Background: About 45.1% of all induced abortions are unsafe and 97% of these occur in low- and middle-income countries (LMICs). Women's abortion decisions may be complex and are influenced by various factors. We aimed to delineate women's abortion decision-making trajectories and their determinants in LMICs. Methods: We searched Medline, EMBASE, PsychInfo, Global Health, Web of Science, Scopus, IBSS, CINAHL, WHO Global Index Medicus, the Cochrane Library, WHO website, ProQuest, and Google Scholar for primary studies and reports published between January 1, 2000, and February 16, 2021 (updated on June 06, 2022), on induced abortion decision-making trajectories and/or their determinants in LMICs. We excluded studies on spontaneous abortion. Two independent reviewers extracted and assessed quality of each paper. We used “best fit” framework synthesis to synthesise abortion decision-making trajectories and thematic synthesis to synthesise their determinants. We analysed quantitative findings using random effects model. The study protocol is registered with PROSPERO number CRD42021224719. Findings: Of the 6960 articles identified, we included 79 in the systematic review and 14 in the meta-analysis. We identified nine abortion decision-making trajectories: pregnancy awareness, self-reflection, initial abortion decision, disclosure and seeking support, negotiations, final decision, access and information, abortion procedure, and post-abortion experience and care. Determinants of trajectories included three major themes of autonomy in decision-making, access and choice. A meta-analysis of data from 7737 women showed that the proportion of the overall women's involvement in abortion decision-making was 0.86 (95% CI:0.73–0.95, I2 = 99.5%) and overall partner involvement was 0.48 (95% CI:0.29–0.68, I2 = 99.6%). Interpretation: Policies and strategies should address women's perceptions of safe abortion socially, legally, and economically, and where appropriate, involvement of male partners in abortion decision-making processes to facilitate safe abortion. Clinical heterogeneity, in which various studies defined “the final decision-maker” differentially, was a limitation of our study. Funding: Nuffield Department of Population Health DPhil Scholarship for PL, University of Oxford, and the Medical Research Council Career Development Award for MN (Grant Ref: MR/P022030/1).
