FormalPara Key Summary Points

Why carry out this study?

Type 2 diabetes mellitus (T2DM) is associated with a significant financial burden and low quality of life in the Kingdom of Saudi Arabia (KSA).

To improve the management of T2DM and reduce the prevalence of T2DM-related complications, there is a need to review the current treatment pathways and assess the economic burden in the KSA.

What was learned from the study?

Analysis of a heterogenous group of patients with T2DM showed that biguanides (monotherapy or combination) were the most commonly used 1st-line therapy.

The lifelong total cost of T2DM care was estimated as 201,377 to 437,371 Saudi Arabian riyal (53,700–116,632 US dollars).

The major cost drivers for the most frequently used 1st-line regimens were the cost of cardiovascular disease complications.

Since T2DM is a progressive disease, treatment intensification of patients with inadequate glycaemic control should be implemented immediately,thereby avoiding the therapeutic inertia that might delay the glycaemic control and imposes a substantial economic impact on the healthcare system in the KSA.

Introduction

Diabetes mellitus (DM) is a chronic condition characterised by elevated glucose levels in the blood (hyperglycaemia). Type 2 diabetes mellitus (T2DM), accounting for around 90% of all cases of DM, is characterised by hyperglycaemia as a result of inability of the body to respond sufficiently to insulin (insulin resistance) followed by an inadequate production of insulin [1]. The global prevalence rate of diabetes (including all types) in adults was estimated as 536.6 million (10.5%) in 2021 [1] whilst that of T2DM was 6059 per 100,000 individuals in 2017, which continues to rise despite various public health measures [2]. The International Diabetes Federation reported the 2021 world population standardised prevalence of diabetes (including all types) as 18.7% for adults in the KSA, which is projected to reach 21.4% by 2045 [1]. Similar to global patterns, the major drivers of this rise in T2DM in the KSA could be increased life expectancy (LE) and lifestyle changes (physical inactivity and unhealthy eating) due to rapid urbanisation and industrialisation [3].

Patients with T2DM have an increased risk of complications, which can lead to decreased quality of life (QoL) especially if the disease is undiagnosed or poorly managed [1, 4]. The chronic T2DM-related microvascular complications include nephropathy, neuropathy and retinopathy [4]. Furthermore, T2DM-related chronic macrovascular complications include coronary artery disease, peripheral vascular disease and cerebrovascular disease [1]. Cardiovascular diseases (CVD) including myocardial infarction (MI) are the major cause of death and disability amongst patients with T2DM [4, 5]. These complications contribute to an increase in direct medical costs (treatment, diagnostics, hospitalisations, and inpatient and outpatient visits) and indirect costs (absenteeism, reduced productivity, lost productivity due to early mortality), and a reduced QoL [4, 6,7,8]. The drugs used for diabetes rank first among the most commonly used medications for chronic conditions in the KSA, thereby posing a significant financial burden [9]. A systematic review of published studies up to 2018 estimated total per-patient direct cost of 1947 US dollars ($) in the KSA for patients with T2DM [10]. The annual per-person diabetes-related healthcare expenditure in the KSA was $1743 in 2021 [1]. It was also estimated that the per capita healthcare expenditure due to diabetes for people aged 15−44 years is almost double compared to people without diabetes [11].

In order to assist healthcare providers in determining the most appropriate treatment options, Saudi National Diabetes Center (SNDC) developed Saudi Diabetes Clinical Practice Guidelines (SDCPG) in 2021 [12]. The guidelines suggest a glycosylated haemoglobin (HbA1c) level below 7% (< 53 mmol/mol) as the overall goal of management of T2DM [12]. Higher or lower targets (6.5–8%) may be appropriate in certain conditions and the targets can be individualised [12]. The SNDC guidelines prefer lifestyle modification, followed by metformin (drug class biguanide), if the disease is not controlled, as the first-line (1st line) treatment for T2DM [12]. In patients without atherosclerotic cardiovascular disease (ASCVD) or chronic kidney disease (CKD), second-line (2nd-line) combination therapy with metformin and sulfonylureas or thiazolidinediones or dipeptidyl peptidase 4 (DPP4) inhibitors or sodium-glucose co-transporter 2 inhibitors (SGLT2i) or glucagon-like peptide 1 receptor agonist (GLP-1 RA), or basal insulin is preferred if the HbA1c target is > 1.5% above the glycaemic goal. The choice of agent is selected on the basis of individual patients’ needs, cost and adverse event profile. In patients with elevated risk for ASCVD, heart failure, or CKD, a GLP-1 RA or SGLT2i is favoured. In patients who require a third-line (3rd-line) agent (with HbA1c > 9%) to achieve glycaemic goals, similar considerations are applied.

These treatment patterns for T2DM may differ in real-world (RW) practice as a result of several factors such as drug efficacy and safety, medicine cost, patient’s choices, drug side effects and physicians’ choices for drugs based on their individual experience [13, 14]. Anyhow, only about a quarter of patients with T2DM achieve the recommended HbA1c level (< 7%) in the KSA [15]. Therefore, there is the potential to curtail the economic burden of T2DM with effective diabetes management [7].

The Saudi Ministry of Health (MoH) aims to improve the management of T2DM and reduce the disease burden in the country by offering better treatment and blood glucose measurement devices, which necessitates extensive review of current treatment pathways and assessment of the associated costs. Therefore, recent baseline characteristics of patients newly diagnosed with T2DM, assigned treatments, and effects of these treatments are required. The current study aimed to collect these baseline characteristics and treatment patterns from a heterogeneous group of patients with T2DM using a retrospective patient chart review approach. Furthermore, the economic burden of T2DM in the KSA was estimated for the most commonly used 1st-line regimens (such as biguanides, sulfonylureas, insulin) and using the IQVIA Core Diabetes Model (CDM).

Methods

Study Design and Setting

This study was a longitudinal, observational study conducted at 15 participating centres (within general hospitals and specialised tertiary diabetes centres under the MoH) in the KSA over a span of 5 years. This study was conducted in accordance with the ethical principles derived from international ethics guidelines, including the Declaration of Helsinki, and the International Council for Harmonisation (ICH) guidelines for Good Clinical Practice (GCP), all applicable laws, rules, and regulations. Consistent with local regulations and prior to the enrolment of patients, the study received approval from the Saudi MoH, independent ethics committee/institutional review board (IEC/IRB). Since this research presented no more than minimal risk of harm to patients and involved no physical procedures with patients, informed consent was not required.

Study Population

The inclusion criteria for this study included a new diagnosis of T2DM over a 10-year study period (between the date of the first site initiation visit [SIV, 2021/2022] and 10 years prior [2011/2012]), with a minimum follow-up period of 5 years in any of the participating centres. Diagnosis of T2DM was based on measured HbA1c ≥ 6.5% (48.5 mmol/mol) and/or fasting plasma glucose (FPG) of ≥ 126 mg/dL.

The exclusion criteria included participation in an investigational study in which patient assessment and/or treatment was dictated by a protocol over the 10-year study period.

Data Sources and Collection

Patient data were extracted from medical records starting at the date of T2DM diagnosis (defined as baseline). Data were collected using INES (electronic Case Report Form), an online data capture and data quality management tool. Cost data from the MoH were collected to assess the costs of healthcare resource use for T2DM.

Data collected at baseline included patient demographics, risk factors and existing complications (cardiovascular, renal, ocular and foot ulcer). Follow-up data included all relevant information in the medical records after the date of diagnosis (with a minimum observation period of 5 years). This included requirement of screening tests (for eye diseases, microalbuminuria [MA], and gross renal proteinuria [GRP] and foot care programme), T2DM treatment patterns, other treatment details, newly diagnosed complications and cost of drugs and screening tests. For treatment patterns, all drugs prescribed to patients and dates of the prescriptions were recorded. A line of therapy (LoT) included all drugs prescribed to a patient within a 3-month duration. Drugs prescribed after 3 months from the previous drug were considered a new LoT.

Data Processing

The effects of treatment on the risk factors (HbA1c, lipids, and estimated glomerular filtration rate) were analysed during the 5-year follow-up period. Annual changes in risk factors were calculated using the closest follow-up data available within ± 6 months of the baseline date + follow-up year.

The incidence rate (per 100 patient-years) of each of the T2DM-related complication was calculated by dividing the number of new complications that occurred after baseline by the total number of patient-years at risk. Complications present at baseline (at date of diagnosis of T2DM) were not included in the incidence rate calculation.

Statistical Analyses

All analyses were performed using R version 3.6.1 or higher (R Foundation for Statistical Computing, Vienna, Austria). According to the nature of this study, the analyses were mainly descriptive and did not require statistical testing. Categorical variables were described by frequencies and proportions. Continuous variables were described by mean and standard deviation (SD). The calculations and proportions did not include the missing/invalid category in the denominators.

Model Description and Inputs

The IQVIA CDM version 9.5 Plus (http://www.core-diabetes.com/) was used to assess the burden of illness, which included long-term projections of clinical and cost outcomes of the most commonly used 1st-line regimens for T2DM from a payer perspective in the KSA. CDM is a proprietary, interactive computer simulation model that includes a series of interdependent Markov sub-models to perform real-time simulations of the progression of diabetes-related complications and associated mortality. CDM and its validation studies have been previously described in literature [16,17,18].

Clinical inputs for the model were retrieved from the retrospective phase of this study (baseline patient demographics, risk factors, ethnicity, proportion of patients using concomitant medications and screening tests, and rate of existing complications) or from published sources (cost of complications, sensitivity and specificity of screening tests [Supplementary Tables 1 and 2]). CDM default values were used for those baseline parameters which were not collected during the retrospective study. Rates of hypoglycaemia events were retrieved from published sources [19,20,21]. The treatment-specific efficacy data obtained during the retrospective phase of this study was used as input for the CDM.

Treatment-specific progression of risk factors (HbA1c, blood pressure, lipids and body mass index [BMI]) were applied as obtained during the 5-year retrospective phase of this study. No progression was assumed afterwards. For cardiovascular risk prediction, the general Framingham risk equation was used which is default for CDM v9.5 Plus. For mortality, combined (the United Kingdom Prospective Diabetes Study 82 [UKPDS 82]) approach was applied.

Utility values, calculated for every patient in each simulation year, were used to estimate the average quality-adjusted life-years (QALYs). The utility and disutility values were obtained from previously published sources (Supplementary Table 3) [22,23,24].

Only direct medical costs (associated with pharmacy, management, and T2DM complications) were included. The annual pharmacy costs for each treatment regimen were calculated using unit costs for drugs, retrieved from National Unified Procurement Company (https://www.nupco.com/en/) and drug dosage from World Health Organization Collaborating Centre for Drug Statistics Methodology (https://www.whocc.no/atc_ddd_index/). Unit costs associated with management and complications (including hospitalisation) associated with T2DM were obtained from published sources (Supplementary Table 1). Costs of complications obtained from published sources were inflated to 2021, where necessary, using consumer price index from The World Bank (https://data.worldbank.org/). The cost outcomes were also presented in US $ using an exchange rate of $1 = 3.75 SAR (Saudi Arabian riyal) [25].

The projected outcomes of the model included LE, QALYs, per-patient annual or lifelong (50 years) costs and event rates of diabetes-related complications per 1000 patient-years. The total lifelong costs were projected on the basis of the cost for treatment (drugs and glucose monitoring), management (screening tests), complications and adverse events (hypoglycaemic events). The projected annual costs were calculated by dividing the undiscounted lifelong costs by the estimated undiscounted LE.

Results

Patient Disposition

Data were collected from a total of 2226 patients (main population). The baseline demographics, risk factors and existing complications of this population are presented in Supplementary Table 4. However, as a result of many missing variables (specifically, baseline HbA1c or start date of treatment) in the collected data, the main population could not be used to provide necessary inputs for the CDM.

Therefore, for further analysis, a subpopulation was defined on the basis of the following criteria: (1) availability of baseline HbA1c, (2) baseline HbA1c ≥ 6.5%, and (3) at least one treatment with available start date (Fig. 1). This subpopulation consisted of 638 patients and all further analyses were performed for this subpopulation.

Fig. 1
figure 1

Criteria for selecting study subpopulation. N number of patients

Full size image

Baseline Demographics and Clinical Characteristics

The subpopulation of 638 patients showed an equal gender representation (50% male) (Table 1). The mean (SD) age at diagnosis of diabetes was 49.1 (11.6) years, with a follow-up time of 8.0 (1.4) years. The majority of the patients belonged to Arabian/Saudi ethnicity (71.0%), followed by White (22.6%).

Table 1 Baseline characteristics of the subpopulation
Full size table

Furthermore, patients had a mean (SD) BMI of 30.5 (6.1) kg/m2, with 44.1% categorised as obese, and an additional 40.1% categorised as overweight (Table 1), re-emphasising that obesity is an important modifiable risk factor for onset and progression of T2DM. The rather high mean (SD) HbA1c of 9.7% (2.0%) at baseline suggests a delay in seeking medical care by these patients due to possible lack of awareness about the disease.

A small proportion of patients reported CVD (5.5%), renal (0.6%), ocular (7.7%) and foot ulcer (1.6%) complications at the diagnosis of diabetes (Table 1).

Treatment Patterns During Follow-up

Table 2 describes the treatment patterns for patients with T2DM by drug class, LoT and information about monotherapy or combinations. Whilst all 638 patients reported data for 1st LoT, a total of 352 patients initiated a 2nd LoT after a mean (SD) duration of 30.6 (22.4) months from the beginning of 1st LoT. The overall analysis of all monotherapy or combinations showed that biguanides (81.5%), sulfonylureas (51.6%), DPP4 inhibitors (26.2%) and fast-acting insulins (17.2%) were the most prescribed 1st-line agents, whilst GLP-1 analogues and SGLT2i were used in only 2.0% of patients each.

Table 2 Description of T2DM treatments including monotherapy or combinations in 1st or 2nd LoTs
Full size table

The analysis of 1st-line regimens (Table 1) showed that the most common therapies were biguanides + sulfonylureas (regimen 1, 29.6%), biguanides (regimen 2, 16.1%), biguanides + sulfonylureas + DPP4 inhibitors (regimen 3, 12.9%), fast-acting insulins + long-acting insulins (regimen 4, 7.5%), biguanides + DPP4 inhibitors (regimen 5, 5.3%), premixed insulins + biguanides (regimen 6, 3.0%), fast-acting insulins + long-acting insulins + biguanides (regimen 7, 2.4%) and long-acting insulins (regimen 8, 1.9%). Of these, regimen 7 showed the longest mean (SD) treatment duration (66.2 [38.8] months), followed by regimen 3 (36.5 [22.1] months) during 1st LoT. The regimens used at 1st-line imply that the patients were in different stages of T2DM which is a progressive condition.

Details of 2nd-line regimens and the overall therapies used for 1st + 2nd LoT are shown in Supplementary Table 5). GLP-1 analogues and SGLT2i were used in 2.3% and 1.1% patients, respectively at 2nd LoT. Supplementary Table 6 describes the proportion of patients using other treatments (such as statins and aspirin) and screening tests.

Evolution of Risk Factors During Follow-up

The evolution of risk factors was analysed over the follow-up period and data were presented as change from baseline during each year (Table 3). After 5 years of follow-up, there was a decrease in all risk factors compared to baseline, except high-density lipoprotein (HDL), suggesting that risk of complications decreased with therapy.

Table 3 Annual changes in risk factors between baseline and first to fifth year of follow-up per patient
Full size table

The impact of the most prescribed treatments on the risk factors (HbA1c and lipids) is shown in Table 4. The maximum mean (SD) annual change from baseline HbA1c was observed with regimen 6 after first year of follow-up (− 3.2% [2.4%]) and with regimen 8 at fifth year of follow-up (− 0.68% [0.21%]).

Table 4 Annual changes in risk factors between baseline and first to fifth year of follow-up, according to most common 1st LoT regimens
Full size table

Newly Diagnosed Complications During Follow-up

Evaluation of T2DM-related complications showed that the event rates per 100 patient-years of cardiovascular, renal, ocular, foot ulcer and neuropathy complications during the minimum 5-year follow-up period for each patient were 0.142, 0.222, 1.496, 0.283 and 0.505, respectively (Table 5).

Table 5 Incidence rates of T2DM-related complications and proportion of other complications newly diagnosed during follow-up
Full size table

Cost of Treatments During Follow-up

The annual per-patient costs of all treatments (T2DM treatments and other treatments) since T2DM diagnosis were evaluated (Table 6). Diabetes drugs were used by 98.8% of patients, resulting in a mean (SD) per-patient per-year cost of 1810 (4010) SAR ($483). The other concomitant medications recommended for patients with T2DM were lipid-modifying agents (73.1%), agents acting on the renin–angiotensin system (31.8%), vitamins (26.9%) and antithrombotic agents (57.0%).

Table 6 Annual cost of all treatments (DM treatments and other treatments) since T2DM diagnosis per patient
Full size table

The proportion of patients requiring eye screening, foot care, MA and GRP tests was 61.3%, 59.6%, 2.0% and 0.8%, respectively (Supplementary Table 7). The guidelines by SNDC and the American Diabetes Association suggest performing an eye examination annually (patients with retinopathy) or at least every 3 years (without any evidence of retinopathy). Also, proteinuria screening tests are recommended to be performed annually for patients with T2DM.

Burden of Illness Analysis for T2DM Using CDM

To be used as input in the CDM, the baseline physiological parameters were stratified by the eight most used 1st-line regimens (Table 7). It was observed that patients who started biguanide alone (regimen 2) had higher BMI at baseline followed by patients who initiated biguanide with insulin (regimens 6 and 7). The efficacy data describing the impact of all therapy regimens on HbA1c and lipids are presented in Table 4. The annual treatment costs of the eight most used 1st-line regimens were calculated and shown in Supplementary Table 8.

Table 7 Baseline physiological parameters according to the most common 1st LoT regimens
Full size table

The burden of illness analysis using CDM showed that the LE and QALYs with the eight most used 1st-line regimens were 25–28 years and 18–21 years, respectively (Table 8). In addition to these health outcomes, the total lifelong disease burden varied from 201,377 to 437,371 SAR ($53,700–116,632) with the cost of CVD complications (141,973–177,466 SAR/ $37,859–47,324) being the major driver (Table 8). Besides the lifelong total cost, the model also estimated annual total costs in the range between 7368 and 16,003 SAR ($1965–4267) (Table 8).

Table 8 Projected health and cost outcomes for eight regimens of T2DM treatments using CDM
Full size table

CDM analysis of T2DM-related complications per 1000 patient-years showed highest rates of non-severe hypoglycaemia events (Supplementary Table 9).

Discussion

This RW study was the first of its kind to perform a comprehensive analysis on the patients suffering from T2DM in the Gulf countries and to project long-term burden of illness across different 1st-line therapy options. The studied subpopulation based on the availability of HbA1c data and 1st LoT information consisted of 638 patients. Of the main population (2226 patients), baseline Hb1Ac data was available for only 1053 patients (47%), as some patients might have been diagnosed at different healthcare facilities before being referred to the study centres. Among these patients, 40 patients reported HbA1c < 6.5% which is not considered diabetes. However, these patients may have been included in the main population because their diagnosis was based on FPG level, which is a less accurate measurement than HbA1c. Therefore, we defined a subpopulation based on HbA1c levels. Availability of the 1st LoT information with non-missing start date (1760 patients among the main population) was another criterion used to define the subpopulation, which helped us assess the evolution of risk factors with different types of therapy, as it was vital for performing the analysis using CDM.

The baseline patient characteristics from our study are comparable with other RW studies conducted in the KSA [14, 15, 26, 27]. The 2013 Saudi Health Interview Survey (SHIS) reported equal numbers of male and female individuals among patients with T2DM [27]. Although a 2017 study conducted at diabetes centres in three cities in the KSA showed a higher proportion of female patients (65%) [15], our study was conducted at 15 diabetes centres and therefore covered a larger number of patients across the KSA. The participants from the KSA in a global, prospective, observational DISCOVER study reported a mean age of 52.4 years and a BMI of 31.9 kg/m2 [14]. Similar baseline characteristics were obtained in a cross-sectional analysis of patients with T2DM by Alramadan et al. and the prospective DAR-MENA T2DM study including patients from Middle East and North Africa (MENA) countries [15, 28]. The higher BMI present at baseline in patients with T2DM in the current as well as previous studies re-establishes the fact that obesity is a risk factor for T2DM and therefore there is a need to implement lifestyle changes and individualised programmes to manage obesity [29]. The rather high HbA1c at baseline in the current study corroborates with previous studies in the KSA which reported HbA1c at baseline between 8.7% and 8.8% [14, 15] and suggest the necessity to raise awareness about the disease.

The rates of complications at baseline obtained in our study are comparatively lower than a few other RW studies conducted in the KSA [13, 15, 30] and in the MENA countries (including the KSA) [28]. The DISCOVER study reported 20% (microvascular) and 11% (macrovascular) rates of complications at baseline among non-insulin users initiating a 2nd LoT [13]. AlSlail and Akil reported rates of complications varying between 3.8% (nephropathy) and 12.3% (CVD) [30]. The DAR-MENA T2DM study showed baseline rates of 32.8%, 9.2% and 8.9% for neuropathy, retinopathy and nephropathy, respectively [13, 28]. These opposing results could be attributed to the differences in study design and patient eligibility criteria. Another reason is that in our study, complications were treated at other centres because diabetes centres only provide outpatient services and, therefore, not all complications were captured.

The analysis of treatment patterns showed that biguanides (monotherapy or combination) were the most common 1st LoT (81.5%). Among the most common 1st line regimens, biguanides + sulfonylureas was used by a majority of the patients as also seen with the national data from the 2013 SHIS survey [31] and the DISCOVER study [14]. The DISCOVER study also reported that 89% of patients were using biguanides as 1st LoT [14]. The overall skewed utilisation of biguanides and sulfonylureas can be partially explained by their comparatively lower cost than newer medications such as GLP-1 RA and SGLT2i [32]. Drug cost is a major factor that influences the choice of a particular agent by patients, physicians and policy makers [32]. Furthermore, GLP-1 RA and SGLT2i were approved in the KSA only after 2014 [33] and 2018 [34], respectively. Moreover, clinical inertia, which includes inhibition to initiate or intensify treatment when indicated [35], may contribute to the preferential use of biguanides and sulfonylureas observed in the current study.

Analysis of our retrospectively collected data showed that the annual per-patient cost of drugs used for T2DM was highest amongst other drugs analysed (agents for lipid modification or acting on the renin–angiotensin system etc.). The low rates of screening tests for eye and renal disease observed in our study suggest a need for improved adherence to the guidelines to avoid diabetes-related complications. The incidence of newly diagnosed T2DM-related complications during follow-up in our study ranged between 0.222 (renal complications) and 1.496 (ocular complications) per 100 patient-years across the different groups of complications. A retrospective study conducted among adult patients with T2DM in the KSA showed frequencies of newly diagnosed complication during an average follow-up of 8 years were 32% (renal), 14% (MI), 14% (neuropathy), 5–23% (ocular) and 4% (foot infections) [4].

To the best of our knowledge, this is the first study conducted in one of the Gulf countries that collected recent baseline and follow-up clinical data for patients with T2DM and provided a comprehensive overall burden of illness using CDM in the KSA. The CDM has been previously used for cost-effectiveness analyses of therapies for T2DM across KSA [36, 37] and globally [38,39,40,41]. Moreover, a systematic review of the literature between 2008 and 2013 showed that CDM is the most widely used choice among the models used for cost-effectiveness analysis of treatments for T2DM and is routinely used for reimbursement decisions and health technology assessments [42]. A number of studies have used CDM to quantify burden of illness of T2DM [43,44,45,46,47]. In the USA, CDM was used to measure the clinical and economic burden associated with poor glycaemic control due to a delay in therapy for patients with T2DM [43]. CDM has also been used in China to evaluate current and future population level burden of illness for T2DM based on the current standard of care versus hypothetical alternative management strategies [44]. However, none of these studies analysed burden of illness according to different treatment regimens as considered in the current analysis.

Our burden of illness analysis showed that the most common 1st-line treatments resulted in LE of 25 to 28 years for people diagnosed at age of 49 years. Given the LE for the general population aged 49 years is 29 to 33 years [48], patients with T2DM seem to lose 4 to 5 years. These results are confirmed by another study that suggested a loss of 5.8 to 6.8 life-years if T2DM was diagnosed at the age of 40 years [49].

When we performed the burden of illness analysis including the cost of complications from published sources, the costs for complications were the major driver of the total direct cost of T2DM. Our results also showed that the higher costs were mainly driven by CVD complications. Cost of CVD complications (40−71% of total costs) was the most expensive category among other complications as also found in a previous cost of illness study performed in Iran in 2009 [50]. It has also been estimated in the USA that costs due to CVD complications accounted for 57% of total complication cost among patients with T2DM in 2012 [51]. Our burden of illness analysis showed that besides complications, the projected drug costs were highest for regimens containing DPP4 inhibitors (37−40% of total cost), fast-acting insulin (32%) or premixed insulin (21%). In Arab countries, insulin and oral drugs (mainly metformin) have been shown to contribute to 7−13% and 9−17% of the total direct cost of diabetes, respectively [52]. The per-person cost of medication for diabetes (both type 1 and type 2) have been estimated as 600−17,000 SAR whilst the costs of visits and laboratory tests varied between 1151 and 134892 SAR in the KSA in 2014 [31]. Likewise, the study conducted in Iran reported that drug costs were the major contributor of the total national direct cost in patients without complications [50].

This study has a few limitations, which are inherent to any retrospective observational analysis. Incomplete or missing data was one of the main challenges in this study, which limited the reporting of all planned analyses of the main population with the required cohort size. The study was conducted at specialist diabetes centres in the country; demographics, risk factors and disease treatment patterns of patients attending these centres might be different from those attending other general facilities. As with all studies utilising retrospective data collection, not all data relevant to the current study were recorded in the medical records, which could have resulted in information bias. The information pertaining to treatment (drug doses, start and end dates) could not be completely captured from patient charts which may have led to underestimation of costs of diabetes-related drugs. Furthermore, there is a likelihood of patients attending other general facilities for T2DM-related complications and hence the diabetes centres may not have captured the complete patient data. As a result, the cost of complications to be input in the CDM could not be collected and these had to be sourced from published articles. For analysis of burden of illness, we only included direct medical costs and did not evaluate indirect costs.

Conclusions

This RW study addresses gaps in the current research by providing a complete landscape of baseline demographic and clinical characteristics of patients with T2DM, their treatment patterns, associated costs (resource use), T2DM-related complications and clinical impact of different therapy regimens in the KSA. Additionally, the burden of illness analysis using CDM in this heterogenous group of patients showed substantially higher cost of T2DM care from a payer perspective in the KSA, with the cost of cardiovascular disease complications being the major driver.