Management of eating, drinking and swallowing difficulties

National Guideline Alliance (UK)

13Management of eating, drinking and swallowing difficulties

Review question: In children and young people with cerebral palsy, what interventions are effective in managing difficulties with eating, drinking and swallowing?

13.1. Introduction

For most children and young people, eating and drinking is an enjoyable experience, undertaken several times a day, usually in the company of family or friends. Meals and snacks serve the purpose of obtaining nutrition and hydration, but also provide a context for social interaction. Children typically progress from a liquid diet, via breast or bottle, through a soft diet to foods that need chewing. They also achieve increasing levels of independence.

Cerebral palsy can disrupt the motor control and coordination of sucking, drinking, biting, chewing and swallowing, particularly in children and young people with severe functional disabilities. This can lead to problems with inadequate intake, the risk of food or drink going into the lungs (aspiration), prolonged dependence on immature food textures (single textures and/or puree) and on being fed by others. Mealtimes may be lengthy, distressing, emotional and unproductive in terms of achieving adequate or perceived adequate intake.

Appropriate management of eating, drinking and swallowing difficulties is important for maintaining respiratory health, optimising nutritional status, maximising independence and supporting social participation. The Committee was interested in reviewing the evidence relating to interventions that are commonly suggested by professionals who are supporting families in this area of everyday functioning. These included postural management, modification of food and fluid textures, feeding techniques and equipment, therapies aimed at improving oral-motor skills and reducing the risk of aspiration.

The aim of this review is to identify clinical and cost-effective interventions for managing difficulties with eating, drinking and swallowing in children and young people with cerebral palsy.

13.2. Description of clinical evidence

Four randomised trials (Gisel 1995, ¹⁹⁹⁶, Ottenbacher 1981, Sigan 2013) and 4 observational studies (Adams 2012, Baghbadorani, Clawson 2007, Gisel 2001) were included in the review.

Evidence from these studies is summarised in the clinical GRADE evidence profile below (Table 50, Table 51, Table 52, Table 53, Table 54, and Table 55). See also the review protocol in Appendix E, the study selection flow chart in Appendix F, forest plots in Appendix I, study evidence tables in Appendix J and the exclusion list in Appendix K.

Table 50

Oral sensorimotor therapy versus routine treatment.

Table 51

ISMAR versus no ISMAR treatment.

Table 52

Multi-component intervention compared to routine physiotherapy.

Table 53

Parent and/or carer training sessions.

Table 54

Multi-component intervention (including Beckman oral exercise training, behavioural intervention and parenting training).

Table 55

Oral sensorimotor stimulations.

Studies were carried out in Bangladesh, Canada, Iran, Turkey and USA. Duration of studies ranged from 5 weeks to 12 months.

With regard to the population considered, 1 study population was diagnosed with moderate impairment (Gisel 2001), whereas the populations in the other included studies were diagnosed with moderate to severe impairment (Adams 2012, Baghbadorani 2014, Clawson 2007, Gisel 1995, ¹⁹⁹⁶, Ottenbacher 1981, Sigan 2014). One study considered mixed populations of participants with cerebral palsy and other neurological conditions, but this study was included as more than two-thirds of the population had cerebral palsy (Ottenbacher 1981).

With regard to the interventions, 4 randomised studies looked at participants who received oral sensorimotor therapy compared with routine therapy in children and young people with cerebral palsy (Gisel 1995, ¹⁹⁹⁶, Ottenbacher 1981, Sigan, 2014). One cohort study looked at children with cerebral palsy who received the Innsbruck sensorimotor activator and regulator (ISMAR) intra-oral appliance compared to children who had no ISMAR appliance (Gisel 2001). One cohort study looked at oral sensorimotor treatment (Baghbadorani 2014) in children with cerebral palsy. One cohort looked at a training programme delivered to children and their caregivers (Adams 2012). One cohort study looked at a multicomponent intervention, including carer training, behavioural interventions and Beckman oral motor exercises in children with cerebral palsy (Clawson 2007).

No evidence was retrieved for the following interventions: postural management, feeding techniques (such as jaw support, food placement and pacing), feeding equipment or pharmacological interventions.

Of the outcomes listed in the protocol and agreed by the Committee, studies reported critical outcomes including weight and height as mean percentiles and mean kilograms or centimetres (Adams 2012, Clawson 2007, Gisel 1995, ¹⁹⁹⁶^,²⁰⁰¹, Ottenbacher 1981). Duration of meal times was reported by 2 randomised studies and 1 non-comparative study (Gisel 1995, ¹⁹⁹⁶, Clawson 2007). One non-comparative study reported the frequency of chest illness once every 3 months, but only the p-value was reported (Adams 2012). Eating times of standard food textures were reported by 2 randomised studies (Gisel 1995, 1996). Outcomes, including oral-motor function or competency in feeding, were reported by 1 randomised study and 1 non-comparative study using the modified Functional Feeding Assessment (FFAm) and Oral Motor Assessment Scale (OMAS), respectively (Baghbadorani 2014, Sigan 2014).

13.2.1. Summary of included studies

A summary of the studies that were included in this review are presented in Table 49.

Table 49

Summary of included studies.

13.3. Clinical evidence profile

The clinical evidence profiles for this review question are presented in Table 50, Table 51, Table 52, Table 53, Table 54 and Table 55.

13.4. Economic evidence

No economic evaluations of interventions relevant to managing eating, drinking or swallowing difficulties were identified in the literature search conducted for this guideline. Full details of the search and economic article selection flow chart can be found in Appendix E and Appendix F, respectively.

This review question was not prioritised for de novo economic modelling. To aid consideration of cost effectiveness, relevant resource and cost-use data are presented in Appendix G.

13.5. Evidence statements

13.5.1. Oral sensorimotor therapy versus routine treatment

13.5.1.1. Nutritional status and/or changes in growth (weight)

Very low-quality evidence from 2 randomised studies with 43 children showed that there was no clinically significant difference between oral sensorimotor therapy and routine therapy for the outcome of weight (kg) in percentiles for age at 10 weeks follow-up.

Very low-quality evidence from 1 randomised study with 20 children showed that there was no clinically significant difference between oral sensorimotor therapy and routine therapy for the outcome of weight (pounds) at 9 weeks follow-up.

13.5.1.2. Duration of meal times (lunch or snack)

Very low-quality evidence from 2 randomised studies with 43 children showed that there was no clinically significant difference between oral sensorimotor treatment and routine treatment for the outcome of lunch duration in patients with cerebral palsy at 10 weeks follow-up.

Very low-quality evidence from 1 randomised study with 20 children showed that that there was no clinically significant difference between oral sensorimotor treatment and routine treatment for the outcome of snack duration in patients with cerebral palsy at 10 weeks follow-up.

13.5.1.3. Eating time of different food textures and/or a change in diet consistency a child is able to consume

Very low-quality evidence from 1 study with 20 children showed that there was no clinically significant difference between oral sensorimotor treatment and routine treatment for the outcome of eating time of pureed food texture in patients with cerebral palsy at 10 weeks follow-up.

Very low-quality evidence from 1 study with 20 children showed that there was no clinically significant difference between oral sensorimotor treatment and routine treatment for the outcome of eating time of viscous (raisin or gelatine) food texture in patients with cerebral palsy at 10 weeks follow-up.

Very low-quality evidence from 1 study with 20 children showed that there was no clinically significant difference between oral sensorimotor treatment and routine treatment for the outcome of eating time of solid (cereal ring or biscuit) in patients with cerebral palsy at 10 weeks follow-up.

Very low-quality evidence from 1 study with 23 children showed that there was a clinically significant beneficial effect of oral sensorimotor treatment compared with routine treatment for the outcome of eating time of pureed food texture in patients with cerebral palsy at 10 weeks follow-up.

Very low-quality evidence from 1 study with 23 children showed that there was no clinically significant difference between oral sensorimotor treatment and routine treatment for the outcome of eating time of viscous or solid food textures in patients with cerebral palsy at 10 weeks.

13.5.1.4. Psychological wellbeing of parents and/or carers

No evidence was retrieved for this outcome.

13.5.1.5. Acceptability of programme

No evidence was retrieved for this outcome.

13.5.1.6. Survival

No evidence was retrieved for this outcome.

13.5.2. Multi-component intervention versus routine physiotherapy

13.5.2.1. Physical function of the oropharyngeal mechanism

Low-quality evidence from 1 randomised study with 81 children showed that there was a clinically significant beneficial effect of oral sensorimotor treatment, as part of a programme that also included postural management, texture modification, changes to feeding techniques and parent education, compared with routine physiotherapy treatment for the outcomes of spoon feeding, drinking and swallowing in patients with cerebral palsy at 6 months (final).

Very low-quality evidence from 1 cohort study with 12 children showed that there was a clinically significant beneficial effect of oral sensorimotor treatment for the outcome of mouth closure but not lip closure onto utensil or lip closure, control of food or liquid during deglutition, or straw suction at 2 months follow-up.

13.5.2.2. Psychological wellbeing of parents and/or carers

No evidence was retrieved for this outcome.

13.5.2.3. Acceptability of programme

No evidence was retrieved for this outcome.

13.5.2.4. Survival

No evidence was retrieved for this outcome.

13.5.3. ISMAR appliance versus no ISMAR appliance

13.5.3.1. Anthropometric measure (weight)

Very low-quality evidence from 1 comparative cohort study with 17 children showed that there was a clinically significant beneficial effect of the ISMAR appliance compared with no ISMAR appliance for the outcome of weight in patients with cerebral palsy at 6 or 12 months follow-up.

13.5.3.2. Anthropometric measure (height)

Very low-quality evidence from 1 comparative cohort study with 17 children showed that there was no clinically significant difference in change between ISMAR appliance and no ISMAR appliance for the outcome of height in patients with cerebral palsy at 6 months but there was a clinically beneficial effect at 12 months.

13.5.3.3. Respiratory health

No evidence was retrieved for this outcome.

13.5.3.4. Physical function of the oropharyngeal mechanism and/or competency in feeding (percentage)

Very low-quality evidence from 1 comparative cohort study with 17 children showed that there was no clinically significant difference between ISMAR appliance for the outcome of spoon feeding and cup drinking skills in patients with cerebral palsy at 12 to 18 months or at 18 to 24 months follow-up.

Very low-quality evidence from 1 comparative cohort study with 17 children showed that there may be a clinically significant benefit of no ISMAR appliance compared with ISMAR appliance for the outcome of swallowing and clearing but there was uncertainty around the estimate at 12 to 18 months. However, low-quality evidence from the same study showed that there was a clinically significant beneficial effect of no ISMAR appliance compared with ISMAR appliance for the outcome of swallowing and clearing at 18 to 24 months (final).

Very low-quality evidence from 1 comparative cohort study with 17 children showed that there was no clinically significant difference between ISMAR appliance and no ISMAR appliance for the outcome of spoon feeding, cup drinking, swallowing or clearing at 12 to 18 months, but there was a clinically significant beneficial effect of no ISMAR appliance compared with ISMAR appliance for the outcome of cup drinking at 18 to 24 months

13.5.3.5. Psychological wellbeing of parents and/or carers

No evidence was retrieved for this outcome.

13.5.3.6. Acceptability of programme

No evidence was retrieved for this outcome.

13.5.3.7. Survival

No evidence was retrieved for this outcome.

13.5.4. Multi-component intervention, (including Beckman oral exercise training, behavioural intervention and parenting training)

13.5.4.1. Nutritional status and/or changes in growth (weight and height)

Very low-quality evidence from 1 pre- and post-intervention cohort study with 8 children showed that there was a clinical benefit of behavioural intervention (including Beckman oral exercises and parent education) for the outcome of weight and height percentiles in centimetres at 1 year, but the 95% confidence intervals were not reported, therefore, the uncertainty around this effect was unclear.

13.5.4.2. Time taken to feed

Very low-quality evidence from 1 pre- and post-intervention cohort study with 8 children showed that children were able to tolerate a longer meal session over the course of treatment after the behavioural intervention (including Beckman oral exercises and parent education), but the 95% confidence intervals were not reported, therefore, the uncertainty around this effect was unclear.

13.5.4.3. Psychological wellbeing of parents and/or carers

No evidence was retrieved for this outcome.

13.5.4.4. Acceptability of programme

No evidence was retrieved for this outcome.

13.5.4.5. Survival

No evidence was retrieved for this outcome.

13.5.5. Six session training programme

13.5.5.1. Nutritional status and/or changes in growth (weight)

Very low-quality evidence from 1 cohort study with 22 children showed that there was a clinical benefit of a 6-session training programme for the outcome of weight z scores at 4 to 6 months but the 95% confidence intervals were not reported, therefore the uncertainty around this effect was unclear.

13.5.5.2. Respiratory health

Very low-quality evidence from 1 cohort study with 22 children showed that there was a clinical benefit of a 6-session training programme in reducing the frequency of chest-related illness occurring at least once every 3 months at 4 to 6 months.

13.5.5.3. Psychological wellbeing of parents and/or carers

No evidence was retrieved for this outcome.

13.5.5.4. Acceptability of programme

No evidence was retrieved for this outcome.

13.5.5.5. Survival

No evidence was retrieved for this outcome.

13.5.6. Oral sensorimotor stimulation

13.5.6.1. Physical function of the oropharyngeal mechanism

13.5.6.2. Change in diet consistency a child is able to consume

No evidence was retrieved for this outcome.

13.5.6.3. Psychological wellbeing of parents and/or carers

No evidence was retrieved for this outcome.

13.5.6.4. Acceptability of programme

No evidence was retrieved for this outcome.

13.5.6.5. Survival

No evidence was retrieved for this outcome.

13.6. Evidence to recommendations

13.6.1. Relative value placed on the outcomes considered

The aim of this review was to assess the clinical and cost effectiveness of interventions for managing difficulties with eating, drinking, and swallowing in children and young people with cerebral palsy. The Committee indicated the following to be the critical outcomes of this evidence review: change in height and weight, respiratory health, and duration of meal times/child’s participation in meal. All other outcomes were reported as important if retrieved from the search.

13.6.2. Consideration of clinical benefits and harms

The Committee recognised that the results were inconclusive and most of the studies were of low- and very low-quality evidence. The Committee agreed that the interventions included in some of the studies would not be able to be replicated in everyday clinical practice as a very high level of staff training was needed; the programmes were of an intensity and duration that would be atypical; and they were mainly conducted in research settings and some used outdated methods (e.g., Ottenbacher 1981). Most of the studies focused on improving oral motor skills, with variation in the attention given to other important outcomes such as growth, weight gain, nutrition, respiratory health, independence skills and the time taken for meals. The Committee noted that some studies focused their analysis on improving physical function, using results such as mouth closure, lip closure onto utensil, lip closure and control of food during eating, drinking and swallowing, and straw suction. However, the Committee highlighted this as a possible limitation of the evidence, as these outcomes are only clinically meaningful if other aspects of feeding (nutrition and respiratory status) are also improving. Also, they pointed out that improvements in areas of physical function did not necessarily reflect the time taken to eat each spoonful. No studies explored the impact of intervention on the experience of children and young people, or their caregivers.

The Committee discussed that, in clinical practice, management of eating and drinking typically involved multiple professionals and addressed several aspects of eating, drinking and swallowing, often simultaneously. Based on the Committee’s experience and by consensus they agreed that it was common to address positioning and postural management. Following this, further interventions may be considered, including modifying food and fluid textures, feeding techniques and other strategies to reduce the risk of aspiration, to improve the efficiency of eating, drinking and swallowing and to promote the development of oral motor skills. Behavioural and emotional aspects of eating and drinking contribute to the family experience and so need consideration. This could be psychological interventions but the Committee thought it was more about allowing feeding and mealtimes to be a pleasant and sociable experience for the family – not painful, lengthy, disrupted and stressful.

The Committee recognised that the most relevant training programme in the studies included in the evidence review was a 6-session training programme that addressed many of the aspects listed above. However, the study included only provided 6 sessions and the Committee noted that, in clinical practice, a longer duration of professional involvement was likely as a child or young person’s abilities and needs would change over time. As a child moved though nursery, school, further education and social care, the circle of people supporting an individual’s eating and drinking would change, with a need for training.

The Committee noted that the interventions addressed in the studies were often focused on improving specific oral motor skills involved in eating or drinking such as mouth closure or lateral tongue movement, while in clinical practice, multiple aspects of eating, drinking and swallowing are considered. For example, moving a child onto more challenging food textures, through focusing on strategies to improve chewing, may represent progress with regard to oral motor skills, but may increase the length of the mealtime or compromise energy intake as each mouthful will take longer and need more effort.

Based on their clinical experience and by consensus, the Committee unanimously agreed to highlight the importance of working in partnership with children and young people with cerebral palsy, their parents and/or carers to address skill development, health-related outcomes (growth and weight gain, respiratory health) and improve the mealtime experience. The Committee recognised the risk of faltering growth in the cerebral palsy population, arising from eating, drinking and swallowing difficulties. Monitoring changes in height and weight was noted as a key outcome in management programmes. Based on specific principles retrieved in the evidence and on their clinical experience, the Committee recommended the development of an individualised plan tailored towards the people involved in the feeding of children and young people with cerebral palsy and eating, drinking and swallowing difficulties and listed a range of management techniques that could be taken into account, such as postural management, texture modification, specialised feeding utensils, communication and the training needs of caregivers.

The Committee noted that studies included in the evidence review did not provide clear support for the use of intra-oral devices. Therefore, they agreed that parents and/or carers should be made aware of this and they agreed a recommendation to that effect. Further to this, the Committee noted that intra-oral appliances such as ISMAR were not widely provided in clinical practice and there was low-quality evidence showing that ISMAR was less effective than control (no ISMAR). Additionally, the importance of considering oral-motor skills within the broader context of health and social aspects of eating and drinking was not widely understood.

The Committee considered that there was very limited evidence to guide care. There were few comparisons of interventions and limited data on the natural history of eating and drinking and swallowing disorders. Furthermore, the research on oral-motor therapies was conflicting. Therefore, the Committee agreed to develop a research recommendation on interventions to improve eating, drinking and swallowing in children and young people with cerebral palsy.

13.6.3. Consideration of economic benefits and harms

The Committee were highly concerned that the oral motor sensorimotor regimens used in some trials included in the clinical evidence review were very intensive, involving hospital admission, or taking time out of the school day. These regimens were considered too burdensome to undertake outside of the research setting, especially if they were done for many months or years. Moreover, the Committee noted that oral motor treatments included in the trials did not aim to manage all aspects of eating, drinking and swallowing difficulties.

The Committee agreed that parents, carers or school staff could be trained to use some oral motor techniques, which would reduce the cost of healthcare professionals, but the efficacy of such training packages has not been assessed. The acceptability to children, families, carers and schools has also not been explored.

It should be noted that the ISMAR intra-operative device is a specific type of intra-oral device, and in the UK the equivalent would be a palatal training aid/device (PTD), which is made individually for each child by the orthodontist or at a specialist centre. The high cost and high skill to fit this device was recognised by the Committee. The SLT and orthodontist would be involved in reviewing PTD (to check functional impact and the fit, approximately every 4 to 6 months) long term. The study included in the review that looked at ISMAR was conducted over 12 to 24 months, but it is possible that in clinical practice some children may abandon the PTD quickly if it is not comfortable or is not effective, whereas others may use the appliance for a number of years. The Committee agreed that a high proportion of patients who try intra-oral appliances will not tolerate them, consequently, the cost effectiveness of oral motor devices will depend largely on patient preference. Furthermore, the Committee believed oral motor devices are primarily used to manage saliva control and are not typically used to manage eating, drinking and swallowing difficulties and questioned their cost effectiveness for this indication.

Overall, the Committee were unable to recommend any specific intervention because any plan would be individualised to the child and young person and may involve several interventions. Despite this, the Committee agreed that factors such as postural management, texture modification and feeding techniques would be considered before initiating exercise programmes or pharmacological treatment. As a result, the least expensive and intensive interventions would be implemented first.

The plan would be reviewed regularly by a healthcare professional to ensure the child and young person and family and/or carer were satisfied with the plan and to modify the plan as necessary. Ideally this would take place in the child and young person’s home and school environment. Consequently, monitoring costs would be incurred regardless of the interventions included in the plan.

The Committee was aware that a plan to manage eating, drinking and swallowing difficulties could not include an intervention if there was no-one there to implement it accurately. For this reason, training costs may be incurred by the school, carers and/or families if the healthcare professional believes the skills of the people who support the child’s eating and drinking are inadequate.

13.6.4. Quality of evidence

Four randomised trials and 4 cohort studies were included in the evidence review. The quality of the evidence for this review ranged from very low to low. Main reasons of bias were: lack of information on the randomisation method used, concealment of allocation unreported or unclear, lack of blinding of investigators and also the blinding of participants due to the type of intervention being administered. The sample sizes of most of the studies was small, except 1 randomised trial that included 81 participants, which increased the uncertainty around the effects of the reviewed interventions In addition, it was not possible to conduct a meta-analysis of studies because of the differences in interventions.

13.6.5. Other considerations

It was noted that the validity of 1 of the outcome measures used in several studies was questioned. The Functional Feeding Assessment is a subscale of the Multidisciplinary Feeding Profile.

The Beckman oral motor exercise programme is not common in UK practice but other oral-motor exercise regimens do exist (for example TalkTools, or individually tailored oral motor programmes developed by speech and language therapists). There is considerable variation in practice across speech and language therapists in the UK with regard to the degree of individualisation, frequency of review, intensity of practice, and duration of such oral motor programmes. Data on compliance with treatment and acceptability of programmes to children, young people and families and/or carers is lacking.

The recommendations related to this evidence review were based on the evidence and the Committee’s clinical experience.

13.6.6. Key conclusions

The Committee concluded that interventions to improve eating, drinking and swallowing in children and young people with cerebral palsy need input from multiple professionals. Studies in this area have largely focused on the development of oral motor skills with limited attention to the impact of interventions on other outcomes such as growth, weight gain, nutritional status, respiratory health, independence, the time taken for meals and the experience of the children and young people and their families and/or carers. The key areas of postural management, food and fluid modification, environmental adaptations and carer training have received limited consideration to date.

13.7. Recommendations

50.: Develop strategies and goals in partnership with the child or young person with cerebral palsy and their parents, carers and other family members for interventions to improve eating, drinking and swallowing.

51.

Create an individualised plan for managing eating, drinking and swallowing difficulties in children and young people with cerebral palsy, taking into account the understanding, knowledge and skills of parents, carers and any other people involved in feeding the child or young person. Assess the role of the following:

postural management and positioning when eating
modifying fluid and food textures and flavours
feeding techniques, such as pacing and spoon placement
equipment, such as specialised feeding utensils
optimising the mealtime environment
strategies for managing behavioural difficulties associated with eating and drinking
strategies for developing oral motor skills
communication strategies
modifications to accommodate visual or other sensory impairments that affect eating, drinking and swallowing
the training needs of the people who care for the child or young person particularly outside the home.

52.: Advise parents or carers that intra-oral devices have not been shown to improve eating, drinking and swallowing in children and young people with cerebral palsy.

53.

Use outcome measures important to the child or young person and their parents or carers to review:

whether individualised goals have been achieved
the clinical and functional impact of interventions to improve eating, drinking and swallowing.

13.8. Research recommendations

3.: What is the clinical and cost effectiveness and safety profile of interventions to improve eating, drinking and swallowing in children and young people with cerebral palsy?

Table 56

Research recommendation rationale.

Table 57

Research recommendation statements.

Publication Details

Copyright

Publisher

National Institute for Health and Care Excellence (NICE), London

NLM Citation

National Guideline Alliance (UK). Cerebral palsy in under 25s: assessment and management. London: National Institute for Health and Care Excellence (NICE); 2017 Jan. (NICE Guideline, No. 62.) 13, Management of eating, drinking and swallowing difficulties.

Table 49Summary of included studies

Study	Intervention/comparator	Population	Outcomes	Comments
Adams 2012	Six sessions of training programme: consisted of education on dietary intake, ease and efficiency of eating, utensils, behaviour of caregiver towards feeding child, postural and physical support for positioning and self-feeding. Each training session included educational content as well as supervised feeding. Teaching methods included traditional pedagogy, discussion, participation and experimental activities, use of visual aids, including a 20-minute video drama created especially for the programme.	Children with moderate to severe (levels III to Von GMFCS) CP and their caregivers N:37 caregivers and their children Age of children (range): 19 to 129 months.	Weight for age (WAZ score). Frequency of chest related illness (n).	Study was conducted in slums of Dhaka, Bangladesh for 4 to 6 months. Cohort study
Baghbadorani 2014	Oral sensorimotor stimulation: focused on tongue lateralisation, lip control, and vigour of chewing. Treatment lasted 15 minutes daily, 3 days a week. Assessments were carried out at 4 and 8 weeks. Tongue lateralisation: a small amount of jam was placed on 4 corners of the lips alternatively (left and right corner and middle of upper and lower lips so the tongue had to remove the stimulus from outside the oral cavity). In order to stimulate the tongue in the mouth, the stimulus was placed in the cheek pocket so that the tongue had to remove it from the cheek in order to swallow.	Children with moderate to severe motor impairment who scored at or below 10 scores on an initial assessment of the Oral Motor Assessment Scale. N:12 Age of children (range): 2 to 7 years.	Effect of oral motor stimulation on oral- motor skills.	Baseline, 4 and 8 weeks Cohort study
Clawson 2007	A hospital-based 6 hour-per day programme, Monday through Friday, for an average of 29 treatment days (5.8 weeks). The focus of the study was behavioural interventions and parent education in addition to an oral-motor exercise component, to address the child’s food refusal. Behavioural interventions: presentation of food near child’s lips until child opened and accepted the bite into their mouth (accepting food, chewing, swallowing). Parent training: involved training in food preparation and calorie boosting (puree, texture grading and food allergies). Beckman oral motor exercise: each therapeutic meal included oral motor exercises followed by oral feeding. The day programme was provided by the MDT. Beckman oral motor exercises were done (by the same staff members throughout admission) for 20 to 30 minutes before each oral feeding. The aim was to increase functional response to pressure and movement (increase range, strength, variety and control of movement for lips, cheeks, jaw and tongue). Caregivers fed the child in the room alone and were observed by the therapist via video and instructed the parent via a wireless communication system.	The diagnosis was moderate to severe feeding difficulties, all children had spastic quadriplegic CP. There is no information about the severity of the CP (no GMFCS level). N:8. Age of children (range): 18 months to 4.7 years.	Mean height and weight (percentiles). Patients were scheduled to return for assessment at 1,4, 7 and 12 months following discharge from the day of the feeding programme. Other measures were reported at discharge, but not follow-up, for example, food acceptance, mouth clearance, inappropriate behaviours, duration of meal, grams per meal, calories consumed, and percent tube fed.	Cohort study
Gisel 1995	Oral sensorimotor therapy versus routine therapy: based on children’s performance on the modified Functional Feeding Assessment (tailored to children’s individual needs). Treatment lasted 5 to 7 minutes daily, before lunch or snack. Tongue lateralisation, lip control and vigour of chewing were the main focus of oral-motor functioning. Small food stimuli were used to elicit a natural eating reaction. Demonstrations of sucking motions were given and children were encouraged to imitate the motion and to suck a liquid. Children with poor sucking control were given thickened liquids. Vigour of chewing: children were encouraged to chew by the therapist placing small pieces of biscuit (medium to strong resistance) over the molars (alternatively right and left).	Children with a diagnosis of CP with moderate to severe motor impairment. N:27. Age of children (mean, SD): Group 1:4.8 (1.4) Group 2: 5.0 (1.9)	Mean weight (percentiles for age). Duration of lunch/snack at school. Time taken to eat foods of standard texture.	Open label trial. Outcome data reported at 10 weeks.
Gisel 1996	Oral sensorimotor therapy versus routine therapy: based on children’s performance on the modified Functional Feeding Assessment (tailored to children’s individual needs).	Children with a diagnosis of CP with moderate to severe motor impairment. N:35. Age of children (range): 4.3 to 13.3 years.	Weight in kg and percentiles for age. Eating time for 3 standard food textures. Duration of lunch at school.	Open label study. Outcome data reported at 10 weeks.
Gisel 2001	ISMAR appliance versus no ISMAR appliance: ISMARs were fabricated and, if satisfactory, were then fitted on the child in school environment, in the presence of caregivers. Care and written wear instructions were provided. During the first week, the research assistant contacted caregivers to ensure safety and correct wear. – Treatment phase I: onset of phase I was determined by ISMAR wear for 20 minutes of wear daily. – Treatment phase II: children were evaluated for mobilisation of oral structures, and goals were determined for each child according to their needs. Grooves were drilled into the lingual part of the occlusal shelves or heads attached to different loci on the ISMAR appliance to stimulate tongue movement.	Children and young people had a diagnosis of CP with tetraparesis and moderate motor impairment. N:17. Age of children and young people (range): 6.6 to 15.4 years.	Weight and height. Competency in feeding: spoon feeding, biting, chewing, cup drinking, straw drinking, swallowing, clearing.	Follow-up study (cohort study).
Ottenbacher 1981	Oral motor therapy: each participant received 30 to 40 minutes of therapy daily, 5 days a week for 9 weeks. Some participants received therapy just prior to or in conjunction with their meals, and others were scheduled for therapy at various times during the day. There were 3 major components to the treatment: inhibition of abnormal oral and postural reflexes facilitation of normal muscle tone desensitisation of the oral region. Control group: Participants received their regular programme of therapy and education. No specific treatment of oral-motor dysfunction or feeding disorders was administered.	Children with profound or severe neuromotor disorder, with dependency in most areas of self-care and feeding N=18/20 participants with CP Age of children (mean, SD): 11.5(4.38)	Weight at pre-therapy and post-therapy	RCT 9 weeks
Sigan 2013	Multi-component intervention: postural management, texture modification, feeding techniques and oral sensorimotor treatment versus routine physiotherapy. Feeding therapy group: 1-hour therapy sessions with a physiotherapist once a week for 6 months (12 sessions in total). Parents continued techniques between therapy sessions. Routine treatment group: Children diagnosed with CP and oral motor dysfunction were called for the first evaluation and then for and evaluation at 6 months. During this time, routine physiotherapy was continued. All patients attended routine physiotherapy according to the established programme during the 6 months.	Children with CP (bilateral UL=LL and LL<UL; unilateral CP, hypotonia, and ataxic). N:81 (consecutively chosen). Age of children (range): 12 to 42 months.	Physical function (spoon feeding, drinking and swallowing subscales of the modified Functional Feeding Assessment).	Single-centre RCT conducted in Turkey, for 6 months’ duration.

: CP cerebral palsy, RCT randomised controlled trial, SD standard deviation, GMFCS Gross Motor Function Classification System, ISMAR Innsbruck sensorimotor activator and regulator, Kg Kilograms.

Table 50Oral sensorimotor therapy versus routine treatment

: CP cerebral palsy, RCT randomised controlled trial, SD standard deviation.

*: The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval [CI]) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

1: Open label randomised trial.

2: The evidence was downgraded by 2 because of selection bias and performance bias.

3: The evidence was downgraded by 2 because of very serious heterogeneity (Chi-squared p<0.1, I-squared inconsistency statistic of 75%) and no plausible explanation was found with subgroup analysis.

4: Evidence was downgraded by 1 because of 95% confidence interval crossing 1 default MID (-0.5 to +0.5 SD).

5: Majority of evidence has only 1 indirect aspect of PICO (population)

6: Evidence was downgraded by 1 because of serious heterogeneity (chi-squared p<0.1, I-squared inconsistency statistic of 50%-74.99%) and no plausible explanation was found with sensitivity analysis.

7: The evidence was downgraded by 2 because of 95% confidence interval crossing 2 default MIDs −0.5 and +0.5 SDs.

8: The evidence was downgraded by 1 because of performance bias.

Table 51ISMAR versus no ISMAR treatment

Outcomes	Illustrative comparative risks^* (95% CI)		No of Participants (studies)	Quality of the evidence (GRADE)	Comments
Outcomes	Assumed risk	Corresponding risk	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	No	ISMAR
Weight (kg)	-	The mean change in weight in the intervention group was MD 0.87 higher (0.2 to 1.54 higher)	17 (1 study)	Very low¹^,²^,³	Change at 6 months Cohort study
Weight (kg)	-	The mean change in weight in the intervention group was MD 1.44 lower (1.89 to 0.99 lower)	17 (1 study)	Very low¹^,²^,³	Change at 12 months Cohort study
Height (cm)	-	The mean height in the intervention group was MD 0.15 lower (2.06 lower to 1.76 higher)	17 (1 study)	Very low¹^,²^,³	Final at 6 months Cohort study
Height (cm)	-	The mean height in the intervention group was MD 2.68 higher (1.21 to 4.15 higher)	17 (1 study)	Very low¹^,²^,³	Final at 12 months Cohort study
Competency in feeding (%) - spoon feeding	-	The mean percentage competency in the intervention group was MD 5.8 lower (16.64 lower to 5.04 higher)	17 (1 study)	Very low1,2	Final at 12 to 18 months Cohort study
Competency in feeding (%) - cup drinking	-	The mean percentage competency in the intervention group was MD 1.9 lower (10.09 lower to 6.29 higher)	17 (1 study)	Very low¹^,²^,³	Final at 12 to 18 months Cohort study
Competency in feeding (%) - swallowing	-	The mean percentage competency in the intervention group was MD 16 lower (32.08 lower to 0.08 higher)	17 (1 study)	Very low¹^,²^,³	Final at 12 to 18 months Cohort study
Competency in feeding (%) - clearing	-	The mean percentage competency in the intervention group was MD 15.5 lower (31.03 lower to 0.03 higher)	17 (1 study)	Very low¹^,²^,³	Final at 12 to 18 months Cohort study
Competency in feeding (%) - spoon feeding	-	The mean percentage competency in the intervention group was MD 2.5 lower (14.97 lower to 9.97 higher)	17 (1 study)	Very low¹^,²^,³	Final at 18 to 24 months Cohort study
Competency in feeding (%) - cup drinking	-	The mean percentage competency in the intervention group was MD 2.5 lower (14.97 lower to 9.97 higher)	17 (1 study)	Very low¹^,²^,³	Final at 18 to 24 months Cohort study
Competency in feeding (%) - swallowing	-	The mean percentage competency in the intervention group was MD 19 lower (32.66 to 5.34 lower)	17 (1 study)	Very low¹^,²^,³	Final at 18 to 24 months Cohort study
Competency in feeding (%) - clearing	-	The mean percentage competency in the intervention group was MD 13.9 lower (24.27 to 3.53 lower)	17 (1 study)	Very low¹^,²^,³	Final at 18 to 24 months Cohort study
Competency in feeding (%) - spoon feeding	-	The mean percentage competency in the intervention group was MD 2.7 higher (2.85 lower to 8.25 higher)	17 (1 study)	Very low¹^,²^,³	Change at 12 to 18 months Cohort study
Competency in feeding (%) - cup drinking	-	The mean percentage competency in the intervention group was MD 3.3 higher (6.26 lower to 12.86 higher)	17 (1 study)	Very low¹^,²^,³	Change at 12 to 18 months Cohort study
Competency in feeding (%) - swallowing	-	The mean percentage competency in the intervention group was MD 3.5 lower (15.62 lower to 8.62 higher)	17 (1 study)	Very low¹^,²^,³	Change at 12 to 18 months Cohort study
Competency in feeding (%) - clearing	-	The mean percentage competency in the intervention group was MD 4 lower (15.89 lower to 7.89 higher)	17 (1 study)	Very low¹^,²^,³	Change at 12 to 18 months Cohort study
Competency in feeding (%) - spoon feeding	-	The mean percentage competency in the intervention group was MD 0.8 higher (6.96 lower to 8.56 higher)	17 (1 study)	Very low¹^,²^,³	Change at 18 to 24 months Cohort study
Competency in feeding (%) - cup drinking	-	The mean percentage competency in the intervention group was MD 9.6 lower (14.23 to 4.97 lower)	17 (1 study)	Very low¹^,²^,³	Change at 18 to 24 months Cohort study
Competency in feeding (%) - swallowing	-	The mean percentage competency in the intervention group was MD 2.2 lower (11.43 lower to 7.03 higher)	17 (1 study)	Very low¹^,²^,³	Change at 18 to 24 months Cohort study
Competency in feeding (%) - clearing	-	The mean percentage competency in the intervention group was MD 3.6 higher (7.96 lower to 15.16 higher)	17 (1 study)	Very low¹^,²^,³	Change at 18 to 24 months Cohort study

: CP cerebral palsy, RCT randomised controlled trial, SD standard deviation, MD mean difference, GMFCS Gross Motor Function Classification System, ISMAR Innsbruck sensorimotor activator and regulator.

*: The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval [CI]) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

1: The evidence was downgraded by 1 because of performance bias.

2: The evidence was downgraded by 1 because of 95% CI crossing 1 default MID (-0.5 to +0.5 SDs).

3: The evidence was downgraded by 2 because of 95% CI crossing 2 default MIDs (-0.5 to +0.5 SDs).

Table 52Multi-component intervention compared to routine physiotherapy

Outcomes	Illustrative comparative risks^* (95% CI)		No of Participants (studies)	Quality of the evidence (GRADE)
Outcomes	Assumed risk	Corresponding risk	No of Participants (studies)	Quality of the evidence (GRADE)
	Routine physiotherapy	Multi-component intervention
Physical function - spoon feeding FFAm Follow-up: 6 months	-	The mean physical function - spoon feeding in the intervention groups was 8.85 higher (1.55 to 16.15 higher)	81 (1 study)	Low¹^,²
Physical function - swallowing FFAm Follow-up: 6 months	-	The mean physical function - swallowing in the intervention groups was 8.4 higher (1.54 to 15.26 higher)	81 (1 study)	Low¹^,²
Physical function - drinking FFAm Follow-up: 6 months	-	The mean physical function - drinking in the intervention groups was 4.13 higher (1.12 to 7.14 higher)	81 (1 study)	Low¹^,²

: FFAm Modified Functional Assessment Scale

*: The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval [CI]) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

1: The evidence was downgraded by 1 due to performance bias.

2: Evidence was downgraded by 1 due to 95% confidence interval crossing 1 default MID (-0.5 to +0.5 SD).

Table 53Parent and/or carer training sessions

Outcomes	Illustrative comparative risks^* (95% CI)		No of Participants (studies)	Quality of the evidence (GRADE)	Comments
Outcomes	Assumed risk	Frequency/mean	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	Placebo	Training session
Frequency of chest-related illnesses at least once every 3 months	-	6	22 (1 study)	Very low¹^,²^,³^,⁴	Outcome at 4 to 6 months
Weight for age (WAZ score)	-	mean 4.07 (SD 2.45)	22 (1 study	Very low¹^,²^,³^,⁴	Final Outcome at 4 to 6 months

: WAZ weight for age, SD standard deviation.

*: The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval [CI]) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

1: The evidence was downgraded by 2 due to performance, attrition and detection bias.

2: The evidence was downgraded by 1 due to study setting in Bangladesh.

3: Not calculable.

4: The absolute risk could not be calculated as there was no comparator group in the study.

Table 54Multi-component intervention (including Beckman oral exercise training, behavioural intervention and parenting training)

Outcomes	Illustrative comparative risks^* (95% CI)		No of Participants (studies)	Quality of the evidence (GRADE)
Outcomes	Assumed risk	Corresponding risk	No of Participants (studies)	Quality of the evidence (GRADE)
	Control	Multi-component intervention
Height Percentile Follow-up: 1 years	The mean height in the control groups was measured using an Infantometres height board	The mean (SD) height in the intervention groups was 16.13 higher (17.08)	8 (1 study)	Very low¹^,²
Weight Percentile Follow-up: 1 years	-	The mean (SD) weight in the intervention groups was 10.28 higher (15.41)	8 (1 study)	Very low¹^,²
Length of food time/time taken to feed Minutes Follow-up: 5.8 weeks	-	The mean (SD) length of food time/time taken to feed in the intervention groups was 17.83 higher (2.06)	8 (1 study)	Very low¹^,²

: SD standard deviation.

*: The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval [CI]) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

1: The evidence was downgraded by 1 due to performance bias.

Table 55Oral sensorimotor stimulations

Outcomes	Illustrative comparative risks^* (95% CI)		No of Participants (studies)	Quality of the evidence (GRADE)	Comments
Outcomes	Assumed risk	Corresponding risk	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	Control	Oral motor assessment
Mouth closure	-	Mean 1.33 (SD 1.15)	12 (1 study)	Very low¹	Change at 2 months
Lip closure onto utensil	-	Mean 0.66 (SD 0.77)	12 (1 study)	Very low¹	Change at 2 months
Lip closure during deglutition	-	Mean 0.5 higher (SD 0.67)	12 (1 study)	Very low¹	Change at 2 months
Control of food during deglutition	-	Mean 1 (SD 0.73)	12 (1 study)	Very low¹	Change at 2 months
Straw suction	-	Mean 0.41 (SD 0.51)	12 (1 study)	Very low¹	Change at 2 months
Control of liquid during deglutition	-	Mean 0.75 (SD 0.45)	12 (1 study)	Very low¹	Change at 2 months
Mastication	-	Mean 1 (SD 0.85)	12 (1 study)	Very low¹	Change at 2 months
Mouth closure	-	Mean 2.41 (SD 0.51)	12 (1 study)	Very low¹	Final at 2 months
Lip closure onto utensil	-	Mean 1.75 (SD 0.62)	12 (1 study)	Very low¹	Final at 2 months
Lip closure during deglutition	-	Mean 1.66 (SD 0.49)	12 (1 study)	Very low¹	Final at 2 months
Control of food during deglutition	-	Mean 1.91 (SD 0.28)	12 (1 study)	Very low¹	Final at 2 months
Straw suction	-	Mean 0.83 (SD 0.93)	12 (1 study)	Very low¹	Final at 2 months
Control of liquid during deglutition	-	Mean 1.5 (SD 0.52)	12 (1 study)	Very low¹	Final at 2 months
Mastication	-	Mean 1.91 (SD 0.28)	12 (1 study)	Very low¹	Final at 2 months
Overall score	-	Mean 12 (SD 1.59)	12 (1 study)	Very low¹	Final at 2 months

: SD standard deviation.

*: The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval [CI]) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

1: The evidence was downgraded by 1 due to performance and detection bias.

Table 56Research recommendation rationale

Research question	What is the clinical and cost effectiveness and safety profile of interventions to improve eating, drinking and swallowing in children and young people with cerebral palsy?
Why this is needed
Importance to ‘patients’ or the population	Children and young people with cerebral palsy may have eating, drinking and swallowing (EDS) difficulties leading to poor nutritional status, aspiration, respiratory infections, hospital admissions and reduced life expectancy. This can impact significantly on families and carers, although they are often keen to maintain oral intake.
Relevance to NICE guidance	The research is essential to inform future updates of key recommendations in the areas of eating, drinking and swallowing, optimising nutrition, and social care needs in the guidance.
Relevance to the NHS	Children and young people with severe swallowing difficulties or inadequate oral intake may require non-oral tube feeding. Surgical placement of gastrostomy/jejunostomy carries a level of risk. Behavioural interventions include postural management, modification of food and fluid textures, feeding techniques and equipment, therapies aimed at improving oral-motor skills and reducing the risk of aspiration. Such interventions require specialist input from speech and language therapists and other members of the MDT. Carer/family training and support also requires resources.
National priorities	NHS Outcomes Framework (April 2016). The management of eating, drinking impacts on the following domains: Preventing people from dying prematurely: children and people with learning disability Enhancing quality of life for people with long term conditions: improving functional abilities, and enhancing quality of life for carers
Current evidence base	Very limited evidence is available to guide care. Both surgical and behavioural intervention studies are typically small, short term and with uni-dimensional outcomes. There are few comparisons of interventions and limited data on the natural history of eating and drinking and swallowing disorders. The research on oral-motor therapies is conflicting. No data is available on the impact of behavioural interventions on participation.
Equality	For most children and young people eating and drinking is an enjoyable experience, undertaken several times a day. Meals and snacks serve the purpose of obtaining nutrition and hydration, but also provide a context for social interaction. Difficulties with eating and drinking and swallowing results in medicalisation of an everyday activity and reduces opportunities for participation, with families reporting reluctance to eat in public. Swallowing difficulties contribute to health inequalities.
Feasibility	Longer term, comprehensive case series and prospective cohort studies exploring the impact of behavioural as well as surgical interventions are required. Multidimensional outcomes should explore changes in nutritional status, respiratory health, and psychosocial impact on children or young people and families. Multi-site, school-based studies, with carer/family involvement, may be more feasible than studying outpatient feeding clinic populations. The main ethical issue would be no or delayed treatment for children and young people at risk of respiratory compromise or under-nutrition.
Other comments	The NIHR Research for Patient Benefit Programme has supported the development of an Eating and Drinking Abilities Classification System in cerebral palsy, and a study on how services meet the psychosocial support needs of children and young people with feeding difficulties and their families is currently funded.

Table 57Research recommendation statements

Criterion	Explanation
Population	Children and young people with cerebral palsy
Intervention	Change from background MDT input to individualised, targeted programmes including a variety of the following: Behavioural interventions include postural management, modification of food and fluid textures, Feeding techniques and equipment, Specific dysphagia therapies used in current clinical practice aimed at improving oral-motor skills and reducing the risk of aspiration. Pharmacological
Comparator	Population without the above
Outcome	Anthropometrics (including linear growth, weight, skin fold thickness) and developmental outcome Reduced risk of aspiration Change in diet consistency a child is able to consume QoL including enjoyment of eating and drinking and swallowing Child’s level of participation in mealtime Psychological wellbeing of parents or carers Acceptability of programme
Study design	Comparative cohort
Timeframe	2 years