Background
Policymakers and implementers are often challenged with making decisions regarding the implementation of complex interventions to ensure high coverage and uptake [
1‐
3]. Complex interventions, by definition, have numerous interrelated elements that impact both implementer processes and recipient responses, including their intervention components, implementation strategy, and contextual features [
4,
5]. Therefore, there is a critical need for approaches that evaluate the relationships between these elements and determine their influence on effective intervention delivery in order to identify under what circumstances complex interventions are successful [
1,
6,
7]. Outcomes from these assessments can support evidence-based decision-making for policymakers and implementers, especially those working within health systems in low- and middle-income countries (LMICs) who often face issues of limited resources and capacity [
8‐
11].
One such complex intervention is mass drug administration (MDA) for neglected tropical diseases (NTDs) – a group of parasitic, viral, and bacterial diseases that affect billions of individuals, with disproportional prevalence in LMICs across sub-Saharan Africa, South Asia, and Latin America [
12,
13]. Untreated NTD infections negatively impact health, learning and productivity outcomes, diminishing quality-of-life and reinforcing cycles of poverty amongst the world’s most disadvantaged populations [
12,
13]. Worldwide, the most prevalent NTDs are soil-transmitted helminths (STHs), a group of parasitic intestinal worms that infect approximately 1.5 billion people globally – of these, an estimated 900 million are pre-school and school-age children [
13‐
15]. Current World Health Organization (WHO) guidelines recommend geographic areas that meet pre-defined STH prevalence thresholds to implement MDA, with a particular focus on deworming pre-school and school-age children in schools. In these programs, all at-risk children receive anti-helminthics, such as albendazole, regardless of their infection status [
12,
13]. This school-based MDA approach is the standard-of-care for STH control across numerous LMICs, reaching hundreds of millions of children annually [
13,
14,
16]. However, this approach does not target adults who serve as infection reservoirs in communities, thus contributing to rapid pediatric reinfection [
12,
16]. A community-wide MDA approach that targets community members of all ages shows promise of interrupting the transmission of STH [
17‐
19]. However, even with this intensified treatment strategy, MDA must be delivered with high coverage, with at least 80–90% of the targeted population dewormed, to achieve transmission interruption [
17‐
21].
Significant evidence suggests best practices for delivery of MDA with high treatment coverage include: a reliable drug supply chain to ensure the adequate allocation and distribution of anthelmintic drugs; a robust training cascade to build implementer capacity; a far-reaching community sensitization strategy to inform and mobilize recipients about MDA; and a well-executed, yet rapid distribution strategy [
15,
22‐
29]. Thus, effective MDA delivery requires a significant investment of material, financial, and human resources as well as adaptable implementation strategies that are feasible, appropriate, and acceptable for heterogeneous implementation settings. However, current MDA operational guidelines issued by the WHO and national Ministries of Health in NTD-endemic countries lack context-specific recommendations that take into account variation in key implementation factors, such as disease epidemiological profiles, community preferences, and health system capacities [
16,
23,
24,
28,
30]. Furthermore, guidelines do not distinguish which implementation activities are most essential for achieving MDA delivery with high coverage. This information is necessary for implementers who are planning MDA at scale in resource constrained environments, and who may not be able to incorporate all best practices into an implementation plan.
Using coincidence analysis (CNA), a cross-case analytical method, we systematically identify the various configurations of intervention delivery activities – known as implementation pathways – that result in high coverage of community-wide MDA for STH. This analysis aimed to characterize the “core components” of MDA delivery – activities that are necessary for achieving high coverage and need to be implemented with fidelity [
31]. Such evidence may help policymakers define the required resources for implementing MDA with high coverage as well as shape implementer decisions regarding implementation that balances fidelity with flexibility.
Discussion
This analysis indicates that, while insufficient or necessary by itself, efficient duration of MDA delivery (within 10 days) in combination with at least one other influential implementation activity – including a conducive implementation context, early arrival of albendazole before the start of MDA, or a flexible community sensitization strategy – consistently led to high coverage of community-wide MDA for STH.
Thus, our results suggest that efficient MDA delivery duration was one of the most influential implementation activities for producing high treatment coverage and MDA delivery within 10 days appears to be an optimal delivery timeframe for community-wide MDA for STH. Current evidence shows that the number of scheduled MDA days plays a significant role in the ability of implementers to meet necessary MDA targets, with overly brief durations negatively impacting treatment coverage [
23,
51‐
53]. However, these studies do not determine a definitive timeframe for effective delivery. To our knowledge, this is the first study to assess the effectiveness of a specific duration of MDA delivery.
Our findings mark an important addition to the NTD evidence base, as MDA duration has several implications for policy makers, NTD program managers, and implementers. Primarily, decision-making regarding program duration affects financial and material resources allocated for implementation. Thus, having clarity around the adequacy of scheduled MDA duration may support implementers to more effectively plan for MDA and maintain cost-effectiveness of the intervention, which is a key policy consideration for transitioning to community-wide MDA for STH [
54,
55]. Specific to planning, it is critical that implementers allocate a sufficient number CDDs for the target population size for the given campaign delivery schedule [
23,
56]. Notably, there may be an important relationship between campaign duration and size of the available workforce; a larger number of CDDs may be necessary to deliver MDA over a shorter duration of time as compared to the number of CDDs needed to deliver MDA to the same target population over a longer period of time. While these results do not directly call for implementers to deliver MDA faster than necessary, they do provide additional consideration for the potential diminishing returns if the number of days are over-extended, given the additional financial and opportunity costs of community-wide MDA as compared to school-based MDA [
54,
57]. These findings also provide novel evidence for an intervention component that is not currently highlighted in NTD operational manuals. A potential area of future CNA research could further examine what specific factors distinguish areas that deliver MDA more efficiently to further strengthen these findings.
Although were ultimately unable to single out a single model to explain high MDA coverage due to model ambiguity, the pathways within the candidate models demonstrate how high MDA coverage could potentially be accomplished in various ways, reflecting the utility of flexible implementation strategies, especially for sensitization strategies. Our results indicate that in addition to efficient MDA duration, three components – community sensitization, drug supply chain, and implementation context – consistently appeared across the models, suggesting their influence in achieving high MDA coverage. These three components are heavily outlined in the existing evidence base as key factors that influence the delivery of and demand for community-wide MDA for various NTDs, as summarized below.
It is critical that community members are aware of upcoming MDA campaigns and have trust in their efficacy and safety. Sensitization helps build awareness, demand, and trust amongst recipient community members and also builds buy-in from key political, civic, and community stakeholders [
24,
25]. Numerous studies have demonstrated the importance of well-designed, multifaceted community awareness strategies on MDA coverage [
25,
27,
51,
56,
58‐
60]. Our findings further demonstrate the strong influence of various sensitization approaches and indicate the importance of flexibility when designing sensitization strategies – in some areas, distribution of written materials may be more effective, while in other areas door-to-door sensitization may be more appropriate.
Drug supply chain
Another critical component of STH MDA delivery is having sufficient amounts of deworming drugs. Shortage of drugs is a potential consequence of late drug arrival, as the delay forces implementers to rely on stocks of deworming drugs in local health clinics to initiate MDA, which are likely in inadequate amounts to reach all targeted populations. Thus, timely drug arrival is an indication of adequate planning as well as a functional supply chain, which are both predictors of MDA coverage [
61‐
63]. Therefore, late drug arrival often serves as a barrier to achieving high MDA coverage [
61‐
63]. Our results further illustrate how timely drug arrival serves as a facilitator for effective MDA delivery. These findings define the importance of effective planning to mitigate supply chain issues and the need for robust supply chain management to help ensure timely and adequate receipt of necessary drugs.
Implementation context
The implementation context plays a significant role in intervention delivery [
2,
4,
5]. There are a number of contextual challenges that may impede MDA implementation and result in fewer individuals being dewormed. In our study, clusters identified several contextual issues that affected MDA delivery, including other ongoing community health programs, heavy rainy seasons, local cultural festivals and religious events. Our analysis suggests that clusters without these contextual challenges positively influenced MDA coverage. Existing evidence highlights the importance of scheduling MDA around other community events that may negatively affect treatment coverage, or reduce the availability of implementers to deliver MDA, including ongoing public health priorities and programs, important religious or cultural events, community activities, or weather periods [
23,
56]. Our findings emphasize the importance of careful planning and cross-sectorial collaboration and communication that in turn, could increase the likelihood of an optimal implementation context.
Overall, our findings illustrate important findings for inclusion in future operational guidelines for delivering community-wide MDA for STH and other NTDs. Policymakers and national-level NTD program managers can utilize these findings when developing MDA budgets and implementation strategies to ensure there are sufficient financial, human, and material resources. Additionally, implementers can use these findings to ensure they sufficiently invest in intervention planning before intervention delivery.
Our analysis has several strengths. Primarily, we included a large amount of implementation data covering six intervention delivery rounds and three distinct geographic settings. These data were pulled from a multi-country study that followed a comprehensive data collection process. We also applied a rigorous analytical process involving the use of CNA within implementation science [
35]. However, these results have several limitations. Primary is our use of data from an ongoing hybrid trial; therefore, our findings may not be completely generalizable to MDA programs implemented under routine practice. Other limitations of generalizability are specific to the CNA methodology. Specifically, our results are directly affected by our calibration choices and might have differed with other calibration thresholds. Additionally, although we were able to identify key operational variables that influence MDA coverage, there were other contextual factors that could have impacted intervention delivery that were not included as part of our analysis, including: degree of community trust or acceptability towards the intervention, implementer satisfaction and motivation (e.g. with workload or incentives), and level of community migration over time [
24]. The candidate models we reported each had coverage scores around 0.50, indicating a role for other factors and additional pathways to successful intervention delivery. The influence of these factors will be assessed in other planned DeWorm3 coverage analyses. Despite these limitations, we further demonstrated the utility of CNA in modeling complex causality for complex intervention delivery. Thus, our results have substantive implications for the future implementation of community-wide MDA to interrupt the transmission of STH across various low-resource settings.
It is also important to note the impact of the COVID-19 pandemic on study implementation activities. Sites were under country-wide lockdown during a significant portion of the last year of planned MDA delivery (i.e., MDA rounds 5 and 6), and all MDA planning and delivery activities were suspended until as late as July 2020. When field activities resumed, sites captured the impact of the pandemic on MDA implementation on the routine process mapping worksheets, e.g., noting how the pandemic negatively impacted the implementation context in some clusters. Thus, this analysis does capture some of these pandemic-specific challenges. However, we acknowledge the need for additional quantitative and qualitative research that directly assesses the impact of the COVID-19 pandemic on MDA planning and delivery and subsequent transmission interruption progress [
64]. Currently, there are several ongoing DeWorm3 evaluations assessing these impacts and understanding how the study’s existing community-wide infrastructure could be utilized to facilitate rapid community responses to COVID-19.
Conclusion
Using an innovative analytic approach, we identified that efficient duration of MDA delivery within 10 days was a highly influential implementation activity for achieving high coverage of community-wide MDA when co-implemented with other key implementation factors such as a conducive implementation context, early arrival of albendazole, and a flexible community sensitization strategy. These findings can be used by STH-endemic countries implementing MDA programs to develop appropriate operational guidelines and support effective implementation planning. Similar methodological approaches may be extended to evaluate other community-based primary care programs implemented in LMIC health systems.
Acknowledgements
The authors would like to thank all site implementation science staff for their support in instrument adaptation, completing the data collection, including Surya Kulasekaran, Yesudoss Jacob, and Rajeshkumar Rajendiran of Christian Medical College, Vellore; Abiguel Alijan, Alvine Sewade and Aubierge Kpatinvo of the Institut de Recherche Clinique du Bénin. We thank all of the field and data managers, field supervisors and field workers of the DeWorm3 study for logistic support during data collection activities. Lastly, we thank all of the community members in each of the sites for their participation in the study.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.