This systematic review has shown that there are a number of studies relating to interventions that reduce unscheduled hospital care for hypoglycaemic events for adults with diabetes. This is currently particularly important as adults with diabetes have been reluctant to go to hospital during the COVID-19 pandemic, fearful of contracting the virus and consequently missing treatment . The interventions were categorised as telemedicine, education, integrated care pathways, enhanced primary care and care management teams and had varying levels of effectiveness as outlined below.
Telemedicine is the delivery of health care at a distance to optimize and improve health outcomes . There were 3 studies that utilised telemedicine, albeit in different ways. Quan et al. used automated telephone calls with behavioural follow ups from health care staff or a trained lay person for 6 months . Due-Christensen et al. utilised acute telephone counselling from a diabetic specialist nurse (DSN) out of hours over 6 months . Warren et al. used a home monitor that captured clinical measurements and provided additional care from a diabetes care coordinator .
The use of telemedicine showed a decrease in unscheduled admissions and ED visits across all 3 studies although the results were not significant [20, 34, 37]. These studies were based in Australia , USA  and Denmark , showing success in different countries. Both Warren et al. and Quan et al. used a RCT design to compare efficacy [20, 37] whilst Due- Christensen et al. used an observational design thus not allowing for the control of variables . Whilst there were no significant differences in the findings, a larger sample size might have shown significance as there were only 126 participants in , 362 in  and 592 in .
Patient education is offered to people with long term conditions to aid and enhance their self-management of their health and wellbeing . Elliott et al. evaluated the DAFNE (Dose Adjustment For Normal Eating) education course that is used in the UK and Ireland whilst Yeung et al. developed a 2.5 year long-term education program that provided low intensity self- management education for 6 months before a 24 month high intensity self-management support component with a certified diabetes educator and clinical psychologist [19, 30].
Patient education also showed a decrease in the number of unscheduled admissions and ED visits. Elliott et al. found that this decrease in ED visits and unscheduled admissions was significant , however Yeung et al. did not find the decrease in admissions to be significant, nor the reduction in ED visits . This could be due to the small sample size in the study by Yeung et al.  (n = 60) and the older age of the participants in the study by Yeung et al.  (mean age 62 years old) compared with the study by Elliott et al.  (mean age 41 years old). The trials took place in the USA  and UK . These results show that a 2.5 year, long- term educational program does not necessarily provide increased benefits over shorter structured education programs.
Care management teams
Care management teams are a team based approach to helping patients and their support system manage chronic illnesses more effectively . Kaufman et al. used coordinated care management teams composed of a registered nurse, licensed practical nurse, community health worker, health coaches, social work staff, program manager, nurse care manager, program director and associate clinical director to improve care . Ginzburg et al. used a nurse managed care team comprising a physician, nurse, social worker, pharmacist, physical exercise consultant and other medical specialists to achieve optimal diabetes control and management, with telephone reminders included . Kearns et al. compared resource utilization in traditional care with resource utilization in care management teams comprising a physician, medical assistant and care manager, who was a certified diabetes educator .
The use of care management teams had mixed results on their efficacy in reducing unscheduled admissions. Kaufman et al. found a decrease in healthcare utilization in both ED visits and admissions however they had a small sample size (n = 25) . Ginzburg et al. also had a relatively small sample size (n = 100). Their study showed a significant increase in non- acute care visits to physicians and ophthalmologists along with a non- significant increase in dietitian visits. These increases, could explain the decrease in hospitalisations observed . Kearns et al. had the largest sample size (n = 19,696), however found no change in healthcare utilization and admissions relevant to the intervention. There was a decrease in urgent care visits and hospital admissions and an increase in ED visits and readmissions, however these changes were seen in both the intervention and control groups, suggesting influence from outside factors not related to the intervention .
Integrated care is the coordination and integration of health services, to ensure the best patient care . The integrated care studies were broken down into those that involved pharmacists and those that utilised the ambulance service. Brophy et al. used collaborative drug management therapy involving both a pharmacist and care manager for high risk patients treated with polypharmacy, providing health coaching, education, transportation assistance and prescription refill assistance . The study by Chung et al. used a clinical pharmacy program under a collaborative drug therapy management program enabling patients to receive a 30 min visit with the pharmacist, when needed, to maintain patient safety and achieve the patients’ goals . Sampson et al. implemented a new integrated care pathway for managing severe hypoglycaemia that involved providing patients with written information on avoiding severe hypoglycaemia and a diabetes education follow up session with an educator within 3 days of their call, unless the patient actively opted out . Bennett et al. used community paramedicine to shift care from ED and inpatients to outpatient and medical home based care by community paramedics implementing a care plan devised by a liaison nurse over a number of visits . Whilst ambulance care is considered unscheduled care, these studies were relevant as they were aimed to prevent hospital admissions and ED attendances.
The pharmacy interventions by both Brophy et al. and Chung et al. showed a statistically significant decrease in hospitalisations upon their implementation. Brophy et al. found that there was a decrease in ED visits although it was not significant . This was contradicted by Chung et al. who found that there was an increase in ED visits, however this increase was less in the intervention group than the comparison group . Whilst both Chung et al. and Brophy et al. had large sample size [23, 24], the sample size in the control and intervention groups in  were unevenly matched (557:225) . These studies both used retrospective data analysis and so confounding variables would not have been controlled.
Studies by both Sampson et al. and Bennett et al. involving ambulance services showed a decrease in unscheduled admissions [25, 31]. Bennett et al. showed a decrease in ED visits and ambulance calls, however an increase in those requiring hospital transport for higher levels of care. The control group also reflected these outcomes, however at a lower rate than the intervention group, indicating that the intervention group utilised care more appropriately than the control group . This contradicts the study by Sampson et al. who showed a decrease in hospital transports upon implementation of the new clinical pathway .
These studies suggest that better co-ordination across health care sectors and professional groups has the potential to, and has been shown to, improve outcomes in adults with diabetes.
Enhanced primary care
Enhanced primary care is increased clinical and social support in the community provided by nurses, care coordinators, support workers and others who work alongside GP’s to help patients learn more about and improve their condition management . Zurovak et al. used integrated teams of physicians, nurses, medical assistants, practice managers, behavioural health therapists, registered nurse health coaches and panel managers to provide patient-centred care to improve behavioural health, care management of chronic illnesses and improved technology . MacKay et al. enhanced the role of the medical office assistant to carry out key tasks in diabetes care to find out if it improved the effectiveness of care . The study by Seidu et al. compared practices providing enhanced primary care with practices providing core care. Enhanced practices had a primary care physician and practice nurse with an interest in diabetes who identified patients who could be discharged from secondary care and managed in primary care with monthly meetings discussing care for the complex cases . Wong et al. implemented a patient empowerment program to give patients greater control over their health care decisions utilizing a collaborative approach between the patient and healthcare provider . In Goff et al. ’s study they implemented a team care model, consisting of 2 registered nurses, 2 medical assistants trained as outreach workers and a case manager and compared the outcome and resource utilization with matched controls who did not receive the enhanced care . The study by Peterson et al. utilised nurses to work with the patient and their primary care physician to develop and implement care plans, contacting patients approximately once a week, to measure association with extending CareFirst’s BlueCross BlueShield commercial health insurance program to Medicare Fee-for-service patients on outcome and resource use . Multiple interventions were used by McLendon et al. to enhance diabetes care including nurse care management involving doctors, physician assistants and nurses, telemedicine and education .
Enhanced primary care was the most common intervention identified in this review with six studies focused on it. Three studies were based in the USA [26,27,28], one in the UK , one in Hong Kong  and one in Canada . Zurovac et al. found a slight, non- significant increase in the number of admissions for ambulatory care sensitive conditions but no overall change in hospitalisations . This contradicted the other five papers that showed a decrease in the number of unscheduled admissions, albeit all non- significantly. MacKay et al. found that whilst there was a decrease in admissions, the control group also had a decrease suggesting external factors not related to the intervention were at play . ED visits were found to be decreased or remained similar in four of the six studies. Seidu et al. did not measure ED visits but commented that they would be unlikely to increase  and Zurovac et al. showed an increase in ED visits although at a lower rate than the control, despite the program not employing specific strategies to reduce ED visits . Wong et al. showed an increase in specialist outpatient clinic visits but a significant decrease in general outpatient clinic visits  with Seidu et al. showing a decrease in the number of non- elective bed days .
Reducing unscheduled admissions for diabetes will help reduce overcrowding and clinical pressures in the emergency department, leading to reduced waiting times. This will also reduce costs on the health services by providing the right care, in the right place, at the right time, by placing the patient at the centre of the model which aligns with the Transforming Your Care (TYC) strategy in the UK . In addition, better co-ordination across health care sectors and professional groups has the potential to, and has been shown to, improve outcomes in adults with diabetes.