Evaluation of health utility in trial-based cost–utility analyses for major cardiovascular disease: protocol for a systematic review

Abstract

Introduction

The global incidence of cardiovascular disease (CVD) is high, and the medical costs associated with its management have been increasing. Cost–utility analyses (CUAs) are essential for understanding the value of healthcare interventions and for decision-making. A majority of the CUAs for CVD are model based and have cited health utilities from previously published data; standard health utilities for the CUAs of CVD have not been established yet. Thus, we aim to identify the standard utilities according to the patients’ condition and disease severity in patients with major CVDs.

Methods and analysis

We will search Medline and Evidence-Based Medicine Reviews for trial-based CUA studies that have reported on quality-adjusted life-years using original health utilities for patients with three major forms of CVD (coronary artery disease, heart failure and atrial fibrillation). Papers on trial-based CUAs will be included, while those on model-based CUAs will be excluded. No restrictions will be made in terms of intervention type. The main outcome comprises the health utilities calculated on a scale of 0–1 (irrespective of the measurement methods) at baseline and after treatment. Two independent investigators will screen the eligibility of articles; they will extract data, including health utilities, from the eligible articles for further analysis. The quality of the included studies will be assessed using the Consolidated Health Economic Evaluation Reporting Standards checklist. We will describe the means and SDs of the health utilities from all the included studies. The mean utility weights for individual studies will be combined through meta-analyses using a random-effects model to obtain the representative health utility value for each disease. Subgroup analyses will be conducted according to the severity and duration of each disease.

Ethics and dissemination

Ethical approval is not required. The review will be submitted to an appropriate peer-reviewed journal.

PROSPERO registration number

CRD42022316278.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• We will systematically identify and critically appraise published cost–utility analyses using a rigorous and extensive search strategy.

• This systematic review will identify the standard values and ranges of health utilities concerning three major forms of cardiovascular disease (CVD): coronary artery disease, heart failure and atrial fibrillation.

• The main limitations include difficulties in identifying all relevant publications from the literature and ensuring that the synthesis of evidence is appropriate and meaningful.

• The included patients may not represent all patients with CVD in the real-world clinical setting.

Introduction

Cardiovascular disease (CVD) is associated with high morbidity and mortality rates. In 2015, 18 million global deaths were attributed to CVD and 423 million people were estimated to have the disease worldwide.^{1 2} Ischaemic heart disease and stroke are particularly concerning in healthcare.³ The WHO reported that ischaemic heart disease and stroke are the first-leading and second-leading causes of death worldwide, respectively; they account for approximately 16% and 11% of the total global deaths, respectively.³ Moreover, CVD has a significant impact on the quality of life (QOL) through disability, lost work time and premature death. Therefore, the financial and human burden of CVD have also been increasing. In the USA, the medical costs and productivity loss associated with CVD are expected to increase to US$1.1 trillion by 2035.⁴ Among the variety of CVDs, coronary artery disease (CAD), heart failure (HF) and atrial fibrillation (AF) have been considered as major forms of CVD that need to be addressed because ischaemic heart disease and stroke are the first and second causes of death worldwide.^{3 5} Accordingly, these major CVD-associated medical costs and healthcare-access issues challenge the sustainability of healthcare systems.^{1 2 4}

Cost–utility analysis (CUA) is an economic evaluation analysis that evaluates healthcare interventions by integrating objective improvement in patient QOL and/or length of life with costs using the cost–utility ratio of dollars expended per quality-adjusted life-years (QALYs) gained.⁶ CUAs help in the making informed decisions regarding the allocation of healthcare resources.^{7 8} In the majority of CUAs, QALYs are widely used to estimate QOL benefits in health. QALYs are calculated by multiplying each utility value by the amount of time an individual spends there, which incorporates changes in both the quantity (longevity and mortality) and quality (morbidity and psychological, functional, social, and other factors) of life (figure 1). In CUAs, health utility values range from 0 (representing death) to 1.0 (representing full health).^{9 10} They play an important role in health economic evaluations because health utilities are used as QOL measures to calculate the QALYs in CUAs.^9–11 Previous studies have reported that health utilities may significantly impact the results of CUAs.^{12 13} In trial-based CUAs, health utilities were originally measured directly using preference elicitation techniques or indirectly by administering utility-based instruments to target respondents.^14–17 Conversely, in model-based CUAs, which represent a common economic evaluation method that uses simulation cohorts of patients with a variety of diseases, health utilities are often cited from published studies reporting the QOL of patients with the target disease. However, a recent study indicated that there were concerning discrepancies in the published health utilities for CUAs on CVD.¹⁷ This previous study also revealed that there has been an increase in the incorrect usage of non-health utility studies as a source of health utilities for CUAs despite the fact that health utilities significantly affect the results of CUAs.¹⁷ Ideally, health utilities should be measured for each CUA if the target population is different; however, this is not practical in terms of the time, effort and cost required. Therefore, summarising health utilities in patients with different forms and severities of CVD is desirable for improving the accuracy of CUAs.

Figure 1.

A schematic representing a QALY calculation for a comparison of two treatments. The QALYs are calculated by multiplying each utility value by the amount of time an individual spends there, which incorporates changes in both the quantity and quality of life. The health utility value is measured from 0 (representing death) to 1.0 (representing full health). The area under the curve represents QALYs. In this case, the new treatment provides benefit in terms of better utility and increased survival. The difference in the coloured areas under each curve represents the additional QALYs gained by providing the new treatment instead of the conventional treatment.

Thus, this systematic review aims to describe the distribution of health utilities in trial-based CUAs on three major forms of CVD (CAD, HF and AF) and to identify the standard utility values according to patient condition and disease severity.

Methods and analysis

The present protocol has been registered within the PROSPERO database (CRD42022316278). This systematic review and meta-analysis will be conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocol (PRISMA-P) statement (online supplemental table 1: PRISMA-P checklist).¹⁸

Criteria for considering studies

Types of studies

We will include papers on economic evaluation studies (in which both the costs and outcomes or only the outcomes are compared) based on data from randomised controlled trials. Furthermore, we will only include papers in English. Papers on partial economic evaluation studies (in which only the costs are compared) will be excluded. We will also exclude papers on model-based studies (non-trial-based studies) and studies on subgroups of previously reported datasets.

Participants

Papers on adults (age ≥18 years) diagnosed with three forms of CVD (namely CAD, HF or AF) will be included in this systematic review. Patients with CAD will include those with coronary atherosclerosis, stable angina pectoris, the acute coronary syndrome, unstable angina pectoris, myocardial infarction, ST-segment elevation myocardial infarction (STEMI) and non-STEMI (NSTEMI). Patients with HF will include those with acute HF, chronic HF, HF with a reduced ejection fraction, HF with a mildly reduced ejection fraction and HF with a preserved ejection fraction. Patients with AF will include those with paroxysmal, persistent and permanent AF. Papers on patients under 18 years of age will be excluded. No restrictions will be made in terms of the sex or country/region of the study.

Intervention and comparators

Papers on all types of interventions (including pharmacological, non-pharmacological and interventional procedures) will be included.

• Pharmacological interventions include medications used to treat the three target CVDs. All types of standard drugs for CVD may be considered, such as antiplatelets, anticoagulants, beta blocker, calcium channel blockers, antiarrhythmics, statins, nitrates, ACE inhibitors, angiotensin-Ⅱ receptor blockers, diuretics, mineralocorticoid receptor antagonists and sodium/glucose cotransporter 2 inhibitors.

• Non-pharmacological interventions may include alcohol abstinence, smoking cessation, exercise, stress and body weight reduction, improvements in dietary life, etc.

• Interventional procedures may include all established interventions for the three target CVDs. For patients with CAD, papers on those who underwent percutaneous coronary intervention and coronary artery bypass grafting will be included. For patients with HF, papers on those who underwent cardiac resynchronisation therapy and implantable cardioverter defibrillator will be included. For patients with AF, papers on those who underwent catheter ablation will be included.

• We will not limit the comparator in the studies: placebo, no treatment, treatment as usual, active control, etc.

Outcomes

We will evaluate the baseline utilities of patients with three major forms of CVD (CAD, HF and AF). We will include papers on studies that have measured and reported QOL health utilities on a scale of 0–1 irrespective of the measuring methods. The process of calculating the associated utility employs either direct or indirect methods. The direct methods are time trade-off (TTO), the standard gamble and the Visual Analogue Scale. The indirect methods are EuroQoL 5-Dimension 3-Level (EQ-5D-3L), EuroQoL 5-Dimension 5-Level (EQ-5D-5L), Health Utility Index, or Short-Form Six-Dimension (SF-6D).^{19 20} Patients with CVD may have several risk factors (eg, hypertension and dyslipidaemia in CAD) as comorbidities. Therefore, utilities will be collected in patients with CVDs, including risk factors. Papers on studies that have used health utilities derived from other studies will be excluded. Average or mixed utilities during the period under observation are also excluded.

Search strategy

This literature search will be conducted using the Ovid Medline and Ovid Evidence-Based Medicine (EBM) Reviews; using combinations of medical subject headings and keywords as the search terms, we will screen for papers published from database inception to 30 June 2022. The draft search terms are presented in table 1, online supplemental tables 2 and 3. Searches will be supplemented with the International Clinical Trials Registry Platform managed by WHO and through relevant existing literature reviews. We will also evaluate similar systematic and narrative reviews to find potentially eligible papers.

Table 1.

Search strategy

Search number	Search detail
#1	"cost benefit analysis"[MeSH Terms] OR ("cost benefit"[All Fields] AND "analysis"[All Fields]) OR "cost benefit analysis"[All Fields] OR ("cost"[All Fields] AND "benefit"[All Fields] AND "analysis"[All Fields]) OR "costs and cost analysis"[MeSH Terms] OR ("costs"[All Fields] AND "cost"[All Fields] AND "analysis"[All Fields]) OR "costs and cost analysis"[All Fields] OR ("cost"[All Fields] AND "utility"[All Fields] AND "analysis"[All Fields]) OR "cost utility analysis"[All Fields] OR ("cost"[All Fields] AND "effectiveness"[All Fields] AND "analysis"[All Fields]) OR "cost-effectiveness analysis"[All Fields]
#2	"quality of life"[MeSH Terms] OR ("quality"[All Fields] AND "life"[All Fields]) OR "quality of life"[All Fields] OR "quality adjusted life years"[MeSH Terms] OR ("quality adjusted"[All Fields] AND "life"[All Fields] AND "years"[All Fields]) OR ("quality"[All Fields] AND "adjusted"[All Fields] AND "life"[All Fields] AND "years"[All Fields]) OR "quality adjusted life year*"[All Fields] OR (“QOL”[All Fields]) OR (“QALY”[All Fields])
#3	"heart failure"[MeSH Terms] OR ("heart"[All Fields] AND "failure"[All Fields]) OR "heart failure"[All Fields] OR ("congestive"[All Fields] AND "heart"[All Fields] AND "failure"[All Fields]) OR "congestive heart failure"[All Fields]
#4	"myocardial ischemia"[MeSH Terms] OR ("myocardial"[All Fields] AND "ischemia"[All Fields]) OR "myocardial ischemia"[All Fields] OR "coronary artery disease"[MeSH Terms] OR ("coronary"[All Fields] AND "artery"[All Fields] AND " disease"[All Fields]) OR ("myocardial"[All Fields] AND "ischemia"[All Fields]) OR ("coronary"[All Fields] AND "artery"[All Fields] AND " disease"[All Fields]) OR "coronary artery dis*"[All Fields] OR ("angina pectoris"[MeSH Terms] OR ("angina"[All Fields] AND "pectoris"[All Fields]) OR "angina pectoris"[All Fields]) OR ("myocardial infarction"[MeSH Terms] OR ("myocardial"[All Fields] AND "infarction"[All Fields]) OR "myocardial infarction"[All Fields])
#5	"atrial fibrillation"[MeSH Terms] OR ("atrial"[All Fields] AND "fibrillation"[All Fields]) OR "atrial fibrillation"[All Fields]
#6	#1 AND #2 AND (#3 OR #4 OR #5)

Study screening and selection

Two investigators (YN and SM) will scan the titles and abstracts of all retrieved articles independently. The eligibility of the included articles will be further evaluated through a full-text review by the two aforementioned investigators. The reasons for exclusion of articles during the full-text review will be documented. The process of study selection will be illustrated in a PRISMA flow diagram.²¹ Any disagreements between the investigators will be resolved via discussions; a third investigator (HK) will moderate any disagreements if consensus is not reached.

Data collection process and data items

From the included publications, we will extract the following data for the systematic review:

• Publication dates.

• Country/region(s) of the study.

• Sample sizes.

• Demographic characteristics, such as age, sex, race, ethnicity, occupation, education and socioeconomic level.

• Disease types (CAD, HF or AF).

• Disease severity: We will classify patients with CAD into groups with STEMI/NSTEMI and stable/unstable angina according to the recent guidelines.^{22 23} We will classify patients with HF using the established HF classifications (for eg, the New York Heart Association functional classification, American Heart Association/American College of Cardiology classification and HF classification by the left ventricular ejection fraction).²⁴ We will classify patients with AF by the AF duration and the presence or absence of comorbidities (lone AF or not).

• Treatment received and comparators.

• Follow-up periods.

• QOL health utilities: mean and SD of the health utilities at baseline and follow-up measurements.

• Survey instruments used for the health utilities (eg, EQ-5D, SF-36 and TTO).

We will develop and pilot the data extraction form for this systematic review. Two reviewers will independently summarise the above-mentioned data items. Discrepancies will be resolved via discussions within the review team. We will contact the relevant authors whenever study data are incomplete.

Quality assessment

Quality assessment will be conducted using the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist.²⁵ This checklist contains 24 items; each item is scored a 1 or 0 depending on whether it is achieved or not in the study, respectively.²⁶ Each of the 24 items in the CHEERS checklist will be independently assessed with respect to the included studies by two investigators (YN and SM). Any disagreements will be resolved via discussion with the third investigator (HK).

Data synthesis and analysis

This review will analyse the mean and range of health utilities in patients with three major forms of CVD (CAD, HF and AF). The outcomes of interest (health utilities) will be continuous values. We will first describe the means and SD of the health utilities at baseline and after treatment in all the included studies. Assuming that all the studies were performed in the same population, we will use a random-effects model to combine the means from the included studies to estimate the standard mean health utility and its uncertainty. We will assess heterogeneity by: (1) comparing the clinical settings, patients and interventions descriptively; (2) visualising the forest plot and (3) performing a statistical test and computing the I² statistic. The study summary will include relevant tables and figures to aid in data presentation. The systematic review and meta-analyses will be performed using the ‘meta’ package of R (V.4.2.1).

Subgroup analysis

Considering that possible heterogeneity may exist, we will prespecify the following subgroup analyses. If sufficient studies are included, we will estimate the standard mean health utility according to the disease severity and duration. For CAD, the subgroup analyses will be conducted for patients with STEMI, NSTEMI and stable/unstable angina.^{22 23} For HF, the subgroup analyses will be conducted according to the severity of HF (using the New York Heart Association functional classification and the left ventricular ejection fraction).²⁴ For AF, the subgroup analyses will be conducted according to the type of AF, including paroxysmal, persistent and permanent AF.

Patient and public involvement

No patient or public involvement.

Ethics and dissemination

This systematic review is exempt from ethics approval and the requirement of consent for participation because it will be carried out using published documents. The results of this systematic review will be presented at international conferences and will be published in a peer-reviewed journal.

Discussion

We have presented the study protocol for a systematic review and meta-analysis to summarise and identify the standard and range of health utilities in patients from trial-based CUAs on three major forms of CVD (CAD, HF and AF).

Cost-effectiveness data are essential for understanding the value of healthcare interventions and for making decisions on the prioritisation of interventions in jurisdictions where healthcare resources are limited. Accordingly, the number of published CUAs on CVD has been increasing every year, and a majority of these CUAs are based on models. Zhou et al evaluated 585 papers on CUAs for CVD that were published between 1977 and 2016; approximately 89% of the CUAs were based on models.¹⁷ These model-based CUAs commonly use published original health-utility studies or previous CUAs that have reported on health utilities as the sources of health utilities. However, Zhou’s group also revealed that many of these model-based CUAs used utilities that were different from those reported originally and did not provide any explanation for such differences.¹⁷ In order to ensure the equitability for CUAs, the standard values and ranges of health utilities in patients with different forms and severities of CVD are required. Zhou’s group also conducted a systematic review of cross-sectional studies to summarise the health utilities elicited from the EQ-5D-5L for patients with different diseases worldwide.¹⁶ Their results included a summary of the health utilities of patients with CVD in general; however, there were no specific data for each form and severity of CVD. To accumulate more details of health utilities for patients with CVD, our study will analyse the health utilities for the three major forms of CVD (CAD, HF and AF) in greater detail. We anticipate that our results will contribute to improving the accuracy of CUAs in the field of CVD, particularly for model-based analyses.

We acknowledge several limitations in our study. First, the number of search engines used for the literature search (ie, Medline and EBM Reviews) is relatively small. Furthermore, the diversity of literature and the necessity for narrow search terms may mean that some potentially relevant papers will be missed. To reduce this risk, we will manually search the identified papers for other potentially relevant articles. Second, papers in languages other than English will be excluded; therefore, a language bias may be existed. Third, our main outcomes comprise the QOL health utilities calculated on a scale of 0–1 regardless of the measurement methods, and we will calculate their weighted means and SDs. However, different measurement methods for health utilities may affect the results. Fourth, we plan to collect data for the utilities described in the economic evaluation studies. Therefore, the included patients may not represent all patients with CVD in real-world clinical practice.

Ethics statements

Patient consent for publication

Not applicable.