Identifying characteristics that enable resilient immunisation programmes: a scoping review

Search strategy and selection criteria

Our review protocol was written according to Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA) guidelines15 16 and was originally registered on the Open Science Framework (OSF)17 on 11 January 2022 (https://doi.org/10.17605/OSF.IO/TRXK2). Following peer review, changes to the protocol were made to clarify the review aims and broaden the literature search. An updated protocol indicating all updates from the original version as tracked changes is available on OSF (https://doi.org/10.17605/OSF.IO/6N2G3).

We conducted the review using the Arksey and O’Malley framework18 and reported information in line with PRISMA for Scoping Reviews guidelines.19 The research questions that we addressed were as follows: (1) is the Health System Resilience Index14 suitable for assessing immunisation programme resilience? and (2) what are the major gaps in primary empirical research in assessing immunisation programme resilience? Our review questions are presented in PICO (Population, Intervention, Comparison and Outcome) format in online supplemental information S1.1, table S1.

Our eligibility criteria are tabulated in online supplemental information S1.2 (table S2 inclusion criteria and table S3 exclusion criteria). We included primary empirical research studies with any study design20 that were published in peer-reviewed journals. We did not review grey literature. We assume that the immunisation programme resilience characteristics discussed in primary empirical research papers are representative of characteristics published in the grey literature. As the qualitative descriptive outcomes considered in this review are likely less subject to selective reporting and publication bias than quantitative and hypothesis-based outcomes, the impact of this approach is lessened. We also restrict our search to the decade preceding the commencement of this review (from 1 January 2011 onwards). The concept of resilient health systems is relatively new, as recently outlined,12 14 and discussion of its application to immunisation programmes is as recent as the development of the 2017 Health System Resilience Index.14 No further limits were applied.

We searched MEDLINE (Ovid), Embase (Ovid), PsycINFO (Ovid), CINAHL (EBSCO), CENTRAL (Wiley), Web of Science (SCI-EXPANDED and SSCI; Clarivate Analytics), and the top 200 hits from Google Scholar.21 The search was run on 2 September 2023, thus including publications from 1 January 2011 to 2 September 2023. The search strategies were developed to identify publications that mention both the keywords ‘immunisation programme’ and ‘resilience’ (as well as several variations on these keywords), filtered out ineligible publication types (eg, reviews, editorials), and applied a publication date limit from 1 January until the date the search was run. No other limits or filters were applied, and the review was not limited to non-English language publications. Our search strategies were independently peer-reviewed using the Peer Review of Electronic Search Strategies (PRESS) checklist22 by Kat Steiner, an Outreach Librarian at the Bodleian Health Care Libraries, University of Oxford. All search strategies are presented in full in online supplemental information S1.3.

The search results were deduplicated and screened using EPPI-Reviewer Web (ER-Web).23 Study selection was a two-stage process: screening on title and abstract followed by screening on full text. The screening was carried out in duplicate by two independent reviewers and disagreements were settled by discussion between both reviewers. Finally, we performed citation searching on the reference lists of all studies included in the review after screening.24

Data extraction

From each study, we extracted basic publication information data (eg, publication year), study design, crisis type, geographical location, disease targeted by immunisation programme and resilience characteristic discussed based on the 2017 Health System Resilience Index.14 The data extraction form was developed in an iterative process and involved an initial piloting stage where five studies were independently data extracted by two reviewers, with disagreements and refinements to the data extraction form agreed through discussion. Data extraction on the remaining studies was first performed by a single reviewer and then validated by a second reviewer, with disagreements settled by discussion between both reviewers. Data extraction was performed using ER-Web.

The study design was assessed at two levels.20 The first level depended on whether the study investigators assigned the exposures: affirmative for experimental studies; negative for observational studies. The second level specified the experimental and observational studies further. For experimental studies, this was determined by whether the exposure allocations were random: affirmative for randomised controlled trials (RCTs); negative for non-RCTs. For observational studies, this was determined by whether a comparison group was included: affirmative for analytical studies; negative for descriptive studies.

We classified the crisis type in each study using the PESTLE (Political, Economic, Social, Technological, Legal, Environmental) taxonomy25 and by specific crisis described. PESTLE crises were categorised as political if related to government interventions and actions; economic if related to, for example, inflation, exchange rates or economic growth; social if related to public demand, traditional or social media coverage, or armed conflict; technological if related to Research and Development (R&D) activity, automation, or IT; legal if related to the law or changes in the law; and environmental if related to disease outbreak, natural disasters or other ecological or environmental events. PESTLE categories were treated as non-exclusive.

Geographical location was extracted by continent (Africa, Asia, Australia, Europe, North America or South America; defined geographically) and country. The disease type targeted by the immunisation programme was extracted as reported in the publication. Lastly, the resilience index being considered was listed as described by the original investigators.

Data analysis

We charted the primary outcome variables using the Health System Resilience Index outlined by Kruk et al,14 which contains five resilience characteristics:

• Aware: Tracks population health threats; maps systems strengths and weaknesses; knows available resources.

• Integrated: Coordinates between government, global and private actors; works across sectors; involves communities.

• Diverse: Addresses range of health problems; provides quality services that meet population needs.

• Self-regulating: Isolates health threats; minimises disruption to essential services; can access reserve capacity.

• Adaptive: Transforms operations to improve function; acts on evidence and feedback; encourages flexible response to fit situation.

We used an adaptive approach to charting this outcome variable based on the studies reviewed. Adaptations were discussed by authors and the data charting form was updated using an iterative process.

The primary results were tabulated in detail, displayed in summary using bar charts and synthesised narratively. An evidence gap map was generated (using EPPI-Mapper V.2.2.4) to highlight evidence gaps in the literature to draw attention to potential blind spots.

Role of funding source

Employees of the funder were involved in the design and conduct of the study; collection, management, analysis and interpretation of the data; preparation, review or approval of the manuscript and decision to submit the manuscript for publication.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Study selection and study characteristics

We identified 37 studies26–62 that were eligible for this review: 29 from databases and 8 from citation searching. The study selection process is displayed in figure 1. Specific characteristics for each included study are presented in online supplemental table S4, with summaries of the study characteristics presented in figure 2.

Figure 1

PRISMA flow diagram of study selection process. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Figure 2

Summary of study characteristics. The details for each individual study are provided in online supplemental table S4. (A) Number of studies published per year. Our literature search spanned from 2011 to 2023. The red dot on 2015 indicates the year ‘What is a resilient health system? Lessons from Ebola’12 was published. This paper is the precursor to the Health System Resilience Index published in 2017.14 The novelty of the concept of resilient health systems is outlined in these publications. Applying the concept of resilience to immunisation programmes appears to reflect the novelty of applying the concept to health systems, with only a single relevant publication being identified pre-2016. (B) Number of studies adopting observational (descriptive or analytical) or experimental (RCT or non-RCT) study designs. (C) Number of studies reporting on crisis type, categorised according to the PESTLE taxonomy. (D) Number of studies per continent. RCT, randomised controlled trial.

Over 97% (36/37) of identified studies were published from 2016 onwards (figure 2A). The majority of these adopted either a descriptive (21/37) or analytical (15/37) study design, with only a single RCT30 design identified (figure 2B).

Crises that occurred were predominantly environmental (26/37): 18 were due to the COVID-19 pandemic, 5 were due to the 2013–2016 Ebola epidemic, 2 were due to hurricanes and 1 was due to an earthquake. There were nine studies on social crises, two were on legal and one was on political crises. We did not identify any publications on either economic or technological crises impacting immunisation programmes (figure 2C).

Most studies reported on crises in Africa, Asia or Europe (12 studies in each continent). Six studies reported on crises in North America, and only three studies originated from both South America and Australia (figure 2D).

While some studies considered the resilience of a specific immunisation programme, several studies considered the resilience of a national immunisation programme. As a result, charting studies by specific diseases were not possible.

Immunisation programme resilience characteristics

Based on our research, we identified five resilience characteristics that may form the basis of an Immunisation Programme Resilience Framework: (1) Integration, (2) Adaptiveness, (3) Resource Availability and Access, (4) Self-regulation and (5) Awareness (figure 3 and table 1). Four characteristics (Integration, Adaptiveness, Self-regulation, Awareness) are consistent with the Health System Resilience Index characteristics; one was identified based on the literature search as a key characteristic that should be taken into account. To clarify the scope and distinction between each of these characteristics, we put forward a one-sentence description outlining their scope when applied to immunisation programmes, which we used for this review (figure 3 and table 1).

Figure 3

Immunisation programme resilience framework. NITAG, National Immunisation Technical Advisory Group; SOPs, standard operating procedures.

Factors identified within the Integration category primarily focused on the need for cohesion within and between immunisation programmes and stakeholders. This included integration of community-based healthcare programmes, interdepartmental and interagency integration, and the importance of education, transparency, and trust between stakeholders. For Awareness, having well-developed data collection systems was considered core to successful monitoring of immunisation programmes, which included careful monitoring of media reporting, vaccine uptake, and response and recovery activities. Resource Availability and Access was also critical to ensure that the required capacity for maintaining stable performance can be met, as well as the ability to access these resources during times of crisis. Examples of critical resources included financing (eg, performance-based financing), having appropriate supply chain infrastructure, staffing, personal protective equipment and standard operating procedures and guidelines in place to deal with unexpected scenarios. The importance of having an adaptive system was also highlighted, such that programmes are updated based on new circumstances and lessons learnt. Primarily this was achieved through rapid local decision-making and accountability for misconduct. Finally, Self-regulation was considered important to immunisation programmes by ensuring a continuous approach to monitoring, evaluating, reviewing, investigating and resolving new issues as they arise.

The five resilience characteristics were specifically refined in the context of immunisation programmes. For example, communication is repeatedly identified as a crucial element in a resilient programme. The reviewed literature frequently highlighted that communication plays a fundamental role in education and outreach efforts in integrating immunisation programmes, with a particular focus on public communication strategies to address vaccine fears, hesitancy and mistrust. Based on these observations, we classified communication under the Integration characteristic. We limited the Awareness characteristic to information collection, which can be viewed as encompassing monitoring, surveillance, tracking, etc. Thus, the Awareness characteristic encompasses monitoring systems such as immunisation information systems (confidential, population-based, computerised databases that record all immunisation doses administered by participating providers to persons residing within a given geopolitical area)63 and adverse events following immunisation systems (any untoward medical occurrence which follows immunisation and which does not necessarily have a causal relationship with the usage of the vaccine).64

We distributed the aims of the Diverse characteristic of the Health System Resilience Index among (1) the Integration and (2) the Resource availability and Access characteristics. The Health System Resilience Index outlines two aims for the Diverse characteristic: (1) to effectively respond to a range of health needs and (2) to adequately finance health systems; prevent financial harm.14 The first aim was incorporated into the Integration characteristic of the Immunisation Programme Resilience Framework, while the second aim was incorporated into the novel Resource Availability and Access characteristic.

A frequently identified and discussed characteristic in the reviewed studies, which was not present in the Health System Resilience Index, was Resource Availability and Access (26/37 studies). Examples of topics included in the Resource Availability and Access characteristic are outlined in table 1 and figure 3, and include healthcare workers’ access to adequate supplies of PPE48 or the public’s access to medical facilities, which can be hindered due to movement or travel restrictions.46 48

The overall characteristics, proposed in figure 3 and table 1, comprehensively incorporate all factors identified in this scoping review as being essential features of a resilient immunisation programme. Of the five Immunisation Programme Resilience Framework characteristics, Integration (26/37 studies) and Resource Availability and Access (26/37 studies) were most often discussed, followed by Awareness (24/37 studies), Adaptiveness (21/37 studies) and Self-regulation (11/37 studies) (table 1).

Gaps in primary empirical research into immunisation programme resilience

We displayed evidence gaps using an online interactive evidence gap map: https://lukebax.github.io/resilience. This map displays studies by continent on the x-axis versus crisis on the y-axis, with experimental design indicated by colour coding. In each cell of the two-by-two grid, the size of each circle is proportional to the number of studies, and gaps in the research are visible as empty cells. Titles, abstracts and links to all publications are embedded in the map.

Evidence on environmental crises was identified across all continents. Social crises were identified in only Asia and Europe. Studies on legal crises originated from Europe only, and a study on political crises originated from Asia only. From this map, it is clear there are significant gaps in our evidence base for understanding the impact of crises on immunisation programmes across the globe.

Conclusion

Conducting this scoping review provided an opportunity to outline a framework that may be relevant to creating resilient immunisation programmes. Nevertheless, for this framework to be useful in practice, it needs to have input from multiple stakeholders, which include national governments, regional bodies, vaccine developers, healthcare professionals, civil society and individuals involved in immunisation programme implementation. By seeking expert advice from such groups, an over-reliance on top-down expert opinion can be avoided and evidence gaps filled. Developing this framework may allow the creation of a tool to inform evidence-based decision-making and secure appropriate funding to ensure immunisation programmes not only optimise functioning during times of crisis but identify weak points needing further enhancement to ensure long-term resilience.