Epidemiology of cerebral palsy on Tanna Island, Vanuatu: A population‐based register study

Abstract

Aim

To describe the epidemiology of cerebral palsy (CP) and access to related services on Tanna Island, Vanuatu, a low‐income country in the Pacific region with limited data on childhood disability, including CP.

Method

A population‐based register of children (aged 0–18 years) with CP was established on Tanna Island. Children with suspected CP were identified using the key informant method and assessed by a multidisciplinary team for clinical diagnosis. Data were collected on demographics, aetiology, clinical characteristics, rehabilitation, and education using a culturally adapted case record form from the Global Low‐ and Middle‐Income Country CP Register.

Results

Of the 98 children assessed, 52 had CP and two acquired brain injury. The observed prevalence of CP was 3.4 per 1000 in the paediatric population; mean age at diagnosis was 8 years 1 month (SD = 5 years 1 month). Sixty‐three per cent (n = 34) of children had primary motor‐type spasticity, 43% (n = 23) were classified in Gross Motor Function Classification System levels III to V, 43% (n = 21) were severely underweight, 83% (n = 45) had never received any rehabilitation, and 79% of school‐age children (n = 30 of 38) did not attend school.

Interpretation

The prevalence and motor severity of CP is alarmingly high on Tanna Island, Vanuatu. Delayed diagnosis and lack of access to rehabilitation and education highlights the urgent need for early diagnosis, intervention, and rehabilitation services for children with CP in Vanuatu.

Prevalence and motor severity of cerebral palsy (CP) is alarmingly high in Tanna Island, Vanuatu. Delayed diagnosis and lack of access to rehabilitation services and education highlight the urgent need for early diagnosis, intervention, and rehabilitation services for children with CP in Vanuatu.

This original article is commented on by Reynolds and Day on pages 1181–1182 of this issue.

Abbreviations

ABI

acquired brain injury

CFCS

Communication Function Classification System

GLM‐CPR

Global Low‐ and Middle‐Income Country CP Register

GMFCS

Gross Motor Function Classification System

HIC

high‐income country

LMIC

low‐ and middle‐income country

MACS

Manual Ability Classification System

MoH

Ministry of Health

NGO

non‐governmental organization

VCPR

Vanuatu CP Register

VHW

village health worker

What this paper adds.

• The prevalence of cerebral palsy (CP) is high on Tanna Island, Vanuatu, at approximately 3.4 per 1000 in the paediatric population.

• The diagnosis of CP is substantially delayed.

• In Tanna, Vanuatu, 83% of children with CP have never accessed any rehabilitation services.

• In Tanna, Vanuatu, 28% of school‐age males with CP and 7% of school‐age females access education.

• Undernutrition is widespread among children with CP, with 43% severely and 18% moderately underweight.

Cerebral palsy (CP) is an umbrella term for a spectrum of motor impairments caused by non‐progressive brain injury or malformation occurring during early development. ¹ CP is the most common condition causing childhood physical disability with a prevalence of 1.6 per 1000 live births in high‐income countries (HICs). ² Prevalence trends in low‐ and middle‐income countries (LMICs) are approximately 3.4 (95% confidence interval [CI] = 3.0–3.9) per 1000 live births. ² Preliminary findings from the Global Low‐ and Middle‐Income Country CP Register (GLM‐CPR) network, including Bangladesh, Nepal, Indonesia, and Ghana, show that there is a higher proportion of severe cases of CP in LMICs and that children with CP lack access to rehabilitation and educational services. ³ In addition, the causes of CP in developing countries are different, some with potentially preventable risk factors. ⁴

Population‐based studies are important to understand the prevalence, risk factors, clinical profile, and rehabilitation access for children with CP and their families. Currently there is no information on the prevalence of children with CP in any LMIC in the Pacific region (i.e. Fiji, Solomon Islands, Vanuatu, Kiribati, Samoa, Tonga, Micronesia, and Nauru). These Pacific nations and territories have unique challenges with small, culturally diverse populations living across remote islands, and are regularly affected by natural and climate‐related disasters. ⁵

The Republic of Vanuatu is an island nation, consisting of 83 islands (65 inhabited) with an estimated population of 320 409 in 2023. ⁶ Little is known regarding the number of children with a disability in Vanuatu; however, 5% of the total population have a disability. ⁷ This study aimed to define the epidemiology, clinical characteristics, and risk factors of children with CP living on the island of Tanna, Vanuatu, to enhance the visibility of children with disabilities, aligning with the United Nations Children's Fund policy of ‘Seen, Counted, Included: Using data to shed light on the well‐being of children with disabilities’. ⁸ We aimed to report population data on the prevalence and burden of CP among children (aged 0–18 years), including the level of impairment, the causes of CP, and rehabilitation service use to address potential gaps in the knowledge of CP in Vanuatu. We hypothesize that the establishment of a CP register will enable accurate estimates of disease burden (e.g. prevalence, specific needs), facilitate clinical surveillance, and promote international collaboration to research the translation of evidence‐based care into affordable service delivery models for children with CP and their families in the Pacific region.

METHOD

We established the first population‐based register of children with CP in Vanuatu (i.e. Vanuatu CP Register [VCPR]) and the Pacific region. The VCPR activities commenced on Tanna Island of the TAFEA Province in Vanuatu. The TAFEA Province consists of five islands: Tanna, Aniwa, Futuna, Erromango, and Aneityum. The island of Tanna encompasses 549km ² and includes an active volcano on the southeast coast. The total population of the island is approximately 31 057 including approximately 15 483 children aged 0 to 18 years (2020 Census). ⁹ This island was chosen because of an existing relationship with the local health service, TAFEA Health, Ministry of Health (MoH), Vanuatu through an Australian non‐governmental organization (NGO), the TAFEA Health Care Association.

Ethics

Ethics approval was gained on 22nd June 2023 from the Vanuatu Health Research Ethics Committee, MoH. Written informed consent (signature or thumbprint) was obtained from the primary caregivers of children. Caregiver information and consent sheets were available in Bislama and English. If carers were illiterate, the information sheet was explained in the local language by MoH staff or a village health worker (VHW). There are more than 100 indigenous (‘local’) languages in Vanuatu; the official languages of Vanuatu are Bislama, English, and French.

Study participants and identification of children with CP

This study used the key informant method to identify children with suspected CP. ¹⁰ , ¹¹ , ¹² This involved training local volunteers known as key informants to identify children with suspected CP from the community and refer them for subsequent assessment. Key informants are people who live or work in their local community, have a social role, and know the local context and people in the area. Key stakeholders, identified by MoH staff, were invited to attend day‐long training about CP. Training sessions were facilitated by the principal investigators (AB, MHAI, GK), with translation performed by staff from the MoH. VHWs and other key community stakeholders (hospital and disability staff, tribal chiefs, pastors, representatives from local government councils, and a women's group from Tanna) were trained using the key informant method. In Vanuatu, VHWs are ideally suited to the key informant role because they deliver basic primary health care at the village level from first‐aid posts and serve a large proportion of Vanuatu's population in their role as volunteer lay health workers. ¹³ Two education days were held. The first education day was held on the 14th October 2022 (15 participants, 10 of whom were VHWs). After ethics approval on 22nd June 2023, the education day was repeated on the 18th August 2023 (32 participants, 23 of whom were VHWs, with six VHWs having attended the first day). Education focused on ‘What is cerebral palsy?’ and provided simple information on basic care. At the conclusion of the education day, participants compiled a list of children in their locality who they thought may possibly have CP in their local villages (35 children listed on 14th October 2022, 75 children listed on 18th August 2023). Defining and explaining ‘cerebral palsy’ was important as there was no word in the local language for CP. Co‐authors and other staff employed by the MoH on Tanna, and community members who attended the education session, reported that they had used descriptive words to convey the idea of ‘not working properly’ to describe children who have a disability such as CP. To support understanding, videos of children with CP of different ages and different Gross Motor Function Classification System (GMFCS) levels were shown.

Assessments and data collection

Children with suspected CP identified by trained key informants were invited to participate in medical assessment clinics to confirm the diagnosis and for registration to the VCPR. A multidisciplinary team consisting of an occupational therapist, physiotherapist, or medical doctor conducted detailed neurodevelopmental assessments of each child attending the camps. Medical assessment ‘clinics’ were held on 22nd to 30th October and 5th to 15th December 2023. The medical assessment clinics were facilitated by MoH staff, who coordinated with VHWs and families of children with potential CP. Assessment clinics were held at several locations around the island, including village first‐aid posts, health centres, school, Lenekel Hospital, and village communal areas and private homes. Language translation was provided by MoH staff, VHWs, or family or community members.

The clinical definition of CP from the Australian Cerebral Palsy Register and the Surveillance of Cerebral Palsy in Europe was used, as in previous CP registers affiliated with the GLM‐CPR. ¹¹ , ¹² Data were collected using the case record form adapted from the GLM‐CPR ³ , ¹¹ in consultation with VCPRs and MoH staff to be appropriate for Vanuatu. Four questions were added to the questionnaire as recommended by Vanuatu MoH staff members: (1) Who cares for the child? (2) If it is not the parent caring for the child, what is the reason? (3) Did the mother have traditional or custom medicine for abortion during the pregnancy? (4) Did the mother have kava or alcohol during the pregnancy?

Caregivers were interviewed to gather detailed history on aetiological factors relating to biological parents (e.g. age, consanguinity), pregnancy (e.g. multiple birth, miscarriage, maternal illness during pregnancy), birth (e.g. gestational age, birthweight, birthplace, complications during birth delivery, signs of birth asphyxia) and the probable timing of brain injury or damage (prenatal period, from conception to onset of labour; perinatal, to 7 days; postnatal, to 2 years of life). The child's age at assessment was calculated based on the date of birth on their identification card, or the parent or carer's verbal recall. When only the year of birth was known, the 1st January of that year was used as the date of birth.

Clinical details collected included motor type and topography, and functional abilities as described by the GMFCS, ¹⁴ the Manual Ability Classification System (MACS), ¹⁵ and the Communication Function Classification System (CFCS). ¹⁶ The presence and severity of secondary impairments, such as epilepsy, intellectual impairment, visual impairment, hearing impairment, and speech impairment, were recorded. The presence of associated impairments was assessed through interview with the primary caregiver and observation of the child. For example, an intellectual impairment was documented if a child had difficulties in adaptive behaviours or intellectual functioning on the basis of statements provided by caregivers or observation of the child. Epilepsy was documented based on a history of single or recurrent episodes of unprovoked seizures in the 3 months before the time of the assessment.

Anthropometric measurements (weight [kg], height [cm], head and mid‐arm circumference [cm]) were taken and converted to z‐scores using the Center for Disease Control and Prevention growth charts to allow comparison to age‐based and sex‐based reference norms for growth. ¹⁷ Moderate underweight, wasting, and stunting were defined as z‐scores between −3 and − 2 for weight‐for‐age, weight‐for‐height, and height‐for‐age respectively. Z‐scores below −3 for weight‐for‐age, weight‐for‐height, and height‐for‐age indicate severe underweight, wasting, and stunting respectively.

Caregivers were interviewed to collect information about their child's access to rehabilitation and education services, and responses were documented in the case record form.

Statistical analysis

Descriptive statistics (e.g. mean, median, proportion) were used to summarize variables such as sociodemographic factors, aetiological characteristics, motor type, motor distribution, functional classification systems, nutrition status, rehabilitation, and education status. Data were analysed using Stata v18 (StatCorp, College Station, TX, USA). ¹⁸ The observed prevalence was estimated by dividing the total number of children with CP by the estimated population of children on Tanna (n = 15 483) taken from the Vanuatu Bureau of Statistics 2020 Census data. ⁹ Associations between risk factors and outcome measures were examined using linear and logistic regression.

RESULTS

A total of 110 children and young people with suspected CP were assessed during the two medical assessment camps on Tanna, Vanuatu. There were 98 children and 12 adults. Of the 98 children, 52 had a clinical diagnosis of CP, two had a clinical diagnosis of acquired brain injury (ABI), and 44 children had another clinical diagnosis (Table S1). Of the 12 adults assessed, five had a confirmed clinical diagnosis of CP. The subsequent results report data about children with CP and ABI only.

The observed prevalence of CP was 3.4 per 1000 children, and the prevalence of CP and ABI was 3.5 per 1000 children. The mean age of the children with CP at assessment was 8 years 11 months (SD = 5 years 6 month) and the mean age at CP diagnosis was 8 years 1 month (SD = 5 years 1 month). Of the 54 children, 65% (n = 35) were male and 35% (n = 19) were female. One child with ABI sustained a brain injury secondary to tuberculosis meningitis at age 9 years in 2021; the other sustained an injury at 5 years of age when the family home collapsed on them because of cyclone Pam in 2015. The children with ABI are included in our results and discussion because their ongoing rehabilitation and medical needs are similar to those of children with CP.

Clinical characteristics (GMFCS, Manual Ability Classification System, Communication Function Classification System), predominant motor type, motor distribution, and associated impairments are shown in Table 1 and Table S2. Most children had bilateral CP (69%). Most children (63%) had the spastic motor type and approximately one‐third had hypotonia (n = 18; 33%). Twenty‐three (43%) children were classified in GMFCS levels III to V, 21 (39%) children were classified in Manual Ability Classification System levels III to V, and 28 (52%) children in Communication Function Classification System levels III to V. The severity of gross motor impairment was not associated with age (−0.62, 95% CI = −1.75 to 0.5, p = 0.271).

Table 1.

Clinical characteristics: motor type, motor distribution, and functional classification level.

		Timing of CP (n = 52)
	Children with CP (n = 54) and ABI (n = 2)	Prenatal and perinatal (n = 26)	Postnatal a (n = 16)	Unknown (n = 10)
Sex
Male	35 (64.8)	16 (61.5)	10 (62.5)	8 (80)
Female	19 (35.2)	10 (38.5)	6 (37.5)	2 (20)
Age category
0–5 years	16 (29.6)	10 (38.5)	5 (31.2)	1 (10)
5–10 years	13 (24.1)	5 (19.2)	3 (18.8)	5 (50)
10–18 years	25 (46.3)	11 (42.3)	8 (50)	4 (40)
CFCS level
I	20 (37)	8 (30.8)	8 (50)	3 (30)
II	6 (11.1)	3 (11.5)	2 (12.5)	1 (10)
III	12 (22.2)	7 (26.9)	4 (25)	1 (10)
IV	7 (13.0)	4 (15.4)	1 (6.2)	2 (20)
V	9 (16.7)	4 (3.8)	1 (6.2)	3 (30)
GMFCS level
I	9 (16.7)	3 (11.5)	5 (31.2)	1 (10)
II	22 (40.7)	10 (38.5)	6 (37.5)	5 (50)
III	7 (13.0)	6 (23.1)	0 (0)	1 (10)
IV	7 (13.0)	2 (7.7)	5 (31.2)	0 (0)
V	9 (16.7)	5 (19.2)	0 (0.0)	3 (30)
MACS level
I	18 (33.3)	7 (26.9)	9 (56.2)	1 (10.0)
II	15 (27.8)	7 (26.9)	5 (31.2)	3 (30)
III	11 (20.4)	7 (26.9)	2 (12.5)	2 (20)
IV	4 (7.4)	4 (15.4)	0 (0)	0 (0)
V	6 (11.1)	1 (3.8)	0 (0)	4 (40)
Primary motor type and distribution
Spasticity
Left unilateral	3 (5.6)	1 (3.8)	2 (12.5)	0 (0)
Right unilateral	14 (25.9)	7 (26.9)	4 (25)	2 (20)
Bilateral (diplegia)	8 (14.8)	2 (7.7)	4 (25)	2 (20)
Bilateral (quadriplegia)	9 (16.7)	6 (23.1)	0 (0)	2 (20)
Hypotonia	18 (33.3)	10 (38.5)	5 (31.2)	3 (30)
Dystonia	2 (3.7)	0 (0)	1 (6.2)	1 (10.0)

Abbreviations: ABI; acquired brain injury; CFCS, Communication Function Classification System; CP, cerebral palsy; GMFCS, Gross Motor Classification System; MACS, Manual Ability Classification System.

^a Children whose brain injury was acquired > 28 days after birth but before 2 years of age.

Secondary impairments

Of the 54 children with CP and ABI, 65% (n = 35) had at least one associated impairment (Table 2). The most common included intellectual impairment (59%) and speech impairment (65%), with 32% being non‐verbal. Other impairments included swallowing difficulties (17%), epilepsy (17%), and visual impairment (28%).

Table 2.

Secondary impairments and nutritional status of children with CP and ABI (total n = 54, with 35 children having another impairment).

Secondary impairments	Tanna CP register, n (%)
Visual impairment
Functionally blind	4 (7.4)
Some visual impairment	11 (20.4)
No visual impairment	39 (72.2)
Intellectual impairment
Severe	7 (13.0)
Moderate	10 (18.5)
Mild	9 (16.7)
Probably some impairment	6 (11.1)
Probably no impairment	2 (3.7)
No impairment	20 (37.0)
Epilepsy
Epilepsy present	9 (16.6)
Epilepsy resolved	2 (3.7)
No epilepsy	43 (79.6)
Hearing impairment
Severe deafness	1 (1.9)
Some hearing impairment	5 (9.3)
No hearing impairment	46 (85.2)
Unknown	2 (3.7)
Speech impairment
Non‐verbal	17 (31.5)
Some speech impairment	18 (33.3)
No speech impairment	19 (35.2)
Dysphagia
No	45 (83.3)
Yes	9 (16.7)
Nutrition (by z‐score)
Normal range	19 (35.2)
Moderately underweight (z‐score ≥ −3 and < −2)	9 (16.7)
Severely underweight, (z‐score < −3)	21 (38.9)
Unknown	10 (18.5)

Abbreviations: ABI; acquired brain injury; CP, cerebral palsy.

Nutrition

Of the 49 children with a z‐score, 21 were severely underweight (z‐score 95% CI = −9.4 to −4.9), nine were moderately underweight (z‐score 95% CI = −2.8 to −2.4), and 19 were in the healthy range (z‐score 95% CI = −0.5 to 0.9). The mean weight‐for‐age z‐score was −3.46 (SD = 4.75, 95% CI = −4.8 to −2.1, median = −2.65); this indicates an overall low weight‐for‐age of children because z‐scores are normally distributed with a mean of zero and an SD of 1. Mean weight‐for‐age z‐scores for each GMFCS level are shown in Table S3. Motor severity according to a child's GMFCS level was significantly associated with a child's weight z‐score (β = −1.18, 95% CI = −2.17 to −0.18, p = 0.02), with the most severely underweight children classified in GMFCS level V (GMFCS level I = reference; level II = −1.06, 95% CI = −3.34 to 2.61; level III = −4.18, 95% CI = −9.37 to 1.01; level IV = −2.25, 95% CI = −7.15 to 2.65; level V = −5.05, 95% CI = −9.74 to −0.36). Monthly income was not associated with the mean weight z‐score (β ≤ 0.001, 95% CI = −0.00009 to 0.0001, p = 0.74). Scatterplots showing the distributions of weight z‐scores according to GMFCS level and income are shown in Figures S1 and S2.

Sociodemographic characteristics

Sociodemographic characteristics are summarized in Table 3. Most families (80%) lived in traditional two‐room houses, with a mean monthly family income of VT 9156 (US$72.9). Illiteracy rates among parents were similar to the national average. ¹⁹ Approximately half of the children (30 of 54) lived with both biological parents, seven (13%) with their biological father only, and eight (15%) with their biological mother only. Other children (17%) were cared for by extended family members because of reasons such as abandonment, remarriage, or parental death.

Table 3.

Child and family demographics and socioeconomic factors for the CP register cohort (CP and ABI), children without CP, all children screened in medical camps, and the Ni‐Vanuatu general population.

Sociodemographic factor	CP or ABI, n = 54	Not CP or ABI, n = 44	Total, n = 98	Vanuatu population
Age category
0–5 years	16 (29.6)	11 (25.0)	27 (27.6)	5012 (16) a
5–10 years	13 (24.1)	16 (36.4)	29 (29.6)	4681 (15) a
10–18 years	25 (46.3)	17 (38.6)	42 (42.9)	6985 (24) a
Sex
Female	19 (35.2)	19 (43.2)	38 (38.8)	8567 (51.6) a
Male	35 (64.8)	25 (56.8)	60 (61.2)	8121 (48.4) a
Type of accommodation
Permanent house	3 (5.6)	5 (11.4)	8 (8.2)	1205 (21) a
Semi‐permanent house	8 (14.8)	1 (2.3)	9 (9.2)	1103 (18) a
Traditional house	43 (79.6)	38 (86.4)	81 (82.7)	2784 (49) a
Number of rooms in house Median (IQR), mean (SD), n	2 (1–3), 2.2 (1.4), 54	2 (1–3), 1.9 (1.2), 43	2 (1–3), 2.1 (1.4), 97	2.5 d , e
Number of people in house Median (IQR), mean (SD), n	5 (4–6), 5 (2.2), 54	5 (4–6), 4.8 (1.6), 41	5 (4–6)	5 d
Monthly family income f Vanuatu median, mean (SD)	5000, 9156 (11 911)	5000, 9236 (13 230)	5000, 9191 (12 418)	17 275 d
Source water
Water supply	40 (74.1)	31 (70.5)	71 (72.4)	1875 (74.4) b
Rainwater	7 (13.0)	8 (18.2)	15 (15.3)	456 (18.1) b
Other (pond, river)	7 (13.0)	5 (11.4)	12 (12.2)	164 (6.5) b
Sanitation
No toilet facility	0 (0)	1 (2.3)	1 (1.0)	–
Standard VIP	13 (24.1)	8 (18.2)	21 (21.4)	–
Unimproved toilet/Bush	40 (74.1)	33 (75.0)	73 (74.5)	–
Unknown	1 (1.9)	2 (4.5)	3 (3.1)	–
Child living in care of
Biological mother and father	30 (55.6)	–	–	65.5% b
Biological mother	8 (14.8)	–	–	17.6% b
Biological father	7 (13.0)	–	–	1.8% b
Neither mother nor father	9 (16.7)	–	–	15.8% b
Who does the child live with if not their mother or father? (n = 9)
Grandmother	6	–	–	–
Aunty	2	–	–	–
Uncle	1	–	–	–
Mother's educational level at time of child's birth
Illiterate	16 (29.7)	–	–	148 (30.9) b
Primary (years 1–6)	21 (38.9)	–	–	44 (9.1) b
Secondary (years 7–10)	11 (20.4)	–	–	195 (40.7) b
Diploma/other trade qualification	1 (1.9)	–	–	N/A
University	3 (5.6)	–	–	16 (3.4) b
Unknown	2 (3.7)	–	–	–
Mother's occupation at the time of child's birth
Airport worker/banker/cook	5 (9.4)	–	–	293 (19.6) a , c
Gardener/farmer	7 (13.0)	–	–	351 (23.5) a
Housewife	35 (64.8)	–	–	90 (6.0) a
Student	2 (3.7)	–	–	293 (19.6) a
Teacher	2 (3.7)	–	–	293 (19.6) a , c
Unknown/other	3 (5.6)	–	–	762 (50.9) a
Father's educational level at time of child's birth
Illiterate	17 (31.5)	–	–	29.83 (16.3) b
Primary (years 1–6)	22 (40.7)	–	–	46.48 (25.4) b
Secondary (years 7–10)	8 (14.8)	–	–	65.51 (35.8) b
Senior secondary (years 11–13)	2 (3.7)	–	–	25.07(13.7) b
University	1 (1.9)	–	–	15.74 (8.6) b
Unknown	4 (7.4)	–	–	–
Father's occupation at the time of child's birth
Airline operation/baker	2 (3.8)	–	–	N/A
Carpenter	4 (7.4)	–	–	419 (23.7) a
Driver	4 (7.4)	–	–	388 (22.0) a
First aid nurse/shopkeeper	2 (3.8)	–	–	N/A
Gardener/farmer	31 (57.4)	–	–	322 (18.2) a
Pastor	3 (5.6)	–	–	350 (19.8) a
Other	5 (9.3)	–	–	286 (16.2) a
Unknown	3 (5.6)	–	–	N/A

Note: Data are n (%) unless otherwise stated.

Abbreviations: ABI, acquired brain injury; CP, cerebral palsy; IQR, interquartile range; N/A, not applicable; VIP, ventilated improved pit latrine.

^a Vanuatu Bureau of statistics 2020.

^b Vanuatu multiple indicator cluster survey 2023.

^c Statistics for student, airport worker and teacher were assumed as employee‐ government from the Vanuatu bureau of statistics 2020.

^d Household Income & Expenditure 2019–2020 NSDP Baseline Survey.

^e 2020 Vanuatu National Population and Housing Census–Analytical report Volume 2.

^f VT5000 = Aus$63, VT9000 = Aus$113.

Known risk factors for CP

The mean age of mothers at childbirth was 26 years 1 month (SD = 6 years 11 months). Five mothers (10%) were 18 years or younger and six were 35 years or older. Nineteen (36.5%) children with CP had parents who were related, with 16 (30.8%) being first cousins. Fourteen children had a family member with a disability and five had disabled cousins. In one family, three siblings had spastic bilateral CP (Table 4).

Table 4.

Risk factors for children with cerebral palsy.

Risk factor, n = 52		Risk factor, n = 52
Assistance with conception		Consanguinity
No	51 (98.1)	No	33 (63.5)
Yes (type unknown)	1 (1.9)	Yes	19 (36.5)
Delivery attendant		Complications during pregnancy
Doctor/midwife	37 (71.2)	No	32 (61.5)
Other family members	7 (13.5)	Unknown	9 (17.3)
Skilled birth attendant/traditional birth attendant	5 (9.6)	Yes	11 (21.2)
Traditional village healer	3 (5.8)
Antenatal care during pregnancy		Birth asphyxia
Unknown	6 (11.5)	Do not know	6 (11.5)
No	5 (9.6)	No	29 (55.8)
Yes	41 (78.8)	Yes	17 (32.7)
Frequency of antenatal care		Early feeding difficulties
Regular	33 (80.5)	Unknown	10 (19.2)
Irregular (<2 visits)	2 (4.9)	No	32 (61.5)
Irregular (≥2 visits)	5 (12.2)	Yes	10 (19.2)
Unknown	1 (2.4)	–	–
Birthplace		Febrile illness during pregnancy
Health centre	4 (7.7)	Unknown	7 (13.5)
Home birth	13 (25.0)	No	35 (67.3)
Hospital	31 (59.6)	Yes	10 (19.2)
Traditional village healer	4 (7.7)	–	–
Delivery plan if home birth		Febrile rash during pregnancy (of n = 10 who reported illness)
Attended planned delivery	5 (38.5)	Unknown	3 (30.0)
Unplanned	3 (23.1)	No	5 (50.0)
Unknown	5 (38.5)	Yes	2 (20.0)
Use of MRI		Febrile illness during labour
No	52 (100)	Unknown	11 (21.2)
Yes	0 (0)	No	38 (73.1)
Unknown	0 (0)	Yes	3 (5.8)
Maternal kava consumption		Swallowing difficulties
No	44 (84.6)	No	43 (82.7)
Yes (type unknown)	5 (9.6)	Yes	9 (17.3)
Unknown	3 (5.8)	–	–
Maternal alcohol consumption		Neonatal encephalopathy
No	44 (84.6)	No	27 (51.9)
Yes (type unknown)	5 (9.6)	Unknown	16 (30.8)
Unknown	3 (5.8)	Yes	9 (17.3)
Use of traditional abortion medicine		Gestational age at birth
No	48 (92.3)	Very preterm (28–32 weeks)	1 (2)
Yes (type unknown)	1 (1.9)	Moderate to late preterm (32–37 weeks)	6 (11.5)
Unknown	3 (5.8)	Term (37–40 weeks)	39 (75)
–		Unknown	6 (11.5)

Note: Data are n (%) unless otherwise stated.

Abbreviation: MRI, magnetic resonance imaging.

The majority (78.8%) of mothers reported receiving antenatal care during pregnancy. One mother reported consumption of traditional medicine to assist with conception, and another reported consumption of traditional medicine to promote abortion during pregnancy. Five (9.6%) of the 52 mothers reported drinking kava during pregnancy, three of whom also reported drinking alcohol.

With regard to place of birth, 59.6% (n = 31) of children were born at the hospital, 7.7% (n = 4) at a health centre, 25% (n = 13) at home, and 7.7% (n = 4) were born at a traditional village healer. Of those born at home, three were unplanned home births.

The mean gestational age at birth of children with CP was 38.6 weeks (SD = 2.2 weeks, n = 45). One child (2% of the cohort with CP) was born very preterm (28 to less than 32 weeks), with the majority (n = 39; 75%) born at term (Table 4). Mean birthweight was 2691.3 g (SD = 546 g) but was known for only 44% of children (n = 23). Only one child had undergone a computerized tomography scan, which was performed at Port Vila Hospital some weeks after birth. None of the children were reported to be part of a multiple pregnancy, nor were any born via caesarean.

Timing of CP

With regard to the timing of CP, that is, when disturbance occurred to the developing brain and resulted in CP, 56% (n = 29 of 52) were prenatal or perinatal, 25% (n = 13) postnatal, and 19% (n = 10) were unknown. Of the 29 reported as prenatal and perinatal, 17 reported birth asphyxia, with 15 of the 29 reporting birth complications. Of the families who reported birth complications, six had prolonged or obstructed labour, and two had pre‐eclampsia. All the families who reported postnatally acquired CP (n = 13) reported high‐grade fevers and seizures, with two reporting that the fever was associated with meningitis and one with pneumonia. There was no relationship between the timing of CP injury as reported by parents and motor type (β = −0.13, 95% CI = −0.56 to 0.29, p = 0.53) or severity of impairment according to GMFCS level (β = 0.07, 95% CI = −0.34 to 0.50, p = 0.77).

Rehabilitation status

Of the 54 children with CP and ABI, 83% (n = 45) had not accessed any rehabilitation services (Table S3). Among the nine children who had received rehabilitation, most had only one‐off services. Three children received wheelchairs before 2020, while two children were seen twice by the NGO TAFEA Health Care and provided with supported seating, orthoses, and therapy advice. One child received surgery and follow‐up rehabilitation in Australia, supported by Rotary. Among children who had not received rehabilitation, 82% (n = 37) of families reported that ‘no service exists’, while others were unaware of available services (n = 4) or gave no comment (n = 4) (Table S4).

Education status

Of the 38 children who were of school age, eight (21%) attended school. Of these eight children, three were classified in GMFCS level I and five in GMFCS level II (Table S5). Five children were under 12 years of age, and three were between 14 years and 18 years. One female child and seven male children attended school (Table S4).

Of the children aged 5 years and older (n = 30) who did not attend school, parents or carers reported the reasons for non‐attendance as ‘the distance to travel (walk) to school was too great’ (n = 12), or ‘the school did not accept children with disability’ (n = 8), or ‘the child's abilities were such that they would not cope in school’ (e.g. communication difficulties, hearing impairment, intellectual impairment, epilepsy), or ‘lack of money’. Examples of parents' comments include their child ‘got hurt walking to school’, and ‘unable [to go to] school is 2km downhill, rough road’, and ‘school too far—not safe for her to go’.

DISCUSSION

To our understanding, this is the first population‐based study describing the epidemiology of children with CP who live in a Pacific LMIC. The study found a high prevalence of CP, significant delays in diagnosis, severe functional limitations, and limited access to rehabilitation and education. The data from this study add to population‐based surveillance studies of children with CP in LMICs across the world. ¹¹ , ¹² , ²⁰

Our findings indicate that the burden of CP is high in Tanna, Vanuatu compared with HICs, which is consistent with data from other LMICs. ¹¹ , ²¹ , ²² This report is an underestimate because there were children with CP who were not seen by the visiting assessment team because of time limitations, transportation, access issues, and weather events. In 2024 and 2025, more children were identified with CP by the authors of this study and other paediatricians visiting from New Zealand aid programmes; these children were not included in the study period under assessment. In addition, the high mortality rate of children with CP in LMICs is recognized ²³ and we acknowledge that there is an unknown number of children with CP who died before the start of the study. For example, the study team knew of a 3‐year‐old male child with CP (GMFCS level V) who was severely underweight, who passed away in July 2022 before the study began, with his twin reportedly having died in utero.

The mean age at diagnosis of CP was 8 years 1 month, which is delayed compared to HICs ²⁴ and similar to other LMICs. ¹¹ , ¹² Many of the children had not been seen by a medical or allied health professional before this study. Like other CP registers, most children had a predominant motor type of spastic CP. However, the proportion of children with a predominant motor type of hypotonia was greater than that reported in HIC and other LMIC studies. The authors can only speculate on the underlying causes for this, considering factors such as malnutrition and historical consanguinity as potential contributors. Unfortunately, magnetic resonance imaging and genetic testing are not currently available in Vanuatu, which limits our ability to investigate the underlying causes of the conditions described under the umbrella term of CP. In the absence of these diagnostic tools, our classification relies on clinical assessment. A recent study from India reported that 11.7% of children with CP were classified with a primary motor type of hypotonia, highlighting the variability in clinical presentation of children in HICs and LMICs. ²⁵

The proportion of children with CP with moderate and severe malnutrition was very high and is higher than that recorded in the general population of Vanuatu. In Vanuatu, 12.1% of the population under 5 years have a −2SDs weight‐for‐age and 4.2% are below −3SDs. ²⁶ Other population statistics include the estimated prevalence of stunting in children under 5 years, which is 31.4%; the under‐5 mortality rate is 18.2 deaths per 1000 live births. ²⁷ On Tanna, we found that the clinical severity of CP was associated with the weight z‐score, with the most severely impaired (e.g. GMFCS level V) being the most severely underweight. However, there was no association between income and weight z‐score. This lack of association was surprising given the previous literature; for example, a study in rural Bangladesh found that moderate‐to‐severe undernutrition among children with CP was associated with both poverty and the clinical severity associated with undernutrition. ²⁸ The lack of association between income and undernutrition might be because Tanna is a remote island and all families are dependent to a large extent on subsistence farming because of geographical isolation, and because of little variance in income between families. Children in this study were substantially more underweight compared to children with CP sampled using CP‐specific growth charts who live in the UK and USA, ²⁹ underscoring the significant nutritional challenges faced in LMICs.

Most children in this study had not accessed any rehabilitation service. Access to health care is challenging on Tanna because of its remote location, rugged terrain, limited transport, high transportation costs, and limited health services. There is one small hospital on Tanna that serves the five islands in the TAFEA Province. Because of the island's rugged terrain and limited transport options, families rely on VHWs for basic health care needs such as first aid, treatment of burns and wounds, and infections. There is currently no allied health workforce on Tanna Island. There are two paediatricians and eight physiotherapists employed by the MoH to serve all of Vanuatu based at the Port Vila Hospital on the island of Efate, and on the island of Santo. The Vanuatu Society of People with Disabilities, which is based in Port Vila does not have staff on Tanna.

In Vanuatu, school attendance is not compulsory. Only 21% of school‐age children with CP in our study were attending school. The percentage of children in the general population attending school in TAFEA Province is 86.5% and 43.7% for primary and secondary school respectively (Vanuatu Multiple Indicator Cluster Survey 2023). ²⁶ Following this trend, there was a smaller proportion of older children with CP attending school (3 of 20 aged 12–18 years), with one attending high school and the other two attending primary school. Children are allowed to attend primary school despite being older. Only one of the 14 females who were of school age attended school, while 7 of 24 school‐age males attended school. The sex discrepancy in school attendance observed in this study is not seen in the general TAFEA population. ²⁶ School attendance is challenging because of several factors, including distance to travel and poor road infrastructure. For example, one family reported that their able‐bodied child takes 1.5 hours to walk to school (3 hours return) and that it would be impossible for their child with a physical and visual impairment to do this.

Family caring arrangements for children with a disability was raised as an important issue by MoH staff. The minimal cluster survey ²⁶ reported that 65.5% of children in the TAFEA Province live with both of their biological parents; however, only 55.6% of the children in this study lived with both their biological mother and father. A greater proportion of the children lived with their father only (n = 7; 13%), compared to the minimal cluster survey in which only 1.8% lived with their father only. A similar proportion of children lived with their mother only in this CP register cohort and the minimal cluster survey (14.8% vs 17.6% respectively). A similar proportion of children lived with neither biological parent in this study compared to the minimal cluster survey (16.6% in the CP register, 12.9% in the cluster survey); instead, they were cared for by their grandmother and extended family.

Consanguinity has been identified as a risk factor for CP. ³⁰ , ³¹ , ³² , ³³ Studies suggest that the increased risk arises from shared genetic material, elevating the likelihood of autosomal recessive genetic disorders that may contribute to CP. ³² Research in regions with high rates of consanguinity, such as the Middle East and South Asia, have highlighted this link, with one Palestinian study identifying consanguinity as a significant risk factor (OR = 4.62; 95% CI = 2.07–10.3). ³⁰ The proportion (n = 19; 36.5%) of children with CP on Tanna whose parents were related was higher than that reported in children with CP in India (18% born of consanguineous parents) ³³ and Bangladesh (15.8%), ¹¹ and similar to a study of children with intellectual disability in Barwani, India (37% born of a consanguineous relationship). ³⁴ Marriage between first cousins is a cultural tradition on Tanna, which has spanned down generations and has benefits of strengthening family ties and maintaining geographical proximity and tribal lands. MoH staff verbally reported that while consanguineous marriages are practised in some other Vanuatu provinces and islands, most do not share the same marriage traditions. As consanguinity is a modifiable risk factor for CP and other genetic disorders, families may appreciate information regarding increased risks of disability (e.g. congenital limb difference, ³⁵ intellectual disability ³⁴ ) because of consanguinity.

The use of kava during pregnancy was reported in five children (9.6%). Kava is a beverage made from the roots of a plant (Piper methysticum) native to the Pacific islands and has sedative, anaesthetic, psychoactive, and euphoriant properties. ³⁶ Kava has an important role in traditional medicine and traditional ceremonies (males‐only); however, it has become more widely consumed for recreational and social use. In Vanuatu, females in the TAFEA Province have the lowest rates of kava consumption, with 5.9% of females reporting they had kava in the last month. ²⁶ The World Health Organization recommends potential harm consideration strategies regarding kava. ³⁶ The risks of consumption during pregnancy or breastfeeding are unknown. Engagement of specialists in perinatal pharmacology and fetal drug interactions is warranted to elucidate potential mechanisms linking kava consumption during pregnancy to potential adverse effects on the developing fetus.

All children in this cohort were single‐born infants; no multiple birth pregnancies were reported. There were also no children born by caesarean delivery. At the time this study was undertaken, Lenekel Hospital was unable to provide caesarean births; mothers who required a caesarean were flown to Port Vila Hospital. In Vanuatu, the maternal mortality ratio (the number of females who die from pregnancy‐related and childbirth‐related causes) is 94 per 100 000 live births, ³⁷ and an infant mortality rate of approximately 16 deaths per 1000 live births. ²⁶ , ³⁸ Approximately one‐third of the mothers in this cohort gave birth at home or at a traditional village healer. Like other LMIC studies, there was a smaller number of infants born preterm in our study, probably because of low preterm survival rates. ¹¹ , ²² In contrast, infants born preterm make up a larger proportion of children with CP in HICs, for example, 43% of children with CP in the Australian CP register were born preterm. ²⁴

In this study, just over half of the families reported prenatal and perinatal causes, with many families reporting birth asphyxia and birth complications. One‐quarter of families reported a CP cause in the postnatal period, with a period of typical development and then an episode of illness with high fevers and seizures. Most families were unable to provide information on a specific diagnosis of the illness, but meningitis (due to bacterial infections and tuberculosis) is a concern in Vanuatu and the Pacific Islands; risk factors such as malnutrition increase the risk of death or disability. ³⁹ , ⁴⁰ Collection of data regarding the timing and cause of CP was challenging because of the low‐resource setting and limited diagnostic resources.

Of the children not diagnosed with CP, 91% had a clinical diagnosis or impairment. Records were kept such that surgical and rehabilitation services may be sought for children who have club feet, cleft palate, or limb deficiency. Medical, nutrition, audiology, and optometry services are all needed. The presence of night blindness indicates vitamin A deficiency; vitamin A deficiency is present in 16% of preschool children in Vanuatu, indicating that vitamin A is lacking in their daily diet. ⁴¹

Challenges experienced by Ni‐Vanuatu families

Vanuatu is recognized as one of the most vulnerable countries in the world because of the impact of climate change. ⁴² While Tanna is known as a fertile tropical island, volcanic ash and acid rain from the volcano, together with weather events such as cyclones and extreme rainfall, cause food insecurity; ⁴² for example, the twin cyclones of February 2023 devastated crops. The Vanuatu economy is based predominantly on the tourism industry; consequently, the lengthy COVID‐19 travel restrictions (2020–September 2022) heavily affected the economy. Subsistence agriculture and issues of poverty and undernourishment are prevalent. ⁴² Most caregivers in our study reported their occupation as a gardener or farmer, which mirrors the overall Tanna constituency in that 65% of households on Tanna report that subsistence living is their main income. ⁹ The proportion of children with CP who were moderately and severely underweight in this study was high; undernutrition puts children at greater risk of dying from common infections. ⁴³ Nearly half of all deaths in children under 5 years are attributable to undernutrition. ⁴³ Lack of health and rehabilitation services and the practical difficulties of accessing services were the two main points families repeatedly raised when asked about the main challenges they face. Travel is expensive in Vanuatu, with interisland and intraisland logistics challenging. Land travel on the island is mostly only possible using four‐wheel drive. Phone and internet connection is not available in many parts of Tanna, making communication difficult.

Study limitations

The reported prevalence of CP and ABI is an underestimate because we were unable to complete assessment of all children with possible CP within the assessment clinical time frames. This is because of a range of factors, including time and travel constraints, for example, we did not have time to visit the north‐eastern part of the island, which could only be accessed by boat, and wet weather made other roads impassable. Unfortunately, not all VHWs were able to attend the education sessions and more cases of children with CP have been identified by visiting volunteer medical professionals since the study completed data collection. Case ascertainment to determine the level of coverage is tricky. Tanna is divided into seven districts that changed boundaries during the study period. The ‘East’ district was recorded as having a population of 2981 in 2022; if the proportion of children was the same as the census data (0.498), there would be approximately 1485 children in ‘East’. We diagnosed six children with CP in ‘East’, which gives an estimated prevalence of 4.0 per 1000. Different districts ranged considerably in prevalence rates, possibly reflecting our coverage of the district from 1.7 per 1000 in ‘West’ to 7.3 per 1000 in ‘South’.

Previous studies of the key informant method show that it has a slightly lower case ascertainment rate compared to door‐to‐door surveys, with a positive predictive value of 77.6%. ⁴⁴ In addition, we suggest that the results are an underestimate of the incidence of CP because of the high mortality rate of infants and children with CP in LMICs. ²³ It is also possible that children with mild impairments are not noticed by VHWs. Because of the remote setting and limited diagnostic tools, we relied predominantly on parent and caregiver interview for medical history.

Recommendations

The Cerebral Palsy Register in Tanna, Vanuatu is a snapshot of one island in a Pacific LMIC. There are potentially modifiable risk factors for CP. Recommendations should be prioritized for effective prevention of CP, such as easier access to perinatal and postnatal care. Consideration of public health education regarding the increased risk of disability in children from consanguineous marriages may be appropriate in regions where consanguinity is common.

We found that a high proportion of children experience undernutrition and recommend urgent nutritional education and support services.

Establishment of early diagnosis and intervention will improve individual outcomes for children with CP. We recommend workforce capacity building of allied health, medical, and nursing professionals to perform screening and early detection, and provide evidence‐based services to children with CP. International collaborations (university, professional associations, government, NGOs) with health professionals in Vanuatu focused on building and extending paediatric rehabilitation services are recommended. Support from larger NGOs and governmental organizations for rehabilitation services and equipment (medical, surgical, allied health, assistive devices) is needed. Technical infrastructure to support telehealth is recommended to help bridge geographical barriers.

For future CP registers in comparable remote LMIC contexts, establishing strong partnerships is essential and should be prioritized. Appropriate time frames for data collection are recommended, allowing for adverse weather conditions, transportation services (air, boat, and land) subject to resource constraints, and environmental factors and difficult terrain that may unexpectedly impede or restrict travel. Sustainable strategies would involve training local personnel, such as community nurses, to maintain the register in a part‐time capacity; however, this approach necessitates dedicated financial support and personnel.

Conclusions

On Tanna Island, Vanuatu, the burden of CP is high, diagnosis is delayed, most children with CP have never accessed any rehabilitation service, and most do not attend school.

Rehabilitation services are limited and not easily accessible. Reducing preventable risk factors and facilitating early diagnosis and intervention should be prioritized. Early diagnosis and intervention services for childhood disabilities warrant urgent public health intervention.

Supporting information

Figure S1: Distribution of weight‐for‐age z‐score by GMFCS

Figure S2: Distribution of weight‐for‐age z‐score by monthly mean income (Vatu)

Table S1: Primary diagnoses of all children and adults seen in the medical camps.

Table S2: GMFCS level of children in different age groups.

Table S3: Mean weight‐for‐age z‐score by GMFCS levels.

Table S4: Access to rehabilitation services and education services.

Table S5: GMFCS level of children attending school.

Burgess A, Jeremiah L, Moiteau S, Al Imam MH, Jahan I, Das M, et al. Epidemiology of cerebral palsy on Tanna Island, Vanuatu: A population‐based register study. Dev Med Child Neurol. 2026;68:1292–1304. 10.1111/dmcn.70137

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.