Persistent poor glycaemic control in individuals with type 2 diabetes in developing countries: 12 years of real-world evidence of the International Diabetes Management Practices Study (IDMPS)

Diabetes affects approximately 463 million people worldwide, of whom 90% have type 2 diabetes, and the prevalence is expected to increase by 51% by 2045. A large number of people with diabetes live in developing regions, with estimates suggesting that 55 million people in the Middle East/North Africa, 32 million in South and Central America, 19 million in sub-Saharan Africa and 88 million in South East Asia have diabetes [1].

Poor glycaemic control can lead to an increased risk of blindness, end-stage renal disease, cardiovascular disease and lower limb amputations [2]. In 2012, 1.5 million deaths worldwide were directly caused by diabetes [2]. A further 2.2 million deaths were due to cardiovascular disease, chronic kidney disease and tuberculosis, associated with high blood glucose levels [2]. Optimal blood glucose control is therefore needed to reduce the risk of complications and premature death in developing regions, which have a high burden of diabetes and possess fewer resources to treat end-stage disease, creating a considerable impact on healthcare systems.

Over the last decade, there have been major advances in diabetes management, resulting in improved outcomes for individuals with type 2 diabetes. These include the introduction of novel technologies, newer oral glucose-lowering drugs (OGLD; e.g. sodium–glucose cotransporter [SGLT]₂ inhibitors, dipeptidyl peptidase 4 inhibitors [DPP-4i] and glucagon-like peptide-1 receptor agonists [GLP-1 RA]), insulin therapies and delivery systems/devices.

While the majority of guidelines consider an HbA_1c goal of 53 mmol/mol (<7%) to be appropriate in most individuals with diabetes, use of individualised HbA_1c goals are recommended, based on patient preferences, characteristics, comorbidities and risk of adverse events [3‐9]. These individualised goals can range from <48 mmol/mol (<6.5%) in individuals with low hypoglycaemia risk and <64 mmol/mol (<8%) in high-risk patients, such as those with a history of severe hypoglycaemia, extensive comorbidities or complications, older individuals or those with long-standing diabetes [3‐9].

Self-management is a cornerstone in diabetes care and provision of diabetes education can improve self-management behaviours and glycaemic control [10‐13]. Most professional bodies recommend the use of structured diabetes education and support programmes delivered by trained healthcare providers (HCPs), such as nurses, to improve self-management, and that these education programmes should be given at the time of diagnosis [3‐9]. According to the International Diabetes Federation guidelines, for every primary care facility, at least one HCP should be trained as a diabetes educator, and these facilities should provide regular, structured group education to support individuals when needed [5].

In anticipation of increasing disease awareness, better care standards and technological advancements, the International Diabetes Management Practices Study (IDMPS) was designed to document and track patient profiles and patterns of care across time in developing countries, where data are limited. The IDMPS is the largest international observational study with the participation of over 6000 physicians from 49 countries across Africa, the Middle East, South Asia, Latin America, Asia and Eurasia. Data were collected using structured case report forms in a series of yearly ‘waves’, with each wave recruiting a different cohort of participants. The first wave of data collection began in 2005 and the most recent wave (wave 7) was completed in 2017. Apart from a standard dataset collected in all waves, each wave had a particular theme, such as understanding factors involved in glycaemic control (wave 1), healthcare resource utilisation (wave 2), barriers to insulin therapy (wave 3), hypoglycaemia (wave 4), symptoms of depression (wave 5), self-management (wave 6) and insulin discontinuation (wave 7) [10, 14‐17].

Results from previous waves have demonstrated the low attainment of treatment goals for LDL-cholesterol (<2.6 mmol/l), BP (<130/80 mmHg) and HbA_1c (53 mmol/mol [<7%]) with only 25% of participants with type 1 diabetes and 36% of participants with type 2 diabetes achieving HbA_1c of 53 mmol/mol (<7%); <8% of participants achieved all three goals [14]. Data analysis from other waves has demonstrated the positive associations of good glycaemic control with patient education and self-monitoring of blood glucose (SMBG) [10], in addition to the positive association of diabetes-related complication rates and increased healthcare resource utilisation [15, 17].

The present analysis of real-world data describes glycaemic goal achievement, therapy use and care management practices in people with type 2 diabetes over time, from the first wave of IDMPS data collection (2005) to the most recent wave (2017).

Glycaemic control in developing countries has been persistently poor over the past 12 years and is growing steadily worse, based on real-world data captured in this large international observational study involving over 66,000 individuals with type 2 diabetes. This situation has developed despite multiple advances in the field of diabetes management, including the development of new medications proven to improve diabetes control and clinical outcomes [18].

Most participants in this study were aged 40–65 with 8–9 years of diabetes duration, with over 80% of participants receiving OGLD. While the use of sulfonylurea monotherapy declined over time, use of metformin monotherapy increased over time and metformin was the most commonly used monotherapy, in line with guideline recommendations [3‐9]. Few participants received newer agents, such as SGLT₂ inhibitors, although it should be noted that these therapies were approved relatively recently in many developing countries; as such, data on their use are only available for wave 7. The proportion of individuals receiving insulin also increased over time, with use of basal + prandial insulin showing a marked increase; despite this rise, however, there was no improvement in glycaemic control. Newer insulin analogue therapies can provide clinical benefits in terms of reduced glycaemic variability or hypoglycaemia risk [19], which may facilitate individuals to achieve glycaemic control; however, in developing countries, these analogues may not be readily available due to issues of access or cost. Data from wave 7 highlighted that a substantial proportion of participants still received human intermediate-acting insulins (24.4%), although long-acting analogue use increased between waves 6 and 7. Participants receiving prandial insulin were mainly using human regular insulins or premix insulins.

Overall, <50% of participants achieved a glycaemic goal of HbA_1c <53 mmol/mol (<7%), and <70% achieved HbA_1c <64 mmol/mol (8%). Furthermore, only 15–25% or 45–50% of participants treated with insulin achieved either the HbA_1c <53 mmol/mol (<7%) goal or the HbA_1c <64 mmol/mol (8%) goals, respectively. These results might be ascribed to a combination of delayed and inappropriate insulin regimen prescription. Logistic regression analysis was used to confirm the declining trend of HbA_1c <64 mmol/mol (<8%) goal attainment over time. The odds of goal attainment compared with wave 1 were significantly lower for waves 2, 5, 6 and 7 (p < 0.05 for all). Other patient groups (including younger individuals [≤40 years], women, those with longer disease duration [>10 years] and those from Europe/Eurasia) were also less likely to achieve the HbA_1c goal compared with their counterparts. Visits to general practitioners or specialists did not seem to influence achievement of glycaemic goal; any potential difference may have been attenuated by study selection criteria requiring all physicians to have prior experience in prescribing insulin. Additionally, individuals with poor glycaemic control or advanced disease progression might tend to seek specialist care.

Although there was an increase in the proportion of individuals receiving diabetes education from physicians, very few received structured diabetes education courses delivered by nurses, dietitians or certified diabetes instructors. This may potentially result in a lack of sufficient contact time to help individuals deal with day-to-day concerns.

These original findings from developing countries concur with other reports based on IDMPS data [14, 20], and are in line with data from developed countries indicating poor rates of glycaemic goal attainment (<53 mmol/mol [<7%]; ~20–40%) [21, 22], indicating that the challenge of attaining good glycaemic control is universal. It is concerning that glycaemic control remains poor, given the increased use of insulin over time; it should be noted that the insulin dose of 0.3–0.7 U/kg was comparable if not higher than that used in clinical trial settings [23‐25]. In this survey, the mean time to insulin initiation was 8 years, similar to that reported in developed countries [26, 27].

Considering that >50% of participants in this study receiving OGLD and/or insulin treatment displayed an HbA_1c value >53 mmol/mol (>7%), this is a clear indicator that the initiation and intensification of insulin remain a major barrier in real-world practice. The discrepancy between doses prescribed and glycaemic control suggests possible patient non-adherence, although this was not formally tested. In this regard, quality improvement programmes implemented at a system level have been shown to improve control of cardiometabolic risk factors and clinical outcomes in community settings; such programmes include multidisciplinary care (with training programmes for physicians and nurses), risk-stratified care planning and regular structured/scheduled assessment of metabolic control and vascular complications [28‐31]. The provision of professional diabetes education for HCPs starting in medical colleges, together with postgraduate training, is critically important to build capacity to help HCPs in educating patients and improving care. An alternative (ideally complementary) approach would be to make changes to the practice environment to provide more integrated multidisciplinary care, to ultimately provide sustained improvements in care outcomes. Such an approach would be conducive to promoting trustworthy relationships and communication between patients and HCPs, and should encompass empathetic listening, tuition of self-management skills and ongoing support [32].

According to professional guidelines, individuals with diabetes should receive structured diabetes education and support programmes [3, 5]; such programmes have been shown to improve glycaemic control, as reported in previous waves of the IDMPS [10]. In this study, over 70% of physicians reported having offered education to their patients; however, physicians are often time-poor in comparison with other HCPs (e.g. nurses, dietitians or certified diabetes instructors), and time is a major factor for a successful diabetes education programme. A meta-analysis of the efficacy of self-management education on glycaemic control in individuals with type 2 diabetes reported that approximately 24 h of contact time (e.g. face to face visits, phone calls) with HCPs are needed to sustain a 1% reduction in HbA_1c over a 12 month period [33]. There are also grounds to argue that regular follow-ups are needed, as results from a recent meta-review indicate that post-intervention improvements in HbA_1c persist until 6 months but tend to attenuate after 12 or 24 months [34]. Consultation times may also be inadequate, perhaps partly due to the pressures of the increasing global population of people with diabetes which may increase the number of patients each physician treats. A recent meta-analysis of primary care physician consultation time (spanning developed and developing countries) reported that consultation time was less than 5 min in 18 of the included countries, accounting for approximately 50% of the global population, with consultation length proportional to per capita health spending [35]. Such data as these indicate a strong need for additional support to engage patients and emphasise the importance of time spent talking to individuals about their needs and concerns. However, in our survey, <20% of participants received a structured educational diabetes programme from any source, which could be a contributing factor to the poor glycaemic goal achievement observed herein. We suggest that promising methods to improve glycaemic control would include training of nurses and dietitians in the provision of diabetes education, a general increase in the number of certified diabetes instructors and the delivery of such education through structured programmes.

The affordability and ownership of blood glucose monitoring accessories (e.g. blood glucose monitors and test strips) are important factors for improving self-management. In this study, an increasing proportion of participants possessed SMBG monitors over time, but there was also a rise in the number of individuals citing cost as a limiting factor for regular SMBG; in those treated with insulin, where a greater frequency of SMBG is needed, the high cost of strips may be a deterrent. Considered together, clinical inertia, insufficient access to structured diabetes education, infrequent SMBG and high cost of monitoring accessories/medications may all contribute to persistently poor glycaemic control in individuals with type 2 diabetes. A combination of factors ascribed to patients (e.g. fear of hypoglycaemia/injections, complex treatment regimen, polypharmacy), physicians (e.g. poor/ineffective communication, insufficient knowledge and support) or healthcare systems (e.g. lack of time/resource for physicians, lack of tools for patient/physician to monitor insulin titration, lack of medical coverage), may be particularly relevant in developing countries where development of infrastructure and capacity cannot cope with the rapid rate of increase in diabetes [36, 37]. Other patient-related factors, such as older age, higher education level and short disease duration, have also been reported to be associated with improved glycaemic control [38, 39]; it is possible that lifestyle factors, acceptance of diagnosis and adherence may all contribute in these cases.

The present study has some limitations. The cross-sectional nature of the survey provides a ‘snapshot’ of practice at any one time; therefore, these observations only allow us to form hypotheses, and cannot infer causality. The self-selecting nature of the patient population should be considered, as those with the poorest control are most likely to visit the clinic; therefore, glycaemic control in the general population may be higher than that shown here. All data were completed by the attending physicians with potential bias in interpretation and recall. Due to the pragmatic nature of the survey, no adjudication was conducted for the reported complications. The varying degrees of local support in implementing the survey also mean that different countries/regions were included/excluded in different waves. These variables add to the heterogeneity of practice, although we have adjusted for major variables (notably age, sex, disease duration and region) when analysing secular trends. Despite these limitations, we believe that aspects of this study such as its large, global population, long duration (12 years) and the practice of structured data collection using predefined variables have provided valuable real-world evidence from areas hit hardest by the diabetes epidemic, where data are lacking to inform practice and policies.

In conclusion, in this 12 year study, poor and worsening glycaemic control was observed in individuals with type 2 diabetes. These real-world data highlight an urgent need for improvement in practice environments, workflow and team structure; such amendments would allow early assessment of patients and identification of unmet needs, thus empowering individuals to improve self-management and consequently glycaemic control. These changes will need to be supplemented by institutional support through capacity building, and policies that promote good diabetes care; these would ideally encompass improved accessibility to and affordability of medications and monitoring accessories, in both developing countries and subpopulations of patients with poor literacy/low incomes in developed countries [5, 40]. Altogether, these could lead to a successful improvement in diabetes care and clinical outcomes in individuals with type 2 diabetes living in developing countries.