The electronic databases – ProQuest One, Scopus, EBSCOhost, Sabinet, Science Direct and Google Scholar – were searched systematically for studies on the implementation of workplace clinical training of undergraduate optometry students. Searches were conducted from May 2022 to July 2022. The search strategy was tailored to meet the requirements of each database by using a combination of keywords related to optometry student clinical training or supervision and clinical practice. For example, keywords: ‘optometry’ AND ‘student’ AND ‘clinical supervision’ AND ‘clinical practice’ were used for ProQuest One search.

The review considered studies that reported on implementation of clinical training at the external training sites or workplace (Table 1), taking into consideration content, quality, perception, impact, process and/or delivery of student clinical training, practice or supervision. The pool for optometry studies was very limited, leading reviewers in extending the search to include other health professions.

Guided by the inclusion and exclusion criteria (Table 1), two reviewers independently conducted the searches extracting resources by title, abstract and full text. The reviewers discussed and compared the papers included at every stage to determine their accordance and discrepancies. Where they differed, reviewers engaged in further discussion, until they reached consensus.

The database search yielded 450 studies and examination by title yielded 74 studies, then 72 studies after removal of duplicates (Figure 2). Examination of potential studies by abstract yielded 33 studies and examination of full texts of the remainder of the studies yielded 21 studies. The 12 studies excluded based on full-text examination were found to not be related to student clinical training and/or supervision and practice. The 21 articles were re-examined, and after thorough review, 7 articles were not in accordance with the aims and objectives of the study and were excluded. A total of 14 eligible studies were critically appraised in order to determine their level of evidence and quality.

Different critical appraisal tools were used to assess the methodological quality of 14 eligible studies (Appendix 1). The level of evidence and quality of studies were graded according to the American Association Critical Care Nurses (AACN) evidence levelling system (Figure 1) (Armola et al. 2009:70–73). Eleven studies were appraised according to the JBI critical appraisal checklists (Whiting et al. 2003:1–3); of which, four were appraised according to JBI systematic reviews checklist (Aromataris et al. 2015:132–140), four according to qualitative studies checklist (Lockwood, Munn & Porritt 2015:179–187), one according to experimental studies checklist (Tufanaru et al. 2020:3–10) and two studies according to text and opinion (descriptive) checklist (McArthur et al. 2015:188–195). Mixed Methods Appraisal Tool (MMAT) sourced from Hong et al. (2018a:285–291) and Hong et al. (2018b:1–11) was used to appraise one study, and two other studies were appraised according to the Non-Research Literature Appraisal Tool (Yale New Haven Health 2016:1–3). Out of 14 studies, 13 studies were found to be reputable sources of evidence (Figure 2) and were included in this systematic review (Table 2).

Based on the level of evidence (Figure 1), the reviewers agreed to grade the 13 studies that were found to be reputable sources of evidence according to a methodological quality grading of: low quality for studies that scored below 40%; good or medium quality for those that scored 40% to 80% and high quality above 80%. In all, 12 studies were found to be of high quality, scoring above 80%, whereas one study was found to be of good or medium quality, scoring 80%. Most studies scored 90% and above; however, some studies that were graded to be of high and medium quality did not meet all critical appraisal criteria against which they were measured (Appendix 1).

Thirteen studies were found to be potential resources to guide clinical training and supervision of undergraduate optometry student training, outside the training institutions (Table 2). Eight studies discussed external clinical training in optometry, four of which were purely about optometry, whereas the other four discussed various disciplines, including optometry. Five studies discussed external clinical training in other health professions.

A standardised tool was developed to extract relevant information from the selected studies, and the extracted data are summarised in a table format (Appendix 2). A qualitative analysis was conducted, whereby texts from the summarised studies’ results were extracted and systematically transformed into highly organised and concise summary of key results (Erlingsson & Brysiewicz 2019:94), to describe a specific phenomenon (Downe-Wambolt in Bengtsson 2016:9). Data were then condensed and labelled to formulate codes (Figure 3), whose patterns were compared and organised into categories leading to themes. That assisted this review in identifying examples of student clinical supervision systems or approaches, enablers, barriers and outcomes as shown in Figure 3. This process required several reiteration to validate data before establishing themes and subthemes emanating from data (Bradshaw, Atkinson & Doody 2017:5).

Educational and clinical supervision are the two key elements for effective supervision, and they have complementary roles that offer distinct but connected contributions to student learning and support (Hindi et al. 2022:7). Clinical supervision support learners in developing their proficiency and confidence, and in translating the knowledge and skills gained, into clinical practice, whereas educational supervision support learners in navigating the clinical pathways (Hindi et al. 2022:7). Both are mainly conducted by qualified professionals (Hindi et al. 2022:7), but there are other clinical training approaches where senior students can supervise junior students in a clinical setting, namely peer-assisted learning (PAL) (Van Vuuren 2017:10). According to Boud in Van Vuuren (2017:9), PAL is a two-way reciprocal learning activity involving the sharing of knowledge, and it is a vehicle to help undergraduate healthcare students learn how to teach (Van Vuuren 2017:9). Peer-assisted learning has been implemented effectively in nursing and medicine; it is fairly new in the allied health professions (Van Vuuren 2017:9).

Coronavirus disease 2019 (COVID-19) has proven that e-learning is possible and moreover essential; it shifted physical teaching to online teaching and in doing so reformed education in general (Jonuscheit et al. 2021:154–155). Some training centres introduced pooling of resources and implemented joint appointments (dual academia-clinician appointments), while others introduced a variety of web-based online platforms, such as simulation, dry and wet labs, virtual reality and e-learning (Chee et al. 2021:262; Dean et al. 2020:1077–1078; Jonuscheit et al. 2021:152; Lee et al. 2020:1737–1738). Simulation was found to be of benefit during clinical training because trainees displayed confidence when they are not practising on real patients, significantly improving their performance (Flanagan & De Souza 2018:428). Tele-medicine and/or consultation, on the other hand, have emerged and been adopted by most practitioners as the preferred mode of providing care in hospital and private practice settings, as well as during clinical skills training (Sehgal et al. 2021:962). Remote supervision, like tele-consultation, allows students to be placed at the external facilities to expose them to other health skills and practice opportunities, while simultaneously addressing staff shortage at those facilities (Seaman et al. 2022:5363).

In sub-Saharan Africa, there is a move for the national regulator to oversee training of ophthalmologists, the examination standards and accreditation of training centres, for programmes to be implemented successfully (Dean et al. 2020:1067–1073). This is enabled by policies, guidelines and standards being available for the countries and institutions to use during training (Dean et al. 2020:1079–1080). Although there is a system in place to standardise ophthalmology training in the sub-Saharan countries, there is, however, a general shortage of ophthalmologists at accredited district facilities used as the training centres, resulting in a shortage of trainers (Dean et al. 2020:1074; Flanagan & De Souza 2018:432–433). Staff shortage is not only encountered in ophthalmology, but other professions do have similar problems, including optometry. Authors, Dean et al. (2020:1077) and Jonuscheit et al. (2021:152) agree and suggest that to manage a shortage of trainers and not disadvantage trainees, institutions should explore other training modes through the development of broader collaborative training programmes with external or international stakeholders.

Irrespective of the facility or sector, it is important that clinical training of professionals at workplaces is structured and delivered in a suitable, equitable and transparent manner, with roles, responsibilities and competencies of trainers and trainees clearly defined (Schafheutle, Jee & Willis 2018:1025–1026). Those responsible for clinical placements should prepare and provide both students and supervisors with clear guidance covering detailed training expectations (Kirkman et al. 2022:9). Moodley and Singh (2020:30) emphasise the importance for academics to ensure students meet the required competencies through community-based clinical training assessments. Numerous queries arose related to assessments traditionally conducted within the formal academic setting; including appropriateness of methods for community-based settings and capacity of clinical supervisors (Moodley & Singh 2020:30–32). In such instances, designing assessments for external settings is a challenge as the learning environments are not standardised, and students would be assessed by a number of tutors, with varied levels of academic skills, making clinical assessments difficult to control (Moodley & Singh 2020:33).

It is also important for the universities to implement student support systems for students to access help before relationships with supervisors are strained, as most students highlighted a lack of support from the university, resulting in a degree of anxiety (Kirkman et al. 2022:9). Student training can offer facilities an increase in human resource compliment and reduced patient waiting period for health services, and high patient numbers at public sector facilities offer students an opportunity to access large patient numbers and a wide range of clinical cases, which is in line with the regulatory authorities (Ebrahim et al. 2019:3–4).

Individual factors (student and trainer characteristics), environmental factors (policies, curriculum) and clinical factors (trainer experience and competence, clinical surrounding and resources, training style), should be identified and planned for as they affect student clinical education, and if not well considered, the expected quality of training may be affected negatively (Pashmdarfard & Shafarood 2018:120). If the clinical training environment is poor, with deficiencies such as a lack of educational or training centres, poor patient numbers and limited resources (funding, trainers, space and/or equipment), trainers may feel overburdened and reluctant to train, ultimately affecting the clinical competence of their assigned students (Dean et al. 2020:1079–1080; Ebrahim et al. 2019:5–6; Pashmdarfard & Shafarood 2018:118–119). This can have both a positive and negative effect on student practice preference post-qualification (Seaman et al. 2022:5381–5382). Under-resourced environments, may on a positive note, as seen with COVID-19, serve as opportunities for institutions and training facilities to explore alternative modes of training that are cost-effective, contextually appropriate, accessible, offer standardised training and improve patient safety through introduction of stricter hygiene practices (Flanagan & De Souza 2018:429; Lee et al. 2020:1754–1755; Sehgal et al. 2021:962).

In a study by Kirkman (2022:9–10), students reported that, they experienced times when relationships with supervisors were strained, resulting in them being anxious, without university support. Student and supervisor characteristics (individual factors) should be taken into consideration during planning and implementation stages of clinical training as they can affect the outcome (Pashmdarfard & Shafarood 2018:120). It is important that supervision models are flexible and adaptable to the settings, learner needs and the healthcare workforce skills (Hindi et al. 2022:7).

Lack of clarity with respect to the roles, responsibilities and accountability of trainees placed at private sector external sites was found to limit their exposure to a wide range of activities during training and that made them feel isolated (Schafheutle et al. 2018:1025). Their counterparts at the public sector sites were found to have had more benefits because they were allocated with other trainees for peer support, qualified professionals as role models, senior staff as supervisors, support staff and other health professionals and thus, not confined to one area, but interacted with a variety of professionals from different disciplines (Schafheutle et al. 2018:1025). There is, however, no conclusive evidence on whether shorter or longer placement periods positively or negatively impact upon the preferred practice location, post-qualification (Seaman et al. 2022:5382).

To maximise student clinical competence, the gap between theoretical knowledge and clinical practice should be minimised (Pashmdarfard & Shafarood 2018:119). Educators must have appropriate clinical qualifications, be competent in clinical practice and have the required experience, as they play a critical role in ensuring that students are competent (Pashmdarfard & Shafarood 2018:119). It is important for universities to be aware of levels of supervisor competence and experience to be able to allocate them or place students accordingly; this is supported by Ebrahim et al. (2019:6), who found junior optometrists to be knowledgeable in theory, but lacking clinical experience, requiring further maturity to serve as clinical teachers. On the other hand, senior optometrists have the necessary experience and maturity to train students, but lack adequate knowledge on current theoretical aspects (Ebrahim et al. 2019:5). Experienced trainers could help facilitate the professional identity development of trainees (Hindi et al. 2022:8).

Training the trainer programmes were found to equip the trainers in teaching, and cascading knowledge and skills to trainees (Dean et al. 2020:1073); however, the transfer and assessment of desirable character, attitude, ethics and responsibilities are yet to be proven (Dean et al. 2020:1078). Peer-assisted learning, ongoing professional development and pedagogical training of supervisors, including assessment methods, are recommended in order to keep supervisors informed and up-to-date with new developments that current students are being taught (Kirkman et al. 2022:9–10; Moodley & Singh 2020:32; Van Vuuren 2017:10–11). This is important, as most of the students participating in community-based education programmes are exit-level students, needing to meet graduate competencies in preparation for the work environment (Moodley & Singh 2020:32).