A total of 1676 citations were identified through the initial search (Fig. 1). Of these, 48 full-text studies were assessed for eligibility and 20 (two of the same program) [13, 26] met the inclusion criteria. Of the included studies, seven were cohort studies [13, 14, 17,18,19, 27, 28], seven mixed-methods [26, 29,30,31,32,33,34], three qualitative [16, 35, 36] and three cross-sectional [37,38,39] (Table 1). There was a total of 4591 participants (n = 533 mentors; n = 1625 mentees; n = 2878 unspecified) with a range of 9–2362 participants in each study. When including only studies which specified year level of mentees/mentors, most mentees were in their first clinical year (n = 898), whilst most mentors were near peers in either their ultimate or their penultimate year (n = 224). Three programs had mentors and mentees from the same year level [30, 33, 34]. In general, mentors and mentees were randomly matched, typically following a selection process for mentors. Six [14, 17, 27, 30, 37, 38] of the included studies considered specific characteristics (such as international students, being taught by the same faculty, personal relationships) when matching mentors and mentees. Mentoring groups were generally small, with one notable exception where a social media platform was used for large group mentoring (> 1000 students within the same year level) [33]. Twelve studies reported a component of peer teaching as part of the mentoring program [13, 14, 16, 26, 28,29,30,31, 36].

Peer mentoring programs could broadly be divided into two types. Fifteen studies (two describing the same program) were ‘longitudinal’ programs, aimed at establishing mentor–mentee relationships to offer ongoing support over a longer period, such as an entire academic year. The remaining five studies were more ‘focused’—generally shorter-term programs with the intent of providing mentorship during significant moments for mentees, such as transition to clinical placement, medical school orientation or exam periods. Accordingly, where programs were named, these names generally reflected the goals—‘BigSibs’ [39] and ‘Mentors in Medicine’ [29] for longitudinal programs, compared to ‘Step Sibling’ [34], which focused on providing (non-academic) support for the USMLE Step 1 exam.

In terms of outcome measurements, 17 studies used quantitative tools, mostly self-developed (piloted or validated in four studies [28, 29, 31, 39]). Three studies used standardised and validated tools: Mentorship Effectiveness Scale [27], DREEM [37] and LCQ, W-BNS and PCS [26]. Assessment timepoints were short-term (start and end of the programs). Only one study [31] assessed additional longitudinal outcomes (12 months post-program). Results were analysed for statistical significance in 12 studies [18, 19, 26,27,28,29,30, 34, 36,37,38,39].

Outcomes for longitudinal and focused programs respectively are presented in the following section, outlining benefits, elements for program success and challenges.

Of the 20 included studies, 15 described longitudinal near-peer programs. Thirteen spanned the duration of the academic year, whilst one was semester-long (12 weeks) [31] and another, 4 years long [36]. Two programs were directed at preclinical students who were transitioning to clinical placement [30, 31]. The remaining 13 studies included first or second year students as mentees, and two involved students across all year levels [29, 36].

Mentors were typically senior students in their second (five programs) or fourth year (two programs), or students from multiple year levels (six programs). Two programs also involved doctors as mentors [30, 36]. One program paired mentees with either mentors in the same year level as the mentee, or two academic years senior [30]. The other program paired student and faculty mentors [18]. The program initially commenced with faculty mentors only, but added student mentors following feedback that mentees were reluctant to meet their mentors. Subsequently, students met more often with both peer and faculty mentors [18]. In contrast, a study by Yang et al. compared the experiences of mentees who had received mentoring from peer mentors to those who had received mentoring on a one-on-one basis with a physician; there were no reported barriers to engagement; however, there was a notable component of teaching in this mentoring program, and attendance was mandated [36].

Studies assessed the impact of their programs on either the mentors, mentees or both. Seven studies [14, 18, 26, 29, 30, 35, 39] assessed outcomes measures for both, whilst three [13, 19, 27] assessed mentee outcomes and three [17, 31, 36] mentor outcomes. Of the seven studies which assessed outcomes for both, all studies reported positive outcomes in confidence, perceived learning environments and psychosocial wellbeing for both mentees and mentors. Where peer mentors or physician mentors were compared, there was no difference in reported mentee benefits [36].

Positive outcomes for mentees were academic development, psychosocial wellbeing and communication skills. Five studies [14, 29, 30, 35, 39] reported mentees felt more confident with their academic skills due to skills gained from mentors, such as academic planning [29] and study techniques [35]. Mentees also perceived the learning environment more positively [13, 26, 31, 36, 37] and reported improvement of social, interpersonal [30, 35] and clinical communication skills [14]. Two studies [29, 30] found improved psychosocial and emotional peer support for mentees, and one study [37] showed mentees were more likely to feel they had a good support system in times of stress. The provision of a ‘safe space’ for mentees was highlighted in three studies [13, 29, 35]. Another study emphasised that mentees felt more competent and autonomous as learners in the learning climate fostered by near-peer mentors [26].

Mentor outcomes were academic and interpersonal skills gained from mentoring, and overall satisfaction with mentoring. Mentors perceived improvement in professional skills, such as responsibility, leadership and communication [18, 30, 31, 35, 36, 38]; moreover, they felt satisfied with their experience [14, 36, 38, 39], and more confident when learning and teaching [