|Definition and measurement of self-harm||▪ Studies with clear definitions of self-harm (or alternative term or concept used) as an intentional act of self-inflicted injury or poisoning, in addition to clear means of case identification, assessment or measurement.||▪ Studies focused on unintended self-harm behaviours (e.g., smoking, drink-driving, eating disorders etc.).|
▪ Studies focused on intended self-harm with socially sanctioned motives (e.g., scarification, manhood rituals, ‘body enhancement’, religious fasting, hunger strikes etc.).
▪ Studies focused on intended self-harm behaviours not approved by the broader sociocultural context but are sanctioned by the subcultures (e.g., cult groups, Goth subcultures, Emo subcultures etc.) within which they occur.
▪ Studies focused on suicidal ideations, self-harm thoughts, or threats, as these do not necessarily translate into or represent acts of self-harm.
▪ Studies focused on suicide (self-inflicted death).
|Prevalence estimate||▪ Studies with specified time frames within which prevalence of self-harm was assessed.||▪ If prevalence estimates cannot be determined within a clear time frame;|
▪ If there is no clear indication of sample size and population denominator.
|Setting||▪ Studies with primary focus on self-harm conducted within non-clinical contexts (i.e., general population, community, school-based, households / neighbourhoods, street-connected settings etc.) in countries within sub-Saharan Africa.|
▪ Studies conducted in clinical contexts focused on self-harm as the main presenting condition.
▪ Clinical studies concerned with self-harm as the primary condition (but not as comorbid condition, e.g., self-harm in HIV/AIDS or epilepsy).
|▪ Studies focused on adolescents in prisons or borstal institutions, unless control groups in such studies allow for the evaluation of risk and protective factors of self-harm in adolescents.|
|Participants||▪ Studies reporting prevalence estimates of self-harm involving participants aged between 10 and 25 years.|
▪ Studies reporting on the associates, risk and protective factors related to self-harm, methods of self-harm used, and reported reasons for self-harm involving participants aged 10 and 25 years with a personal self-harm history at the time of assessment for the study.
▪ Studies with wide age range but majority (90% or more) of the participants are within the age bracket of 10–25 years.
|▪ Adolescents with pervasive developmental disorders, cancer, insulin-dependent diabetes, epilepsy or HIV/AIDS adolescent patients, unless control groups in such studies allow for the evaluation of risk and protective factors of self-harm.|
▪ Studies involving participants within wide age ranges with the study results not disaggregated by age, making it impossible to link specific results to participants age 10–25 years, and where participants are stratified by age but with participants aged 10–25 years constituting less than 90% of the total sample which did not specifically link the reported prevalence estimates, identified risks or associates of self-harm, protective factors, methods of self-harm, or the identified reasons for self-harm to young people aged 10–25 years.
|Study Designs||▪ Studies with focus on self-harm which address at least one of the four specified objectives of this review using: (1) quantitative methods (i.e., school-based, household-based, population/community-based cross-sectional survey; census; retrospective or prospective descriptive cohort designs; case controls; case reports; randomised controlled trials, and analytic cohort designs); or (2) qualitative methods (e.g., interviews, focus groups etc.); or (3) retrospective reviews of clinical records.|
▪ Cross-national studies involving countries in sub-Saharan Africa and other countries outside the sub-region, which stratify and link the results to the included countries. In such instances, the specified results related to the sub-Saharan African countries were included in this review.
|▪ Studies based on the same dataset reported in an earlier publication included in this review.|
▪ Systematic reviews, commentaries, editorials, opinion pieces, correspondence, and articles not based on data.
▪ Where full text of the identified article was unavailable or could not be accessed.
▪ Cross-national studies involving countries in sub-Saharan Africa and other countries outside the sub-region, which did not stratify or link the results to the respective included countries.