quality of life, morbidity, and mortality, and grade of the BOE resulting in strengths of high, moderate, weak, or no recommendation [19].

Results

Fifteen articles were reviewed for the final synthesis. The articles provide Level I evidence, with the exception of one Level III article. Findings were organized into four areas of intervention: oral motor stimulation, non- nutritive sucking, oral support, and co-interventions. Supplemental provide information on risk of bias of articles.

Click to enlarge

All studies exhibited a low risk of bias, with the exception of the meta- analysis by Daley and Kennedy [20], which presented with moderate risk due to lack of information related to data extraction, number of excluded articles, and unclear study appraisal methods.

Figure format from “Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement,” by D. Moher A, Liberati J, Tetzlaff DG, Altman; PRISMA Group [21],

therapists’ hands in incubator without touching infant for

15 mins.

All interventions

provided for 10 consecutive days.

Level III Cohort prospective study

Intervention Measurement of NNS, standardized

N = 51

Bingham, et al.

[23]

M/F = 16/35

doi:10.1136/ad c.2009. 164186 feeding advance schedule, and performance of Inclusion Criteria • infants born between 25 NOMAS.

and 34 weeks’ PMA

Intervention Group 1: Oral

stimulation protocol: cheeks, lips, and tongue were stimulated to breathing comfortably with no respiratory support or with nasal cannula (room improve muscle contractility, and air) only Level I RCT strength and orientation reflexes, Boiron, et al.

[24]

Intervention Group 1, n=9.

inhibit mouth- closing reflex, and initiate sucking and https://doi.org/ M/F = 5/4

10.11 11/j.1469- 8749.2007.0043 Intervention Group 2, n=11.

swallowing (SS) reflexes.

M/F = 4/7 Intervention Group 3, n=12.

Group 2: Oral support: chin, cheek,

9.x M/F = 7/5 Control Group, n=11. M/F =

and lip support to

7/4 aid in lip closure and deglutition and movement of nipple to corner of mouth to stop SS pattern to facilitate breathing.

Level I Meta-analysis

Intervention Sucking apparatus

Daley, et al. [25]

N = 10 studies

with and without NPO, oral feeding with and without doi:10.1097/00 Inclusion Criteria • English-language papers

00523 7- 200012000- NGT in place, various nipple

00006 on nipple feeding, feeding performance, and feeding types, breast feeding vs bottle Infants with higher NNS organization scores transitioned to FOF 3 days PMA at IOF PMA at FOF PMA at SOF earlier (p<0.05) than infants with more chaotic patterns of suck bursts and Transition time IOF to a higher NNS organization FOF.

score resulted in fewer number of days from IOF

to SOF (p <0.10).

Non-nutritive

pacifier attached to a Non-nutritive sucking pressure and sucking activity were statistically significantly increased for catheter connected to a pressure transducer to the stimulation and the stimulation+ support groups compared with the control group at Day 7 and the pressure amplifier, to calculate the Day 14 (p<0.001). Significant improvement was also noted at Day 17 mean maximum non- nutritive and Day 20 for time for sucking pressure.

milk ingestion.

The following areas did

Effects of NPO, NGT,

not have statistically significant results (as indicated by no to small effect size): NGT, nipple type (only the comparison nipples, breast vs bottle, GA, oral support, oral stim, and of Enfamil to Nuk nipple showed a large effect size), NNS.

breast vs bottle, and NNS.

efficiency in premature feeding, oral support, and NNS

infants • Experimental and quasi-

experimental designs Level I RCT

Intervention

Intervention Group 1, n=19.

Group 1: OS consisting of stroking the lips, M/F = 12/7

Intervention Group 2, n=18.

cheeks, gums, tongue and sucking M/F = 11/7

on a pacifier Intervention Group 3, n=18.

Group 2: Tactile/kinesthetic

M/F = 10/8

Control Group, n=20. M/F =

Fucile [4]

(T/K) involving stroking the body

16/4 and limbs and passive range of Inclusion Criteria

• Clinically stable preterm motion to limbs Group 3: OS + T/K infants, born between 26 and 32 weeks GA
• Appropriate size for GA
• Receiving all tube feedings Control Received no
• No chronic medical stimulation complications Intervention Pre-feeding oral Level I RCT Intervention Group, n=16.

stimulation program based on M/F = 7/9

Beckman’s

Control Group, n=16. M/F =

principles consisting of 12 minutes stimulation

6/10 Fucile, et al. [26]

Inclusion Criteria

• Preterm infants, born between 26 and 29 weeks doi:10.1067/mp to cheeks, lips, gums, and tongue, d.200 2.125731 followed by 3 minutes of sucking GA
• Appropriate size for GA
• Receiving full tube feedings on a pacifier routinely used in the nursery (120 kcal/kg/day)
• No chronic medical Control Sham stimulation The following areas did have statistically significant results: length of NPO prior to feeding, gestational age, oral support, and oral stimulation.

Independent oral feeding was achieved significantly earlier in all three intervention groups than Time to attainment of the control group (p<0.001). Proficiency and independent oral feeding volume transfer were significantly greater in the three intervention groups, Volume of milk taken over first 5 rate of transfer was significantly greater in the minutes OS and multi- OS+T/K groups, and there was less Sucking skills.

volume loss in the OS group only compared to the control group (all tests p<0.042).

Independent oral feeding was attained significantly Time to attain independent earlier in the Exp. group than the control group, 11 oral feeding Number of days to reach ± 4 days (mean ± SD) versus 18 ± 7 days, respectively (P = .005). Overall intake and rate of one and 4 successful oral feedings milk transfer were significantly greater over time in the Exp. group than per day Overall intake the control group (P = .0002 and .046, respectively). There was no difference in length of hospital stay between the 2 groups, although the Exp.

Rate of milk

transfer Length of hospital stay.

group was discharged an average of 5 days sooner Gaebler, et al.

[27]

Intervention Group, n=9.

stroking protocol a higher percentage of nipple/partial nipple feeds Motor

doi:10.5014/ajo M/F = 6/3

and an additional 2- minute perioral and t.50.3. 184

Control Group, n=9. M/F =

intraoral stimulation

8/1 Inclusion Criteria • Preterm infants, born between 30 and 34 weeks program.

Control 5 min pre-feeding

GA

• Medically stable and in stroking protocol only isolette or open crib
• Fed via gavage or nasogastric tube
• No history of cardiac or gastrointestinal disorders, or central nervous system dysfunction Level I RCT Intervention Group 1, n=19.

M/F = not identified

Intervention Group 2, n=18.

Intervention

Group 1: Nonnutritive oral

M/F = not identified

Intervention Group 3, n=18.

motor therapy (NNOMT) M/F = not identified Group 2: Infant massage therapy Control Group, n=20. M/F = Lau, et al. [28]

doi:10.3233/NP not identified (iMT) Inclusion Criteria

• Preterm infants, born between 26 and 32 weeks Group 3: Combined M- 1262612 interventions (NNOMT + iMT) GA
• Appropriate size for GA
• No congenital anomalies or Control No intervention chronic medical (sham) complications
• Characterized as “feeders and growers”
• Clinically stable Level I Intervention RCT Received developed Lessen (2011) Level I Intervention Received developed Premature Intervention Group, n=10. (PIOMI) - specifically designed fo 182115a2a M/F = 4/6 infants Premature infant Lessen [12] 10.1097/ANC.0 oral Moter Intervension (PIOMI) -specifically b013e3 as young as 29 weeks GA based on Beckman’s protocol designed fo 182115a2a M/F = 4/6 infants as young and Control Group, n=9. modified Assessment Scale during a than those in the control group (t[16] = 1.77, p <

1-min trial of .05; scored higher on normal characteristics of the nutritive suck scale (U nonnutritive sucking and a

5-min trial of = 25, p = .08); were discharged from the hospital earlier, (t[16] = - nutritive sucking

2.4, p = .01); and gained more weight than their counterparts (t[16] = 1.49,

p = .07.

Number of

days from Infants in the NNOMT+

start to independent iMT group attained independent OF significantly earlier than oral feeding.

Overall transfer (%

controls (p<0.001) with shorter day intervals from start of OF to 3–5 daily volume taken/volume oral feedings.

Infants in both NNOMT

prescribed) Proficiency

and NNOMT+ iMT groups transitioned faster from 3–

(% volume taken at 5 min/volume

5 daily oral feedings to independent OF (p≤0.003).

prescribed) Rate of transfer over

Infants in all intervention

groups demonstrated a faster rate of oral feeding the entire skill maturation than the feeding (ml/min).

control group.

For feeding progression,

although there was a statistically significant decrease in transition from Feeding progression gavage to oral feedings for Length of hospital stay the PIOMI group of 5 days sooner, infant birth weight covariant eliminated the statistical significance. Although infants in the for this younger GA for M/F = 3/6 mins of finger stroking to cheeks (internal as 29 weeks GA based on Beckman’s protocol and modified for this and external), lips, gum Inclusion Criteria tongue, and younger GA for M/F = 3/6 mins of palate.

• Preterm infants, born finger stroking to cheeks (internal and between 26 and 29 weeks PMA Control
• Appropriate size for GA external), lips, gum tongue, and palate.

Provider’s hands inside

Control Provider’s hands inside isolette

isolette f • Clinically stable per medical staff, but 5 mins, not touching f5 mins, not touching infant. Each intervention infant. could be receiving oxygen per was provided f nasal Each intervention was cannula 7 consecutive days.

provided f cannula Level I Randomized

crossover design Intervention/Control, n=10

Control During control observation, infa was positioned in right side-lyin for 2

(crossover) M/F = not identified Inclusion Criteria •

Pickler, et al.

[29] doi:10.1016/j.a dnc.20040.05.0 Preterm infants, born prior to 32 Control weeks GA • No known minutes prior to

5 cognitive, neurologic, feeding wi no pacifier offered.

cardiovascular, gastrointestinal, or craniofacial disorders Level I RCT Intervention Group, n=49 Intervention 15 minutes perioral with 2 losses in follow-up resulting in 47 analyzed. M/F and intraor stimulation using = not identified Control Group, n=49. Control gloved finger a pacifier during Pimenta, et al.

[30] doi:10.2223/JPE M/F = not identified gavage until preterm infants started oral diet for Inclusion Criteria • Preterm infants, born between 26 and 32 feeding.

D.183

a period of at least weeks, 6 days GA • Adequate or small for GA

10 days. Control Sham procedure with no form of oral • Birth weight <1500 g PIOMI group were discharged 2.6 days sooner than the controls, this was not a statistically significant difference.

NNS and NS measured by

NS waves were smoother

and more regular than NNS waves. Time to onset a stretch sensitive chin and duration of the first non-nutritive suck burst were positively correlated strain gauge for measuring sucking rate with time to onset for the first nutritive suck burst (r and rhythm during feeding; effects of pre- = 0.79, P = 0.01) and (r =

0.94, P = 0.01).

Prefeeding NNS had no

feeding NNS on breathing

statistically significant effect on characteristics of measured with a nasal Thermistor, breathing or on any other characteristics of NS. Behavioral state during and on behavior state feedings and feeding before, during, and efficiency were not affected by prefeeding NNS after Length of hospital stay for infants in the experimental Length of hospital stay group was significantly lower than for the control and breastfeeding group, which was discharged 10.8 days later.

rate at discharge, 3- month and 6- month follow- The length of stay in the hospital for the control group was 52.37±19.51. The length of stay for the up.

experimental group was

41.81±17.7 (p=0.007). Fifty-nine infants (61.5%)

No malformations. Severe asphyxia, or presence of 3rd

stimulation during gavage feeding.

or 4th degree intracranial hemorrhage Intervention Oral support (The

first author) held the infant’s cheeks inward and forward Level I Randomized crossover design by placing her right Intervention/Control, n=20 ring finger on the infant’s left cheek (crossover) M/F = 7/13 and the thumb of the other hand on the infant’s opposite Yea Shwu, et al.

Inclusion Criteria • Preterm infants, born 25 to [31]

cheek to assist the infant in sealing the doi:10.5014/ajo

36 weeks GA • Allowed to take 15 ml by lips around the t.2010. 09031

nipple. Simultaneously, the

mouth per feeding, • Were inefficient feeders author placed her (i.e., unable to consume an average of 4 ml of feeding right little finger under the infant’s chin to stabilize the intake per min in a 5-min lower jaw.

feeding assessment) Control No oral support Level I RCT

Intervention 15 mins oral

Intervention Group, n=10

Younesian, et al.

stimulation according to [32]

M/F = 5/5

doi:10.5812/irc Control Group, n=10. M/F =

Beckman’s principles once per

mj.17( 5)2015.13515

5/5 day for 10 successive days 20-

Inclusion Criteria

40 mins prior to were breastfeeding at the time of hospital discharge,

31 (36.9%) at 3 months, and only 18 (20.5%) at 6 months of corrected age. At discharge, 46.9% of the control group and 76.5% of the experimental group were breastfeeding (p =

0.003). There were statistically significant differences between rates of breastfeeding at discharge (47 vs. 76%), 3 months (18 vs. 47%) (p = 0.003) and 6 months after discharge (10 vs. 27%) (p = 0.029).

For the intervention condition, statistical significance was noted in higher intake rate during Feeding performance the initial 5-min feeding period (p=0.046), lower % including duration, percentage leakage for initial 5-min (p=0.040), shorter feeding ingested, percentage duration (p=0.044), and higher intake rate for entire feeding (p=0.023) leakage, intake rake, than during control suck frequency, condition.

No statistical differences were noted in prescribed and mean volume ingested per volume consumed (p=0.11) or overall % of suck.

Physiological

leakage (p=0.84).

No significant differences

status was also assessed.

were noted in physiological status between conditions.

Number of

Statistical significance was

days to transition to SOF and FOF

achieved for SOF in Exp group over controls for 1 (P < 0. 001), 4 (P < 0.001), and 8 (P < 0.001) feedings (8 oral feedings per per day. Infants in Exp group were discharged day for 2 consecutive days), length ~1 week earlier than controls (p<0.05). Both

There were 13 articles of Level I evidence and one of Level III evidence reviewed that provide supportive data of preterm infants [15]. There is also a high BOE/high strength of evidence for recommending intra-oral stimulation immediately following peri-oral stimulation prior to oral feeding of preterm infants.

Findings support that oral motor stimulation techniques, particularly those based on Beckman Oral Motor Intervention principles, can be used to promote a more organized suck-swallow-breathe coordination, improve latching, and increase suction strength and endurance, which may therefore lead to safe and successful oral feeding. The use of peri- and intra-oral pre-feeding stimulation was associated with p values showing statistically significant positive outcomes of shorter time to full oral feedings, increased volume intake, improved feeding efficiency, shorter length of hospital stays, and/or increased weight gain [25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36].

Before initiating oral stimulation with preterm infants in preparation for oral feeding, the infant should be physiologically stable, demonstrate hunger cues, and exhibit neurobehavioral states of being quiet and alert [10, 27, 29, 37]. Once aforementioned stability, cues, and states are achieved, it is recommended that the infant receives peri-oral stimulation to cheeks, upper and lower lips, and the jaw followed by intra-oral stimulation to internal cheeks, gums and [4, 9, 10].

Non-Nutritive Sucking

Twelve articles of Level I evidence provided support of the use of NNS as a pre-feeding intervention. According to LEGEND criteria, there are both a high BOE and high strength of evidence for recommending NNS as a preparatory method to promote successful feeding in preterm infants [15]. Non-nutritive sucking may increase strength, endurance, and suction and may help with organization of infants’ physiological and behavioral states as well as suck-swallow-breathe coordination [4, 5]. Multiple studies report that NNS using a pacifier or finger resulted in statistically significant improvement and moderate to large effect sizes in suck organization, coordination, and/or strength and endurance [3, 12, 10, 25, 28, 30, 33, 37]. NNS combined with oral stimulation also produced large positive effect sizes on oral feeding [22, 34]. One study by Pickler and Reyna [29] found no statistically significant effect of NNS on breathing characteristics or feeding efficiency. However, they did find that first NNS suck burst positively correlated with onset of first nutritional suck burst. The authors identified several limitations including small sample size and only two observations per infant. Fucile, Gisel & Lau [26] and Pimenta, et al. [30] found that preterm infants who received NNS in conjunction with oral stimulation were discharged earlier than those who did not receive NNS by a mean average of 5 days and 10.8 days respectively.

It is recommended that NNS be performed by placing a pacifier or gloved fifth finger in infant’s mouth during gavage feedings and 2-3 minutes prior to feeding following intra-oral stimulation [4, 10, 20, 26, 30, 37]. The finger should be placed at the midline, center of the palate, gently stroking the palate to elicit a suck [10]. If using a pacifier, a standard pacifier should be used rather than an orthodontic, flat, or bulb shaped pacifier [9].

Oral Support

Consistent with LEGEND criteria, there is a high BOE and high strength of evidence for providing oral support during oral feeding. Five articles of Level I evidence supported the provision of oral support during feeding of preterm infants to safely maximize oral intake [15]. According to this review of evidence, oral support is beneficial to infants who have poor suck performance, strength, and/or endurance, but once suck performance improves with the provision of NNS, oral support is no longer necessary. Oral support decreases fluid loss, provides cheek and jaw stability, and aids in coordination of deglutition [4, 24, 31, 33, 34]. It is recommended that oral support be provided during oral feeding to provide stability and ameliorate the sucking pattern by placing the middle finger under the chin providing pressure at the mandible, the thumb and index fingers compressing the cheeks toward lips, and the fifth digit compressing the floor of the oral cavity under the chin to reinforce the swallowing [4, 31, 34, 36].

Co-Interventions

Both the BOE and strength of evidence are high according to LEGEND criteria to recommend combining interventions in preparation for and during oral feeds of preterm infants [15]. The systematic review by Arvedson, et al. [34] and studies by Asadollahpour, et al. [22], Fucile, Gisel & Lau [26] and Zhang, et al. [33] provided level I evidence that when combined with oral support and/or NNS, oral stimulation resulted in statistically significant p values over oral stimulation alone for weight gain and transition to full oral feedings. Results from Fucile [4], Gaebler & Hanzlik [27], and Lau, Fucile & Gisel [28] reported that subjects who received touch therapies, such as stroking or massage along with oral stimulation gained significantly more weight and were discharged significantly earlier than those who received only one intervention or no intervention.

The infant must be monitored for apnea, oxygen desaturation, and bradycardia during oral stimulation. Lack of suck-swallow-breathe coordination, the effort required to actively respond to stimulation, and the infant’s immature body systems contribute to this risk [4, 10, 34]. These adverse events are uncommon during oral stimulation and did not result from oral stimulation during any of the studies examined.

Applicability Issues

There are initial costs to consider when implementing the recommended pre-feeding oral motor interventions, primarily related to personnel salaries. These include time to train the therapists and NICU nursing staff, time to provide intervention to the infants, and time for parent/caregiver training. No specialized equipment is needed to carry out these interventions. The pre-feeding oral motor methods and oral support recommendations can be incorporated into established or on-demand feeding schedules.

Implications for Practice

It should be noted that OT literature on preterm feeding is lacking in the area of preterm oral motor stimulation and feeding of preterm infants. There are minimal systematic reviews and meta-analyses on the topic. This systematic review adds to the available evidence in an effort to promote best practices. Benefits and risks related to the finding are identified below.

Benefits

Benefits of following these best practices of oral motor stimulation, NNS, and oral support include:

• improvement of suck-wallow-respiration coordination
• increased volume intake
• improvement in efficiency of feeding and decreased time required for oral feeding
• decreased time to transition to full oral feeding
• weight gain
• shorter length of hospital stay

Risks

Risks associate with these methods of oral motor stimulation, NNS, and oral support include:

Although the procedures for oral feeding, other than recommendations for oral support, are not included in this systematic review, any time oral feeding is introduced to a person with swallowing difficulties, there is the risk of aspiration.

The infant must be monitored for apnea, oxygen desaturation, and bradycardia during oral stimulation. Lack of suck-swallow-breathe coordination, the effort required to actively respond stimulation, and the infant’s immature body systems contribute to this risk. These adverse events are unlikely and did not result from oral stimulation during any of the studies examined.

The infant may experience physical discomfort during oral stimulation, although this risk is minimal and unlikely.