Study design and procedure

We conducted a prospective, observational mixed-methods pilot study from May 2021-September 2022 in seven GP practices across two urban London boroughs. The study was designed as a pilot to test processes and approaches which may inform a future large-scale study or trial. The overall objectives were to measure routine vaccination coverage among migrants presenting to UK primary care and establish and test new referral pathways for catch-up vaccination. The study procedure was as follows: following recruitment, participants were asked about their vaccination history (including for routine childhood immunisations including MMR, Td/IPV, and other vaccines including tuberculosis/bacille Calmette-Guerin vaccine (TB/BCG) and HPV), which was coded into their electronic patient record and the study database. In the absence of a written vaccination card or record documenting a completed vaccine course, or if patients said they had not had a vaccine or were unsure, patients were referred for catch-up vaccination for each eligible vaccine (following the UK algorithm for vaccinating individuals with uncertain or incomplete immunisation status [21]) and invited to attend an appointment(s) with their practice nurse. Eligible catch-up vaccines were MMR, Td/IPV, HPV (aged 11–25 years) and MenACWY (aged 10–25 years). Practice nurses followed the UK algorithm to administer missing vaccine doses, boosters, and courses and recorded the data into the patient’s electronic record and the study database. A standardised data collection tool was designed to facilitate the collection of data, which then prompted referrals for catch-up vaccination (see Data collection and referral pathway for catch-up vaccination).

PICOTS criteria for the study are shown in Table 1. In addition to collecting quantitative data from migrant patients, we explored the views of practice staff on catch-up vaccination and current guidance, including barriers to implementation, suggestions, and areas for improvement and support, through focus group discussions (FGDs), which were carried out in August 2022. During the study, we also decided to conduct an in-depth interview with two staff members to explore examples of good practice from the most successful (in terms of recruitment and uptake) participating practice. We carried out in-depth interviews with a diverse range of recently arrived (≤ 10 years) migrants to explore views and concerns around catch-up vaccination after arrival in the UK. The study tool, recruitment, and data collection pathways are shown in Fig. 1. The reporting of this study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [22].

Ethics and PPI

This study received ethics approval from the NHS Health Research Authority Yorkshire and Humber—South Yorkshire Research Ethics Committee (20/YH/0342) on 18 December 2020. The qualitative in-depth interview study with migrants received ethics approval from the St George’s, University of London Research Ethics Committee (REC reference: 2020.0058). Migrants with lived experience of the UK immigration and healthcare systems were involved in the design of this study through our National Institute for Health and Care Research (NIHR)-funded Patient and Public Involvement and Engagement (PPIE) Project Advisory Board and were compensated for their time and contributions.

Population, recruitment, and informed consent process

The study was conducted with support from the NHS North Central London Research Network (NoCLoR) and the North Thames Clinical Research Network (CRN). GP practices in areas with a high proportion of migrant residents were purposively invited to join the study. We aimed to recruit up to 10 GP practices across two boroughs (Barnet and Tower Hamlets) in North and East London (referred to henceforth as sites 1 and 2), with a target sample size of 100 participants. Boroughs were selected for their high proportion of migrant residents (estimated to be approximately half, according to 2021 Census data [23]). Both rank in the top 50% of most deprived local authorities in England, based on the English indices of deprivation 2019 [24], although Tower Hamlets ranks as significantly more deprived than Barnet. In practice, seven GP practices were recruited, with six across site 1 and one practice belonging to site 2.

Patients registered at participating practices were eligible for the study if they were (a) aged 16 years or older, (b) born outside of the UK (our migrant definition excluded those born in North America, Australia, New Zealand, or Western Europe, as defined by the UN maximal definition of Western Europe [25]), and (c) capable of giving informed consent. Recruitment procedures differed between the two sites (see Fig. 2).

Fig. 2

Figure showing standardised data collection tool (left) and referral pathways implemented in study sites 1 and 2 (right). VPDs, vaccine-preventable diseases; PN, practice nurse; HCA, healthcare assistant; CRN, clinical research network

We held site initiation visits with all practice sites, inviting GPs, practice managers, healthcare assistants (HCAs), and nurses involved in immunisation. Alongside delivering training on the current UK primary care catch-up vaccination guidelines [21] and the referral pathway to implement upon identifying under-vaccinated patients, these visits covered the study processes and procedures, approaches to identifying the study population and recruiting participants, and use of the standardised data collection tool.

At site 1 (n = 6 GP practices), clinical practice staff were originally going to recruit and consent patients. However, the recruitment pathway was modified as clinical staff were under intense pressure from the COVID-19 pandemic, so the CRN led the recruitment and consenting process. Practice nurses and HCAs first identified patients who met the eligibility criteria, filtering patient records by ethnicity or notes on migrant status (where recorded) to identify those potentially eligible and sent an SMS/text message with a link to the study website, from which patients could download the study documents (participant information sheet [PIS], consent form, and leaflets about catch-up vaccination and HPV vaccination, all available in the six dominant local languages, which were Arabic, Farsi, Pashtu, Romanian, Urdu, English). A researcher at the CRN (DF) then followed up with patients by a telephone call enquiring whether they would like to join the study and to take informed consent. Practice nurses also mentioned the study opportunistically to patients during routine appointments, who would then be referred to the CRN researcher (DF) for consent. At site 2 (n = 1 GP practice), the practice nurses HCAs invited and consented participants to the study opportunistically during routine appointments, as per the original recruitment pathway. Participants were given hard copies of the study documents and given the opportunity to ask questions and decide whether they wanted to participate. We gave practice and CRN staff a copy of a form detailing the names of common childhood vaccines in multiple languages, to support taking vaccine history during appointments (see Supplementary files). Telephone interpreters (via Language Line) were available on request at both sites during recruitment and data collection.

Data collection and referral pathway for catch-up vaccination

We developed a standardised data collection tool using Microsoft Excel, which was used to collect specific sociodemographic information (such as country of birth, which is not routinely recorded in patient records), immunisation history, and monitoring and uptake data when patients were referred for catch-up vaccination (Fig. 2). We documented participants’ rates of under-vaccination for MMR, Td/IPV, and other key vaccines in the UK routine immunisation schedule, history of VPDs, and uptake rates of MMR, Td/IPV, MenACWY, and HPV vaccines following referral to the practice nurse for catch-up vaccination. We also explored sociodemographic factors associated with under-vaccination in the study population. Immunisation history was based on self-reporting or vaccination records (via the primary care computer system or hand-held vaccination cards) where available.

Data collection and referral pathways and procedures differed between sites and are outlined in Fig. 2. In site 1, the CRN researcher collected core data via telephone call with the participant, which were recorded in the patient’s electronic medical record and on the password-protected study database. The CRN researcher determined the participant’s need for catch-up vaccinations based on the study training and the UK catch-up vaccination guidelines [21] and, if accepted by the participant, contacted the practice nurse (at the practice where the patient was registered) to arrange an appointment. Once the CRN staff had facilitated an appointment for first doses, they then left practice nurses to follow-up patients for subsequent doses as per routine care. Subsequent catch-up vaccination doses (uptake data) were recorded by practice staff in the patient’s medical record at the time of administration and these data were later extracted by the CRN researcher (see Data management, follow-up, and statistical analysis). In site 2, the practice nurse collected core data (recorded in the patient’s medical record) during face-to-face appointments, administered first doses where vaccine stocks allowed, and booked patients for any necessary follow-up appointments for catch-up vaccinations and subsequent doses. Anonymised study data (core, monitoring and uptake data) were extracted from electronic patient records by the practice manager at site 2 and securely transferred to the CRN researcher, who added them to the aggregate study database.

Data management, follow-up, and statistical analysis

We aimed to follow-up patients for a minimum of 6 months at both sites to allow for all doses (Td/IPV is 3 doses, with a 4-week gap in between each; Fig. 1). At the end of follow-up, the CRN researcher securely extracted monitoring and outcomes data from participants’ electronic medical records and updated the aggregate study database. A de-identified, anonymised version was then transferred securely to the study team at St George’s for data cleaning and analysis.

Data cleaning and analyses were done using STATA 12. All tests were two-tailed and p values less than 0.05 were regarded as significant. We used descriptive statistics to describe the sociodemographic characteristics, vaccination history, VPD history, and catch-up vaccine uptake of participants. We summarised continuous data with mean and standard deviation (SD) and described categorical responses using the frequency and percentage. Comparisons between categorical variables were calculated using Pearson’s chi-squared test, and comparisons between continuous variables were calculated using unpaired t-tests.

Bivariable and multivariable logistic regression analyses were chosen to model the relationship between a binary outcome and predictor variables and were used to look for factors associated with being un-vaccinated (received zero doses) or under-vaccinated (received at least 1 dose, but not full schedule) for key vaccines at the time of study enrolment. Outcomes included un-vaccinated for MMR vaccine, un-vaccinated for Td/IPV vaccine, un-vaccinated for MMR vaccine and Td/IPV vaccine, unvaccinated for any polio—combined or single vaccines, unvaccinated for any measles—combined or single vaccines, and under-vaccinated for MMR vaccine or Td/IPV vaccine. Explanatory variables were age, sex, birth region, region lived prior to the UK, years in the UK, and study site (migration reason and occupation were only recorded in site 1 and were therefore not included in the regression analyses). Multivariable models were built in a forward, stepwise fashion. Age, sex, and birth region were adjusted for in each multivariable model; certain variables were removed from the final model to reduce collinearity.

Qualitative component

Our qualitative component included FGDs and an in-depth interview conducted with practice staff from participating practices and in-depth interviews conducted with recently arrived migrants. Topic guides were developed by the research team. The interviews with migrants were done remotely (either over the phone or through video call) across 17 months. Migrant participants were recruited using purposive and snowball sampling, with the aim of recruiting participants from a broad range of nationalities, migration statuses, and age groups. Adverts for the study and participant information sheets were circulated to 20 UK-based migrant support groups (mostly based in South London and chosen for their locality around St George’s, University of London) and on social media. Those who expressed an interest in taking part were contacted by telephone, and the study was explained to them with interpreters available on request. Translated participant information sheets were circulated, and written informed consent was obtained from all participants prior to carrying out an interview (methods reported in full elsewhere [26]). We did three FGDs which were scheduled to take place at the end of routine practice meetings conducted on Microsoft Teams (most convenient for participants). Participants were practice nurses, HCAs, and practice managers (roles involved in vaccination delivery/scheduling) from the participating practices. An in-depth interview was conducted with two staff from site 2 (due to timing, these staff had not participated in FGDs). For the FGDs, all staff received information about the study and how their data would be used in advance, which was reiterated at the start of the meeting, and staff were able to make an informed decision about their participation. Participants were asked to imply consent by remaining on the call, which was considered appropriate because the topic was low risk, not audio recorded, and anonymised summary feedback (broad views) was collected. All participants received a PIS and provided written informed consent prior to participating. Both the FGD and staff interviews followed a semi-structured topic guide, which explored participants’ experiences of implementing the study, current barriers and challenges to delivering catch-up vaccinations, and suggestions for improving the tool, referral pathways, and engaging migrant patients/promoting catch-up vaccination among these groups. Broad views and selected short-hand quotations (non-attributable) were collected during FGDs in the form of hand-written and typed notes (by SH and LPG). The staff interview was conducted by AFC with two staff participants in a private room, audio-recorded and transcribed verbatim by a professional transcription service.

Qualitative data were analysed deductively using a flexible and rapid thematic analysis and evaluation approach [27]. Notes from the FGDs which were reflected on and discussed afterwards by AFC, SH and LPG, and AFC then independently coded and grouped the findings into broad barrier and facilitator concepts using a matrix method (by hand and in Microsoft Excel). The data in the matrix were corroborated and discussed again by the three researchers, to ensure rigour and coding reliability. The same approach was used to analyse the transcript of the key informant interview. Migrant interviews were analysed using the thematic framework approach in NVivo 12. Triangulation occurred when the qualitative and quantitative data were combined but also by the interaction between the three researchers during data collection and analysis and through the contributions of their own perceptions, beliefs, and academic disciplines to the collection and interpretation of data.