A randomized controlled trial of the effectiveness of a community-based rabies vaccination strategy

  1. Felix Lankester
  2. Ahmed Lugelo
  3. Joel Changalucha
  4. Danni Anderson
  5. Christian Tetteh Duamor
  6. Anna Czupryna
  7. Kennedy Lushasi
  8. Elaine Ferguson
  9. Emmanuel S. Swai
  10. Hezron Nonga
  11. Maganga Sambo
  12. Sarah Cleaveland
  13. Sally Wyke
  14. Paul C. D. Johnson
  15. Katie Hampson

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Background

Approximately 60,000 people die from dog-mediated rabies annually. Low and heterogeneous coverage reduces the effectiveness of dog vaccination campaigns that can eliminate rabies. Campaigns typically involve teams travelling annually to villages to deliver cold chain stored vaccines from centralized headquarters. Thermotolerant vaccines enable novel decentralized delivery of locally-stored vaccines by communities throughout the year. We compared the effectiveness of annual team-based versus continuous community-based dog vaccination strategies.

Methods

We conducted a cluster randomized controlled trial across Mara region, Tanzania. Trial clusters were administrative wards (112, on average four villages each). For the team-based arm vaccinators hosted annual static-point clinics, whilst for the community-based arm, a ward-based animal health officer with a village community leader managed vaccinations using vaccines stored within the ward. We measured vaccination coverage, the primary outcome, twice annually per cluster (month 1 and 11) through household surveys over three years (November 2020 to October 2023) and examined spatial and temporal coverage variations as secondary outcomes.

Findings

Community-based delivery achieved significantly higher coverage (49-62%) than team-based delivery (22-46%), and consistently exceeded the critical threshold for herd immunity (40%), Odds ratio (OR): 1.48-3.49. The lower less uniform coverage achieved through team-based delivery had a higher monthly probability of falling below the critical threshold (0.6, 95% CI: 0.38-0.81) vs 0.18 (95% CI: 0.04-0.40). Greater declines in coverage over the year were recorded in the team-based arm compared to the community-based

Conclusion

Community-based mass dog vaccination achieves higher more consistent coverage than team-based delivery across settings typical of many sub-Saharan African countries. This approach could play an important role in national rabies elimination programmes aiming to end human rabies deaths by 2030 as part of the global ‘zero by 30’ strategy.

Funding

Department of Health and Human Services of the National Institutes of Health (R01AI141712), Wellcome Trust (207569/Z/17/Z, 224520/Z/21/Z) and MSD Animal Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Article activity feed

  1. PREreview

    This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/15535406.

    CONSORT Peer Review Report 

    Title of the preprint:

    A randomized controlled trial of the effectiveness of a community-based vaccination strategy

    Preprint platform:

    BioRxiv

    Review type:

    Peer review report based on the CONSORT 2010 extension for cluster randomized trials

    Status:

    ✓ Approved with minor revisions

    1.    Summary of the Study

    This study presents the results of a cluster randomized controlled trial evaluating a community-based rabies vaccination strategy in rural Tanzania. Two delivery approaches were compared across 448 villages over a three-year period: standard centralized team-based and decentralized community-based delivery. Using a thermotolerant rabies vaccine and local coordination, the community-based model aimed to achieve higher and more consistent vaccination coverage. Coverage data were collected through biannual household surveys, and statistical analyses assessed temporal stability, spatial variation, and adherence to critical herd immunity thresholds. The findings suggest that the community-based delivery achieved higher and more consistent coverage and may be a valuable strategy for global rabies elimination efforts [1].

    2.    General Comments

    This manuscript reports a well-designed cluster randomized trial comparing community-based and team-based rabies vaccination strategies in rural Tanzania. The study addresses a relevant challenge to global rabies elimination efforts, with a well-organized research design and methodology. Reporting aligns well with CONSORT 2010 cluster guidelines [2].

    While the overall design and execution are strong, improvements could be made to enhance clarity and presentation. Certain technical terms and methods would benefit from clearer definitions early in the text, and several figures could be better contextualized to support interpretation. Minor issues with language consistency, grammar, and formatting are also present.

    Further detailed comments on reporting and design are provided in the section-by-section review below.

    3. CONSORT-Specific Comments

    a.     Title and Abstract

    • CONSORT 1a: The title refers to the study as a 'community-based' trial, which implies that it is a cluster trial. However, the cluster trial design could be stated more clearly in the title in order to improve clarity and ensure that the trial is easily found through electronic search databases.

    b. Introduction

    • CONSORT 2a: The scientific background and rationale for the overall study are clear, but the rationale for the cluster design itself is not. This could be clarified by briefly describing why a traditional individual-level randomization design is not feasible or adequate in this case.

    • CONSORT 2b: A general objective of evaluating the effectiveness of the two strategies is given, but a clear and specific hypothesis pertaining to the individual, village, and ward levels of each cluster would clarify the study's aims.

    • General: The geographical focus of the study is clear. However, the manuscript alternates between using national and regional when referring to rabies elimination in Tanzania. Consistent terminology here would eliminate confusion.

    c. Methods

    • CONSORT 3a: The study design—a stratified parallel cluster randomized trial—and the allocation ratio (1:1) are clearly stated. The application of the design features at the cluster level is defined, but is scattered across the section. This could be improved by clearly outlining this in the introduction of the section before going into further detail later in the section.

    • CONSORT 6a: Outcome measurements are clearly described and contamination avoidance efforts are cleverly described through the 'fried-egg' metaphor. A simple schematic could complement this metaphor well.

    • CONSORT 8ab: The randomization allocation method is clearly stated as being a custom R script. This method could be clarified further by briefly describing the rationale and methods used to create the script.

    • CONSORT 9: The manuscript explains how clusters were randomized using a custom R script but does not describe whether allocation was concealed before assignment. A brief description of the allocation concealment mechanism would strengthen transparency and ethical reporting.

    • CONSORT 10a: It is noted that randomization was implemented using a custom R script written by the trial statistician. However, it is not specified who enrolled the clusters or who assigned them to interventions. Clarifying these roles would improve reporting transparency.

    • CONSORT 10c: Community engagement and consent for household surveys are mentioned, but it is unclear whether consent was obtained from community representatives, individual dog owners, or both. Stating from who and when consent was obtained would strengthen transparency and ethical reporting.

    d. Results

    • CONSORT 13ab: While the text and tables report the number of clusters and participants analyzed, a participant flow diagram is not included. Adding one would help clarify the progression of clusters and dogs through the trial stages.

    • CONSORT 14a: The recruitment and follow-up dates (November 2020 to October 2023) are stated in the Methods section, but not in the Results. Including this information within the Results section as well would make the trial timeline easier to follow.

    • CONSORT 14b: It appears that the trial ended as planned, but a brief note on the ending would clarify the trial timeline.

    • CONSORT 17a: Primary and secondary outcomes are presented with effect sizes and 95% confidence intervals at multiple time points, and clustering is appropriately accounted for using a GLMM. While variance components are reported in the sample size section, an explicit intracluster correlation coefficient (ICC) for the primary outcome is not provided. Including this would support the study design.

    • General: The large number of tables and figures placed back to back makes the section visually dense. Therefore, reorganizing them in a logical order with text in between would improve readability.

    e. Discussion

    • CONSORT 20: While potential sources of bias, such as variation in survey response accuracy and implementation consistency, are noted in earlier sections, further elaboration in the discussion section would help contextualize the study's findings.

    • General: The use of subheadings throughout the section is helpful, but the final section would benefit from an additional subheading to separate the content on vaccine thermotolerance from the final concluding remarks. Further, the manuscript states that fewer than 11% of dog bites involve a member of the same household, based on unpublished data. Clarifying the source or context of this figure would improve transparency and help support the rationale for community-level interventions.

    f. Other Information

    • CONSORT 23/24: The manuscript does not report a trial registration number or indicate whether a full trial protocol is publicly available. Including this information, if available, would improve transparency.

    4. Final Recommendations

    This manuscript presents a well-designed and relevant cluster randomized trial that offers strong evidence for improving rabies vaccination coverage in rural settings. Both the community-based and team-based strategies are clearly described, with the community-based approach demonstrating greater effectiveness. In addition to the specific CONSORT-based comments provided, several broader points related to presentation, clarity, and terminology should also be addressed to strengthen the manuscript to a publishing standard.

    The manuscript would benefit from improvements in clarity and presentation of visualizations. Tables and figures should be more clearly integrated into the text, with brief in-line summaries and logical placement to guide the reader's interpretation. A comparative table outlining the key features of the two delivery strategies—community-based and team-based—would also enhance clarity, as this information is currently dispersed across the manuscript.

    Furthermore, clearer definitions of technical terms, such as pulsed centralized team-based delivery strategy, decentralized community-based strategy, and cold-chain conditions, could be briefly defined at first mention to improve accessibility and clarity.

    Finally, to further improve accessibility and clarity, English-variant inconsistencies, minor grammar errors, and formatting issues should be addressed.

    Note: This peer review report was prepared as part of a structured peer review writing sample developed during training in peer review strategy. Artificial intelligence tools were used to support the screening and interpretation of complex content, as the reviewer is not a subject matter expert in rabies epidemiology.

    References:

    1.     Lankester F, Lugelo A, Changalucha J, et al. A randomized controlled trial of the effectiveness of a community-based rabies vaccination strategy. bioRxiv. Published October 31, 2024. doi:10.1101/2024.10.28.620430

    2.     Campbell MK, Piaggio G, Elbourne DR, Altman DG; CONSORT Group. CONSORT 2010 statement: extension to cluster randomised trials. BMJ. 2012;345:e5661. doi:10.1136/bmj.e5661

    Competing interests

    The author declares that they have no competing interests.

    Use of Artificial Intelligence (AI)

    The author declares that they used generative AI to come up with new ideas for their review.

  2. Version published to 10.1101/2024.10.28.620430v1 on bioRxiv