Economic burden of type-2 diabetes in Peru: a cost-of-illness study valuing cost differences associated with the level of glycemic control

ABSTRACT

Objectives

Type 2 diabetes mellitus (T2DM) represents an increasing public health problem in Peru. This study aims to estimate the national economic burden of this disease for the public funder, the social security, and private sector insurers.

Methods

Direct healthcare costs were estimated for a cohort of 45-to-75-year-old adults diagnosed with T2DM in 2019, over a 20-year period. Disease progression was modeled using PROSIT Models and literature, including acute and chronic microvascular and macrovascular complications. Three scenarios of glycemic control were considered: current levels of 35.8% of the population controlled (HbA1c < 7%) (S1); 100% controlled (S2) and; 100% uncontrolled (S3). The impact of diabetes prevalence on overall costs was evaluated in sensitivity analysis.

Results

Total national economic burden was estimated at $15,405,448,731; an annual average per patient of $2,158. Total costs would decrease to $12,853,113,596 (−16.6%) in S2 and increase to $16,828,713,495 (+9.2%) in S3. Treating patients with complications and risk factors could cost 6.5 times more, being stroke the complication with the highest impact. Up to a 67.6% increase in total costs was found when increasing T2DM prevalence.

Conclusions

T2DM places a heavy burden on the Peruvian healthcare budget that will be even greater if poor glycemic control is maintained.

KEYWORDS:

• Type-2 diabetes mellitus
• Peru
• cost analysis
• economic burden
• glycemic control level
• diabetes complications

1. Introduction

Diabetes mellitus (DM) is a non-communicable chronic disease that represents a significant burden in terms of morbidity, disability and mortality and, therefore, it has a considerable impact on the costs of health finances and society in general. According to the International Diabetes Federation (IDF), in 2021, around 10.5% of people worldwide suffered from this disease; more than triple the estimate in the year 2000. This implied a total expenditure of approximately $966 billion. If the trend continues, it is expected that this disease will affect 12.2% of the population in 2045, at a cost of $1,054 billion [1].

Peru is not an exception. According to the Demographic and Family Health Survey (Endes), 3.9% of Peruvians older than 15 years old were diagnosed with DM in 2019, a percentage that increased to 4.5% and 4.9% in 2020 and 2021, respectively [2]. Furthermore, other investigations carried out in the country, based on clinical criteria and not self-reports, estimated that the prevalence of this disease is 7%; being 96.8% cases of type 2 diabetes [3].

Compared with type 1 diabetes mellitus (T1DM), the onset of T2DM is silent; that is, it may not generate symptoms. Therefore, patients are usually diagnosed late, when they already have complications [1]. In addition, in many cases, diagnosed patients do not follow adequate treatment, which increases the risk of complications in the long term [4]. According to a systematic review conducted in the country, less than 30% of diabetic patients presented a glycosylated hemoglobin (HbA1c) level < 7%, the optimal control standard at the international level [3]. Similarly, the country’s Diabetes Epidemiological Surveillance System reported 35.8% of diabetic patients with optimal glycemic control in 2019 [5].

Consequently, diabetes is the sixth leading cause of blindness in Peru and the leading cause of chronic kidney disease and non-traumatic amputation of limbs. Likewise, it is associated with 31.5% of acute myocardial infarctions and between 9.7% and 15.5% of strokes [3]. For this reason, from a budget perspective, diabetes can represent a high cost for the Peruvian health system; yet, few studies in the country have evaluated its impact.

This study aims to estimate the economic burden associated with T2DM in the country, focusing on key health funders, including the public funder (Public Health Insurance-SIS, Ministry of Health-MoH and Regional Governments), the Social Health Insurance (Seguro Social de Salud, Essalud) and private Health Insurance Companies (Entidades Prestadoras de Salud, EPS). The estimation encompassed direct healthcare costs over a 20-year period, considering different scenarios of glycemic control, with sensitivity analysis conducted on overall results to assess prevalence rate variations.

2. Methods

2.1. Target population and time horizon

A population cohort of 443,683 adults between 45 and 75 years of age diagnosed with T2DM (5.9% prevalence) in 2019 was evaluated over a 20-year period. The age range selected is in which the Peruvian diabetic population is concentrated and in which complications begin to be seen.

The diabetic population was identified using healthcare service records reported by the National Health Superintendency (Superintendencia Nacional de Salud, Susalud) for each health subsystem: public subsystem, Essalud and the five private EPS that operate in the country [6]. All individuals who registered at least one diagnosis of diabetes (International Classification of Diseases, Tenth Revision – ICD-10 code from E-10 to E-14, according to the latest review published by the World Health Organization – WHO) [7] were included in the study: 242,702 in the public subsystem (54.7% of the total population), 183,878 in Essalud (41.4%) and 17,103 affiliates to private insurers (3.9%).

2.2. Glycemic control scenarios

Clinical evolution of patients with T2DM depends to a large extent on their ability to adhere to treatment and their level of control of the disease from the beginning [8]. For this reason, in addition to the current scenario, two hypothetical scenarios were evaluated to visualize the impact on costs of presenting less or greater glycemic control.

The scenarios evaluated were the following:

• Baseline current scenario (S1): According to the Peruvian Diabetes Epidemiological Surveillance System, 64.2% of the population with suboptimal glycemic control (HbA1c > 7%) and; 35.8%, with optimal control (HbA1c < 7%) during the entire time landscape [5].

• Scenario with optimal glycemic control (S2): Current scenario for the first year and; 100% of the population with HbA1c < 7%, from the second year of analysis.

• Scenario with suboptimal glycemic control (S3): Current scenario for the first year and; 100% of the population with HbA1c > 7%, from the second year of analysis.

As the models employed for disease progression projection required an exact value of Hba1c rather than a reference range, a level of 6.9% was utilized for the optimal glycemic control cases and a level of 9.1% for the suboptimal control cases.

2.3. Clinical variants

Four general clinical profiles for a patient with diabetes were examined: (i) T2DM without complications and risk factors; (ii) T2DM without complications and at least one risk factor, (iii) T2DM without risk factors and at least one complication, and (iv) T2DM with at least one risk factor and at least one complication. Risk factors and complications under study were defined by taking as a reference the ICD-10 classification published by the WHO, the version currently used in the country (Table 1).

Table 1. Risk factors and complications associated with T2DM under study.

Group	Subgroup	ICD-10 Code
DM		(E10-E14).9
Risk Factor	Hypertension	E10-E14 + I10
	Obesity	E10-E14 + E65-E68
	Dyslipidemia	E10-E14 + E78.0-E78.5
Acute complication	Hypoglycemia^1,2	(E10-E14).0
	Hyperglycemia^1,2	(E10-E14).0
	Ketoacidosis	(E10-E14).1
Microvascular complication	Nephropathy^3	(E10-E14).2
	Retinopathy	(E10-E14).3
	Neuropathy	(E10-E14).4 + (G99.0)^4
Macrovascular complication	Stroke – Hemorrhagic	(E10-E14).5 + I60-I62
	Stroke – Ischemic^5	(E10-E14).5 + I65-I66
	Stroke – Other	(E10-E14).5 + I63-I64/I67-I69
	Heart failure	(E10-E14).5 + I50
	Coronary heart disease (Angina and myocardial infarction)	(E10-E14).5 + I20-I22
	Peripheral vascular disease (PVD)	(E10-E14).5
Microvascular and macrovascular complication	Diabetic foot (with and without amputation)^6	(E10-E14).4 + (G99.0)^4 + E10-E14.5

DM: Diabetes mellitus, ICD-10: International Classification of Diseases, Tenth Revision.

1/It includes only severe cases that require medical attention in a health facility.

2/Given that the same ICD-10 coding is assigned to cases of hypoglycemia and hyperglycemia, the proportion of the total cases reported by Susalud corresponding to each complication was approximated using the paper by Amiel S.A. et al., which was used in the PROSIT model of hypoglycemia [9].

3/It includes microalbuminuria and chronic kidney disease. Cases requiring kidney transplantation are not considered.

4/Although the ICD-10 autonomic neuropathy registry code (G99.0) was included during the initial design, given its low frequency compared to other complications, the prioritized variants did not include these cases.

5/Although the ICD-10 registration codes for ischemic stroke were included during the initial design, given their low frequency compared to other complications, the prioritized variants did not include these cases.

6/All patients having neuropathy and PVD were considered to have diabetic foot. The proportion of patients amputated from this total was calculated taking as a reference the paper by Rodrigues B. et al. [10] and the proportion of patients amputated reported in the SIS records for Metropolitan Lima and other representative regions (Lima provinces, Arequipa, Junín, Piura and Ucayali) in 2019.

Considering the predominant possible combinations at the national level, according to what was reported in the Susalud 2019 healthcare service records, a total of 192 specific clinical variants associated with the disease were defined. Additionally, 4 clinical variants of severe cases that lead to death were added when the diabetic patient suffered from the following complications: (i) myocardial infarction, (ii) nephropathy, (iii) stroke and (iv) septicemia associated with diabetic foot; and 1 additional variant that brought together cases of death due to other clinical conditions.

In total, 197 variants of clinical presentation were obtained. Each of them with a particular resource use profile, according to the corresponding signs and symptoms (Supplementary Table S1).

2.4. Population by clinical variant

For the base year, the population by clinical variant was identified based on the diagnoses of complications and risk factors registered for the target population in the 2019 Susalud databases. Additionally, the number of severe cases resulting in death was approximated based on data reported by the National Death Information System (Sistema Informático Nacional de Defunciones, Sinadef) in 2019.

For the following years, the progression of the diabetes population toward hypoglycemia, nephropathy, retinopathy, stroke, coronary disease and diabetic foot complications was estimated using the open-access PROSIT diabetes modeling tool, which is based on Markov models [11]. Demographic characteristics and parameters related to baseline health conditions and lifestyle in the models were adjusted to the Peruvian context according to 2019 Susalud records, national surveys [12,13], records of medical procedures for SIS affiliates in 2019 and literature for the country [10,14–20].

For the other complications, for which there was no PROSIT model, alternative estimates were made whenever the available information allowed it. Cases of heart failure were projected using data described in the scientific literature regarding its prevalence and survival probability by age and sex [21–23]. Based on these parameters, a Monte Carlo simulation with 10,000 iterations was performed. Cases of neuropathy and peripheral vascular disease (PVD) were calculated based on the behavior of the population with diabetic foot that did not end in amputation, which was estimated as a complementary result of the amputees’ cases projected using the PROSIT amputation model. For hyperglycemia and ketoacidosis, it was assumed that cases would remain constant due to the lack of scientific evidence to make consistent estimates for the country.

Deaths associated with a severe case of myocardial infarction, stroke, nephropathy or diabetic foot, were projected using the PROSIT corresponding to each complication. Meanwhile, deaths from other causes were estimated from the PROSIT of prediabetes, which determined the transition of the average population with T2DM toward their state of death. Based on the literature, PROSIT models accounted for the additional risk of death in the diabetic population and, in the case of complications, included also the additional risk given the severity of the clinical state. All models were adjusted to the Peruvian case using the general mortality table for Peru estimated by the United Nations for the 2015–2020 period [24].

2.5. Costs estimates

Direct costs associated with T2DM were estimated considering the differences between the three health subsystems under study. For the public funder and Essalud, cost estimations were carried out using a standard micro-costing approach, where medical procedures required for managing each clinical variant were identified based on national Clinical Practice Guidelines (CPGs) and international treatment standards, and then its costs estimated following the standard cost methodology established by the MoH [25]. This considered the costs of healthcare personnel using remuneration regulations for each entity; infrastructure costs obtained from public investment projects reported by the Ministry of Economics and Finance (MoF); equipment and medical supplies costs based on public procurement information reported by the State Contracts Supervisory Body (Organismo Supervisor de las Contrataciones del Estado – OSCE), MoH, SIS and Essalud; and utilities and overheads and administrative costs reported by the MoF. In addition, medication costs were included using Information on drug purchases by SIS and MoH.

For private health insurance companies (EPS), given the lack of detailed public information on costs, a ratio reflecting the relationship between the costs of the public financier and the private sector was developed, using the general expenditure information reported by the EPS to Susalud for a reference group of clinical variants [26]. The resulting ratio was then applied to the costs of the public funder for each clinical variant.

Clinical variants costs were calculated for the base year (2019) and projected throughout the study time horizon, considering the inflation estimated by the MoF in the update report on macroeconomic projections 2021–2024 [27]. The total cost of each scenario was estimated based on the number of cases per clinical variant, discounted at 3.5% per year. All costs are reported in US dollars (USD, $), using an exchange rate equivalent to $1 per S/3,355 (PEN, Peruvian soles), as reported by the Central Reserve Bank of Peru for the base year of the estimates [28].

2.6. Sensitivity analysis

A sensitivity analysis was carried out to assess the effect of adjusting the prevalence rate of T2DM in Peru considering the potential underreporting in official data. The analysis considers a prevalence variation of + 0.5%, until reaching a prevalence level of 9.9%. This range includes the different levels of prevalence described by previous studies carried out in the country (around 7%) [29–31], and the South and Central America region prevalence (9.4%) reported by the IDF [32].

3. Results

3.1. Total economic burden

Maintaining the current levels of glycemic control in the country (S1), the economic burden of T2DM during the 20 years of analysis was estimated at $15,405,448,731. The greatest financial burden would be assumed by Essalud, closely followed by the public funder. Calculated costs for these subsystems represented 44.7% and 42.7% of the estimated total cost, respectively; while private insurers accounted for 12.7% (Figure 1). This composition of spending was mainly based on the larger population concentrated in the first two health subsystems.

Figure 1. Total economic burden associated with T2DM by health subsystem, according to glycemic control scenario (billions of USD).

S1: Scenario 1, S2: Scenario 2, S3: Scenario 3. Note: Costs exhibited indicate the present value over the 20-year time horizon.

The estimates show an increasing trend in costs, reaching the maximum cost ($1,110 million) in the last year of the time horizon. This behavior reflects that the cost of treating diabetes is closely linked to the progression of the disease over the years. As the population with T2DM transitions to more severe health states, a greater amount of health services is demanded, as well as more complex ones. As a consequence, treatment costs increase.

3.2. Costs according to glycemic control level

The total economic burden estimated would vary according to the level of glycemic control. If the total diabetic population managed to adhere to the prescribed medical treatment and thus maintain HbA1C levels within the suggested optimal range < 7% (S2), the estimated total cost for the 20 years of analysis would result in $12,853,113,596 (−16.6% compared to the baseline scenario). If the control level of the patients were suboptimal, with all the patients presenting HbA1c levels > 7% (S3), the economic burden would amount to $16,828,713,495 (+9.2% compared to the baseline scenario). Regardless of the level of glycemic control, the greatest burden would be assumed by the public funder and Essalud (Figure 1).

The cost trajectory of the suboptimal glycemic control scenario (S3) would be similar to that of the baseline scenario, consistently increasing. In this scenario, the average annual cost would be $1,156 million; cost that would reach its maximum level in year 20, with $1,224 million. In contrast, in the optimal glycemic control scenario (S2), the estimated average annual cost is $878 million, reaching its highest level in year 18, with $907 million. This reflects that the economic burden would be reduced earlier in scenarios of greater control.

3.3. Costs by clinical profile

The current annual average cost of treating a diabetic patient in the country was estimated at $1,715, $2,296 and $6,970 for the public funder, Essalud and private insurers, respectively. On average, these amounts could increase by 9.1% if the patient had elevated HbA1c levels and decrease by 16.4% if glycemic control was adequate.

Estimated costs vary depending on the clinical profile of each patient. Treatment of diabetic patients with complications and risk factors would be more expensive than for those without additional diagnoses (Figure 2). For the base scenario (S1), the annual average cost of treating a patient with complications and without risk factors could represent up to five times (4.43–5.93; according to the subsystem) more than the cost of a diabetic patient without complications. This would increase to 6.5 times (5.68–7.42; depending on the subsystem) if the patient had risk factors. The trend was similar for all scenarios.

Figure 2. Average annual cost per patient with T2DM according to clinical status and health subsystem.

The cost for the alternative scenarios is presented between brackets. The lower limit presents the cost in S2 (scenario with optimal glycemic control) and; the upper limit, the cost of S3 (scenario with suboptimal glycemic control).

3.4. Costs by complication

Total costs broken down by complication and health subsystem are presented in Figure 3. In the baseline scenario (S1), stroke is the complication that would impose the greatest economic burden on the country across all subsystems. Over the 20 years, it would cost $1,502 million to the public funder, $1,026 million to Essalud, and $338 to private insurers; due to the high cost of medical attention, especially invasive procedures that require highly specialized supplies. In the base year, costs to treat patients with this complication ranged between $22,690 (optimal control) and $23,089 (sub-optimal control), for the public funder; between $23,911 (optimal control) and $24,668 (sub-optimal control), for Essalud; and between $97,676 (optimal control) and $99,394 (sub-optimal control), for private insurers (Supplementary Table S2).

Figure 3. Total economic burden ranking by complication, according to health subsystem (millions of USD).

Costs for the three scenarios are shown in the brackets [S1; S2; S3] and indicate the present value over the 20-year time horizon.

Other complications that could have a significant impact on the financial burden of both the public funder and Essalud because of their high incidence are PVD, diabetic nephropathy, and diabetic neuropathy. Whereas for private insurers, hyperglycemia was the most frequent among patients and therefore contributed considerably to the economic burden.

3.5. Sensitivity analysis

The estimated costs increased significantly as the level of prevalence of T2DM increased (Figure 4). If a prevalence level of 9.9% is considered, the total cost calculated for the 20 years of analysis would increase to $25,825,714,350 in the current scenario (S1); to $21,546,976,400 in S2; and to $28,211,677,257 in S3. This indicates that depending on the level of underreporting in the official diabetes figures for the country, the costs reported in this study could increase by up to 67.6%.

Figure 4. Sensitivity analysis – total economic burden according to the level of T2DM prevalence (billions of USD).

S1: Scenario 1, S2: Scenario 2, S3: Scenario 3. Note: Costs exhibited indicate the present value over the 20-year time horizon.

4. Discussion

4.1. Costs implications

This study broadens the frontier of knowledge of diabetes in Peru by projecting the direct costs associated with this disease for the three largest health financiers at the national level and deepening the analysis according to the level of glycemic control of diabetic patients.

Among the main findings is that this chronic condition places a significant economic burden on the country: $15 billion over the 20 years of analysis. Specifically, for the public funder, estimated costs during the base year (2019) represented 7% of the total public budget allocated to the health sector in that year [33]. These results are similar to those reported by the IDF who estimated a total cost of $1,572 million and a cost per person of $1,135 in Peru in 2019, compared to $958 million and $2,158 estimated in this study [32]. The differences possibly lie in the different methodological approaches and the population evaluated, which in the case of the IDF covers a greater age range (20 to 79 years).

Furthermore, costs would be amplified if glycemic control levels in the population deteriorate, mainly due to the greater predisposition of the diabetic population to develop complications when they do not follow adequate treatment from the beginning. These cost differences are widely documented in previous research conducted in both Latin American and European countries. In Mexico, it was estimated that the average annual cost per patient was $3,193.8. When the patient did not present complications, the cost decreased to $2,740.3 and; when complications were present, it rose to $3,550.2 [34]. The SECCAID study calculated that the annual total direct cost of treating the diabetic population in Spain was EUR 5,809 million, of which EUR 2,143 million was mainly due to complications. This implied that the public funder allocated around EUR 1,770 annually per diabetic patient [35].

Unfortunately, the majority of the Peruvian diabetic population (around 70%) presents suboptimal control, with HbA1c levels > 7% [3,5]. In addition, many are referred late for evaluation of potential complications. Diabetic retinopathy occurs in 1 out of 4 patients and about 20% in advanced degrees [36]. In another study carried out in a public hospital in 2011, of the total number of patients referred for renal evaluation, 68.49% were in stage 4 or 5; the most severe [3].

The current situation represents a serious public health problem for the country, which, if not addressed promptly, could worsen in the coming years, demanding a greater number of budgetary resources. Therefore, prevention and early treatment of diabetes is vital. From the first level of care, it is necessary to reinforce monitoring through regular checkups, to reduce cardiovascular risks and other risk factors, as well as detect eye, kidney and foot damage. In this way, any indication of a complication can be evaluated and treated in time [36,37]. Additionally, policymakers should carefully assess the diverse array of diabetes medications available, from commonly used insulins and hypoglycemic agents to newer, higher-cost technologies. Considering evidence regarding safety and cost-effectiveness across different populations, potentially beneficial interventions should be incorporated into the disease care protocols for the country.

4.2. Impact on life expectancy and life quality

In Peru, diabetes mellitus caused the loss of 289,449 years of healthy life (Disability-adjusted life years – DALYs) in 2019, which translates into 9 years per thousand inhabitants. Of these, 65.7% is attributed to disability (Years of life lost due to disability – YLDs); while the remaining 34.3%, to premature death (Years of life lost due to premature death – YLLs) [38].

Likewise, diabetic patients report having a lower quality of life, measured in Quality-adjusted life years (QALYs). The literature reports that the loss of utility associated with the disease is −0.215, which could increase, on average, to −0.231, if the patient also presents risk factors such as dyslipidemia, hypertension, and obesity. Furthermore, these numbers increase as the patient’s state of health worsens. On average, acute, microvascular, and macrovascular complications generate an additional decrease in the quality of life of −0.035, −0.084, and −0.096, respectively. Diabetic foot amputation represents the complication with the highest additional burden (−0.28) [39–41].

Based on the figures reported in the literature, under the current glycemic control scenario (S1), a loss at the population level of 1,990,007 QALYs was estimated for our country during the study time horizon (including loss of quality of life and premature mortality). This could increase to 2,025,348 in a suboptimal control scenario (S3), losing 35,341 more QALYs; and decrease to 1,926,629 in an optimal control scenario (S2), gaining 63,378 QALYs. Consequently, the cost per QALY in the scenario with suboptimal control ($3,631) was estimated to be 1.10 times higher than the current scenario ($3,306) and 1.33 times higher than the optimal scenario ($2,731) (Figure 5).

Figure 5. Annual cost per QALY (thousands of USD).

PV: Present value, S1: Scenario 1, S2: Scenario 2, S3: Scenario 3.

4.3. Limitations

There are a number of limitations in our analysis. The prevalence of diabetes may be underestimated as diabetes records in Susalud databases did not account for the population unaware of their condition or who did not attend to receive medical care in 2019. To overcome this limitation, a sensitivity analysis was conducted to acknowledge the variation in total costs associated with higher prevalence rates, considering findings from other Peruvian and regional studies.

The PROSIT models used for case projection were developed for a European context and adjustments were only partially feasible, to the extent that available information for Peru allowed it. Key demographic, clinical and lifestyle characteristics, representing the main variables that affect the patient´s disease progression, were modified to reflect the behavior of the Peruvian population, but the transition probabilities remained unaltered. Also, PROSIT models for each complication were independent, so their results were not taken directly but rather used as a reference to determine the trends in the clinical evolution of diabetic patients, and then applied to the target population identified for the base year by clinical variant. Furthermore, not all complications under analysis were modeled by a PROSIT. Yet, these gaps were reduced to some extent by seeking alternative literature and adjusting it to the Peruvian setting, as exemplified in heart failure case projection. For the remaining complications, assumptions were made. Variations in the level of glycemic control could only be made for those complications modeled by a PROSIT. The behavior of the population for the other complications was assumed constant across glycemic scenarios. Likewise, only the HbA1c level was varied within the models. The other parameters were kept constant since there was no exact information for the country of the association between these variables and HbA1c improvement or deterioration.

Concerning cost estimations, the methodology used assumed medical care under optimal treatment standards as outlined by the MoH given the absence of an actual operational cost system in the country, so the results could differ from the actual expenses observed. Moreover, the standard therapeutic management in the country, following the Peruvian CPG, and the most used drugs, covered by each financer’s List of Essential Medicines was considered. Thus, typical treatment with insulins and hypoglycemic agents was assumed, and a potential increase in costs should be noted as new technologies are incorporated into disease care protocols in the country. Further, the cost of clinical variants across glycemic control scenarios only varied based on the different frequency of use of health services. However, since the type of pharmacological treatment of this disease can vary depending on the clinical characteristics of each patient, the amount and frequency of drug administration was assumed to be constant between scenarios.

Finally, this study does not delve into the estimation of the indirect social costs associated with the disease, such as the loss of labor productivity due to absenteeism, presenteeism, inability to work, and premature death. The additional costs for the State and the patients generated by the disability due to the disease are not analyzed either. Future studies are required to complement the analysis carried out in the present investigation.

5. Conclusions

Type-2 diabetes mellitus is currently causing a significant impact on the Peruvian healthcare budget. Furthermore, this burden will increase if the current levels of treatment adherence and glycemic control persist unchanged, so addressing the disease at early stages should be a paramount concern for healthcare systems. By implementing a comprehensive strategy that seeks to prompt prevention, as well as detect and manage diabetes and its risk factors timely, not only the economic burden would be alleviated but also significantly enhance the overall quality of life of the Peruvian population.

Abbreviations

CPG	=	Clinical Practice Guidelines
DALYs	=	Disability-adjusted life years
DM	=	Diabetes mellitus
Endes	=	Demographic and Family Health Survey
EPS	=	Private Health Insurance Companies
Essalud	=	Social Health Insurance
HbA1c	=	Glycosylated hemoglobin
ICD	=	International Classification of Diseases
IDF	=	International Diabetes Federation
MoF	=	Ministry of Economics and Finance
MoH	=	Ministry of Health
OSCE	=	State Contracts Supervisory Body
PVD	=	Peripheral vascular disease
QALYs	=	Quality-adjusted life years
Sinadef	=	National Death Information System
SIS	=	Public Health Insurance
Susalud	=	National Health Superintendency
T1DM	=	Type 1 diabetes mellitus
T2DM	=	Type 2 diabetes mellitus
WHO	=	World Health Organization
YLDs	=	Years of life lost due to disability
YLLs	=	Years of life lost due to premature death

Declaration of interest

J Seinfeld, A Sobrevilla, M Rosales, M Ibáñez, D Ruiz and E Penny are employees of Videnza Consultores who were contracted by Sanofi to conduct this research. S Londoño is a Sanofi employee and may hold shares and/or stock options in the company. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Author contributions

J Seinfeld, A Sobrevilla and S Londoño guided the conceptual design and interpretation of the study. M Rosales, M Ibáñez and D Ruiz contributed to the design of the methods, collected the data and performed the population projection and cost estimations and analysis. A Sobrevilla and E Penny contributed to the medical design and analysis. All authors participated in the drafting and revision of this report and approved the final version for publication.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

A poster displaying the main results of the research was presented at the ISPOR Europe 2023 Conference, which took place in Copenhagen, Denmark, from November 12th to November 15^th. Translation assistance was provided by Lance Venter and was funded by Sanofi. Moreover, the authors would like to thank the Sanofi study team for the support during the conduct of this study and Liliana Silva for editorial assistance.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14737167.2024.2333337.

Additional information

Funding

The study was funded by Sanofi.