Abstract Introduction Research methods and design Results Discussion Conclusion Acknowledgements References About the Author(s) Thembi J. Katangwe Department of Pediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa Mariana Kruger Department of Pediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa School of Psychology, University of KwaZulu-Natal, Pietermaritzburg, South Africa Ronald van Toorn Department of Pediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa Jeanetta van Zyl Department of Pediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa Sandile Ndlovu Department of Pediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa Regan Solomons Department of Pediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa Kirsten A. Donald Department of Pediatrics and Child Health, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa Department of Neuroscience Institute, University of Cape Town, Cape Town, South Africa Citation Katangwe, T.J., Kruger, M., Van Toorn, R., Van Zyl, J., Ndlovu, S., Solomons, R. et al., 2024, ‘Paediatric cerebral palsy in South Africa: Prevention and care gaps at hospital level’, African Journal of Disability 13(0), a1449. https://doi.org/10.4102/ajod.v13i0.1449 Original Research Paediatric cerebral palsy in South Africa: Prevention and care gaps at hospital level Thembi J. Katangwe, Mariana Kruger, Ronald van Toorn, Jeanetta van Zyl, Sandile Ndlovu, Regan Solomons, Kirsten A. Donald Received: 26 Apr. 2024; Accepted: 27 July 2024; Published: 30 Oct. 2024 Copyright: © 2024. The Author(s). Licensee: AOSIS. This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: Population-based data show high proportions of severe cases of cerebral palsy (CP) in resource-poor regions such as sub-Saharan Africa, where most children have potentially preventable risk factors (factors that may increase the likelihood of CP occurrence but can be mitigated through medical interventions). Objectives: This study aimed to describe the demographic and clinical profile of children living with CP accessing services at Tygerberg Hospital over a period of 10 years (2010–2020), identify the potential gaps in care (proportion of individuals in a country requiring but not receiving suboptimal or inadequate care), and comparison with a similar study at the same centre two decades ago. Method: This 10-year retrospective study investigated causes and morbidities in children with CP, attending a central hospital in the Western Cape, South Africa. Results: A total of 613 children with CP were identified. Perinatal causes were predominant, especially in 57.7% (n = 354) of the cohort: perinatal asphyxia (41.1%) and preterm birth (16.6%). Postnatal causes constituted 15.2% (n = 93), which included tuberculous meningitis (3.6%) and bacterial meningitis (3.6%). The most common complications were intellectual impairment (61.8%; n = 379); epilepsy (30.8%; n = 189) and visual impairment (54.7%; n = 234). A third of the cohort had severe CP, classified as Gross Motor Function Classification System IV and V (38%). Conclusion: Most of the previously documented main drivers of CP are still present and the implementation of healthcare prevention strategies remains inadequate. Contribution: This study provides longitudinal evidence to confirm that CP in a South African setting is associated with a high burden of potentially preventable causes. Keywords: cerebral palsy; perinatal asphyxia; prematurity; care gaps; risk factors. Introduction Globally, cerebral palsy (CP) remains an important cause of motor disability. It has a reported prevalence of more than 4 per 1000 live births in low- and middle-income countries (LMICs) (Murugasen et al. 2024), compared to 1.6 per 1000 live births in high-income countries (HICs) (Katangwe et al. 2024). Aetiological differences between different regions have been documented for CP. In HICs, prematurity associated with extremely low birth weight is most frequently documented, while perinatal asphyxia and kernicterus remain prevalent causes in less-resourced settings (Hollung et al. 2018; Smithers-Sheedy 2016). Cerebral palsy registers in HICs have proven to be effective in documenting aetiological and phenotypic trends for individuals living with CP as well as providing data that led to the implementation of targeted strategies to improve prenatal, perinatal and postnatal healthcare, with a subsequent decline in CP prevalence (Johnson, Blair & Stanley 2011; Morgan et al. 2016). For many HICs, besides strengthening neuroprotective strategies for infants born prematurely, the quality of life for children living with CP has also improved (Makris, Dorstyn & Crettenden 2021). Data paucity in LMICs including South Africa, remains a major hindrance in understanding the full spectrum of aetiologies. This directly impacts opportunities for planning preventative strategies alongside motivating for the required health workers and infrastructure to support these approaches. The estimated CP prevalence for South Africa is derived from a few studies, which have reported up to 10 per 1000 live births in certain communities (Murugasen et al. 2024). In many LMICs, perinatal asphyxia, kernicterus and central nervous system (CNS) infections (Katangwe et al. 2024) remain important CP causes. Results from the Global LMIC CP register confirm differences in aetiology of CP between LMICs and HICs (McIntyre et al. 2022). While some risk factors are common across contexts, there are important risk factors that are region-specific. Examples of such factors prevalent in sub-Saharan Africa and far less common in HIC settings include postnatal infections, such as tuberculous meningitis (TBM) and cerebral malaria (Katangwe et al. 2024). A previous study (between 2003 and 2004) of 242 children living with CP and accessing care also at this teaching hospital found a high prevalence of perinatal insults (38% of total cases; predominantly birth asphyxia and prematurity-related complications) (Van Toorn et al. 2007). Of note was also a very high percentage of children with acquired causes of CP (21%), especially secondary to TBM and kernicterus. Therefore, it is essential to investigate the CP aetiology over time to ensure the implementation of evidence-based interventions to address preventable causes, especially as there is an increase in the medico-legal litigation against healthcare professionals for CP caused by preventable conditions. This study therefore aims to investigate whether the risk factors and aetiologies of CP have changed over the last two decades and whether the gaps in healthcare access and specific interventions remained. The study objectives were as follows: (1) to describe the socio-demographic characteristics of children living with CP in this cohort; (2) to describe the identified risk factors for CP, ascertain the timing of the brain insult and/or injury where possible and document the physical classification and the motor function by using the Gross Motor Function Classification System (GMFCS) level; (3) to describe the most common co-morbidities experienced by children living with CP and to describe the gaps in care for children with CP at this tertiary facility; (4) to compare findings to a study performed in the same setting between 2003 and 2004 (Van Toorn et al. 2007). Research methods and design Setting and study population This was a descriptive retrospective cohort study of 613 children (1–13 years of age) living with CP, seen in tertiary-level specialised paediatric developmental and paediatric neurology clinics between January 2010 and December 2020 at Tygerberg Hospital. The hospital is a large state-funded hospital with 302 secondary and tertiary paediatric beds, serving approximately half of the 2.4 million children requiring healthcare services in the Western Cape province, South Africa (Tygerberg Hospital 2022). Children aged between 1 and 13 years (at the time of the data collection) who accessed paediatric neurology and neurodevelopmental services at the teaching hospital with a confirmed diagnosis of CP were included. The diagnosis fulfilled the ‘Surveillance of Cerebral Palsy in Europe and the Australian Cerebral Palsy Register’ case definition that includes the following key elements: (1) an injury to the developing brain that is non-reversible and remains non-progressive; (2) it affects movement, posture, and motor function; and (3) the insult or anomaly occurs in a developing brain (Surveillance of Cerebral Palsy in Europe [SCPE] 2000). Data collection Routinely collected demographic and outcome data, populated through the Enterprise Content Management System (ECM) of the teaching hospital were reviewed. The children’s families’ socioeconomic stata were deduced from the income classification used by the Western Cape Provincial Government of South Africa to determine subsidies for public healthcare usage within the province (Western Cape Government 2022). Patients are either fully subsidised if they are unemployed or pensioners and partially subsidised according to their earnings. A full breakdown of all healthcare subsidy criteria that is used in the province can be seen in the footnote of Table 1. TABLE 1: Baseline characteristics of children living with cerebral palsy at Tygerberg Hospital. Data validation was carried out by listing all children attending the neurology and/or neurodevelopment clinics during the study period with their unique hospital record number. The unique hospital record number was thereafter used to access full patient medical records in the provincial patients electronic database named ECM for the Western Cape province. Where there were discrepancies, the paper-based patient records were retrieved for further data collection and validation. Clinical documentation in case notes, laboratory and special investigation reports were used to determine CP aetiology. Perinatal asphyxia was defined using the 2019 American College of Gynaecology (ACOG) task force criteria on neonatal encephalopathy and CP (The American College of Obstetrics and Gynaecology 2017, 2019). Suggestive history, clinical examination, laboratory findings including auditory brainstem responses and magnetic resonance imaging (MRI) were used to confirm chronic bilirubin encephalopathy (kernicterus). Polymerase chain reaction (PCR) test results, viral cultures and neuroimaging findings confirmed congenital infection. The presence of at least one of the following (1) intraventricular haemorrhage (IVH) with or without ventriculomegaly on cranial ultrasound scan; (2) the presence of periventricular leukomalacia (PVL) on brain imaging confirmed white matter injury of prematurity. White matter injury of prematurity is an umbrella term of all the pathologies that result in cerebral white matter injury of immature brains in preterm babies. These include germinal matrix haemorrhage-intraventricular haemorrhage (GMH-IVH), PVL, and diffuse white matter injury (Lee et al. 2017). Definite bacterial meningitis was confirmed by the presence of a positive gram stain and/or positive bacterial culture on cerebrospinal fluid (CSF). Tuberculous meningitis was identified using the TBM research case definition (Marais et al. 2010) comprising clinical, laboratory and radiological components. Hypoglycaemic brain injury was confirmed by the presence of the following: Plasma CSF glucose <2 mmol/L T2-weighted MRI images showing bilateral white matter signal intensities in the occipital and parietal lobes. Hypovolemic shock was confirmed by the presence of at least two of the following: Documented hypotension Evidence of fluid resuscitation and inotropic support Evidence of ischaemic brain injury on imaging. Prenatal causes of CP were classified as non-progressive brain injuries that occurred in the developing brain in utero, anytime from the post-conception period to the period before birth (Bax et al. 2005). Perinatal causes were non-progressive brain injuries that occurred in the developing brain during birth. Postnatal causes were non-progressive brain injuries that occurred after birth (Bax et al. 2005). The SCPE classification system (SCPE 2000) was used to classify CP. The documented topographic classification was used as supporting evidence for classification (Bax et al. 2005). Motor function was captured using the GMFCS (Paulson & Vargus-Adams 2017). Comorbidities were defined as disorders occurring alongside CP but may also occur independently without the CP (Hollung et al. 2020). The intellectual level of a child with CP in this cohort was determined by using the Molteno developmental screening assessment tool (Springer et al. 2022) and from that the Developmental Quotient (DQ) was calculated and recorded for each child in their files. Children without information on their DQ assessment were categorised in an ‘unknown’ group (Brown, Parikh & Patel 2020). Behavioural problems were identified from the narrative text of the hospital notes, bearing in mind challenges in the identification of behaviourally defined conditions in a child with significant neuromotor disability. A child with CP was deemed epileptic if they had two or more unprovoked seizures more than 24 h apart, with or without documented electroencephalogram and neuroimaging results and confirmation of chronic anti-epileptic usage. The vision was categorised into five main groups: Normal vision Some other visual impairment (strabismus, refractive errors, gaze dysfunction, visual field defects) Confirmed cerebral cortical visual impairment Normal light perception – can fixate and follow Blindness. The hearing was categorised into four main groups: Normal hearing Some hearing impairment (use of hearing aids, conductive hearing loss, and sensorineural hearing loss) Confirmed deafness No data available. Allied health services were defined as healthcare professionals excluding medical professionals (aHPCSA 2018). Examples of healthcare services include physiotherapy, occupational therapy, speech and language therapy, audiology, nutrition and dietetics, social work and health education. A child was deemed to have accessed allied health services if their records documented attendance of at least four of any of these services regardless of the time frame. Statistical strategy Descriptive statistics were used to analyse demographic data, clinical subtypes and motor function (GMFCS). Categorical variables were reported using absolute values (n) and reciprocal percentages (%), corrected to one decimal point. Medians and interquartile ranges (IQRs) were calculated for numerical values, assuming normal distribution. All data were analysed using STATA version 16.0 (StataCorp 2019). This statistical package was recommended by our statistics expert as it offers improved functions that assist with data cleaning as well as having a wide range of statistical techniques, with advanced features for managing missing data, a commonality in retrospective studies. Ethical considerations Ethical approval to conduct this study was obtained from the Stellenbosch University, Health Research Ethics Committee (No. S21/10/195). Results A total of 613 children were identified with CP. The median age at study inclusion was 9 years (IQR 6–10 years). Most of these children were of school-going age (78.8%; n = 483) (see Figure 1). The majority were males (60.4%; n = 370) and the median age of identifying at-risk children for CP was 9 weeks (IQR 3–19 weeks). FIGURE 1: Age distribution of children living with cerebral palsy at Tygerberg Hospital. Characteristics of children living with cerebral palsy Spastic CP was the most common type of CP in this cohort (60.4%; n = 370), followed by mixed CP type (10.8%; n = 66). Within this spastic group, bilateral spastic CP was the predominant CP type (n = 370; 60.4%), with a third having equal involvement of all four limbs (n = 194; 31.6%) and 12.3% (n = 75) having more lower limb involvement. Unilateral spastic CP was present in 16.5% (n = 101). Some of the other CP types included dystonic CP (4.6%; n = 28), dyskinetic CP (2.4%; n = 15) and ataxic CP (0.5%; n = 3) (Table 1). The majority of the cohort had severe motor function, GMFCS IV and V (38.0%; n = 233). The annual income for most of the families in this cohort was < ZAR100 000 (