An ongoing struggle: a mixed-method systematic review of interventions, barriers and facilitators to achieving optimal self-care by children and young people with Type 1 Diabetes in educational settings

Two recent narrative reviews [5, 6] have focused on T1D and school and both have methodological limitations. Wodrich et al.[6] did not use systematic processes or report the characteristics and designs of studies. Tolbert [5] only used the keywords type 1 diabetes, school and management to retrieve 10 quantitative descriptive surveys and 1 mixed-method study. No attempt was made to determine study quality in either review. Although these reviews provide useful background context, neither provide a trusted source of synthesized evidence to inform decision-making and policy and practice development.

It is important that a child or young person’s T1D should be managed effectively in educational settings in order to ensure optimal glycaemic control. In contrast to the previous reviews, we sought to conduct a policy-informed mixed-method systematic review that utilized a comprehensive and systematic search strategy and assessed the methodological quality of the included studies. Findings from the review were then used to inform intervention development in a large United Kingdom (UK) Government funded study [7]. The objectives were:

In the UK [8‐18] and US [19‐22], a number of key guidelines set out the components of safe and optimal T1D care at school. International clinical practice consensus guidance has also been developed by the International Society for Pediatric and Adolescent Diabetes (ISPAD) [14]. The development of T1D medical management plans that address specific needs of the child, specific guidance on general T1D management, and training needs of all those involved in supporting the child with T1D at school are common across all guidelines. We extracted key elements of best practice for children’s T1D management in schools and used this as a conceptual framework to guide analysis and interpretation of evidence (Table 1).

We conducted a mixed-method systematic review. The design was informed by mixed-method synthesis methods developed by the Evidence for Policy and Practice Information (EPPI) Centre [25, 26] and is shown in Figure 1. We followed Cochrane Effective Practice and Organisation of Care Guidance on the inclusion of more diverse quantitative study designs to determine the effectiveness of interventions as our initial scoping review has identified few randomized controlled trials [27]. The EPPI ‘mixed-methods’ triangulation approach maps evidence from effectiveness studies (Stream1: quantitative data) with evidence from studies reporting the attitudes and experiences of participants (Stream 2: non intervention studies including surveys and qualitative studies). We then conducted an overarching narrative synthesis from streams 1 and 2 to determine the extent to which interventions to optimize T1D care and management in educational settings addressed the barriers, and built on the facilitators, identified by children, parents and teachers. The quantitative component of the review (stream 1) adhered as far as possible to PRISMA reporting guidelines (http://​www.​prisma-statement.​org). We developed a detailed protocol which is not publically available.

The search strategy is summarised within a modified Setting, Population/People/Perspective, Intervention/Issue of Interest, Comparison, Evaulation (SPICE) [28] table (see Table 2). The search terms included medical subject headings (MeSH) and ‘free text’ terms in combination and was adapted according to the particular database. A single search was used for both stages of the review with no methodological restrictions (for a sample of searches see Additional file 1). The databases searched for relevant studies were: CINAHL, MEDLINE, Scopus, British Nursing Index, Cochrane Library, EMBASE, PsychINFO and Web Of Science. In addition, reference lists of retrieved papers and published reviews were searched and unpicked for potentially relevant papers. References were managed using Endnote X1.

Studies were included if they focused on children and young people with T1D within an educational setting and included those 3–16 years in preschool or formal education and those 16–25 in post compulsory education. In addition, studies including or focusing on parents, peers, educational setting personnel and health professionals that related to this age group were included. Restrictions were not applied in terms of research design or methods. Unpublished data were not sought from authors. All studies published in the preceding 15 years were included (January 1996-July 2011) that were conducted in any country and published in English in peer-reviewed scientific journal. A 15 year window was selected to capture a reasonably contemporary context. Studies were excluded if there was no before and after measures (stream 1) and if the study did not directly report the views of children and young people, parents, peers, professionals (stream 2).

All studies identified were assessed for relevance by DE and LS to the review based on the title and abstract. For studies that appeared to meet the inclusion criteria, or in cases when a definite decision could not be made based on the title and/or abstract alone, the full paper was obtained for detailed assessment by two researchers against the inclusion criteria. Any disagreement was resolved by consultation with a third independent reviewer (JN).

Figure 2 shows the flow of papers at each stage. A total of 71 papers reporting 66 studies were included.

For stream 1 (intervention studies), randomised intervention studies were assessed on criteria developed by Kirk et al. [29]. A summary of the quality assessment is provided in Table 3.

Key aspects of quality for non randomised intervention studies in stream 1 were based on the work of Deek et al.[33] (see p39 of the Centre for Reviews and Dissemination, University of York guidance on undertaking reviews in health care [34] ). A summary of the quality assessment is provided in Table 4.

The strength of synthesized findings for stream 1 (intervention studies) was assessed using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach [43] where certainty of evidence is reported as being high, moderate or low/very low.

For studies in stream 2 (non-intervention studies) that used a survey design we used the checklist as designed by Rees et al.[44] and for qualitative studies using the appropriate checklist available from the Critical Appraisal Skills Programme (CASP) [45]. These were then incorporated with quality criteria that were adapted from Kirk et al. [29] to provide a summary of quality assessment and available with Table 5.

Confidence in synthesized qualitative and survey findings was assessed using the Confidence in the Evidence from Reviews of Qualitative Research (CerQual) tool developed by Glenton et al.[102], which uses a similar approach to GRADE. The original CerQual approach was designed for qualitative findings and we used the same process but included findings from surveys in the assessment of confidence. Confidence in findings is described as high, moderate or low (See Figure 3). All studies were included unless fatally flawed and study quality is reported for each stream.

Additional study characteristics (Additional files 2 and 3) and results (Additional files 4 and 5) were extracted directly into pre-formatted tables and followed the format recommended by the Centre for Reviews and Dissemination (CRD) [34]. One researcher extracted the data and a second researcher independently checked extracted data for accuracy and completeness [34]. Any disagreements were noted and resolved by consensus among the researchers.

Three types of syntheses were performed. Firstly for stream 1 (intervention studies) meta-analysis was inappropriate due to the heterogeneous nature of the studies in relation to populations, interventions and outcomes. Instead the results from the studies (any summary measure) were reported in a narrative summary within and across studies. Secondly, for stream 2 we used Ritchie and Spencer’s thematic framework synthesis [103] for non-intervention studies. All included studies in stream 2 were then uploaded into the software Atlas Ti and an a priori index coding framework based on the conceptual framework presented in Table 1 and issues of interest mapped against review questions and objectives was applied to studies. Thirdly, a final overarching synthesis of intervention and non intervention studies was conducted. For this final synthesis a matrix was constructed that mapped best practice guidance against the age-related barriers and facilitators identified by children and young people, parents, school personnel and school health professionals and age-related interventions and outcomes in stream 1 (Additional file 6). We were particularly interested to see the extent to which interventions were effective and addressed the barriers identified by children, parents and teachers/health professionals, and built upon the facilitators to providing optimal care and management of children and young people with T1D in educational settings. We also identified gaps in evidence, assessed the robustness of the synthesis by making observations about the quality of included evidence, and looked specifically at the age and context of child participants in interventions compared with child participants in studies of attitudes and experiences.

Sixty-six studies were included in the review. For stream 1 (intervention studies) 11 were included (see Table 4, and detailed tabular summary Additional file 2). Only 3 out of the 11 studies were randomised controlled trials (RCTs) [30‐32]; 1 was a controlled trial [80], 3 were before and after studies [35, 36, 38], 2 were analog experiments [39, 40], 1 cohort study [41] and 1 programme evaluation [42].

Of the 11 interventions, only 2 were explicitly reported as theory based [36, 37]. The study by Wdowik [79] utilised the Theory of Reasoned Action, and Social Learning Theory and developed an expanded Health Belief Model. The conceptual frameworks for the pilot study by Faro [36] were based on social learning and developmental theory.

Sample sizes were small and ranged from 20 to 156 with the exception of the cohort study where the number of nurses attending the continuing education programme was 417 [41]. Follow-up periods ranged from 3 months to 1 year. The majority of studies (9) were conducted in the US and 1 study was conducted in Canada [32].

For stream 2, 55 studies were included (see Table 5, and detailed tabular summary Additional file 3). Thirty-four studies used a survey design, 17 used a qualitative approach, 2 employed a mixed-method design, 1 utilized a survey followed by qualitative group interviews and 1 employed a retrospective survey using case notes.

Studies investigated different types of interventions and used different outcomes to assess their effectiveness and were too diverse to undertake a meta-analysis. A narrative and tabular summaries (see Additional files 2 and 4) are reported. The narrative summary and tables are organized into two groups: interventions focusing on children and young people with T1D at educational settings and interventions focusing on school personnel.

Diabetes quality-of-life was measured in two studies [30, 80] Both studies found significant improvements on the treatment barriers subscale at 12 months (Izquierdo et al.[31], p = 0.039) and Engelke, [35] p = 0.01). Izquierdo et al.[31] also found a significant improvement (at 6 months: p = 0.017 which was maintained at 12 months).

Three studies [30, 31, 36] measured HbA1c levels (measure of glycaemic control). Two studies [30, 31] showed significant improvements in the HbA1c readings over a 3 month [30] (p < 0.05) and 6 month [31], (p < 0.02) period following the intervention, whereas the other [36] showed no significant change. The effect of the intervention on health service use was measured in two studies. This section of the analysis by Izquierdo et al.[31] was poorly reported but showed that urgent visits to the school nurse for diabetes related problems and urgent calls to the diabetes centre decreased significantly (p value not reported) and that there were significantly fewer hospitalisations (p value not reported), and emergency department visits (p value not reported). Whereas Faro et al.[36] did not show any significant differences for the frequency of hospitalization or emergency department visits.

One study investigated the diabetes knowledge of university students and reported that knowledge was significantly improved (p < 0.001) as a direct result of the intervention and was maintained at 3 month follow up (p < 0.001). They also showed a significant increase in the number of university students who knew their recent HbA1c results (p = 0.003) post intervention [37].

Six studies involved school personnel and the samples included both school nurses and school teachers [38], elementary school teachers [32], regular and special education elementary teachers [39], continuing education and pre service teachers [40] and school nurses [41].

The T1D knowledge of school teachers showed a significant improvement (p < 0.004) after the implementation of an education programme [38] whereas there was no significant change after the introduction of a compact disc (CD Rom) teaching tool [32]. This difference in findings could be attributed to the fact that all school personnel already had experience of caring for a child and young person with T1D and therefore already possessed a good level of knowledge of T1D [32] whereas only 38% of teachers in the study by Siminerio and Koerbel [38] had experience of children with T1D.

With regard to confidence, providing education through a CD Rom was found to significantly increase school teachers levels of confidence in managing diabetes (p < 0.016) [32]. Two further studies assessed teachers confidence in attributing class learning and behaviour problems to hypothetical students with T1D and found that a teacher’s level of perceived confidence to manage a child with T1D in their classroom was not related to the amount of disease related information they received [39]. However the more knowledge teachers were given about the consequences in the classroom of chronic health conditions the more confident they were in attributing chronic conditions to behavior (p = 0.007) [104].

Two further studies reported that perceived levels of competence for school nurses and suggested that where diabetes education was part of continuing education programs that school nurses ability to manage students with diabetes would be enhanced [41] and one study showed significant significantly improved results (p = 0.0001).

Best practice guidance (see Table 1) sets out optimal ways for children and young people to self-manage their T1D whilst at school. However, there is no specific guidance on the management of T1D specifically for college/university students but there are recommendations concerning all young people with T1D and alcohol within the NICE guidelines [24].

Less than half (31-46%) of students had a written care plan [55, 65, 69]. School policies generally applied to the entire student body within a particular school and did not often consider the child and young person with T1D and their needs to perform T1D self-management at school [52, 60]. School nurses felt that a care plan for emergencies was important for facilitating the care of a student with T1D in the school environment [99].

Younger students (≤10 years) kindergarten/nursery through to junior/middle school) reported that they needed assistance with blood glucose monitoring during school hours [47, 68]. This was usually the role of school nurse [68]; or a designated member of the school staff [46, 47, 62, 65], peers [62] and in some instances parents [65, 68]. Older students(≥12 years) generally required less assistance once they were attending high/secondary school [65, 68].

Glucagon was found to be available at school for between 34–49% of students [47, 66, 69]. A high percentage of both children (60%) [47] and parents (64 percent) [66], felt that that glucagon should be readily available, together with a person who was aware of how to administer it. Only 10% of children had experienced a serious hypoglycaemic episode at school [66]. In only a very small number of cases was a call to emergency services made (3% [67]), or glucagon administered [68].

Students considered having accessible test kits and snacks available whenever they needed them as important [46]. Healthcare professionals felt that schools should rethink policies which prevented children having easy access to their medical equipment [92]. Students reported that storage of medication or items related to T1D was in a variety of places i.e. with the student [52‐54], another room in school, in classroom, in the school office [52], the nurse’s office [54], secretary/teachers office/desk [54]. School nurses felt that they could better support students if they could have ready “access” to snacks and testing kits as well as appropriate medical supplies [98, 99]. Most students with T1D took a snack to school [50] but some reported that they were not allowed to eat snacks when they needed to [54, 60]. However, some students reported that they it difficult it have a snack before physical education (PE) lessons [58].

When insulin was administered in school, between 46–97% [47, 51, 52, 63, 68] self-injected, especially older students (11 years and older) who attended secondary/high school [65, 68]. For a small minority of students it was the school nurse (18%) [54] or a member of school staff (1-6%) [47, 54, 65, 67]. For younger students (6–10 years [47], ≤ 12 years [65],:≤10 years [68]) (2-32%) it was the young student’s parent who came into school to give an injection or administer a bolus if the child needed insulin if no one at school was trained and/or allowed to administer insulin [47, 65, 67, 68]. In certain instances, however, students occasionally had to go home if nobody was available to administer insulin [65] and when this was not possible a small minority reported treatment modifications were made because of a lack of cooperation from the school [47, 66]. A small number of students were also not allowed to inject insulin whilst in school [50].

Some students who self-injected were supervised (20–49%) whilst taking their insulin [54, 65]. This was usually the role of a school nurse or a designated member of the school staff [68]. Younger children (6–10 years) required more support [47] especially if they were in kindergarten/nursery (5–6 years) or infant/elementary school (7–10 years) [68]. Students generally appreciated being reminded by the teaching staff to administer their insulin [54].

Only 30–54% of students were permitted to check their own blood glucose in the classroom [65, 68, 69] increasing to 74% for students at high school [68]. For those not allowed to perform blood glucose monitoring in the classroom, a number alternative locations are provided which include medical room [65], school office [65], head teachers office [65], anywhere [65]. Students have been shown to demonstrate significantly better glycaemic control when they are given flexibility to decide where to perform self-care behaviours [61].

With regard to insulin administration, students reported problems with a lack of a private location within the school where they could administer injections [50, 52, 69]. Locations provided or chosen by the students included first aid/medical room/health office [50, 52, 60, 65], toilets/cloakroom [50, 52, 65], classroom [50, 69], “Wherever I have my lunch”, school dining room [53], cupboard in the school office [53], school office [65], locker between classes [61] and head teachers office [65]. However some students reported being happy with their usual place [53]. Most students with T1D took a snack to school [50] but some report that they were not allowed to eat snacks when they needed to [52, 60].

Whilst some schools provided strategies so that students with T1D could participate in sport [52, 66], some parents stated that their assistance and presence was required during and/or after school sport [67] especially for younger children (age not specified) [46] and in some instances their children were not allowed to play sports such as football [70]. Children felt that participation in such activities could be facilitated if a nurse was still on site [49] as T1D supplies were often locked in the nurse’s office [58]. Older children said that their coaches needed to be more knowledgeable about T1D [46].

The timing of school lunches was a commonly reported problem [52] with only 25% of primary school teachers and 38% of secondary school teachers appreciated that students with T1D should not be late for a meal [86].

Some students, parents and school nurses however felt that food choices provided in the school canteen, vending machine and classrooms, which were conducive to healthy T1D management, were limited [60, 98] and that ensuring snacks and appropriate foods were available can reduce barriers to good control at school [99]. Students reported that they would benefit from more healthy food and drink options [57, 66, 70], prominent and consistent information about prepared food and developing nutritional analyses for all foods available in the cafeteria as a way of helping in choosing meal options in the cafeteria [60]. As a solution, some students in some instances had to take their own lunch to school [55]. Some parents reported that schools were not able, or did not consider it their responsibility, to modify diets to enable children with T1D to eat a school lunch [66].

Parents sometimes reported having trouble getting nutritional information about foods served and portion sizes from their child’s school. This made it difficult to plan ahead whether using a constant carbohydrate approach or insulin to carbohydrate ratio [23]. Only 7% of students reported that their school cafeteria made carbohydrate content of prepared foods available [61].

Parents reported that their child’s T1D affected their decisions regarding extracurricular activities [61]. In some instances parents/guardians were asked to act as chaperone on field trips, especially for younger children [68], but parental attendance should not be considered a prerequisite for participation by students with T1D [19].

Even though teachers felt that all children with T1D should be allowed to go on extended trips with the school [86], between 15 and 20% [47, 62, 65, 66] of parents reported experiencing difficulties with their school over responsibility of the children during 1-day trips especially for children under 10 years of age [47, 66] with greater problems for trips which extend over several days [66]. In some instances, parents reported that their child was not allowed to participate in outside school trips unless accompanied by a parent or a school nurse [70] and for others, the schools had specific policies in relation to medication on school trips and worked with the family to ensure that children and young people could participate [52]. School nurses reported that planning for management during ‘out of town’ trips was critical [99].

Flexible accommodation with exams and educational tests was considered important as older children (>11 years) reported (23-39%) that if they experienced hypoglycaemic events before or during an exam, they did not have the opportunity to do the exam again [47, 62].

Students with T1D often expressed a feeling of ‘being different’ from their peers due to their T1D [50‐52, 55, 92, 99]. They also felt embarrassed if they ‘had a hypo’ [51, 55] or when they had to check their blood glucose or take medication at school [55, 59]. These feelings could act as barriers to positive T1D self-care behaviours in school [98]. In an attempt to not appear different from their peers, young people compromised their T1D self-management by choosing not to alleviate their symptoms [51]. Some students reported that they did not like peers watching them inject [59]. On the other hand, some students stated that they did not mind others seeing them take their insulin [53].

Some parents reported that their child was bullied/picked on at school because of their T1D (26%) [65] and a small percentage of students themselves reported problems with their peers such as diabetes-related bullying or teasing [50, 51, 61]. This was more of a problem for older students (over 12 years) in high/secondary school [65]. As a result, students balanced the need to have peers around them who knew about their T1D and the emergency management strategies by telling one or two close friends they felt they could trust [52] to keep the T1D a secret from most of their peers [46, 51]. Some talked about such peers as a T1D “buddy”, who with training would be able to recognize hypoglycaemia, alert staff, prompt self-care, buffer teasing, and escort the student to the nurse [61]. When age and level of HbA1c was taken into consideration, students who received help from trained peers were found to have significantly higher quality-of-life in the school environment [61].

School healthcare personnel were more able to facilitate optimal management when confident in providing support to children, able to communicate with healthcare providers, possessed T1D knowledge and skills and undertook regular education and training. The main difference between the UK and US policy was the provision of a school nurse. There are no specific recommendations regarding the role of the school nurse within UK guidelines. The US guidelines recognise the school nurse as the most appropriate person in the school setting to provide care for the student with diabetes.

Despite the policy intent, many US schools did not have a full-time nurse, and sometimes a single nurse covered more than one school [105‐107]. The numbers of students reporting that their schools had a school nurse on site varied widely (Spain: 21 to 48% [47, 62] and US: 65 to 95% [54, 68, 69] with a smaller percentage working full time [46, 68], and some students reported that they received support from school counsellors (57%) for non-medical, T1D related problems [61]. US children reported that even though they had school nurses assigned to their schools that these nurses did not come every day, and they worried about what might happen if they “got very low” and no one was there to help them [46]. Those nurses who did come every day were not always on site all day and this caused problems for some children as supplies were often locked in their office [46]. Just under 50% of children from Spanish schools [47, 66] felt that a nurse should be available daily during school hours to help with the management of T1D. Parents also expressed concern about the qualifications and training and preferred the presence of a daily onsite nurse [70] as opposed to a health aide who they felt was unable to provide adequate care [60]. Over 80% of school nurses felt that numbers of school nurses that were available across schools for students with T1D were inadequate and that a school nurse should be available on school premises during the school day if a student with T1D was enrolled [55]. Guttu et al.[97] demonstrated that a significant correlation existed between increased presence of school nurses and care and services provided to children with T1D.

The greatest support students received at school came from teachers [46, 47, 62, 66]. A high proportion of school personnel however (65%), expressed concern about the potential liability and child safeguarding issues when caring for students with T1D in school [55], which was related to concern surrounding exposure to and interaction with children’s bodies, especially someone else’s child [92].

The most supportive teachers were those who (1) were flexible in allowing children and young people to test their blood glucose [46], (2) allowed them go to the nurses’ office in the middle of a class or test [46], (3) included snack times for the entire class based on the schedule of students with T1D [60], (4) kept a supply of juice or snacks for students with T1D to use during an emergency [60]. Some students (14–45%) reported that they were not allowed to snack in class when they needed to [50, 61, 69] or that teachers delayed the student attending the nurse’s office to treat their hypoglycaemia [61].

Some young people felt that school nurses were well educated about T1D [60], whereas others felt the nurses’ knowledge could be improved [60]. A third of school nurses perceived their own levels of T1D knowledge as being low to average [98] and rated themselves as moderately confident in providing T1D care and education [96]. Self-efficacy was significantly higher if they were currently participating in the care of children with T1D, when there were students with T1D in the school system, and when they were supervising students with blood glucose meter testing [96]. Only 20% felt adequately prepared to assist a child with hypoglycaemia [55].

Parents reported concerns about Continuous Subcutaneous Insulin Infusion (CSII) therapy in school, specifically testing, bolusing, and pump management [64]. As well as being completely unfamiliar with pumps or CSII therapy [64], the biggest challenge faced by school nurses was learning to count carbohydrates when a student was on CSII therapy [101]. When they first encountered a student on CSII therapy, school nurses were scared, intimidated and overwhelmed as a consequence of their lack of education and experience with this new technology [101].

The majority of school nurses (94%) had up-to-date T1D reference materials in their offices [96] with some nurses obtaining information about T1D from the internet and professional books and magazines [98]. Just over a third (36% of school nurses reported that they had attended a conference on T1D during the past year [96]. Barriers to acquiring new information by school nurses were time constraints (37%), lack of access to education/regular updates and inadequate training (28%) [98].

Children and young people reported that they would like teachers to be better informed about T1D and to have better knowledge about T1D in order to help the students manage their T1D in school [46, 60, 62]. The lack of education received by school personnel was described as being problematic by parents [60]. Most teachers had received written information about children’s T1D (82%) [65], whereas only 22% of both regular education and special education teachers indicated that they felt well informed regarding T1D [93]. Students and parents felt that teachers should receive written instructions in order to improve the management of T1D and to improve the child’s integration at school [47, 66], in particular information about the symptoms and steps to be followed in case of a hypoglycaemia, more information about T1D in general, information regarding the optimal management of emergencies [62] that should be kept in their class and in the common areas [47]. School nurses however felt that school staff needed to improve their T1D knowledge and this then could reduce barriers to good control at school [99]. Both children and parents felt teachers had a basic knowledge about T1D and that they were adequately trained to care for children to manage T1D [55, 62, 66] although sometimes they reported concerns that there was confusion at school between Type 1 and Type 2 diabetes [65].

Diabetes healthcare professionals’ biggest concern was about the ability of teachers to spot the onset of hypoglycaemia and react quickly [92]. Teachers were confident in their ability to be able to recognise the signs of hypoglycaemia (70-71%) [87, 95] but were less confident that they would be to cope with emergencies that arose from high or low blood sugar levels (42-63%) [87, 95].

Having regular appointments with healthcare providers and written communication between the healthcare providers and school nurses regarding management needs for school and increased information exchange between the two was seen as beneficial [92, 99]. However, communication between the healthcare team and the school nurse was reported to only occur often or very often for a quarter of nurses [55]. Healthcare providers were often difficult to reach and were too busy to respond to questions [98].

Important facilitators to optimal self-care and management included the ability to balance T1D and student lifestyle and to have specific diabetes management strategies in place when drinking alcohol.

College/University students found it challenging trying to juggle all aspects of T1D. Students said that lack of a perceived routine at university was a barrier to effective self-management [80] with many reporting little or no time to engage in practices such as blood glucose testing [76, 79, 80, 83, 85], exercising [79, 80, 85], eating snacks during the day [79, 83] and injecting in a suitable environment [83]. Inadequate finance was also cited as a barrier to successful T1D management with students studying in the US as they worried about the cost of blood glucose monitoring strips [80]. Whereas in the UK this is not an issue and students reported that being on a student budget did not affect their control because they get free prescriptions and supplies, they either lived in halls of residence where meals were provided, or they borrowed money from their parents, or built up a student debt to be paid off later [85]. Having to eat in class or carry food or supplies around and having to test blood glucose were seen as barriers to diabetes management [80]. Motivating factors were being physically able to keep up with their peers and the of long term consequences [80]. When significant barriers or negative emotions were present Wdowik et al. [79] reported that students with positive attitudes and good intentions however, may be unable to engage in desired self-care behaviours when complications arose. Whilst some students wanted their friends to know about their diabetes and to be able to help them in an emergency others were concerned about what their peers would think if they knew they had diabetes and wanted to avoid being treated differently.

The majority of students reported that they drank alcohol whilst at university [74, 82, 85]. Drinking excessive amounts of alcohol can interfere with metabolic control and can induce hypoglycaemia and in the longer term it can worsen or increase the risk of diabetes complications [108].

Students reported using several strategies when drinking alcohol which were in keeping with current guidelines for young people [24] and general advice for students [109]. These included the following: eating before and/or during drinking [74, 80, 82, 84]; keeping track of how many drinks you were having [82]; determining, in advance, not to exceed a set number of drinks [80, 82, 84]; avoiding drinking games [82]; having a friend let you know when you have had enough [82]; alternating non-alcoholic with alcoholic beverages [82]; pacing drinks to 1 or fewer per hour [82]; choosing not to drink alcohol [84]; drinking an alcohol look-alike (non-alcoholic beer, punch etc.) [82]; avoiding or diffusing peer pressure [84]; limiting the frequency of going out to parties and bars [84]; drinking with trusted friends [80, 84] who understood the symptoms of low blood sugar that would mimic intoxication [80], and checking or monitoring blood glucose levels [76, 84]. Students reported that they usually tested their blood glucose at home, or in more private spaces, before and after going out, rather than while they were out [76].

Students felt that they did not receive adequate support from their college or university that would enable them to balance the demands of further education and management of their T1D. An example was being allowed to manage their glycemic control during examinations [83].

Although some students preferred to continue to receive care from their diabetes team in their home town [85] others felt unsupported by their T1D team [83]. For example they reported that care had to stop with their old diabetes team now that they were at university and they experienced difficulties assessing a new local diabetes team [83], whilst others felt that they had not received enough information particularly in relation to managing their diabetes and drinking alcohol [84].

Additional file 6 shows the juxtaposition of guidance imperatives mapped against barriers and facilitators to optimal T1D management in educational settings identified by children, parents and professionals, mapped against interventions to promote optimal T1D management in educational settings.

Overall, interventions aimed to provide additional targeted help and support for children to self-manage some specific aspects of their T1D in schools, or interventions to increase knowledge and awareness of school nurses and staff, and better communication with diabetes professionals. These foci broadly match with children’s and parents views and experiences of what needs to be done to optimize children’s T1D self-management in education settings, which in turn map onto interventions outlined in guidance as shown in Table 1.

In the following section, where reciprocal guidance, quantitative and qualitative evidence exists addressing the same issues, a synthetic line of argument derived from integrating intervention effectiveness and views evidence is presented along with an assessment of the certainty/confidence in evidence.