Assessing routine childhood vaccination acceptance, hesitancy and refusal in Cape Town, Western Cape, South Africa: a mixed-method study protocol

Abstract

Introduction

Vaccines have proven to be one of the most effective ways to prevent infections. Since the introduction of the Expanded Programme on Immunisation worldwide by WHO in 1974, vaccines have saved more than 150 million lives and prevented countless illnesses and disabilities. In South Africa, vaccination has contributed substantially to the decline in under-five childhood deaths from 89 418 in 2005 to 47 409 in 2013. However, the country still has limitations in achieving optimal uptake of vaccines.

Methods

This study will use a mixed-methods design to identify the factors associated with acceptance, hesitancy and refusal of childhood vaccines in three subdistricts of the Cape Town Metropolitan Health District in South Africa. In phase I of the study, we aim to estimate the uptake of childhood vaccines and assess the factors associated with uptake among a representative sample of at least 236 caregivers of children aged 0–2 years in a cross-sectional study. For phase II of the study, we will explore the behavioural and social drivers of childhood vaccination from the perspective of caregivers of children aged 0–2 years using a purposively selected sample of 20–25 participants from phase I. Finally, for phase III of the study, we will explore the behavioural and social drivers of childhood vaccination from the perspective of 20–25 healthcare workers providing vaccinations to children aged 0–2 years purposively sampled from primary healthcare facilities.

Ethics and dissemination

Ethical approval was obtained from Stellenbosch University (S23/10/232). Various steps will be taken to ensure the anonymity and confidentiality of participants. Findings of the study will be shared with various healthcare stakeholders through conferences and presentations at relevant meetings.

Conclusion

Evidence from several studies conducted in South Africa shows a significant increase in vaccine hesitancy within the country. The findings of this study will provide insight into the characteristics associated with vaccine acceptance, hesitancy and refusal in Cape Town. The results will help inform the development of evidence-based interventions that can be implemented to improve vaccination coverage.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Our mixed-methods study will focus on understanding both structural and psycho-social factors that may be influencing childhood vaccination which has rarely been studied in the South African context.

• The rigorous methods, including the sampling strategy, of this study will ensure that it provides a comprehensive understanding of the barriers and facilitators to vaccination uptake.

• This study will focus on children aged 0–2 years of age; therefore, the results of the study may not be generalisable to all children receiving vaccinations.

• There is a possibility of social desirability bias where some participants may want to give acceptable answers regarding vaccinations.

• There is also a possibility of recall bias where participants may have to recall about vaccinations in the absence of a vaccination card.

Introduction

Since the introduction of the Expanded Programme on Immunisation (EPI) worldwide by WHO in 1974, vaccines have saved millions of lives and prevented countless illnesses and disabilities.¹ In South Africa, vaccination has contributed substantially to the decline in under-five childhood deaths from 89 418 in 2005 to 47 409 in 2013.¹ The first introduction of the EPI programme in South Africa in 1974 included 6 antigens, namely measles, polio, tuberculosis, pertussis, diphtheria and tetanus; and has increased to 11 antigens which now include rotavirus, haemophilus influenzae type B and hepatitis B, pneumococcal conjugate and human papillomavirus vaccines.^{1 2} Vaccination of children starts at birth, and subsequent doses scheduled at 6 weeks, 10 weeks, 14 weeks, 9 months, 12 months and 18 months for children under 2 years of age and at 6, 10 and/or 12 years for older children.¹ Although the EPI programme has improved vaccination coverage considerably in South Africa, vaccination coverage has remained suboptimal.² Over the last few years, vaccination coverage has fluctuated between 80% and 90% in South Africa for fully vaccinated children.³ The 2019 EPI national coverage survey indicates that vaccination coverage stands at 76.8% for fully vaccinated children by 18 months of age.³ Low vaccination coverage has serious consequences, such as re-emergence of previously controlled vaccine-preventable diseases in children.⁴

Diseases like diphtheria and measles have shown signs of re-emergence, attributed to low vaccination coverage. Recent outbreaks of these diseases have been reported in Gauteng, KwaZulu-Natal and Western Cape provinces.^{4 5} The National Institute for Communicable Diseases recorded 1151 laboratory-confirmed measles cases across eight provinces between epidemiological week 40 of 2022 and week 35 of 2023.⁶ As of 2023, UNICEF report indicated there is a 30% decline in confidence in childhood vaccines in South Africa.⁷ Reasons for low vaccination coverage globally are commonly attributed to supply and demand issues which relate to the provision of vaccines and vaccination services and population decision to vaccinate.⁸ Structural barriers such as low vaccine availability and poor access to health facility are well known to reduce vaccination coverage in South Africa; however, research suggests that social and psychological factors can also affect vaccine uptake.⁹

National estimates of vaccination coverage often do not reflect the variability within the country. In addition, concerns about the safety of vaccines and their effectiveness have been shown to greatly influence decisions around vaccination.^{10 11} Whether a child receives the intended vaccination in a timely manner is often influenced by parental decision.¹² Studies show that parental decisions to vaccinate their child or avoid vaccination are often complex and multidimensional.^{9 12} Unvaccinated children tend to cluster in communities, increasing the possibility of outbreaks of vaccine-preventable diseases.^{9 12} In these communities, some of the reasons why some parents do not vaccinate may be related to psychological beliefs that oppose the value of vaccinating children.¹¹

Research shows that common themes associated with psychological beliefs leading to vaccine refusal or vaccine hesitancy include beliefs that vaccine(s) cause more harm than the disease that they are meant to prevent.¹¹ A prominent example is the persistent misinformation linking the measles-mumps-rubella (MMR) vaccine to autism.¹¹ While these theories were debunked, the misinformation about MMR is still causing vaccination scares, delays and refusal.^{11 12} Additionally, the promotion of mandatory vaccination policies has historically faced resistance and may further fuel vaccine hesitancy.¹³ In South Africa, like many countries, parents have to consent to have their child vaccinated and their refusal can cause missed opportunity for children who would benefit from receiving the vaccination.¹⁴ The growing body of evidence in South Africa shows that vaccine hesitancy is a developing phenomenon; however, understanding of its extent and determinants, including strategies of how to prevent or address, is still limited.¹⁵

Most research in the country has focused on structural barriers to childhood vaccination such as supply and demand; however, not much has been done to understand behavioural and social factors that may be influencing childhood vaccinations. Various conceptual frameworks have been established focused on vaccine hesitancy and psychosocial factors influencing the decisions to vaccinate. These include (1) ‘C’ framework for determinants of vaccine hesitancy which draws on psychological models of decision-making behaviours; (2) the WHO Strategic Advisory Group of Experts (SAGE) vaccine Hesitancy Determinants Matrix which draws from adaptations of ecological models of health behaviours; (3) Cooper’s alternative model of vaccine hesitancy determinants, which suggest four main categories of factors influence vaccination behaviours; and (4) a framework by the WHO SAGE group, Measuring Behavioural and Social Drivers of Vaccination (BeSD), that may influence vaccination uptake.^{8 16}

The initiative around the vaccine hesitancy models developed and various working groups has ensured a growing evidence base around vaccine hesitancy.⁸ While it is known that vaccine hesitancy is on the rise in African countries, including South Africa, its direct effects on vaccine coverage and herd immunity are not known.⁸ This study aims to identify the factors associated with childhood vaccination acceptance, refusal and hesitancy that may be influencing vaccination uptake and in turn vaccine coverage using WHO’s framework BeSD).

Objectives

The aim of the study is to identify and understand behavioural and social drivers of childhood vaccination in the Cape Town Metropolitan District of the Western Cape, South Africa. The specific objectives are:

1. Estimate the rate of uptake of childhood vaccines and assess factors associated with the uptake of childhood vaccines among a representative sample of caregivers of children aged 0–2 years in the Cape Town Metropolitan Health District.

2. Explore the behavioural and social drivers of childhood vaccination in the Cape Town Metropolitan Heath District from the perspective of caregivers of children aged 0–2 years.

3. Explore the behavioural and social drivers of childhood vaccination in the Cape Town Metropolitan Health District from the perspective of healthcare workers providing vaccinations to children aged 0–2 years.

Methods

This study will use a mixed-methods design to identify the factors associated with vaccine acceptance, hesitancy and refusal influencing the uptake of childhood vaccines in three subdistricts of the Cape Town Metropolitan Health District of the Western Cape province of South Africa where vaccine coverage is routinely low. According to the Cape Town Metro health plan 2018/2019 to 2020/2021, vaccination coverage for children under 1 year of age was the lowest in Cape Town Eastern, Mitchells Plain and Tygerberg Health Sub-Districts on the 2018/2019 target with a coverage of 73.8%, 74.7% and 75.0%, respectively.¹⁷ The Cape Town Metropolitan Health District has the largest urban communities of all municipalities in the Western Cape and performs functions of both district and local municipalities.¹⁸ The Cape Town Metropolitan Health District serves a population of 4.1 million persons.¹² The study will be conducted in three phases as depicted in figure 1. We aim to start recruitment from 1 November 2024 and anticipate results to be available by 31 October 2025.

Figure 1. Study phases.

Conceptual framework

The conceptual framework for our study will be based on WHO’s BeSD conceptual model for behavioural and social drivers of vaccine uptake. The model has four domains: (1) thinking and feeling, (2) social processes, (3) motivation (or hesitancy) and (4) practical factors (see figure 2).¹⁹ These domains, to a great extent, follow the increasing vaccination model (IVM). Motivation, for example, is part of the IVM model; however, the BeSD conceptual model has separated it to clearly illustrate where hesitancy fits within the model (figure 2).¹⁹ Each domain has constructs with indicators that can be calculated in each region. The constructs and indicators selected are suited to provide region-specific actionable information to inform their immunisation programmes.¹⁹ The items in each construct from the BeSD model may need to be modified and refined to identify behavioural and social drivers for the demand and uptake of childhood vaccines in the Cape Town Metropolitan Health District, Western Cape, South Africa.

Figure 2. The Behavioural and Social Drivers of Vaccination model.

Phase I: estimating the rate of vaccination and identifying associated factors influencing childhood vaccination uptake

This study will use a cross-sectional study design. We will be recruiting caregivers of children aged 0–2 years via randomly selected Early Childhood Centres (ECDs) in the Cape Town Metropolitan Health District of the Western Cape Province with routinely low vaccination rates. Sampling for the survey will be done in three subdistricts where vaccination rates are routinely low.¹⁷ The names of ECDs in the Western Cape will be downloaded from the National ECD Emis database into an Excel sheet.²⁰ We will conduct a simple random selection of 25–30 Early Development Centres from the three subdistricts using Stata software.²¹ We will use a purposive sampling strategy to select participants to ensure a diverse sample. Participants will be selected, taking into consideration their gender, socio-economic status and geographical location.

Sample size calculation

Rate of vaccination uptake in the Western Cape was used for sample size calculation. The national EPI coverage survey 2019 indicates only 77% of children had received all 14 vaccine doses recommended for children up to 18 months of age nationally.³ In the Western Cape, the proportion of children with Road to Health Booklet indicating that they had received the recommended 14 vaccine doses by age 18 months was 81%.³ A sample size of at least 236 is required to estimate the rate of vaccination uptake to achieve a precision of ±6%. Sample size estimation was done using WINPEPI.²²

Recruitment

We will visit randomly selected Early Development Centres and recruit approximately 250 eligible participants from the three chosen subdistricts. To be eligible for inclusion, participants must be a parent or caregiver of children 0–2 years and at least 18 years and older. Should a parent have more than one child within the age range, we will conduct the survey based on one child who is the youngest of the children within the age range to determine both the vaccination status and the intention to vaccinate the child. The study will be introduced to the participants by a study principal investigator or field research assistants, and a description of the study will be provided to the participants in an information sheet in one of the three languages: English, IsiXhosa or Afrikaans. Consent forms for the caregiver will be provided and explained in one of the three languages: English, IsiXhosa or Afrikaans. Participants will be included in the study once consent is given.

Data collection

We will conduct surveys using a structured questionnaire drawing from the BeSD Child Immunization Survey and the 2019 EPI national coverage survey (online supplemental material 1).^{3 14} We will pilot-test the study questionnaire among three participants, one participant for each language to ensure survey questions and response options are understood as intended, are adequately adapted to the local context and measure what they are supposed to measure. We will recruit the two participants from the target population to ensure questions are suitable for the intended population. Data will be collected using current best practices to ensure high-quality responses in the surveys. A computer-assisted face-to-face interview style will be applied via field research assistants who will be equipped with portable electronic devices for data collection.²³ Each survey should take approximately 20 min, and participants will answer questions with the assistance of the field research assistants who will translate questions into the language of the participants. Data collection will take approximately 1 month. Field research assistants will carry out survey data recording in real time on predesigned and pretested electronic forms using portable electronic devices. Electronic devices will be equipped with Bluetooth and Global Positioning System technology to enable remote electronic transfer of files and allow tracking of the data collection process. We will use Open Data Kit, a free and open-source software for collecting data in resource-constrained areas.²⁴ Field research assistants will be trained in data protection and management procedures, for example, storage and transmission of data. Data transmission will be encrypted to guarantee data protection.²³

Data management

We will collect data using password-protected REDCap-enabled mobile tablets. The questionnaire, which is in English, will be translated into IsiXhosa, Afrikaans languages on REDCap. Raw data will be stored in a secure database, and only the principal investigator and coinvestigators will have access to the raw data. Data collected in the mobile tablets will be saved as a data file which will be exported to REDCap as a CSV file and then to the Stata software for analysis. Identifying information will be removed from all transcripts before analysis and will be destroyed once the study is completed. Data management will be in accordance with the Stellenbosch University data regulation policy which includes the Protection of Personal Information Act (4 of 2013) (‘POPIA’).

Data analysis

We will report the proportion of immunisation uptake and intention with the corresponding 95% CIs. Log-binomial regression analysis will be used for univariate and multivariable analyses to assess the association between uptake of immunisation and exposure variables. Exposure variables with a p-value<0.1 in univariate analysis will be entered in multivariable analysis. Exposure variables with a p-value<0.05 will be considered statistically significant in multivariate analysis. We will report risk ratios as measures of association, with the corresponding 95% CIs. All analyses will be conducted using Stata V.17 or higher (Stata Corp, Texas, USA).²¹

Phase II: identifying behavioural and social drivers of childhood vaccination uptake from the perspective of caregivers

This study will use a qualitative method. We will be recruiting caregivers of children aged 0–2 years who will be identified from phase I for in-depth interviews. We will conduct semistructured interviews using the BeSD Childhood Immunization In-Depth Interview Guide (CIDI) for caregivers.¹⁴

Sampling

We will begin with a sample size of 20–25 participants. However, sampling will continue if we have not reached saturation of the data through this initial sample size.²⁵ A sample of participants who took part in the survey study (phase I) will be invited to take part in an in-depth interview. The target population will be caregivers of children ages 0–2 who are eligible to receive vaccines. Drawing on initial findings from the survey (phase I), sampling will be purposive with the aim of ensuring caregivers with a range of childhood vaccination views and practices are included, including those who accepted vaccination for their children, those who are hesitant and those who have refused vaccination for their children. Sampling will also seek to ensure that the sample reflects a range of sociodemographic characteristics, including in relation to education, gender, socio-economic status and geographical location.

Recruitment

We will invite selected participants to participate in the study via email and/or telephone. On invitation to the study, participants will be provided with an information sheet about the study and provided an opportunity to contact the principal investigator or field research assistants with questions. To be eligible for inclusion, participants must be 18 years or older caring for a child who is between 0 and 2 years. The study will be introduced to the participants by the principal investigator or field research assistants in English or interpreted in either IsiXhosa or Afrikaans according to participants’ language of choice. An information sheet describing the study will be provided to the participants in one of the three languages: English, IsiXhosa or Afrikaans. Consent forms for caregivers will be provided and explained to the participants in the language of choice between the three languages. Participants will be included in the study once consent is given.

Data collection

We will conduct semistructured interviews using BeSD CIDI for caregivers/parents (online supplemental material 2) with caregivers of children under the age of 2 years. We will pilot-test the topic guide among three participants, one participant for each language to test the flow of the questions, the relevance of probes and the feasibility and acceptability in terms of time and length of the discussion. We will use the results of the pilot study to inform adaptation of the guide if needed. Each interview will be conducted in a private room at a convenient time and place for the participant. Interviews will be conducted using the in-depth interview guide by the principal investigator or research assistant who is fluent in either English, Afrikaans or IsiXhosa according to th participant’s language of choice. Interviews will be approximately 20–45 min long, with open-ended questions to allow participants to be fully expressive, and to understand and respond to questions. Participants will be asked questions that are structured around the four domains ‘thinking and feeling’, ‘social processes’, ‘motivation’ and ‘practical factors’ to understand their experiences with, and perception about, childhood vaccinations. The discussions will be recorded in a portable recorder device and transcribed in the language of the interviewee’s languages. The transcriptions that are in IsiXhosa and Afrikaans will be translated into English.

Data management

The data sets from interviews will be stripped of all personal information to deidentify the data. Each participant will be allocated identification numbers that will be used instead of their names on hard copies and electronic documents. Study notes from discussions and recording devices will be securely stored in a locked suitcase. The audio recordings will be uploaded to a secure password-protected location iCloud at the end of each interview. Electronic files such as transcripts will be password-protected and stored in password-protected computers of the principal investigator and a backup of electronic data will be saved on the password-protected iCloud location. All hard copies such as handwritten interview notes and printed transcripts including signed consent forms that include information that identifies participants will be stored in a locked cabinet that only the principal investigator has the key to. All records will be kept in a secure location by the principal investigator after the study has ended until the passing of the period stipulated by the university to keep these records; after that period, the records will be destroyed. Data collected from this study will only be accessible to the principal investigator and coinvestigators.

Data analysis

Thematic analysis will be used to code and organis transcribed data segments for analysis.²⁶ A codebook will be developed both deductively and inductively, drawing on the main thematic findings from the 2021 Background Paper on Behavioural and Social Drivers of Vaccine Uptake as well as codes that will emerge inductively from collected data.²⁷ The codes generated will be independently verified against raw data for accuracy and consistency by a second researcher. The analysis will seek to develop themes related to the caregivers’ experiences and perspectives of childhood vaccination, their attitudes towards childhood vaccination uptake and the behavioural and social factors influencing their decisions to vaccinate or not. We will use illustrative quotes to develop analytical summaries. Data analysis will be done using NVivo, a software for qualitative data analysis.²⁸

Phase III: identifying behavioural and social drivers of childhood vaccination from the perspective of healthcare workers providing vaccinations to children

This study will use a qualitative survey study design. We will recruit healthcare workers providing vaccination to children aged 0–2 years. We will conduct semistructured interviews using the BeSD CIDI for frontline health workers.¹⁴

Sampling

We will begin with a sample size of 20–25 participants. However, sampling will continue if saturation of the data has not been reached through this initial sample size.²² We will use a one-stage cluster sampling technique using the Western Cape health directory containing all health facilities in the Western Cape to randomly select 8–10 primary health facilities from the three subdistricts.¹⁸ We will use a purposive sampling strategy to select healthcare workers from primary healthcare facilities. Participants will be selected, taking into consideration their roles, gender, socio-economic status, and geographical location.

Recruitment

We will seek approval from the Western Cape Department of Health and Wellness and the City of Cape Town Municipality City Health Directorate before the recruitment of participants begins. We will approach facility managers to refer healthcare workers who are providing vaccination services within the facility. From the names provided by the health facility manager, we will select and recruit healthcare workers to participate. To be eligible for inclusion, healthcare workers must be providing vaccination services to children as part of their daily responsibilities. The study will be introduced to the participants by a principal investigator or field research assistants in English, IsiXhosa or Afrikaans according to the participant’s language of choice. The description of the study will be provided to the participants in an information sheet in one of the three languages. Consent forms for healthcare workers will be provided and explained to the participants in one of the three languages. Participants will be included in the study once consent is given.

Data collection

We will follow the same data collection process from the phase II study to conduct semistructured interviews using BeSD CIDI for front-line health workers (online supplemental material 3).¹⁴

Data analysis

We will follow the same data analysis strategy from phase II of the study. The analysis will seek to develop themes related to healthcare workers’ experiences and perspectives of childhood vaccination, and how they view attitudes towards childhood vaccination uptake, and the behavioural and social factors influencing the decisions of caregivers to vaccinate their child(ren) or not.

Data management

We will follow the same data management process from phase II of the study.

Ethics and dissemination

This study was submitted and obtained ethics approval (S23/10/232) from the Health Research Ethics Committee at Stellenbosch University and permission will be sought from the Western Cape Department of Health and Wellness and the City of Cape Town Municipality City Health Directorate to conduct assessments in Cape Town Metropolitan health facilities. We will seek informed consent from participants who wish to take part in the study. This study will be conducted according to the ethical guidelines and principles of the international Declaration of Helsinki, South African Guidelines for Good Clinical Practice and the Medical Research Council Ethical Guidelines for Research. Participants will be provided with an information sheet for additional information about the study.

Description of risks and benefits

Potential risks

There is a low probability of harm or injury (physical, psychological, emotional) occurring because of study participation and low potential risks in the time needed to participate. We will minimise disruptions to routine or usual work schedules for healthcare workers and caregivers. The probability and magnitude of harm or discomfort anticipated in this study will not be greater than that ordinarily encountered in the routine work or daily activities of participants. Potential risks for beneficiaries may include feelings of unease or anxiety related to answering the interview questions. Should participants feel uneasy or have anxiety while being interviewed, the principal investigator or research team member will discuss with participants if they wish to stop or pause the interview and get permission to reschedule the interview at another time. If the participants require further psychological help, the participant will be referred to their nearest primary health facility for counselling.

Potential benefits

Study participants may indirectly benefit from this study by being able to identify their own attitudes, social drivers and behaviours that may be influencing their decisions to provide and/or accept childhood vaccines. This study can potentially improve childhood vaccination uptake for participants by providing useful childhood vaccination and related health information to parents/caregivers and the importance of timely vaccination. The study may also potentially influence healthcare workers to learn more about the benefits of vaccination and raise their awareness about childhood vaccination to caregivers/parents. Findings from the study will also provide the opportunity to inform local stakeholders and policymakers about how best to achieve optimal childhood vaccination coverage.

Informed consent process

Signed informed consent from all participants will be obtained before surveys or interviews are conducted. Study description will be provided, and the informed consent form will be explained to the participants in English, IsiXhosa or Afrikaans on recruitment. Should participants not understand English, an interpreter will engage with participants in their preferred language (IsiXhosa or Afrikaans) to provide the study description and explain the consent form. The principal investigator or field research assistants will clearly explain to all study participants that they are free to leave the study at any point and that they may choose not to answer any question asked. Participants will be assured that should they choose to leave the study or decide not to answer any question, it shall be of no risk to them.

Privacy and confidentiality

In addition to following the ethical principles for research conducted in humans, data collected in the study will be used with respect to the POPIA 2013 act. Field research assistants will be required to sign a research confidentiality agreement before the commencement of data collection. For interview records, external transcription services will be contracted to ensure compliance with adequate data protection rules and regulations. All recordings will be deleted on completion of the study.

Dissemination

The findings will be shared with study participants (caregivers, facility managers, healthcare workers) through a report summarising the main results and published in peer-reviewed journal articles. Participants will receive these and other updates through their provided contact details. Additionally, results will be communicated to various healthcare stakeholders through conferences and presentations at relevant meetings.