You are here

  • Home
  • Archive
  • Volume 13, Issue 10
  • Prognostic prediction models for clinical outcomes in patients diagnosed with visceral leishmaniasis: protocol for a systematic review

Article Text

Article menu
Infectious diseases
Protocol
Prognostic prediction models for clinical outcomes in patients diagnosed with visceral leishmaniasis: protocol for a systematic review
  1. James Wilson1,2,
  2. Forhad Chowdhury1,2,
  3. Shermarke Hassan1,2,
  4. Elinor K Harriss3,
  5. Fabiana Alves4,
  6. Prabin Dahal1,2,
  7. Kasia Stepniewska1,2,
  8. Philippe J Guérin1,2
  1. 1Infectious Diseases Data Observatory (IDDO), University of Oxford, Oxford, UK
  2. 2Centre for Tropical Medicine and Global Health, University of Oxford, Oxford, UK
  3. 3Bodleian Health Care Libraries, University of Oxford, Oxford, UK
  4. 4Drugs for Neglected Disease Initiative, Geneva, Switzerland
  1. Correspondence to Dr James Wilson; james.wilson{at}iddo.org

Abstract

Introduction Visceral leishmaniasis (VL) is a neglected tropical disease responsible for many thousands of preventable deaths each year. Symptomatic patients often struggle to access effective treatment, without which death is the norm. Risk prediction tools support clinical teams and policymakers in identifying high-risk patients who could benefit from more intensive management pathways. Investigators interested in using their clinical data for prognostic research should first identify currently available models that are candidates for validation and possible updating. Addressing these needs, we aim to identify, summarise and appraise the available models predicting clinical outcomes in VL patients.

Methods and analysis We will include studies that have developed, validated or updated prognostic models predicting future clinical outcomes in patients diagnosed with VL. Systematic reviews and meta-analyses that include eligible studies are also considered for review. Conference abstracts and educational theses are excluded. Data extraction, appraisal and reporting will follow current methodological guidelines. Ovid Embase; Ovid MEDLINE; the Web of Science Core Collection, SciELO and LILACS are searched from database inception to 1 March 2023 using terms developed for the identification of prediction models, and with no language restriction. Screening, data extraction and risk of bias assessment will be performed in duplicate with discordance resolved by a third independent reviewer. Risk of bias will be assessed using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). Tables and figures will compare and contrast key model information, including source data, participants, model development and performance measures, and risk of bias. We will consider the strengths, limitations and clinical applicability of the identified models.

Ethics and dissemination Ethics approval is not required for this review. The systematic review and all accompanying data will be submitted to an open-access journal. Findings will also be disseminated through the research group’s website (www.iddo.org/research-themes/visceral-leishmaniasis) and social media channels.

PROSPERO registration number CRD42023417226.

  • Prognosis
  • Neglected Diseases
  • INFECTIOUS DISEASES
  • Systematic Review
  • STATISTICS & RESEARCH METHODS
  • Protocols & guidelines
https://creativecommons.org/licenses/by/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

https://doi.org/10.1136/bmjopen-2023-075597

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Strengths and limitations of this study

    • We present a protocol for a robust and comprehensive systematic review of visceral leishmaniasis (VL) prognostic models, using current best-practice guidelines on data extraction, risk of bias assessment and reporting.

    • Inclusion criteria are designed to identify a broad range of VL prognostic model studies, including all patients with a VL diagnosis, and with no exclusions based on treatment setting, type of clinical outcome or prediction horizon.

    • We describe a comprehensive and evidence-based search strategy to identify a broad range of prognostic model studies across five large bibliographic databases, with no limitations on language or initial publication date.

    • Unpublished, non-peer-reviewed studies, such as conference abstracts and educational theses, are not included in the eligibility criteria.

    • A systematic assessment of the current use and impact of VL prognostic models is considered outside the scope of the planned review.

    Introduction

    Visceral leishmaniasis (VL), a parasitic infection transmitted between mammalian hosts via the bite of an infected sandfly, is a disease mostly prevalent in tropical regions that disproportionately touches vulnerable people affected by forced migration, malnutrition and poverty.1 The disease often presents insidiously with fever, splenomegaly and weight-loss, and is almost always fatal without effective treatment.2 The WHO estimates an incidence of 50 000–90 000 cases per year,3 resulting in approximately 400 000 disability-adjusted life years lost and over 5000 deaths.4 However, accurate estimates of disease burden are obfuscated by limited country-level reporting, evolving dense foci of infection in remote areas, and a paucity of active surveillance.1 Despite progress made over the last 20 years, successful treatment remains challenged by high drug costs, prolonged treatment courses requiring hospitalisation and frequent drug side effects.5 Patients with previous treatment failure or immunosuppressive comorbidities such as advanced HIV suffer from particularly high relapse and mortality rates.1 6

    To optimise individual patient care and effectively balance the distribution of constrained resources, identification of patients at high risk of treatment failure and subsequent death is crucial. Risk stratification is also important on a population level; elimination programmes in endemic areas can use risk prediction tools to strategically target patients prone to treatment failure and hence reduce the infectious reservoir driving onward transmission. Prognostic prediction models (referred henceforth as prognostic models) play a central role in VL risk stratification; informing healthcare providers, policymakers and patients on the treatment setting, treatment regimen and intensity of follow-up.7–9

    Systematic reviews of prognostic models have been published across a range of infectious diseases,10–12 serving not only to inform healthcare providers on available risk stratification tools, but also as a research tool to identify candidate models for external validation or updating (recalibration) with data from new settings. Indeed, the lack of external validation studies is considered the greatest barrier to the broader acceptance of prognostic models as a reliable and acceptable clinical tool.13–15

    Aim

    We will perform a systematic literature review to identify, summarise and appraise prognostic models in patients diagnosed with VL. Specifically, we focus on models that predict clinical outcomes such as treatment failure (initial failure or relapse) and death, developed subsequent to VL diagnosis.

    The review will serve two principal purposes:

    • Inform stakeholders, such as policymakers and healthcare workers directly involved in the treatment of VL patients, on the available prognostic models and their setting-specific clinical utility, strengths and limitations.

    • Inform researchers interested in using their own data for the development, validation or updating of VL prognostic models.

    Methods and analysis

    This systematic review will adhere to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA); a reporting guideline for systematic reviews,16 and Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis: checklist for Systematic Reviews and Meta-Analysis (TRIPOD-SRMA); a reporting guideline for systematic reviews of prediction models.17

    Important protocol amendments will be documented on PROSPERO. We use the PRISMA Protocols (PRISMA-P) checklist to guide reporting of the protocol (online supplemental material 1).18

    Study eligibility

    We follow a PICOTS (Population, Index model, Comparator model, Outcomes, Timing, Setting) approach to frame our review question and inclusion criteria (table 1).19 20

    View this table:
    Table 1

    PICOTS approach to frame the research question

    Our population of interest includes all patients with a confirmed or suspected diagnosis of VL as reported by the study authors. We include all published, peer-reviewed studies that develop, externally validate, update, or any combination thereof, a prognostic model with the intention of predicting individual clinical outcomes following VL diagnosis.

    In accordance with expert guidance on the methodology of prediction model research,14 21 22 we define a prognostic model as a multivariable model (including two or more predictors) where the intention is to predict outcomes at the individual patient level. Prognostic model studies are distinguished from predictor finding or prognostic factor studies, where the intent is to investigate the effect of a single or group of factors on an outcome of interest.23 We therefore exclude all studies that present models where the aim is not to predict risk at the individual patient level. We also exclude unpublished and non-peer-reviewed studies, including conference abstracts and educational theses, studies that only report diagnostic prediction models, and animal studies.

    To complement the systematic review, using the same search strategy we will identify (1) systematic reviews and meta-analyses of prognostic models, and (2) any impact studies that investigate the clinical outcomes of using vs not using an eligible prognostic model. These studies will not be subject to formal data extraction, but will be summarised in a narrative review.

    Search strategy

    An information specialist (EKH) created the search strategies to retrieve relevant records from the following databases: Ovid Embase; Ovid MEDLINE; the Web of Science Core Collection, SciELO and LILACS. The databases were searched from inception to 1 March 2023 with no language restriction. Where necessary, records published in languages not spoken by the authors will be translated using the Google Translate service (http://translate.google.com), or otherwise using a professional translation service. The search strategy used text words and relevant indexing terms to retrieve studies describing eligible prognostic models. The Ingui search filter was combined with an additional search string developed by Geersing et al,24 25 and adapted for Ovid Embase (table 2) and remaining bibliographic databases (online supplemental material 2).

    View this table:
    Table 2

    Search strategy initially developed in Ovid Embase and searched on 1 March 2023 (search queries were subsequently adapted to other bibliographic databases)

    Selection process and data extraction

    All references were exported to Covidence for deduplication and screening (Covidence systematic review software, Veritas Health Innovation, Melbourne, Australia. Available at www.covidence.org).26 Google Scholar will be used as a grey literature source once screening is complete to identify additional relevant studies. Citation searching of all included studies will be performed to identify further studies for screening.

    Studies identified from the search strategy are being independently screened by two reviewers (JW, FC). Preliminary screening is at title and abstract level followed by full-text screening. A third reviewer (PD), an experienced statistician, will make the final judgement on study inclusion if discordance remains following discussion between the two screening reviewers. A flow diagram will be presented as per the PRISMA 2020 checklist.16

    Study information will be extracted, collected and managed using the REDCap electronic data capture tools hosted at the University of Oxford.27 A data extraction form will be created based on CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies (CHARMS) for data extraction and Prediction model Risk Of Bias ASsessment Tool (PROBAST; https://www.probast.org/) for assessing risk of bias (table 3).20 22 A pilot form will be trialled prior to data extraction. Two independent reviewers (JW, SH), both with expertise in prediction modelling, will independently extract the study information. Where discordance remains after discussion, a final decision will be made by a third expert reviewer (PD). Study authors will not be contacted to obtain information not reported in the study.

    View this table:
    Table 3

    Information for data extraction and subsequent summary and appraisal

    Risk of bias

    Risk of bias will be summarised and reported separately for each model development and external validation using PROBAST.22 Two reviewers (JW, SH), will answer 20 signalling questions across four domains (participants, predictors, outcome and analysis), which will be used to judge the overall risk of bias as either ‘low’, ‘high’ or ‘unclear’. Discordance between the two reviewers will be resolved by a third reviewer (PD).

    If impact studies are identified, their risk of bias will be assessed using the Cochrane Risk of Bias tool for randomised comparative designs or the Risk Of Bias In Non-randomised Studies of Interventions tool for studies using a non-randomised comparative design.28 29

    Data synthesis

    We will present a narrative synthesis of our review findings. Key extracted information will be summarised in tabular form. Performance statistics, including measures of calibration and discrimination, where available, will be presented alongside their derivation method.

    Figures will be used to concisely communicate important information, including (1) summary of the candidate and final predictors included in each model and (2) risk of bias assessment across the four domains (participants, predictors, outcomes and analysis).

    Strengths and limitations of the identified models will be considered.

    Given our aim is not to compare multiple external validations of a single model, we will not be performing a meta-analysis.30

    Patient and public involvement

    None.

    Ethics and dissemination

    Ethics approval is not required for this review. The systematic review will be submitted to an open-access journal for peer-review and publication. Findings will also be disseminated through conference presentations, the research group’s website (www.iddo.org/research-themes/visceral-leishmaniasis) and social media channels. All extracted information will be made freely available as supplemental material submitted during publication.

    Discussion

    We present a protocol for the first systematic review of prognostic models for clinical outcomes in patients diagnosed with VL; a neglected tropical disease that affects some of the most disadvantaged communities in the world. Thoughtful risk stratification of patients using prognostic models can assist clinical decision-making and inform policy, guiding the optimal allocation of often-limited resources. By identifying, summarising and appraising the published VL prognostic models, we hope that the planned systematic review will serve as a comprehensive resource for VL stakeholders, including healthcare workers, policymakers and researchers.

    Clinical outcomes in VL are heterogeneous, with rates of initial treatment failure, relapse and mortality varying according to known and unknown factors. Many predictors of poor clinical outcomes have been identified, including extremes of patient age, severity of clinical signs and symptoms, laboratory investigations, the immune status of the patient (including the presence of advanced HIV), the patient’s clinical management, geographical location, parasite genotype and resistance profile.9 31 32 The relative contributions to patient outcomes of these inter-related factors can be described through multivariable modelling and subsequently used to estimate individual patient risk in the form of a prognostic model. However, such a model’s performance in a new population can only be directly assessed through external validation. Indeed, this is considered an essential step prior to model use, but infrequently performed in practice.21 The planned systematic review will concisely summarise key information presented across all identified prognostic model studies. VL healthcare providers and policymakers can then use this information, including performance estimates from external validations, to assess a model’s applicability to their own patient population.

    This review will not only serve healthcare providers and policymakers in identifying relevant risk stratification tools, but also provide a resource for research groups aiming to validate or update existing prognostic models. The Infectious Diseases Data Observatory is developing a data repository of individual participant data (IPD) from VL clinical trials and observational studies (www.iddo.org/research-themes/visceral-leishmaniasis).33 A VL data platform including IPD presents an exciting opportunity for the development, validation and updating of prognostic models.34

    An important strength of the planned review is its broad eligibility criteria; we include models describing all clinical outcomes and impose no restriction on model setting, publication language or prediction horizon (elapsed period between the intended time of model use and the outcome being predicted). Given concerns about study quality and accessibility, we will not be reviewing unpublished nor non-peer-previewed studies, such as conference abstracts or educational theses. A further limitation of the planned review is that we will not be contacting study authors to request unreported information, although we will explicitly report where information is missing. A systematic assessment of the current use and impact of VL prognostic models, including policy guidelines, is considered beyond the scope of the review, however, the review’s findings will be considered in the context of current practice as understood by the authors.

    In summary, we present a protocol for the systematic review of prognostic models of clinical outcomes for patients diagnosed with VL. With the aim of identifying, summarising and appraising the available risk models, we hope to provide a current reference to stakeholders engaged in VL patient care, policy and research.

    Ethics statements

    Patient consent for publication

    References

    1. ↵
      1. Burza S,
      2. Croft SL,
      3. Boelaert M
      . Leishmaniasis. Lancet 2018;392:951–70. doi:10.1016/S0140-6736(18)31204-2
      OpenUrlCrossRefPubMed
    2. ↵
      1. Guerin PJ,
      2. Olliaro P,
      3. Nosten F, et al
      . Visceral leishmaniasis: current status of control, diagnosis, and treatment, and a proposed research and development agenda. Lancet Infect Dis 2002;2:564–73. doi:10.1016/s1473-3099(02)00372-9
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. World Health Organization
      . Leishmaniasis. Available: https://www.who.int/news-room/fact-sheets/detail/leishmaniasis [Accessed 24 Aug 2023].
    4. ↵
      1. Vos T,
      2. Lim SS,
      3. Abbafati C, et al
      . Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet 2020;396:1204–22. doi:10.1016/S0140-6736(20)30925-9
      OpenUrlCrossRefPubMed
    5. ↵
      1. Alves F,
      2. Bilbe G,
      3. Blesson S, et al
      . Recent development of visceral leishmaniasis treatments: successes, pitfalls, and perspectives. Clin Microbiol Rev 2018;31:e00048-18. doi:10.1128/CMR.00048-18
    6. ↵
      1. Alvar J,
      2. Aparicio P,
      3. Aseffa A, et al
      . The relationship between leishmaniasis and AIDS: the second 10 years. Clin Microbiol Rev 2008;21:334–59, doi:10.1128/CMR.00061-07
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Ministry of Health (Brazil)
      . Guide to Health Surveillance. 5th Edition, revised and updated. (Guia de Vigilância em Saúde. 5a edição revisada e atualizada). Brasilia, Brazil, Available: https://bvsms.saude.gov.br/bvs/publicacoes/guia_vigilancia_saude_5ed_rev_atual.pdf [accessed 24 Aug 2023].
    8. ↵
      1. Naylor-Leyland G,
      2. Collin SM,
      3. Gatluak F, et al
      . The increasing incidence of visceral leishmaniasis relapse in South Sudan: a retrospective analysis of field patient data from 2001–2018. PLoS Negl Trop Dis 2022;16:e0010696. doi:10.1371/journal.pntd.0010696
    9. ↵
      1. Belo VS,
      2. Struchiner CJ,
      3. Barbosa DS, et al
      . Risk factors for adverse prognosis and death in American visceral leishmaniasis: a meta-analysis. PLoS Negl Trop Dis 2014;8:e2982. doi:10.1371/journal.pntd.0002982
    10. ↵
      1. Wynants L,
      2. Van Calster B,
      3. Collins GS, et al
      . Prediction models for diagnosis and prognosis of COVID-19: systematic review and critical appraisal. BMJ 2020;369:m1328. doi:10.1136/bmj.m1328
    11. ↵
      1. Peetluk LS,
      2. Ridolfi FM,
      3. Rebeiro PF, et al
      . Systematic review of prediction models for pulmonary tuberculosis treatment outcomes in adults. BMJ Open 2021;11:e044687. doi:10.1136/bmjopen-2020-044687
    12. ↵
      1. Njim T,
      2. Tanyitiku BS
      . Prognostic models for the clinical management of malaria and its complications: a systematic review. BMJ Open 2019;9:e030793. doi:10.1136/bmjopen-2019-030793
    13. ↵
      1. Altman DG,
      2. Vergouwe Y,
      3. Royston P, et al
      . Prognosis and prognostic research: validating a prognostic model. BMJ 2009;338:bmj.b605. doi:10.1136/bmj.b605
    14. ↵
      1. Steyerberg EW,
      2. Moons KGM,
      3. van der Windt DA, et al
      . Prognosis research strategy (PROGRESS) 3: prognostic model research. PLoS Med 2013;10:e1001381. doi:10.1371/journal.pmed.1001381
    15. ↵
      1. Wyatt JC,
      2. Altman DG
      . Commentary: prognostic models: clinically useful or quickly forgotten BMJ 1995;311:1539–41. doi:10.1136/bmj.311.7019.1539
      OpenUrlFREE Full Text
    16. ↵
      1. Page MJ,
      2. McKenzie JE,
      3. Bossuyt PM, et al
      . The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. doi:10.1136/bmj.n71
    17. ↵
      1. Snell KIE,
      2. Levis B,
      3. Damen JAA, et al
      . Transparent reporting of multivariable prediction models for individual prognosis or diagnosis: checklist for systematic reviews and meta-analyses (TRIPOD-SRMA). BMJ 2023;381:e073538. doi:10.1136/bmj-2022-073538
    18. ↵
      1. PRISMA-P Group,
      2. Moher D,
      3. Shamseer L, et al
      . Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015;4:1. doi:10.1186/2046-4053-4-1
      OpenUrlCrossRefPubMed
    19. ↵
      1. Debray TPA,
      2. Damen JAAG,
      3. Snell KIE, et al
      . A guide to systematic review and meta-analysis of prediction model performance. BMJ 2017;356:i6460. doi:10.1136/bmj.i6460
    20. ↵
      1. Moons KGM,
      2. de Groot JAH,
      3. Bouwmeester W, et al
      . Critical appraisal and data extraction for systematic reviews of prediction modelling studies: the CHARMS checklist. PLoS Med 2014;11:e1001744. doi:10.1371/journal.pmed.1001744
    21. ↵
      1. Collins GS,
      2. Reitsma JB,
      3. Altman DG, et al
      . Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. Eur Urol 2015;67:1142–51. doi:10.1016/j.eururo.2014.11.025
      OpenUrlCrossRefPubMed
    22. ↵
      1. Moons KGM,
      2. Wolff RF,
      3. Riley RD, et al
      . PROBAST: a tool to assess risk of bias and applicability of prediction model studies: explanation and elaboration. Ann Intern Med 2019;170:W1–33. doi:10.7326/M18-1377
      OpenUrlCrossRefPubMed
    23. ↵
      1. Riley RD,
      2. Moons KGM,
      3. Snell KIE, et al
      . A guide to systematic review and meta-analysis of prognostic factor studies. BMJ 2019;364:k4597. doi:10.1136/bmj.k4597
    24. ↵
      1. Ingui BJ,
      2. Rogers MAM
      . Searching for clinical prediction rules in MEDLINE. J Am Med Inform Assoc 2001;8:391–7. doi:10.1136/jamia.2001.0080391
      OpenUrlCrossRefPubMed
    25. ↵
      1. Geersing G-J,
      2. Bouwmeester W,
      3. Zuithoff P, et al
      . Search filters for finding prognostic and diagnostic prediction studies in MEDLINE to enhance systematic reviews. PLoS ONE 2012;7:e32844. doi:10.1371/journal.pone.0032844
    26. ↵
      Covidence systematic review software, Veritas health innovation. Melbourne, Australia, Available: www.covidence.org
    27. ↵
      1. Harris PA,
      2. Taylor R,
      3. Minor BL, et al
      . The REDCap consortium: building an international community of software platform partners. J Biomed Inform 2019;95:103208. doi:10.1016/j.jbi.2019.103208
      OpenUrlCrossRefPubMed
    28. ↵
      1. Sterne JA,
      2. Hernán MA,
      3. Reeves BC, et al
      . ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ 2016;355:i4919. doi:10.1136/bmj.i4919
    29. ↵
      1. Sterne JAC,
      2. Savović J,
      3. Page MJ, et al
      . RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ 2019;366:l4898. doi:10.1136/bmj.l4898
    30. ↵
      1. Damen JAA,
      2. Moons KGM,
      3. van Smeden M, et al
      . How to conduct a systematic review and meta-analysis of prognostic model studies. Clin Microbiol Infect 2023;29:434–40. doi:10.1016/j.cmi.2022.07.019
      OpenUrl
    31. ↵
      1. Grace CA,
      2. Sousa Carvalho KS,
      3. Sousa Lima MI, et al
      . Parasite genotype is a major predictor of mortality from visceral leishmaniasis. mBio 2022;13:e0206822. doi:10.1128/mbio.02068-22
    32. ↵
      1. Collin S,
      2. Davidson R,
      3. Ritmeijer K, et al
      . Conflict and Kala‐Azar: determinants of adverse outcomes of Kala‐Azar among patients in Southern Sudan. Clin Infect Dis 2004;38:612–9. doi:10.1086/381203
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Bush JT,
      2. Wasunna M,
      3. Alves F, et al
      . Systematic review of clinical trials assessing the therapeutic efficacy of visceral leishmaniasis treatments: a first step to assess the feasibility of establishing an individual patient data sharing platform. PLoS Negl Trop Dis 2017;11:e0005781. doi:10.1371/journal.pntd.0005781
    34. ↵
      1. Debray TPA,
      2. Riley RD,
      3. Rovers MM, et al
      . Individual participant data (IPD) meta-analyses of diagnostic and prognostic modeling studies: guidance on their use. PLOS Med 2015;12:e1001886. doi:10.1371/journal.pmed.1001886

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Twitter @jamesemaj

    • Contributors The study concept and design were conceived by JW, PJG, KS and PD. EKH developed the search strategy. JW, FC and PD will complete the literature screening. JW, SH and PD will perform data extraction. KS and PD have provided statistical support. FA provided disease-specific expert advice. The first draft of this manuscript was prepared by JW, who will also draft the completed systematic review. All authors critically reviewed and approved the submitted version.

    • Funding JW is undertaking a PhD supported in part by the University of Oxford through the Clarendon Scholarship and Clinician-Investigator Award, and the Infectious Diseases Data Observatory, which is itself supported by a diverse portfolio of grants, including from the Bill & Melinda Gates Foundation (recipient: Professor Philippe Guerin; ref: INV-004713). The funders have no role in the conception, design or conduct of the systematic review.

    • Competing interests None declared.

    • 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.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.