Development of a multicenter cerebral palsy registry in three Arabic-speaking countries: Preliminary results

Overview of study design

The multicenter CP register study was conducted across five institutions in three ASCs. The participating Arabic academic institutions included (1) Faculty of Medicine, Ain Shams University, Cairo, Egypt, (2) Faculty of Physical Therapy, Cairo University, Cairo, Egypt, (3) Faculty of Physical Therapy, October 6 University, Giza, Egypt, (4) National Institute Mongi Ben Hmida of Neurology, Tunis, Tunisia and (5) Hamad Medical Corporation, Doha, Qatar. The project’s administrator/supervisor was responsible for administering and supervising the project, including providing regular critical reviews of the study plan and conduct. Site leaders (local site leaders) lead the project within a participating institution and manage day-to-day queries from local researchers collaborating within that institution. Local site leaders were also accountable for ensuring that research practices were in accordance with the requirements of the project’s scientific protocol and the local ethics committee. Site leaders recruited relevant research collaborators to participate in the register. Research collaborators were responsible for identifying patients, obtaining consent, and collecting data. They were grouped under the “Arabic Cerebral Palsy Register Group.” An online interactive training session, led by a senior physiotherapist/author (MAE), was conducted to explain the study and educate research collaborators on conducting a comprehensive assessment of children with CP in accordance with protocol settings and data collection forms. The aim was to standardize measurements across assessors and increase the reliability of research collaborators. Each participating institution piloted the data collection form on at least one patient before starting actual work.

Surveillance methods and recruitment

We enrolled children/adolescents with CP who regularly attend the outpatient clinics or inpatient wards of the participating institutions, per the inclusion criteria and protocol’s ethical requirements that were set based on international registries of CP criteria. All participating institutions were academic referral centers in their respective countries and are based in the capital cities.

Inclusion and exclusion criteria

The term CP encompasses a group of disorders, which may be patho-etiologically diverse. To be defined as CP, the following items should be fulfilled: (a) an insult or anomaly affecting the immature brain, namely prenatal, natal, or postnatal in early years of life, (b) the primary brain insult or anomaly is permanent, and (c) it involves a disorder of movement and/or posture and of motor function. The diagnosis was based on the Surveillance of CP in Europe Guidelines. An updated description of CP has recently been proposed.^[21] Patients were included if they fulfilled the above definition of CP. Inherently progressive brain insults, such as anomalies associated with neurodegenerative disorders, were excluded. We included children and adolescents aged from 24 months to 18 years with a medical diagnosis of CP.

Data collection, quality, and sharing

A Research Electronic Data Capture (REDCap) was used to create physician-administered de-identified data collection forms.^[22] It was hosted by Ain Shams University. REDCap is a secure web data collection tool for building and managing online surveys and databases. Collected data included both patient- and physician-reported items. The electronic data collection form was discussed, developed, and agreed upon by the authors [Appendix]. Only the REDCap administrator had access to the whole de-identified data set. Site leaders at the other participating institutions were given access to their respective data sets. Computers were protected by strict security measures. Different levels of access were granted to the study researchers, namely the REDCap administrator and the site leaders of the participating institutions. The de-identified data set was shared with authors/researchers across participating institutions in ASCs for data analysis and research purposes per protocol requirements. The identified paper versions of the electronic data collection forms were stored safely by site leaders for clinical management purposes.

Data analysis

The data were analyzed using the Statistical Package for the Social Sciences program, version 27. Quantitative data were presented using minimum, maximum, mean, and standard deviation. Qualitative data were presented using counts and percentages. Student’s t-test was used to compare quantitative data between two independent groups. Chi-square and Fisher’s Exact tests were used to compare qualitative data between different groups. A P ≤ 0.05 was considered statistically significant.

Risk factors

We investigated 390 CP patients, 217 males, in this cross-sectional study. Patient demographics are presented in Table 1. We reported notable potentially preventable or modifiable CP risk factors, particularly those occurring pre- and perinatally [Tables 2 and 3]. Prematurity showed a highly significant correlation with low birth weight, length of stay in the neonatal intensive care unit (NICU), mechanical ventilation, multiple pregnancies, assisted reproduction (P < 0.001), and a significant association with a history of antenatal maternal illness (P = 0.02) [Table 4].

Comorbidities

There was a higher prevalence of epilepsy among children with more severe motor impairments as per Gross Motor Function Classification System (GMFCS) and Manual Ability Classification System (MACS) levels (χ² = 23.31, P = 0.000) and (χ² = 32.50, P = 0.000), respectively [Supplementary Table 1]. We found a similar significant association between the prevalence of cognitive/intellectual disability and motor function impairment as per GMFCS levels (χ² = 31.23, P = 0.000) [Supplementary Table 2]. We found a similar significant association between the prevalence of visual and auditory impairments and motor function impairment as per GMFCS levels (χ² = 11.17, P = 0.02) and (χ² = 8.60, P = 0.02), respectively [Supplementary Tables 3 and 4]. We reported a highly significant correlation between the prevalence of speech impairment and GMFCS and MACS levels (χ² = 99.87, P = 0.000) and (χ² = 77.83, P = 0.000), respectively. In that, children with more severe GMFCS and MACS levels were more likely to experience speech and language impairment [Supplementary Table 5]. We reported a highly significant correlation between the prevalence of anemia and GMFCS and MACS levels (χ² = 10.42, P = 0.04) and (χ² = 15.60, P = 0.000), respectively. In that, children with more severe GMFCS and MACS levels were more likely to have anemia [Supplementary Table 6]. Children with more severe swallowing difficulties (including choking) were more likely to experience nutritional deficiencies and anemia (χ² = 45.96, P = 0.000).

Relation to health service utilization

Several risk factors were significantly associated with disease severity. For example, NICU admission was significantly associated with more severe motor function impairment as per GMFCS levels (χ² = 16.71, P = 0.001). The proportion of children in more severe GMFCS levels III, IV, and V who were admitted to NICU was significantly higher compared to children with less severe GMFCS levels I and II [Supplementary Table 7]. Receiving occupational therapy under regular healthcare follow-up was significantly associated with less severe manual impairments, as measured by the MACS (χ² = 20.65, P < 0.001). The proportion of children in MACS levels I and II who received regular occupational therapy was significantly higher compared to children in MACS levels IV and V [Supplementary Table 8]. However, the correlation between GMFCS and MACS levels, and history of pathologic fractures was not statistically significant (χ² = 4.91, P = 0.08) and (χ² = 4.96, P = 0.08), respectively.

Interestingly, 86.2% of patients did not receive any orthopedic surgery, and 85.6% did not receive any form of Botox injections. Only 59.0% received physiotherapy under regular healthcare follow-up, and only 4.4% received instructed parent/home physiotherapy exercises [Supplementary Tables 9 and 10].

Strengths and limitations

The current study has limitations. Many ASCs, including Egypt, are in the early stages of adopting electronic health record systems in primary healthcare and tertiary referral hospitals.^[39,41] In general, many African and Asian LMICs face similar challenges of implementing electronic health record systems, especially in regard to technical infrastructure, financial restraints, and compliance of physicians and healthcare providers, among others.^[42-45] Longitudinal health records are important to patient care and the establishment of surveillance registers. This could have created enrollment or recruitment bias. Expanding the institutional adoption of electronic health record systems and utilizing Internet-based modalities to engage parents and healthcare professionals in patient recruitment and registration could reduce bias and yield a fairly representative sample of the CP population in ASCs. In addition, this has made the current estimation of CP prevalence decidedly challenging or impractical. Some items on the data collection form were self-reported by the patient or parent. It was not possible to ascertain the level of agreement between patient self-report and the medical records/reports. This could have induced recall bias and influenced the reliability of the collected data. However, previous studies have shown substantial agreement between maternal self-reports and medical records.^[46,47] Moreover, comorbidities such as intellectual disability and audiovisual impairment were assessed dichotomously or categorically. Qatar was the only HIC in our study. However, HICs collectively constitute only 13.5% of the total population of ASCs.^[10] Therefore, this is unlikely to have a remarkable impact on the interpretation of our findings [Figure 1]. The administration time of the data collection form items and the associated respondent-patient/parent-burden may have been linked to the relatively limited sample size. However, this was a preliminary study that is expected to expand.

Close

Figure 1:

Graphical Abstract: The health burden of cerebral palsy in this sample of Arabic-speaking countries was substantial. Perinatal risk factors, comorbidities, and disease severity constituted a notable health burden. Results suggest that the pediatric healthcare provision is inadequate and/or underutilized. CP: Cerebral palsy

Export to PPT

Data availability statement

Raw data were generated at Ain Shams University by REDCap. The derived data supporting the findings of this study are available from the REDCap administrator, author S.M.A.H., on request.

Recommendations:

Health policymakers in ASCs are urged to consider the current interpretations. The feasibility of implementing an Arabic CP register could be improved by expanding the institutional adoption of electronic health records across ASCs. Expanding the coverage of this Arabic CP register by including more participating medical institutions from ASCs can help us arrive at more reliable conclusions. This could better guide our clinical practice.