- Top of page
- What this paper adds
Aim To report the prevalence, clinical associations, and trends over time of oromotor dysfunction and communication impairments in children with cerebral palsy (CP).
Method Multiple sources of ascertainment were used and children followed up with a standardized assessment including motor speech problems, swallowing/chewing difficulties, excessive drooling, and communication impairments at age 5 years.
Results A total of 1357 children born between 1980 and 2001 were studied (781 males, 576 females; median age 5y 11mo, interquartile range 3–9y; unilateral spastic CP, n=447; bilateral spastic CP, n=496; other, n=112; Gross Motor Function Classification System [GMFCS] level: I, 181; II, 563; III, 123; IV, 82; IV, 276). Of those with ‘early-onset’ CP (n=1268), 36% had motor speech problems, 21% had swallowing/chewing difficulties, 22% had excessive drooling, and 42% had communication impairments (excluding articulation defects). All impairments were significantly related to poorer gross motor function and intellectual impairment. In addition, motor speech problems were related to clinical subtype; swallowing/chewing problems and communication impairments to early mortality; and communication impairments to the presence of seizures. Of those with CP in GMFCS levels IV to V, a significant proportion showed a decline in the rate of motor speech impairment (p=0.008) and excessive drooling (p=0.009) over time.
Interpretation These impairments are common in children with CP and are associated with poorer gross motor function and intellectual impairment.
Cerebral palsy (CP) refers to a collection of disorders of movement and posture caused by damage to, or malformation of, the developing brain in early life. It is the leading cause of significant motor impairment in children and can be further complicated by the presence of associated impairments. Communication disorders in CP are thought to be commonplace1 and the result of a complex interplay between motor impairment, swallowing and drooling difficulties, intellectual impairment, and vision and hearing impairment, which can further limit the child’s abilities. Oromotor impairments can also be associated with one another. So difficulties swallowing can be associated with excessive drooling,2 and excessive drooling may be associated with dysarthria.3
The impact of communication problems in the life of the child with CP and their family has far-reaching consequences and is associated with reduced participation in everyday activities,4 reduced quality of life in ‘relationships with parents’,5 and problems with psychological adjustment.6 In addition, poor speech intelligibility is associated with restricted communication between child and parent,7 and excessive drooling with a negative impact on the child’s interaction with peers and on their self-esteem,8 with an increase in their daily care needs.9 Finally, difficulties swallowing and chewing can lead to prolonged feeding times, increased strain on the caregiver,10 and impaired growth.11
Despite this, the prevalence of oromotor and communication disorders in CP is unknown1 and may be related to the severity of motor impairment,10 although not necessarily.3 Many of the studies so far rely on relatively small numbers of children with CP recruited from incomplete sources of ascertainment like clinics12 or they have been based on subsets of children, thus limiting generalizability; or they have included young children where a diagnosis of CP might not have been confirmed. For example Reilly et al.13 found oromotor dysfunction affected 90% of a small sample of children with CP aged 12 to 72 months. Dysphagia has been estimated to occur in 27% of children with CP,14 increasing to 99% in a sample of severely affected children with intellectual impairment.15 Drooling has been reported to occur in 78% of children with CP,14 with an estimated 18 to 28% experiencing severe drooling,16,17 increasing to 49% in children with severe CP and intellectual impairment.15
The aim of this paper is to report on the prevalence of oromotor dysfunction (motor speech problems, swallowing/chewing difficulties, excessive drooling) and communication impairments (expressive speech and language difficulties excluding articulation defects), to quantify associations with other clinical and sociodemographic characteristics, and to investigate trends over time in a geographically defined population of children with CP.
- Top of page
- What this paper adds
This study is based on the analysis of children notified to the Northern Ireland Cerebral Palsy Register (NICPR) during 1992 to 2009. Cases were defined as ‘early-onset CP’ where a lesion or malformation to the developing brain occurred sometime before, during, or soon after birth but no later than the 28th day of life after birth. ‘Late onset CP’ was defined as damage or a lesion to the developing brain occurring sometime after the 28th day of life but by the child’s fifth birthday.
The method of the NICPR has been described elsewhere.18 However, in summary, multiple and overlapping sources of ascertainment are used. Each child notified to the NICPR is followed up with a standardized assessment form for recording motor deficits of central origin.19 Young children are followed up ideally at age five before being confirmed as a patient. When the Register began in 1992, patients born between 1980 and 1986 were ascertained retrospectively whereas those born between 1987 and 2001 were ascertained prospectively. Quarterly returns about children who have died or emigrated are notified to the NICPR by the Central Services Agency (responsible for registrations with general practitioners). Patients who die before the age of 1 year are excluded. For this paper the census date for deaths, emigration, and immigration was 1st June 2009.
CP subtype was defined and classified according to the Surveillance of Cerebral Palsy in Europe project.20 Patients with spasticity were further subdivided into those having unilateral and bilateral spastic CP. Gross motor function was classified using the Gross Motor Function Classification System (GMFCS).21 This was done restrospectively by applying an algorithm to registry data, the validity of which was found to be satisfactory.22 Intellectual impairment was defined as present where the IQ was less than 70, considered ‘moderate’ if the IQ was 50 to 70, and ‘severe’ if the IQ was <50. This assessment was based on IQ where available or best clinical estimate. The presence of seizures related to the 12 months preceding completion of the standardized assessment form. Deprivation quintile (ranging from least deprived to most deprived) was derived from address at residence and by linking postcode to electoral ward, and ward to deprivation scores (Noble index23).
Three questions in the standardized assessment form pertain to the existence of oromotor dysfunction or motor speech problems (articulation defects or dysarthria), swallowing and chewing problems, and excessive drooling, all of which are answered by a response of ‘present’ or ‘absent’. There is a further question about communication impairments referring to expressive speech and language difficulty but excluding articulation defects. Where impairments are present, the method of communication recorded includes ‘able to use speech’, ‘speech and formal methods’ (e.g. Makaton, computer-assisted, signing), ‘formal methods alone’, or ‘unable to communicate’, and are classified as mild, moderate, severe, or profound respectively.
By 1 June 2009 a total of 1357 individuals with actual or possible CP had been notified to the NICPR, born between 1980 and 2001, of whom 33% had been identified retrospectively and 95% (1294/1357) had completed standardized assessment forms. The remaining individuals were either untraceable, had moved out, or died, and no data were accessible. Nine per cent (122/1357) of children had an assessment completed before 5 years of age. A higher proportion of younger and older individuals (born 1980 and 1998–2001) had no standardized assessment forms (p<0.05) and thus these years are excluded in the analysis of trends over time.
Data for live births were obtained from the routine publications of the Northern Ireland Statistics and Research Agency and the Registrar General’s annual reports.
The NICPR has ethical approval (Queen’s University Ethics Committee 263/90) and is governed by an advisory committee, which meets on an annual basis and includes a parental representative.
Prevalence rates (per 1000 live births) with 95% confidence intervals were calculated from the Poisson distribution. Logistic regression was used to investigate the relation between oromotor and communication impairments (dependent variables) and the clinical and social characteristics of the children (independent variables). The results are presented as odds ratios (ORs) with 95% confidence intervals. In univariate analysis, only those independent variables significant at p<0.2 were selected for entry into a multivariable model. In the multivariable analysis, with the addition of each new independent variable the model was checked using the likelihood ratio statistic and only included if p<0.01. Final models were checked using backwards elimination (p<0.01). All models were checked for interaction between GMFCS and intellectual impairment. Individuals with missing data on any of the covariates in the final models were excluded. All models were checked for goodness-of-fit (using the Homer–Lemeshow test) and were found to be satisfactory (p>0.05). Poisson regression was used to test for change in overall prevalence of CP over time (p<0.05), and six periods (with 1981/82 as the baseline period) were studied. Change in the proportion of individuals with oromotor and communication impairments over time was analysed with the χ2 test. All analyses used STATA (version 9; Stata Corp. College Station, TX, USA).
- Top of page
- What this paper adds
Median age at first notification to the NICPR was 4 years 2 months (interquartile range 2–8y) and median age at first assessment was 5 years 11 months (interquartile range 3–9y). The characteristics of the 1357 children born between 1980 and 2001 or resident in Northern Ireland are shown in Table I. There were a total of 533 925 live births during the same period and 1268 individuals with ‘early-onset’ CP, giving a prevalence of 2.4 per 1000 live births (95% confidence intervals 2.2–2.5). Table II shows the number, proportion, and live-birth prevalence of oromotor dysfunction and communication impairments. Overall 17% (95% confidence interval 15–19) of children recorded had no method of communication.
Table I. Characteristics of all cases (n=1357) born between 1980 and 2001
|Population characteristics||Individuals with early-onset CP (n=1268)||Individuals with late-onset CP (n=89)||All cases (n=1357)|
|n (%)||n (%)||n (%)|
| Spastic unilateral||412 (32)||35 (39)||447 (33)|
| Bilateral spastic||471 (37)||25 (28)||496 (37)|
| [Bilateral spastic CP-dyskinetic]||[16/471]||[1/25]||[17/496]|
| Dyskinetic||34 (3)||2 (2)||36 (3)|
| Ataxic||25 (2)||4 (4)||29 (2)|
| Unclassifiable||47 (4)||0 (0)||47 (3)|
| Missing||279 (22)||23 (26)||302 (22)|
| I||169 (13)||11 (12)||181 (13)|
| II||516 (41)||42 (47)||563 (42)|
| III||120 (9)||3 (3)||123 (9)|
| IV||75 (6)||6 (7)||82 (6)|
| V||248 (20)||25 (28)||276 (20)|
| Missing||140 (11)||2 (2)||142 (10)|
| None (IQ>70)||607 (48)||30 (34)||641 (47)|
| Moderate (IQ 50–70)||183 (14)||15 (17)||200 (15)|
| Severe (IQ<50)||326 (26)||40 (45)||371 (27)|
| Missing||152 (12)||4 (4)||156 (12)|
| None ever||686 (54)||24 (27)||713 (53)|
| Past only||177 (14)||19 (21)||198 (15)|
| Currently active||288 (23)||43 (48)||336 (25)|
| Missing||117 (9)||3 (3)||120 (9)|
| <1500g||257 (20)||1 (1)||258 (19)|
| 1500–2499g||274 (22)||6 (7)||281 (21)|
| 2500+g||627 (49)||69 (78)||705 (52)|
| Missing||110 (9)||13 (15)||123 (9)|
| 1 (least deprived)||85 (7)||3 (3)||88 (6)|
| 2||76 (6)||1 (1)||77 (6)|
| 3||114 (9)||12 (13)||126 (9)|
| 4||223 (18)||12 (13)||235 (17)|
| 5 (most deprived)||693 (55)||58 (65)||751 (55)|
| Missing||77 (6)||3 (3)||80 (6)|
| Male||722 (57)||59 (66)||781 (58)|
| Female||546 (43)||30 (34)||576 (42)|
| No||1137 (90)||78 (88)||1215 (90)|
| Yes||131 (10)||11 (12)||142 (10)|
Table II. Presence of oromotor dysfunction and communication impairments (frequency, proportion, and rate with 95% confidence intervals [CIs]) for individuals with ‘early-onset CP’ using live birth denominators, birth years 1980 to 2001
|Oromotor and communication impairments (absent/present)||n||Percentage of 1268||Valid (%)||Rate per 1000 live births||95% CI|
|Motor speech (articulation)|
| Present (Missing data n=100)||419||33||36||0.8||0.7–0.9|
|Swallowing and chewing|
| Present (Missing data n=100)||240||19||21||0.5||0.4–0.5|
| Present (Missing data n=100)||258||20||22||0.5||0.4–0.5|
|Communication (speech and language)|
| Present (Missing data n=139)||471||37||42||0.9||0.8–1.0|
|If communication impairment present (n=471): how does the child communicate?|
| ||n||Percentage of 1268||Percentage of 471||Rate per 1000 live births||95% CI|
|Speech and formal methods||58||5||12||0.1||0.1–0.1|
|No communication (Missing n=0)||214||17||45||0.4||0.3–0.4|
Data about all four impairments (motor speech, swallowing/chewing, excessive drooling, and communication) were available for 88% of individuals (1119/1268), of whom 52% (580/1119) had one impairment or more. Thirty-two per cent (184/580) of children had one impairment only, 30% (175/580) had two, 20% (115/580) had three, and 18% (106/580) had all four impairments present. The most common combinations of impairment, accounting for more than half of those with impairments (n=580), were motor speech problems and communication impairments (20% [115/580]) and all four impairments present (18% [106/580]). Communication impairments alone were present in 17% (96/580).
The final multivariable models showing the relation between oromotor dysfunction and communication impairments and the characteristics of the CP population are shown in Table III. Motor speech impairment was significantly related to CP subtype and was more common in bilateral spastic CP (OR 1.6) and non-spastic CP (OR 5.1) compared with those with unilateral spastic CP. An increased risk for motor speech problems was particularly evident for those in GMFCS levels IV (OR 4.0) and V (OR 8.0) compared with those in level I. Intellectual impairment was significantly related to the risk of having a speech impairment in those in IQ categories 50 to 70 (OR 2.7) and <50 (OR 3.6), and at significantly higher risk than those with an IQ >70. There was no significant interaction between GMFCS and IQ in relation to motor speech impairment (p>0.05). The final model explained an estimated 24% of the observed variation in speech impairment in the population of people with CP.
Table III. Multivariable models showing the relation between oromotor dysfunction and communication impairments and child characteristics (birth years 1980–2001) as adjusted odds ratios (ORs) and 95% confidence intervals (CIs)
| ||Oromotor dysfunction and communication impairments|
|Clinical characteristics||Speech (articulation; no impairment [n=680] coded ‘0’ vs with impairment [n=421] coded ‘1’)||Swallowing and chewing (no impairment [n=883] coded ‘0’ vs with impairment [n=247] coded ‘1’)||Excessive drooling (no impairment [n=835] coded ‘0’ vs with impairment [n=266] coded ‘1’)||Communication (language; none/mild impairment [n=774] coded ‘0’ vs mod/profound impairment [n=297] coded ‘1’)|
|n||OR||95% CI||p||n||OR||95% CI||p||n||OR||95% CI||p||n||OR||95% CI||p|
|Unilateral spastic (reference category)||442||1||–||<0.001|| || || || || || || || || || || || |
|Bilateral spastic||571||1.6||1.1–2.4|| || || || || || || || || || || || || |
|Non-spastic (+17 mixed type)||88||5.1||2.8–9.1|| || || || || || || || || || || || || |
|I (reference category)||156||1||–||<0.001||160||1||–||<0.001||156||1||–||<0.001||656a||1||–|| |
|II||502||2.1||1.2–3.5|| ||520||1.5||(0.6–3.4|| ||502||2.2||1.1–4.6|| || || || ||<0.001|
|III||119||2.5||1.3–4.9|| ||119||1.8||0.7–4.8|| ||119||2.9||1.3–6.7|| ||115||2.3||1.2–4.4|| |
|IV||73||4.0||1.9–8.4|| ||76||4.8||1.9–12.4|| ||73||4.8||2.0–11.4|| ||71||6.1||3.1–11.8|| |
|V||251||8.0||4.1–15.6|| ||255||15.7||6.6–37.4|| ||251||12.9||5.9–28.1|| ||239||20.3||11.5–35.7|| |
|None (IQ>70) (reference category)||578||1||–||<0.001||602||1||–||=0.004||578||1||–||<0.001||568||1||–||<0.001|
|Moderate (IQ 50–70)||181||2.7||1.8–4.0|| ||183||2.2||1.3–3.8|| ||181||2.8||1.8–4.4|| ||176||4.7||2.5–8.9|| |
|Severe (IQ<50)||342||3.6||2.5–5.3|| ||345||3.0||1.8–4.9|| ||342||2.9||1.8–4.4|| ||327||15.9||8.9–28.8|| |
|No (reference category)|| || || || ||1019||1||–||<0.001|| || || || ||967||1||–||<0.001|
|Yes|| || || || ||111||3.5||1.9–6.1|| || || || || ||104||3.1||1.3–7.9|| |
|None ever (reference category)|| || || || || || || || || || || || ||612||1||–||=0.005|
|Past only|| || || || || || || || || || || || ||171||1.8||1.0–3.2|| |
|Currently (past 12mo)|| || || || || || || || || || || || ||288||2.3||1.4–3.8|| |
| ||Total variance explained=24%. Variance explained by GMFCS=18%.||Total variance explained=36%. Variance explained by GMFCS=32%.||Total variance explained=21%. Variance explained by GMFCS=18%.||Total variance explained=57%. Variance explained by GMFCS=44%.|
Swallowing and chewing impairments were significantly related to GMFCS levels, with a significantly increased risk of impairment among those in GMFCS levels IV (OR 4.8) and V (OR 15.7) compared with those in level I. The risk was also significantly increased among those with an IQ of 50 to 70 (OR 2.2) and an IQ <50 (OR 3.0) compared with those with an IQ >70. There was no significant interaction between GMFCS and IQ (p>0.05). Finally, the risk was significantly increased among children known to have died (OR 3.5) compared with those who had not. The final model explained an estimated 36% of the observed variation in swallowing and chewing impairment.
Excessive drooling was significantly related to GMFCS and the risk was greatest among children in GMFCS levels IV (OR 4.8) and V (OR 12.9) compared with those with level I. Having intellectual impairment (IQ<70) was significantly related to an increased risk of excessive drooling. There was no significant interaction between GMFCS and IQ (p>0.05). The final model explained an estimated 21% of the observed variation in the impairment.
The presence of moderate to severe communication impairments was significantly related to GMFCS with those in GMFCS levels IV (OR 6.1) and V (OR 20.2) at highest risk compared with those in level I. Children with intellectual impairment (IQ<70) were at significantly higher risk of communication impairment than those without; this was particularly evident for those with IQ <50 (OR 15.8) compared with those without. There was no significant interaction between GMFCS and IQ (p>0.05). Children who had died (OR 3.1) were significantly more likely to have had communication impairment than those still living at the census date. Finally, those with active seizures (OR 2.3) were at significantly higher risk of communication impairment than those without a history of seizures. The final model explained an estimated 57% of the observed variation in communication impairment.
There was no significant trend or heterogeneity in the total prevalence of CP in live births during birth years 1981 to 1997 (p>0.05). Table IV shows the proportion of individuals with oromotor dysfunction and communication impairment stratified by grouped GMFCS levels for birth years 1981 to 1997. The proportion of individuals with oromotor and communication impairments who had CP in GMFCS levels IV to V showed a statistically significant downward trend in speech impairment (p=0.008) and excessive drooling (p=0.009) over time.
Table IV. Trends in the proportion of individuals with oromotor dysfunction and communication impairments (individuals with early-onset CP only) for birth years 1981 to 1997
|GMFCS levels||Impairments||Birth years||χ2 test|
|1981/82 (n=122)||1983/85 (n=180)||1986/88 (n=201)||1989/1991 (n=182)||1992/1994 (n=157)||1995/1997 (n=164)||All years (n=1006)|
|n (%)||n (%)||n (%)||n (%)||n (%)||n (%)||n (%)||Trend (df=1; p)||Heterogeneity (df=5; p)|
|I–III||Motor speech||22 (26)||27 (24)||38 (27)||29 (23)||22 (19)||20 (18)||158 (23)||>0.05||>0.05|
|Swallow/chewing||5 (6)||9 (8)||11 (8)||10 (8)||6 (5)||7 (6)||48 (7)||>0.05||>0.05|
|Excessive drooling||11 (13)||14 (12)||16 (11)||17 (14)||15 (13)||10 (9)||83 (12)||>0.05||>0.05|
|Moderate–profound communication impairments||8 (9)||7 (6)||14 (10)||11 (9)||8 (7)||4 (4)||52 (7)||>0.05||>0.05|
|IV–V||Motor speech||23 (77)||43 (80)||42 (78)||40 (76)||27 (77)||19 (49)||194 (73)||0.008||0.001|
|Swallow/chewing||18 (60)||33 (61)||32 (59)||29 (55)||22 (63)||18 (46)||152 (57)||>0.05||>0.05|
|Excessive drooling||16 (53)||38 (70)||27 (50)||31 (59)||22 (63)||10 (26)||144 (54)||0.009||0.001|
|Moderate–profound communication impairments||20 (69)||39 (78)||45 (83)||44 (83)||25 (78)||22 (61)||195 (77)||>0.05||>0.05|
|All individuals||Motor speech||45 (39)||70 (42)||80 (41)||69 (39)||49 (33)||39 (26)||352 (37)||0.002||0.021|
|Swallow/chewing||23 (20)||42 (25)||43 (22)||39 (22)||28 (19)||25 (16)||200 (21)||>0.05||>0.05|
|Excessive drooling||27 (23)||52 (31)||43 (22)||48 (27)||37 (25)||20 (13)||227 (24)||0.015||0.008|
|Moderate-profound communication impairments||28 (25)||46 (29)||59 (31)||55 (32)||33 (23)||26 (18)||247 (27)||>0.05||0.043|
- Top of page
- What this paper adds
We believe this study makes an important contribution to knowledge of the prevalence of these impairments in children with CP. Overall case ascertainment was estimated to be high (95%), with over 70% of individuals being multiply notified, thus giving confidence about case coverage. In the compilation of the NICPR and for the purposes of this paper, we defined oromotor dysfunction as those defects associated with motor speech problems, difficulties with swallowing and chewing, and excessive drooling. Communication impairments were considered present where there were speech and expressive language problems excluding articulation defects.
We estimate that half the population of children with CP have one or more impairments of oromotor function and/or communication. More specifically, motor speech problems were present in a third of children, swallowing and chewing difficulties and excessive drooling in about a fifth of all individuals, and some degree of impaired communication in 40%. Children in this group had a spectrum of difficulty. However, the data collection tool used in the compilation of the NICPR did not include standardized speech and language tests; rather, it gave a ‘yes’ or ‘no’ response to four possible communication methods and difficulties. These ranged from an isolated expressive speech and language delay but uses speech to communicate, to a requirement for an augmentative communication method including signing, to no formal speech or communication.
Gross motor function and the degree of intellectual impairment were significantly and independently related to the risk of all the oromotor and communication impairments studied. The relationships were as might be expected, with increasing risk for oromotor and communication disorders with decreasing functional and intellectual ability. This was particularly the case for children in GMFCS levels IV and V compared with those in level I; and for children with any degree of intellectual impairment compared with those without (IQ≥70). Clinical subtype was independently related to motor speech problems only where children with non-spastic CP in particular were at higher risk of motor speech problems compared with those with unilateral spastic CP. The variable ‘clinical subtype’ may represent some aspect of neuroseverity that is not captured by the GMFCS like the ‘quality of movement’, as opposed to functional ability. This may be particularly important in relation to motor speech problems. Incoordination and involuntary movements that can affect the oral and facial muscles are a particular feature of dyskinetic CP classified here in the non-spastic subgroup and found to be at higher risk of motor speech problems than any of the spastic subtypes.
The presence of epilepsy was significantly and independently related to the risk for communication impairments after adjusting for intellectual impairment, compared with children who never had seizures. This association has been reported by others.7 Finally, early morality was independently related to an increased risk of swallowing/chewing and communication impairments, although caution must be used in interpreting these analyses as the direction of the relationship cannot be inferred. For example the presence of oromotor and communication impairments may be more usefully considered risk factors for early mortality in children with CP.
The overall rate of CP in Northern Ireland during the 1980s and 1990s was stable, although there were significant declines in the proportion of children and young people with speech impairment and excessive drooling among the ‘severe’ subgroup (GMFCS levels IV–V). We found decreases in the proportion of all cases with bilateral spastic CP, level V CP, IQ<70, and IQ<50 in this population during the same time period.24 This could be part of a more optimistic clinical picture, with a smaller proportion of children with severe and complex CP, although there are other possibilities. First, these findings could reflect differences in ascertainment or data recording over time, but this seems unlikely as reporting methods have remained constant, as have the proportion of children being multiply notified and the overall rate of CP. Also, not all impairments showed a similar decline. Secondly, the results may be due to cohort effects. The particular impairments – motor speech problems and excessive drooling – may be more amenable to treatment with intensive speech and language therapy and medication, which may have been more accessible for younger children. It is also possible over the time period of the study methods to support communication for those with severe problems have made advances. Similarly impaired children in different birth cohorts could be classified in different communication subgroups because of this. However, for most of our analyses we grouped children with ‘moderate to profound’ communication impairments together, so the results are unlikely to be affected. It is not possible to determine if the clinical improvements observed in this study may be due to cohort effects because the NICPR is essentially cross-sectional with data captured ideally at age 5 years. A longitudinal study design would elicit this.
Earlier reports of oromotor and communication impairments in children with CP have often predated the development of the GMFCS and thus have limited the extent to which results across studies are comparable. Although we did not collect data on the GMFCS prospectively, we classified children retrospectively, which has been shown to be reliable.23 The inclusion of this standardized classification of gross motor function is an added strength of our study. We recognize a limitation of the study is lack of information on the severity of oromotor dysfunction. However, there is no agreed, systematic method of capturing this information. The development of a robust classification system of functional communication25 could help to standardize further and improve the quality of data collected by CP registries in the future.