Summary of included studies for a scoping review of the evidence regarding the effectiveness of donor supported leadership development programs intended to promote women's leadership in public health systems in LMICs

Methods:
The scoping review was based on Arksey and O’Malley’s methodological framework which comprises the following key steps: 1) identifying the research question, 2) finding relevant studies, 3) study selection, 4) data charting and 5) collating, summarising and reporting results (Arksey and O’Malley 2005).

Search strategy and selection criteria:
A systematic search for relevant literature was conducted on January 2020 using PubMed, SCOPUS, ProQuest and PsycINFO. A combination of key terms was used, which were grouped according to key concepts – health sector, women, leadership development and LMICs using truncation, proximity and Boolean operators ‘AND’ or ‘OR’. The search string was developed and refined in Scopus in consultation with a librarian and adapted for the other databases (refer to supplementary material for search string). We excluded non-English publications and restricted our search to papers published after 2000. The date range 2000–2020 was chosen as a point of reference as it encompasses a period of proliferation of LDPs intended to promote women’s leadership. Database searches were supplemented by hand searching of reference lists of identified articles. We also searched websites including Google and Google Scholar for non-peer-reviewed literature. Additional grey literature was identified through searching websites of key bilateral and multilateral development aid donors including DFAT, the United Kingdom Department for International Development, USAID and the Swedish International Development Cooperation Agency. These organisations were purposively sampled from among the Organisation for Economic Co-operation and Development top contributors of aid in support of gender equality and women’s empowerment both in relative and absolute terms. Title and abstract screening was conducted by a single reviewer (SM) with cross-validation of a random sample by a second reviewer (NS). The same approach was taken for full text screening of relevant articles. A third independent reviewer (EL) was available for consultation to resolve any conflicts which arose. Studies were included if they were evaluations of donor-supported leadership development interventions, based in public health sectors in LMICs, and included female participants in leadership and/or managerial roles. We excluded studies if they described evaluations of non-donor-funded LDPs, or donor-supported interventions implemented beyond the scope of public health sectors in LMICs. Theses, dissertations, book reviews, commentaries/opinion pieces, editorials and conference papers were also excluded. For the purpose of the review, we defined the public health sector as public entities at the national and sub-national levels which deliver health services including government ministries of health and health care institutions (i.e. hospitals and primary health care facilities) (Ayeleke et al. 2016). Women’s leadership refers to women in formal leadership and/or managerial roles across the health sector, and aid donors refer to agencies which finance development interventions in LMICs including bilateral and multilateral sources

Data extraction and analysis:
Data was extracted from eligible studies by two reviewers (SM, NS) using an agreed extraction template which addressed general study characteristics, and outcomes and impacts of interventions including evaluation challenges. Studies were not assessed for quality or risk of bias as the scoping review approach provides an overview of all relevant literature regardless of study design (Arksey and O’Malley 2005). However, limitations of the included studies were identified and discussed, and future directions for research articulated including the types of studies or evaluation designs which could be used to assess the effectiveness of women’s leadership interventions. A numerical summary of study characteristics was provided, and key outcomes and methodological challenges were summarised using a narrative synthesis. In addition, we analysed the gender composition of included evaluations based on participants demographic characteristics. We developed a theory of change for leadership development and gender transformative programmes to guide our data analysis and synthesis.

研究方法：
本范围综述基于Arksey与O’Malley提出的方法学框架，该框架包含以下核心步骤：1）明确研究问题，2）检索相关研究，3）研究筛选，4）数据制图，5）整理、汇总并报告研究结果（Arksey和O’Malley，2005）。

检索策略与筛选标准：
本研究于2020年1月针对PubMed、SCOPUS、ProQuest及PsycINFO数据库开展相关文献的系统性检索。检索时采用关键词组合策略，根据核心概念划分为四大类——卫生部门、女性、领导力发展以及低收入和中等收入国家（Low- and Middle-Income Countries，LMICs），并使用截词、邻近检索以及布尔逻辑运算符‘AND’和‘OR’进行组配。检索式由研究人员与图书馆员协作在Scopus数据库中拟定并优化，随后适配至其余数据库（检索式详见补充材料）。本研究排除非英文文献，并将检索范围限定为2000年之后发表的论文。选取2000-2020年作为检索时限，是因为该时段涵盖了旨在推动女性领导力发展的领导力发展项目（Leadership Development Programs，LDPs）大量涌现的时期。数据库检索之外，本研究还通过手工检索已纳入文献的参考文献列表进行补充检索；同时针对Google及Google Scholar等网站开展检索，以获取非同行评议文献。此外，通过检索重点双边及多边发展援助捐赠方的官方网站，额外获取灰色文献，这些捐赠方包括澳大利亚外交贸易部（DFAT）、英国国际发展部、美国国际开发署（USAID）以及瑞典国际开发合作署。上述机构均从经济合作与发展组织（OECD）中，按相对及绝对援助规模排名前列的、支持性别平等与女性赋权的捐助机构中，通过目的性抽样选取。由一名评审员（SM）开展标题与摘要筛选，并由第二名评审员（NS）对随机抽取的样本进行交叉验证。相关文献的全文筛选采用相同流程。若出现评审分歧，可由第三名独立评审员（EL）进行协商裁决。纳入标准如下：研究需针对低收入和中等收入国家（LMICs）公共卫生领域内、由捐赠方资助的领导力发展干预措施开展评估，且研究对象包含担任领导力及/或管理职务的女性参与者。排除标准如下：研究若针对非捐赠方资助的领导力发展项目（LDPs）开展评估，或针对低收入和中等收入国家（LMICs）公共卫生领域以外的捐赠方资助干预措施开展评估，则予以排除；同时排除学位论文、书籍评论、评论/观点类文章、社论及会议论文。为本综述之目的，本研究将公共卫生部门定义为：国家级及亚国家级层面提供卫生服务的公共机构，包括政府卫生部及各类医疗卫生机构（如医院与基层卫生保健设施）（Ayeleke等，2016）。本研究中，女性领导力指在卫生领域内担任正式领导力及/或管理职务的女性；援助捐赠方指为低收入和中等收入国家（LMICs）的发展干预措施提供资金支持的机构，包括双边及多边来源机构。

数据提取与分析：
两名评审员（SM与NS）采用统一拟定的数据提取模板，从符合纳入标准的研究中提取数据，模板涵盖研究基本特征、干预措施的效果与影响，以及评估过程中面临的挑战等内容。由于范围综述的方法旨在全面概述所有相关文献，无需考虑研究设计类型，因此本研究未对纳入研究的质量或偏倚风险进行评估（Arksey和O’Malley，2005）。不过，本研究对纳入研究的局限性进行了识别与讨论，并阐明了未来研究方向，包括可用于评估女性领导力干预措施有效性的研究类型与评估设计方案。本研究对研究基本特征进行了量化汇总，并采用叙述性综合法对核心研究结果与方法学挑战进行总结。此外，本研究基于研究对象的人口统计学特征，对纳入评估研究的性别构成进行了分析。同时，本研究构建了领导力发展与性别变革型项目的变革理论（Theory of Change），以指导数据分析与综合过程。