Methods

Study Design

An explanatory mixed‐methods design was used, consisting of a quantitative descriptive survey followed by qualitative semi‐structured interviews. This design enabled the identification of patterns in feeding challenges (Phase 1) and provided in‐depth contextual explanations of these patterns through caregivers' narratives (Phase 2). The quantitative phase included standardized tools assessing socio‐demographic characteristics, functional severity, eating and drinking ability and household food access. The qualitative phase, guided by phenomenological principles, explored caregivers' lived experiences, decision‐making processes and interpretations of the feeding challenges identified in the survey. This sequential approach was appropriate for a pilot study and ensured methodological rigour by linking measurable trends to subjective, context‐dependent experiences. This study was approved by the University of the Witwatersrand Human Research Ethics Committee (HREC) (Non‐Medical) (Approval number: STA_2025_39). Informed consent was provided by all participants prior to completion of the online survey and again before the interview.

Participants and Recruitment

Non‐probability convenience sampling was utilized. Recruitment took place at two physical disability‐focused centres and through various social media platforms (WhatsApp, Instagram and Facebook). Using social media expanded accessibility to caregivers who faced time, transport or caregiving constraints, aligning with research indicating that digital platforms are increasingly valuable for reaching dispersed or under‐resourced populations. This online site played a critical role in the recruitment process and supported inclusivity by enabling participation beyond physical care centres. By incorporating both disability‐focused centres and social media settings, the study captured caregiving experiences situated within varied socio‐economic conditions and service contexts.

Eligible participants were primary caregivers of children aged 2–8 years with a confirmed diagnosis of CP. Caregivers were required to be directly responsible for the child's daily feeding and nutritional decision‐making. Caregivers also needed to be proficient in English to participate in interviews. Children across all levels of the Gross Motor Function Classification System (GMFCS) and Eating and Drinking Ability Classification System (EDACS) were included.

Survey Development

A study‐specific survey (Supporting Information) was developed by the three authors who are all speech‐language pathologists working with children with CP and associated feeding difficulties in South Africa. The third author (SA) is a lecturer at the University of the Witwatersrand and her research focuses on paediatric feeding difficulties and family‐centred approaches to intervention. Research Electronic Data Capture (REDCap) was used for screening, electronic consent and survey administration (Patridge and Bardyn 2018). Caregivers were asked to consult their child, as appropriate. The survey included five data collection tools. Following the completion of the survey, all participants were interviewed online.

Demographic Questionnaire

The demographic questionnaire captured caregiver demographics, socio‐economic conditions and child‐related information, providing essential context for interpreting feeding practices and barriers within the South African food environment.

GMFCS

The GMFCS is a functional classification system composed of five levels, used to assess the motor skills of children with CP. The levels are I: walks without limitations, II: walks with limitations, III: walks with the aid of a manual mobility device, IV: has limited mobility and can use a motorized chair and V: needs to be transported in a manual wheelchair (Ho et al. 2017). Based on this system, scores one and two are defined as mild, three is moderate, and four and five are classified as severe CP (Rosenbaum et al. 2008).

EDACS

The EDACS was also used as an instrument for screening to categorize each child's eating and drinking ability. Scores ranged from minimal difficulty to severe impairment with aspiration risk. This facilitated comparisons between caregiver experiences across levels of feeding severity. The EDACS was studied and proven to be valid and reliable (Tschirren et al. 2018).

Food Inventory Questionnaire

The Food Inventory Questionnaire was used to assess household food availability and children's dietary intake, with analysis focused on whether specific foods were eaten by the child or not. It was adapted to the South African context using locally relevant foods (e.g., pap, amasi, samp, vetkoek), informed by Eating and Drinking in Southern Africa (Du Rand and Fisher 2020), enabling a culturally and environmentally valid assessment of feeding within household food constraints.

Household Food Insecurity Access Scale (HFIAS) for Measurement of Food Access Questionnaire

This tool was used to better understand food security in each of the families' households and was adapted from approaches used in the United States of America (Coates et al. 2007). HFIAS was developed as a guide to understanding the multidimensional concept of food insecurity (Coates et al. 2007). This questionnaire guided interviews related to food security under different domains. This questionnaire was used in Ethiopia and developing countries (with different cultural contexts) in recent years to gather insight into food insecurity and household perceptions of this (Coates et al. 2007). The prevalence of household food insecurity can be determined by the use of the HFIAS. Gebreyesus et al. (2015) found evidence that the HFIAS tool has good internal consistency is a valid tool to measure household food insecurity. It is recommended that adaptations be made to the wording of the tool before its use (Gebreyesus et al. 2015). It was found that the HFIAS was reliable and valid when translated into Arabic and used in rural Lebanon (Naja et al. 2015).

Semi‐Structured Interviews

Semi‐structured interviews were conducted with caregivers following the completion of the online survey. The interview questions focused on caregivers' lived experiences, daily routines, strategies, safety concerns and socio‐cultural influences on feeding. Interviews were conducted online via Microsoft Teams, with 1GB mobile data provided where necessary. Each interview lasted between 45 and 55 min. All participants signed an additional consent form prior to participation.

Data Analysis and Integration

To ensure accuracy, two researchers independently checked and entered the data (TV, RA). The characteristics of the recruited children and caregivers were expressed as frequencies and percentages. The survey data were analysed descriptively. For all participants, GMFCS, EDACS and HFIAS scores were calculated and reported. Interview data were analysed using thematic analysis. Thematic analysis has six steps: (1) familiarization of the data, (2) coding the data, (3) generating themes, (4) review and development of themes, (5) defining and refining themes and (6) writing the report. Two researchers (TV, RA) independently coded all responses. Discrepancies were resolved by consensus with adjudication by a third independent coder (SA). Themes were discussed with all authors, and key themes were triangulated with the survey data. Themes were derived both inductively and through predetermined areas of discussion, based on known feeding challenges from the literature. Quantitative and qualitative datasets were integrated during analysis. Survey findings provided descriptive patterns of feeding challenges and food access, while qualitative data contextualized and explained these patterns. This integrated analysis improved the depth, validity and interpretive richness of findings for a pilot investigation.