Cost-Utility Analysis of High-Intensity Statins; Atorvastatin 40 and 80 mg versus Rosuvastatin 20 mg from Egyptian Public Payer’s Perspective: A Markov Model

Abstract

Background:

Dyslipidemia is a significant burden correlated with the risk of developing cardiovascular morbidity and mortality. However, no available studies compare the cost-utility analysis of high-intensity statins for managing Egyptian patients with dyslipidemia. The study aimed to measure the cost-utility of different high-intensity statins in managing patients with dyslipidemia within the Egyptian public sector.

Methods:

A half-cycle corrected Markov model with six mutually exclusive health states was developed to reflect the real practice of managing patients with dyslipidemia upon the physician’s decision to use atorvastatin 40 and 80 mg vs. rosuvastatin 20 mg. A 1-year cycle length was applied, and a 10-year time horizon was selected to reflect the long-term consequences.

Results:

Over the 10 years horizon, atorvastatin 40 mg was associated with the lowest annual cost per percent LDL-C reduction, followed by rosuvastatin 20 mg and atorvastatin 80 mg. Atorvastatin 40 mg generated a total of 4.152 quality-adjusted life years (QALYs), with incremental 0.0012 QALYs, versus rosuvastatin 20 mg, with savings of 3,245 egyptian pounds (EGP). Atorvastatin 80 mg produced a total of 4,159 QALYs, with an incremental 0.0084 QALYs, versus rosuvastatin 20 mg at an incremental cost of 435 EGPs, representing a more cost-effective option under the willingness-to-pay (WTP) threshold in Egypt of 139,000 EGPs.

Conclusions:

Atorvastatin 40 mg was shown to be strongly dominant over rosuvastatin 20 mg, with numerically higher QALY outcomes and lower associated costs. Atorvastatin 80 mg is considered a cost-effective option versus rosuvastatin 20 mg, with an incremental cost-utility ratio (ICUR) value significantly below the WTP threshold.

Introduction

Cardiovascular diseases (CVDs) are the leading cause of death worldwide, with 17.9 million deaths in 2019 (32% of all fatalities), as per the World Health Organization (WHO) estimates.[1] Heart attack and stroke caused 85% of these losses of life. Moreover, low- and middle-income countries account for more than 75% of all CVD deaths.[2] The economic burden of heart disease in the United States was valued at about $219 billion each year, including the cost of medications, health care services, and premature death.[3] A well-known causal risk factor for CVDs such as stroke, myocardial infarction (MI), and coronary heart disease (CHD) is dyslipidemia, a cluster of lipid abnormalities that include high levels of low-density lipoprotein cholesterol (LDL-C), total cholesterol (TC), triglycerides (TG), and low levels of high-density lipoprotein cholesterol (HDL-C).[4] Dyslipidemia is closely related to an increased risk of mortality, imposing substantial costs on the healthcare system.[5,6,7] Analyzed data from the Global Burden of Disease study revealed rising numbers of deaths and disability-adjusted life years (DALYs) attributed to high LDL-C (1.47 and 1.41 times higher in 2019 compared to 1990, respectively).[8] Controlling lipid levels is thus a crucial step in reducing CVDs,[9,10] and the guidelines recommend treatment with high-intensity statins (atorvastatin 40 and 80 mg, and rosuvastatin 20 and 40 mg) up to the highest tolerated dose to reach the designated goals for LDL-C, corresponding to different risk groups.[4] It is reported that atorvastatin 40 and 80 mg, and rosuvastatin 20 and 40 mg have the ability to reduce LDL-C by at least 50%.[11] Several clinical trials have illustrated that atorvastatin therapy significantly reduced cardiovascular events in CHD patients.[12,13,14] The MERCURY I and II trials[15,16] and the STELLAR trial[17] have demonstrated that rosuvastatin had a potent role not only in reducing LDL-C but also in raising HDL-C.

In Egypt, non-communicable diseases (NCDs) account for 84% of all deaths, of which 40% are due to CVDs.[18] Different figures were reported for the prevalence of dyslipidemia in the general Egyptian population, ranging from 19.2 to 36.8%.[19] Further, ischemic heart disease is ranked first^* as the most common cause of death in Egypt.[20] The pattern of dyslipidemia was previously described in Egyptian patients with the acute coronary syndrome (ACS), where low levels of HDL-C (<40 mg/dL) were found in 66%, and high levels of TG (>150 mg/dl), lipoprotein (a) (>30 mg/dL), TC (>200 mg/dL), and LDL-C (>130 mg/dL) were found in 63.33%, 62%, 60.67%, and 58%, respectively.[21]

Egypt’s healthcare system has struggled for decades to address the public health issue.[22,23] However, extensive efforts have been exerted in recent years to tackle this problem. Through “The Universal Health Coverage Project,” the Egyptian government has prioritized providing health services for all its inhabitants, making coverage essential and eradicating existing inequalities.[24] Moreover, the healthcare system is going through significant changes in Egypt due to the formation of the Unified Procurement Authority (UPA) as a centralized procurement and supply interface. Through evidence-based technology assessments and value-driven procurement techniques, this body aims to promote fair access to pharmaceuticals and health technology products.[25] This nascent landscape in Egypt attempts to endorse the most cost-effective options for treating patients, especially those with chronic diseases such as dyslipidemia. Thus, the objective of this study was to evaluate the cost-utility of branded atorvastatin 40 and 80 mg versus branded rosuvastatin 20 mg for managing patients with dyslipidemia within the Egyptian public sector from the ministry of health (MOH) payer’s perspective. This evidence will hopefully support the decision-making process in Egypt for better allocation of scarce healthcare resources.

Methods

Model overview

A standard half-cycle corrected Markov cohort model with six mutually exclusive health states was developed to reflect the real practice of managing Egyptian patients with dyslipidemia upon the physician’s decision to use high-intensity statins, and risks repeatedly change over time [Figure 1]. It was designed to demonstrate the cost-utility of branded atorvastatin 40 and 80 mg vs. branded rosuvastatin 20 mg as the high-intensity statins in the Egyptian markets. A hypothetical cohort of patients suffering from high LDL-C levels, with or without a prior history of major cardiovascular events (MCVEs), was considered. The hypothetical cohort characteristics were drawn from the Cholesterol Treatment Trialists’ (CTT) Collaboration meta-analysis.[26] A 10-year time horizon was selected to reflect the long-term consequences. The Markov model was populated following the Egyptian MOH payer’s perspective using a 1-year cycle length. All costs and outcomes were discounted at 3.5% annually.[27] The analysis was performed using Excel Microsoft 365 (Microsoft, Redmond, WA, 2020).

Figure 1.

A Schematic diagram for the Markov state-transition

Model inputs

Likelihood of events

As shown in Table 1, patients (either with a history of MCVE or not) entered the model by initiating the high-intensity statin therapy upon the treating physician’s decision. An MCVE was defined as any major coronary event, coronary revascularization (angioplasty or bypass grafting), or stroke.[26] Transitions between health states were modeled to occur at 1-year intervals, and the model transitions patients between states of ‘No cardiovascular (CV) Event’, ‘Post-CV Event’, and ‘Death’. Each transition is based on probabilities that depend solely on the current state, LDL levels, and time from the first or previous event. The baseline transition probabilities (i.e., the risk of MCVE and death) were based on the average starting age and mean LDL-C level of the cohort. The cohort mean age and treatment efficacy were obtained from the meta-analysis conducted by Law et al.[28] The authors used 164 randomized placebo-controlled trials of six statins to determine the reduction in serum concentrations of LDL-C and incidence of ischemic heart disease (IHD) events and stroke, according to drug, dose, and duration of treatment. Treatment efficacy was modeled as the percent reduction in LDL-C level from baseline for each treatment. Reduction in MCVE risk was calculated in terms of the percent of risk decrease per each mmol/L decrease in LDL-C.[28]

Table 1.

Model inputs and data sources for the cost-utility analysis

Item Labels	Item Value	Low Margin	High Margin	Source
Cohort Characteristics
Cohort average age (years)	55	43	71	[28]
Baseline LDL-C level (mg/dL) primary prevention	137	123.3	150.7	[26]
Baseline LDL-C level (mg/dL) secondary prevention	143	128.7	157.3	[26]
Egypt adult mortality (55-64 years)	0.082	0.074	0.09	[33]
Baseline percentage of MCVE in cohort	4	3.6	4.4	[26]
Treatment Efficacy
Atorvastatin 40 mg LDL-C reduction %	49%	44.166%	53.958%	[28]
Atorvastatin 80 mg LDL-C reduction %	55%	48.130%	61.670%	[28]
Rosuvastatin 20 mg LDL-C reduction %	48%	45.833%	50.833%	[28]
MCVE Risk
Primary prevention untreated	1.83%	1.65%	2.01%	[26]
Primary prevention treated with atorvastatin 40 mg	1.16%	1.09%	1.23%	Calc.a
Primary prevention treated with atorvastatin 80 mg	1.08%	0.99%	1.17%	Calc.a
Primary prevention treated with rosuvastatin 20 mg	1.18%	1.14%	1.21%	Calc.a
Secondary prevention untreated	5.68%	5.11%	6.25%	[26]
Secondary prevention treated with atorvastatin 40 mg	3.52%	3.30%	3.73%	Calc.a
Secondary prevention treated with atorvastatin 80 mg	3.25%	2.96%	3.55%	Calc.a
Secondary prevention treated with rosuvastatin 20 mg	3.56%	3.43%	3.66%	Calc.a
Secondary prevention untreated (2+ events)	18.01%	16.21%	19.81%	[26]
Secondary prevention treated (2+ events) with atorvastatin 40 mg	11.15%	10.45%	11.82%	Calc.a
Secondary prevention treated (2+ events) with atorvastatin 80 mg	10.31%	9.37%	11.27%	Calc.a
Secondary prevention treated (2+ events) with rosuvastatin 20 mg	11.29%	10.89%	11.59%	Calc.a
MCVE hazard ratio of 1 mmol/L drop in LDL-C level	0.210	0.189	0.231	[26]
Statin Treatment Cost (EGP) (q.d.)b
Atorvastatin 40 mg price (7 tabs)	75	67.5	82.5	AP
Atorvastatin 80 mg price (7 tabs)	89.5	80.6	98.5	AP
Rosuvastatin 20 mg price (7 tabs)	87.5	78.8	96.3	AP
MCVE Costs (EGP)
Average MCVE cost (per event)	21,553	17,242	25,864	Egy OSF
Utilities
Age-related utilities	0.841	0.7569	0.9251	[30]
Primary prevention first event disutilities	-0.0336	-0.0302	-0.0370	[30]
Secondary prevention disutilities
Stable CVD	-0.018	-0.0162	-0.0198	[31]
MCVE averagec	-0.1211	-0.10899	-0.13321	[31]
Death	0	0	0
Discount rate
Discount rate	0.035	0	0.05	[27]

AP: Acquisition price; EGP: Egyptian pound; Egy: Egyptian; CVD: Cardiovascular disease; LDL-C: Low-density lipoprotein cholesterol; MCVE: Major cardiovascular event; OSF: on-state fund. ^aCalculated equations are detailed in the Supplementary file. ^bThe dosage of the medication is one tablet per day. ^cWeighted average of major adverse cardiovascular events (MACE) events disutilities

The baseline LDL-C levels (mg/dL) in primary & secondary prevention and annual MCVE risks were brought from the CTT’s Collaboration study.[26] It reports the results of meta-analyses from 170,000 participants in 26 randomized trials to calculate the average risk reduction per 1·0 mmol/L LDL-C reduction at 1 year after randomization.[26] The MCVE risk reduction for each drug was calculated based on the decrease in LDL-C and the baseline LDL-C level of primary or secondary prevention [Supplementary File].

Costs

From the payer’s perspective, only the total cost incurred by the payer was taken into account [Table 1]. The prices of statin therapy (acquisition prices) were obtained from the Egyptian drug databases listed by the Egyptian Drug Authority,[29] and the annual costs were calculated accordingly. The costs of managing different MCVEs were brought from the Egyptian On-State Fund (OSF) price list for the year 2021, and the average cost was populated in the model. All costs were calculated in Egyptian pounds (EGP).

Outcomes

The outcomes were expressed in terms of quality-adjusted life years (QALYs). Data inputs for utilities were retrieved from previous studies,[30,31] as reported in Table 1. Due to a lack of local data, the age-related utilities were obtained from a published report of nationally representative values for the non-institutionalized US adult population for seven health-related quality-of-life scores [Supplementary Table S1], which included data from more than 54,000 subjects with an instrument completion rate above 85%.[30] The MCVEs effect on utility weights was calculated and subtracted from the age-specific utility, and total QALYs were calculated accordingly. MCVEs disutility were obtained from a study conducted in the USA across more than 38,000 subjects, in which the authors reported the preference-based scores (the EQ-5D_index scores) of chronic conditions and associated sociodemographic characteristics in the US nationally representative Medical Expenditure Panel Survey (MEPS) using regression models.[31]

Sensitivity analysis

One-way sensitivity analyses were carried out as recommended by the Consolidated Health Economic Evaluation Reporting Standards (CHEERS): ISPOR Task Force report[32] to examine the robustness of the model results across variations in input estimates. All estimations of clinical parameters, utilities, costs, and discount rate were used to test the robustness of the base case findings. For each parameter, all inputs were varied between high and low margins, where confidence intervals (CIs) from the literature or reasonable ranges were used based on different published sources.

Results

In Law et al.,[28] atorvastatin 40 mg and rosuvastatin 20 mg were associated with almost the same percentage reduction in LDL-C (mmol/L): 49% and 48%, respectively. Atorvastatin 80 mg was associated with a greater percent reduction in LDL-C, 55%. Using these figures in parallel with the calculated annual treatment costs (acquisition cost), atorvastatin 40 mg had the lowest annual cost per percent LDL-C reduction, followed by rosuvastatin 20 mg and atorvastatin 80 mg [Table 2]. Atorvastatin 40 mg had the least total cumulative discounted costs over the 10-year horizon, followed by rosuvastatin 20 mg and atorvastatin 80 mg. However, the lowest total cumulative discounted QALYs were generated by rosuvastatin 20 mg, followed by atorvastatin 40 and 80 mg. These results yielded an incremental 0.0012 QALYs for atorvastatin 40 mg versus rosuvastatin 20 mg with savings of 3,245 EGPs ($104.83). For the other comparison, atorvastatin 80 mg produced an incremental 0.0084 QALYs versus rosuvastatin 20 mg at a higher incremental cost of 434.6EGPs ($14.05), representing a more cost-effective option versus rosuvastatin 20 mg under a WTP threshold of three times the gross domestic product (GDP) per capita (139,200 EGP) [Supplementary Figure S1 ^{(66KB, tif)} ]. The total cumulative costs, QALYs, and the resulting incremental cost-utility ratio (ICUR) values of the cost-utility analysis are presented in Table 3. For the sensitivity analysis, the deterministic one-way sensitivity analysis confirms the robustness of the baseline results. However, it was only sensitive to the price of medications and atorvastatin 80 mg efficacy with regards to the atorvastatin 80 mg versus rosuvastatin 20 mg comparison [Figure 2].

Table 2.

Annual costs and effects on low-density lipoprotein cholesterol (LDL-C)

Statin and dose	Annual acquisition cost EGP	Percent reduction in serum LDL-Ca	Cost/percent reduction in serum LDL-C (EGP)
Atorvastatin
40 mg	3,913	49%	79.87
80 mg	4,670	55%	84.91
Rosuvastatin
20 mg	4,566	48%	95.12

EGP: Egyptian pounds; LDL-C: Low-density lipoprotein cholesterol. ^aAccording to Law et al.[28]

Table 3.

Total cumulative costs, QALYs, and the resultant ICER value of the cost-utility analysis

Prevention level and treatment dose	Total Cumulative Discounted Cost EGY	Total Cumulative Discounted QALYs	ICER (cost/QALY gained)	Interpretation
Overall
Atorvastatin 40 mg	20,800	4.1522	Dominant
Rosuvastatin 20 mg	24,045	4.1510
Difference	- 3,245	0.0012
Atorvastatin 80 mg	24,480	4.1594	51,732	Cost-effective
Rosuvastatin 20 mg	24,045	4.1510
Difference	435	0.0084

ICER: incremental cost-effectiveness ratio, QALYs: Quality-adjusted life years

Figure 2.

Tornado diagram showing a series of one-way sensitivity analyses. (a) Atorvastatin 40 mg versus rosuvastatin 20 mg and (b) atorvastatin 80 mg versus rosuvastatin 20 mg

Discussion

The comparisons carried out in the present study indicate that branded atorvastatin 40 and 80 mg are likely to be cost-saving and cost-effective, respectively, over branded rosuvastatin 20 mg in terms of reducing cardiovascular mortality and morbidity among Egyptian patients suffering from high LDL-C levels, with or without prior history of major cardiovascular events, over a 10-year time horizon. Further, rosuvastatin 20 mg was associated with the highest cost/percent reduction in serum LDL-C. The dominance of atorvastatin 40 mg was driven by lower costs and higher cardiovascular events and deaths avoided, and thus total QALYs gained. Similarly, a study conducted by Mould-Quevedo JF et al.[34] illustrated that rosuvastatin efficacy in reducing cardiovascular events was minimal compared to its high acquisition cost, especially in Colombia. Therefore, the atorvastatin medications were considered the cost-effective options in Brazil and Colombia. However, Sanjay K Gandhi et al.[35] demonstrated that rosuvastatin 20 mg was cost-effective compared with atorvastatin 40 mg in Swedish patients with a high risk of cardiovascular events.

For atorvastatin 80 mg, the cost-utility analysis over rosuvastatin 20 mg was likely to be attributable to the increased avoidance of cardiovascular events and deaths despite higher costs, which yielded an ICUR value far from the WTP threshold in Egypt. The deterministic one-way sensitivity analyses prove the model’s robustness for both scenarios. However, it was only sensitive to the price of medications and atorvastatin 80 mg efficacy with regard to the atorvastatin 80 mg versus rosuvastatin 20 mg comparison.

The present study includes several strong points. It utilized data obtained from two large meta-analysis studies: MR LAW[28] and CTT Collaboration.[26] Moreover, the model was checked internally for calculation errors by manually changing the values of different cells and testing the effect of these changes on results (which was also performed during the one-way deterministic sensitivity analysis). The model looks at patients prescribed high-intensity statins whenever decided by the physician and explicitly accounts for uncertainties of the epidemiological parameters, MCVE risks, and quality of life by using ranges and CIs based on published sources. However, our study has several limitations since utility and disutility values were collected from international sources due to the absence of local data. Furthermore, due to the unavailability of WTP studies in Egypt, we followed the commonly used threshold by Egyptian authorities (3 GDP per capita).

Conclusions

Branded atorvastatin 40 mg was shown to be strongly dominant over branded rosuvastatin 20 mg, with higher QALYs and lower associated costs. Branded atorvastatin 80 mg is a cost-effective option versus branded rosuvastatin 20 mg with an incremental cost-effectiveness ratio (ICER) value of 51,732 EGPs/QALY, which is significantly below the WTP threshold of 139,200 EGPs as recommended by the WHO guidelines. The deterministic one-way sensitivity analysis proves the results of the Markov model to be sensitive to medication prices only and to atorvastatin 80 mg efficacy in percent reduction of LDL-C level. This model may inform decision-makers about resource allocation for the care of patients with high LDL-C levels in Egypt.

Abbreviations

CVDs: Cardiovascular diseases; WHO: World health organization; MI: Myocardial infarction; CHD: coronary heart disease; LDL-C: Low-density lipoprotein cholesterol; TC: Total cholesterol; TG: Triglycerides; HDL-C: High-density lipoprotein cholesterol; NCDs: Non-communicable diseases; ACS: Acute coronary syndrome; UPA: Unified Procurement Authority; MCVEs: Major cardiovascular events; CTT: Cholesterol Treatment Trialists; IHD: Ischemic heart disease; OSF: On-State Fund; EGP: Egyptian pound; QALYs: Quality-adjusted life years; MEPS: Medical Expenditure Panel Survey; CHEERS: Consolidated Health Economic Evaluation Reporting Standards; AP: Acquisition price; WTP: Willingness-to-pay; GDP: gross domestic product; ICER: Incremental cost-effectiveness ratio; CI: Confidence interval; MACE: Major adverse cardiovascular events; ICUR: Incremental cost-utility ratio; MOH: Ministry of health; CV: Cardiovascular; ICER: incremental cost-effectiveness ratio.

Conflicts of interest

There are no conflicts of interest.

Figure S1

(a) Atorvastatin 40 mg associated costs and generated QALYs versus rosuvastatin 20 mg and (b) atorvastatin 80 mg associated costs and generated QALYs versus rosuvastatin 20 mg

Supplementary

Table S1.

Age-related utilities

Measure	Age group	Mean/Males	Mean/Females
EQ-5Da	40–49	0.887	0.863
	50–59	0.861	0.837
	60–69	0.840	0.811
	70–79	0.802	0.771

Adapted from[30]. ^aThe average of EQ-5D/US figures for ages from 40 to 79 years old were weighted based on Gender, Male 73% : Females 27% (in alignment with the Cholesterol Treatment Trialists’ (CTT) Collaboration, 2010[26] baselines characteristics of gender distribution), and then populated in the model

Supplementary File

Cost-Utility Analysis of High-Intensity Statins; Atorvastatin 40 and 80 mg Versus Rosuvastatin 20 mg from Egyptian Payer's Perspective: A Markov Model

1. MCVE risk reduction calculation:

   MCVE risk reduction is calculated based on a risk decrease of 21% per each mmol/L decrease in LDL-C. You calculate the risk reduction by the decrease each comparator achieves in LDL in mmol multiplied by the baseline LDL of primary or secondary prevention.

2. Medication annual cost calculation:

   Statins costs are the drug cost multiplied by the number of days in a year.

3. Age-related utilities calculation:

   The average of EQ-5D/US figures for ages 40-79 years old was weighted based on gender, males 73%: females 27% (in alignment with the Cholesterol Treatment Trialists' (CTT) Collaboration, 2010[1] baseline characteristics of gender distribution), and then populated in the model.

4. MCVE average calculation:

   The weighted average of MACE events disutilities.

Funding Statement

This study was funded by Viatris Egypt.