Methods and validity indicators for measuring adherence to statins in secondary cardiovascular prevention: a systematic review

Abstract

Background

Adherence to statin therapy is crucial for reducing the recurrence of cardiovascular events. Numerous methods exist to measure medication adherence, including those based on prescription data, patient self-report, medication counting, and direct methods. It is important to determine which of these methods are appropriate for use in clinical practice. This systematic review aimed to identify the methods used to measure adherence and persistence to statins in patients undergoing cardiovascular secondary prevention and to evaluate the validity indicators of these methods.

Methods

This systematic review included studies reporting methods to measure adherence and/or persistence to statins in cardiovascular secondary prevention. Medline, Embase, and Scopus databases were searched from inception to February 2025. Rayyan was used for the study selection and extraction data processes. Validity indicators of the adherence/persistence methods were collected; it was reported. Risk of bias of studies reporting the method validity was evaluated using the COSMIN (Consensus-based Standards for the Selection of Health Measurement Instruments) tool.

Results

A total of 77 studies were included. Regarding adherence measurement, the most frequently used method was prescription refill records (n = 55) and self-report methods (n = 20). Electronic monitoring methods (n = 2), self-perceived adherence by physician (n = 1), and pill counting (n = 1) were less frequently used methods. Direct methods, using HPLC–MS/MS, were used in combination with other indirect methods (n = 5). For measuring persistence, prescription refill records were the predominant method (n = 9), while self-report methods were used in three studies, and one study used a standardized questionnaire. Several of the indirect methods have validity indicators for measuring adherence in different study populations and to different medications. Only one study provides validity indicators for the MAT questionnaire specifically adapted for statins.

Conclusions

The methods for measuring adherence to statins in secondary cardiovascular prevention were predominantly indirect, relying on prescription and supply records and self-report methods. Pill counting, electronic monitoring, and direct measurement via LC–MS/MS were less commonly used. Persistence was primarily measured through prescription refill records. None of the indirect methods was validated; thus, their use for measuring adherence to statins is not recommended. There is a need for new validated tools, incorporating a gender perspective, to measure adherence to statins in this population.

Systematic review registration

PROSPERO CRD42023463981.

Background

Cardiovascular diseases (CVDs) are among the most prevalent conditions worldwide, contributing to significant morbidity and mortality [1, 2]. The World Heart Federation (WHF) [3] estimates that approximately 35 million people experience a cardiovascular event each year. CVDs not only lead to a substantial decline in quality of life but also impose a heavy economic burden on healthcare systems [4]. The pathogenesis of these diseases is influenced by a range of risk factors, including modifiable and non-modifiable ones [5]. The inadequate control of cardiovascular risk factors has shifted attention toward secondary cardiovascular prevention. This approach combines lifestyle changes and pharmacological measures to reduce the risk of recurrence in patients who have already experienced a cardiovascular event [1, 6].

Dyslipidemia is a key focus in cardiovascular secondary prevention, with statin therapy recommended by the AHA/ACCF [7] alongside lifestyle changes [8]. However, despite its benefits, ensuring proper medication use is challenging, as studies show that only about 50% of patients in high-income countries adhere to their prescribed treatments [9, 10]. Poor adherence leads to worse disease management, lower survival rates, higher recurrence risks, reduced quality of life, and increased healthcare costs [11].

The World Health Organization (WHO) defines therapeutic adherence as “the extent to which a person’s behavior—taking medications, following a diet, and/or making lifestyle changes—corresponds with the agreed recommendations from a healthcare provider.” The WHO also emphasizes that improving adherence may be the most cost-effective strategy for managing chronic conditions [10]. Specifically, medication adherence is defined as “the process by which patients take their medications as prescribed, comprising initiation, implementation, and discontinuation” [12]. Medication adherence is a multifactorial phenomenon shaped by five interrelated domains [13] related to patient characteristics such as age, employment status, socioeconomic conditions, culture, educational level, geographic area, and race [14, 15]; social and familial support [16]; disease characteristics; therapeutic regimen; and healthcare system conditions, including healthcare professional characteristics [17, 18].

Treatment efficacy depends not only on daily drug intake but also on long-term continuation. Persistence, which refers to the time between the initiation of treatment and the last dose taken before discontinuation, measures how long a patient continues the medication according to the intended duration. It is typically measured as the proportion of days a patient adheres to the treatment or the average time until therapy discontinuation [12, 19].

Methods for measuring medication adherence are generally classified as direct or indirect. Direct methods include techniques such as directly observed therapy (DOT), therapeutic drug monitoring (TDM), and ingestible sensor-based systems. These methods are objective, specific, and highly accurate but are often costly and impractical for routine clinical practice. Indirect methods, on the other hand, include patient self-report questionnaires, pill counts, calculations of the proportion of days covered (PDC) or the medication possession ratio (MPR) based on dispensing records, and medication event monitoring systems (MEMS), among others [19,20,21]. Patient self-report questionnaires based on clinical interviews are particularly popular in clinical practice. While this approach has limitations—including subjectivity, recall bias, and response bias due to its reliance on self-reported data—it remains widely adopted because of its practicality, simplicity, and cost-effectiveness [20].

Notable self-report questionnaires include the Haynes-Sackett Test [22], the Morisky-Green Test [9, 10], and the 8-item Morisky Medication Adherence Scale (MMAS-8) [19,20,21, 23]. The MMAS-8, in particular, is one of the most widely used tools in clinical practice. However, despite being validated for use in populations and conditions different from its original context [24,25,26], the MMAS-8 has often been applied without prior validation, resulting in evidence of its limitations in certain populations, such as patients with type 2 diabetes in Spain [27]. Furthermore, the original study on the MMAS-8 was recently retracted due to inconsistencies in its reported sensitivity and specificity values [23]. Therefore, although questionnaires like the MMAS-8 are valuable in clinical practice, it is essential to consider their limitations and the need for contextual-specific validations before their application, particularly in diverse populations and conditions different from the original ones [28]. This systematic review aimed to identify the methods used in research to measure adherence and persistence to statins in patients undergoing secondary cardiovascular prevention. It also sought to evaluate the validity and accuracy indicators of these methods.

Materials and methods

The protocol for this systematic review was registered in PROSPERO (Reference: CRD42023463981), and the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [29] guidelines were followed to report the methodology and results. The Office of Responsible Research of the University Miguel Hernández approved the study (Reference: TFG.GME.VFGG.MMM.231103).

Eligibility criteria

This review included studies that measured adherence or persistence to any type of statin and reported the methods used. Studies that evaluated adherence or persistence to statins in combination with other treatments were excluded. Regarding the study population, articles were selected if they included individuals aged 18 and older undergoing secondary cardiovascular prevention. Conditions considered for secondary prevention included ischemic heart disease, acute myocardial infarction, stroke, cerebral hemorrhage, transient ischemic attack, renal failure, heart failure, peripheral artery disease, dissecting aortic aneurysm, and diabetic or hypertensive retinopathy. All participants had to be receiving statin therapy. Eligible study designs included observational studies (cross-sectional, case–control, and cohort studies) and experimental studies. Excluded materials comprised letters, editorials, case reports, reviews, opinion articles, abstracts, conference papers, study protocols, non-scientific studies, and those written in non-Latin alphabet languages.

Sources of information and search strategy

The databases Medline, Embase, and Scopus were searched to retrieve relevant studies. Articles published from the inception of each database until February 28, 2025, were included, with no language restrictions except for the requirement that studies be written in Latin alphabet. The search strategy combined controlled vocabulary and free text terms, including “Treatment Adherence and Compliance,” “Medication Adherence,” “Hydroxymethylglutaryl-CoA Reductase Inhibitors,” “Cardiovascular Diseases,” and “Acute Coronary Syndrome.” Filters were applied for publication type and population age. The complete search strategies for each database are detailed in Supplementary Material 1.

Study selection

Articles identified were exported to the Rayyan platform for screening. After automatic detection of duplicates, manual removal was performed. Two independent researchers conducted a two-stage screening process: (1) title and abstract review and (2) full-text eligibility assessment. Discrepancies were resolved by consulting a third researcher. For studies with restricted access, university library services were utilized; studies that remained inaccessible were excluded from the review.

Data collection

Data from eligible studies were extracted by one researcher and verified by another. Extracted data included author, year, location, study design, population characteristics, whether adherence, persistence, or both were measured (considering adherence as the degree to which patients follow the prescribed dosage frequency and persistence as the continuity of medication use over time without interruption), sample size, study setting, type of statin for which adherence or persistence was measured, methods used for measurement (type and description), criteria for defining a patient as adherent/persistent or non-adherent/non-persistent, validity indicators of measurement methods (if available), and psychometric properties of the adherence questionnaire (if available).

Risk of bias

The primary objective of this review was to identify the methods used to measure adherence and persistence to statins in patients undergoing cardiovascular secondary prevention, without focusing on clinical outcomes or intervention effectiveness. Therefore, a formal risk of bias or methodological quality assessment of the included studies was deemed unnecessary, as these aspects do not directly impact the primary objective of this review, which centers on identifying measurement methods. However, for studies assessing the validity of the method in question, risk of bias was evaluated using the COSMIN (Consensus-based Standards for the Selection of Health Measurement Instruments) tool [30]. This was due to the fact that these studies provide key data (validation indicators) that may be influenced by methodological design quality, which is essential for the reliability of the identified adherence methods.

Data synthesis

A descriptive synthesis summarized study characteristics, and a narrative synthesis detailed the measurement methods for adherence and persistence separately. Validity indicators of validated methods and psychometric properties of questionnaires were tabulated. Due to insufficient studies reporting validity indicators for statin adherence measurement methods, a meta-analysis was not feasible. A meta-analysis was not feasible due to the insufficient number of studies reporting validated methods for measuring statin adherence. The lack of validated methods compromises data reliability and comparability, increasing heterogeneity and the risk of bias. Without standardized, validated measurement tools, pooling data would not yield meaningful or accurate conclusions.

Results

Following the database search, 1488 articles were identified, and after duplicate removal, 1340 titles and abstracts were screened. Of these, the full texts of 144 studies were assessed for eligibility, leading to the inclusion of 77 articles in this systematic review. The most common reason for exclusion was failure to meet the inclusion criteria for the study population. Figure 1 provides the PRISMA flow diagram [29], which details the study selection process.

Fig. 1

PRISMA 2020 flow diagram for new systematic reviews which included searches of databases and registers only. Source: Page MJ, et al. BMJ 2021;372:n71. https://doiorg.publicaciones.saludcastillayleon.es/10.1136/bmj.n71.814.

The general descriptive characteristics of the articles are presented in Table 1. The included articles were published between 2002 and 2023. Most studies were conducted in the USA (n = 25), followed by other countries such as Canada (n = 6), the UK (n = 5), and Taiwan (n = 3). Regarding study design, cohort studies predominated (n = 45), followed by experimental studies (n = 24) and cross-sectional studies (n = 6). Sample sizes ranged from two to 813,887 patients. Most studies were conducted in hospital settings (n = 40). A significant proportion (74.0%, n = 57) focused on adherence, with fewer studies measuring persistence (11.7%, n = 9) or both adherence and persistence (14.3%, n = 11). As for the types of statins evaluated, most studies assessed multiple types, with atorvastatin and rosuvastatin being the most frequently analyzed (Table 2).

Table 1 General descriptive characteristics of the studies included in the review (n = 77)

Table 2 Methods of measuring adherence and persistence, and their validity indicators (n = 77)

Methods for measuring statin adherence

The reviewed studies employed various methods to measure adherence to statins. These included, in order of frequency, review of prescription refill records (n = 55), self-report methods (n = 20), direct monitoring methods via plasma or urine (n = 5), electronic monitoring devices (n = 2), self-perceived adherence by physicians (n = 1), and pill count methods (n = 1). Among prescription-based adherence indicators, the most commonly used were the PDC (n = 30) and the MPR (n = 16), with adherence thresholds typically defined as PDC or MPR ≥ 80%. Some studies further categorized MPR-based adherence into optimal, adequate, and suboptimal levels. Figure 2 summarizes all these adherence measurement methods grouped into six main groups.

Fig. 2

Pie chart grouping adherence measurement methods into six main groups: pill counting methods, dispensation records, direct methods, electronic monitoring methods, self-perceived adherence by physician, and self-report methods

Self-report tools used to measure statin adherence included:

• 8-item Morisky Medication Adherence Scale (MMAS-8) [23]: An 8-item scale scoring adherence from 0 to 8, where lower scores indicate higher adherence.

• 4-item Morisky Medication Adherence Scale (MMAS-4) [105]: A shorter version scoring adherence from 0 to 4, with lower scores reflecting higher adherence.

• 7-day recall: A single-item measure asking patients how many days they took their statin in the past week.

• Medication Adherence Tool (MAT) [108, 110]: A 7-item questionnaire rated on a 6-point Likert scale, evaluating various aspects of adherence from the patient’s perspective.

• Visual analog scale (VAS): A line scale from 0 to 100% divided into 10 intervals, where patients mark their adherence level.

• Gehi et al. questionnaire [111]: A 3-item tool assessing adherence qualitatively, without generating a cumulative score.

• SEAMS Questionnaire [107]: A 13-item scale scored on a 3-point Likert scale, with scores ranging from 13 to 39, where higher scores indicate better adherence.

• 24-h recall: A single-item measure assessing whether the patient took their medication in the last 24 h.

Two studies used electronic devices to quantify statin adherence, without specifying thresholds for classifying patients as adherent or non-adherent:

• Medication event monitoring system (MEMS) [106]: Electronically records each time the medication container is opened, providing precise data on medication access frequency and timing.

• GlowCap®: An electronic cap device that emits visual or auditory reminders for medication intake, while logging the frequency of use.

One study [64] employed pill count methods, defining adherence as the consumption of 85–100% of the expected pills. Lastly, some studies used direct monitoring methods, such as tandem liquid chromatography-mass spectrometry (LC–MS/MS), which measures adherence by detecting drug levels in biological fluids, providing objective verification of recent statin consumption [61].

Regarding validity indicators for indirect methods, none was specifically designed or validated for measuring statin adherence. Table 2 summarizes the psychometric properties previously reported for the MMAS-8 (retracted), MMAS-4, SEAMS, and MAT scales for measuring adherence to other medications and populations. Only in the case of the MAT scale did authors test internal consistency when adapted for statins (Cronbach’s alpha = 0.66) [88]. Direct measurement methods assessing adherence through statin detection (or its metabolites) in the patient’s body provide objective verification of recent statin intake. The reliability of these results depends on the validity indicators of the analytical method, which were reported in five studies (Table 2). Thompson et al. [90] used the HPLC–MS/MS method to evaluate adherence with a detection limit of 1–200 ng/mL, stating that variations in drug pharmacokinetic parameters did not affect relative detection. This suggests the method’s precision is reliable for identifying drug presence in urine, although full analytical validation details were not provided.

Methods for measuring statin persistence

Persistence in statin use was assessed through various methods, predominantly based on prescription refill records, each employing specific criteria to define continuity in medication acquisition (Table 2):

• Interruptions without renewal within a defined period: Patients were classified as persistent if they renewed their prescriptions without exceeding a predetermined interruption pe

• Time between prescription and supply: Persistence was determined by evaluating whether patients refilled their medication within a defined timeframe after the initial prescription.

• Proportion of days covered (PDC): Persistence was defined in several studies as a PDC ≥ 80% during the follow-up period, with lower values indicating prolonged treatment interruptions and classified as non-persistence.

• Medication possession ratio (MPR): In one study, patients were considered persistent if their MPR was ≥ 80% over a 1-year follow-up period.

• Continuity of prescription: Persistence was assessed by verifying whether patients had statins available on a specific date, regardless of prior supply interruptions.

• Supply frequency: Patients were classified as persistent if they maintain a consistent supply frequency or exceed a minimum threshold.

Three studies [31, 54, 71] employed a self-report approach, where patients were asked during follow-up phone calls or home visits about their medication use. Discontinuity was defined as an interruption in treatment lasting more than 90 days. One study [100] used a standardized questionnaire although no further details were provided.

Persistence measurement methods based on prescription refill records generally lack formal validation, as no standardized process ensures their accuracy or consistency across diverse contexts. However, sensitivity analyses were conducted in some studies to justify the chosen cutoff points: Allonen et al. [25] validated a 180-day cutoff for assessing statin use continuity through sensitivity analysis. Another study [97] defined non-persistence as a period exceeding 6 months without a prescription after the last covered day of statin supply, supported by sensitivity analyses conducted by the authors. Figure 3 summarizes all these persistence measurement methods grouped into the two main groups.

Fig. 3

Pie chart grouping persistence measurement methods. Self-report and dispensing records methods are the three main groups

Quality assessment of validation studies for adherence measurement methods

Among all the reviewed studies, only one [88] specifically addresses the psychometric properties of a method for measuring adherence to statins: the MAT adapted for this medication type. This study assessed the internal consistency of the adapted MAT, reporting a Cronbach’s alpha of 0.66, indicating low internal consistency. Using the COSMIN tool for evaluation, this study demonstrates several limitations in the psychometric validation of the adapted MAT. The instrument’s reliability, assessed through internal consistency, is acceptable but low. Regarding content validity, the study did not conduct a comprehensive analysis or confirm the specific relevance of the items adapted for statins. Criterion validity is also insufficient, as the study did not compare the MAT against a reference standard, such as plasma statin levels. Construct validity was partially evaluated through concordance with other self-report methods; however, the low concordance suggests potential differences in the construct being measured, without an in-depth analysis. Finally, in terms of interpretability, the study provides a basic classification of adherence versus non-adherence but lacks validated cutoff points tailored to patients undergoing statin therapy.

Discussion

This systematic review identified various methods for measuring adherence to statins in secondary cardiovascular prevention, including prescription refill records, notably through the use of PDC and MPR, self-report tools (statin-adapted MAT [88], adherence VAS, 7-day recall, 24-h recall, MMAS-8 [23], MMAS-4 [105], SEAMS [107], and Gehi et al.’s adherence question [111]); and, less frequently, pill counting, electronic monitoring (MEMS [106] and GlowCap®), self-perceived adherence by physician, and direct measurement through detection of statins or their metabolites in blood or urine using LC–MS/MS [62]. For persistence, findings reveal that measurement methods are largely based on prescription refill records. Regarding the validity indicators of the methods used, none of the indirect methods included validity indicators specific to measuring adherence to statins, except for the statin-adapted MAT [88], which showed low internal consistency. Direct methods are considered valid as they provide acceptable validity indicators for the analytical technique employed.

Regarding the terminology used for adherence and persistence, it is not always consistent in the literature. Therefore, the nomenclature employed by various studies (adherence or persistence) was considered, regardless of whether it adhered strictly to the standard definitions [9]. Many studies use these terms interchangeably, even though adherence refers to the proportion of prescribed doses taken as directed, while persistence pertains to the continuation of treatment without interruptions. Additionally, other terms such as compliance and concordance have been used to describe different aspects of medication use. However, compliance often carries a negative connotation of subordination to the prescriber [112, 113], and concordance is frequently misinterpreted as synonymous with compliance [114,115,116]. This lack of clarity in terminology and measurement methods complicates the comparison of study results and leads to inconsistencies in conclusions about the effectiveness of adherence interventions. Greater consistency in terminology and methodology would help standardize the literature and facilitate evidence-based healthcare policy decisions.

Prescription and refill records are widely used tools for evaluating medication adherence, particularly for chronic treatments. The most commonly employed methods, PDC and MPR, are often assessed according to the interpretations of different study authors. The PDC is calculated as the percentage of days within a period during which the patient has the medication available, excluding duplicate supply days. This index is considered one of the most robust methods for measuring adherence, as it assesses whether the patient had the medication available each necessary day, excluding “overstocking” due to additional dispensations. Although PDC has not been validated in the traditional psychometric sense, it is an accepted and reliable method in adherence research due to its consistency, broad applicability, and positive correlation with clinical outcomes [117]. In comparison, the MPR measures the proportion of time the patient has had the medication available during a given period but can exceed 100%, indicating surplus medication due to early refills. The primary limitation of these methods is that they cannot confirm whether the patient actually ingests the medication. A study by Márquez-Contreras et al. [118] demonstrates that MPR calculated from electronic prescription data is effective in measuring adherence in hypertensive patients using MEMS as the gold standard (sensitivity of 87% and specificity of 93.7%), although MPR may overestimate adherence when there is refill overlap. In contrast, CMG has been used far less frequently to measure statin adherence, and its correlation with pill count has been weak [119], suggesting limitations in accuracy and use compared to other adherence methods, particularly for different medication types.

Self-report methods are straightforward and practical tools for assessing adherence from the patient’s perspective; however, their validity may be affected by recall bias or social desirability bias [88]. The MMAS-8 and its previous version, the MMAS-4, are widely used questionnaires in chronic conditions, though they were initially developed to measure adherence to antihypertensives. This questionnaire has been studied across numerous populations and contexts, with varying psychometric properties. In some studies, MMAS-8 has demonstrated good validity and reliability [24,25,26], while in others, its internal consistency and predictive adherence ability have been limited [27], suggesting that its accuracy may depend on the specific context and population. Notably, the original study by Morisky, which developed and validated the MMAS-8, has been retracted, raising concerns about the instrument’s validity and the integrity of its psychometric properties [23]. In contrast, the MAT allows not only for assessing adherence levels but also for identifying possible reasons or barriers to non-adherence, such as forgetfulness, side effects, lack of understanding about treatment, or difficulties accessing medication. Although it has been adapted for patients on statins [88], it exhibits moderate internal consistency and does not meet COSMIN [30] quality standards, warranting additional validation. The Gehi method is based on only three questions, which may not capture all aspects of patient adherence behavior [20]. This tool is simple and practical but has limited predictive validity compared to more detailed scales. Although some studies have used VAS to assess adherence and found correlations with other self-report methods, no universal validation confirms its precision and reliability across all contexts or medications. The VAS may be useful as a complementary measure, but its validity for accurately and reliably measuring adherence is often limited [120, 121]. Reminder methods, such as the 7-day and 24-h recalls, have been used in adherence studies to provide a quick and point-in-time picture of patient treatment adherence. While these methods may correlate with other adherence measures, they are less detailed and may suffer from recall bias, limiting their accuracy in long-term adherence assessments.

Pill count is an indirect method used in some adherence studies, although its application in statin adherence evaluation is scarce. It involves counting the remaining pills in the container to infer adherence. While it is a cost-effective method, its validity is limited, as it cannot guarantee that the patient took the recorded doses. Electronic devices like MEMS are considered a reference standard in adherence assessment, offering a detailed record of patient behavior. However, their validity is limited, as they do not confirm ingestion when the patient opens the container. GlowCap® operates similarly, recording openings but not ensuring ingestion. Both devices, while useful as approximations, have significant limitations and are not recommended as the sole adherence reference.

Direct methods are based on detecting the drug or its metabolites in bodily fluids. For statins, this approach allows confirmation of the medication’s presence in the body, ensuring it has been ingested and absorbed. However, this method has significant limitations: due to the half-life of statins, they may be undetectable in the blood shortly after the last dose, making them unsuitable for measuring short-term adherence. Establishing adherence thresholds or plasma concentration cutoff points is crucial to differentiate between adherence and non-adherence, as in the study by Kristiansen et al. [62], which calculated the theoretical plasma concentration range for statins in the steady state, classifying patients into three different adherence levels. Direct methods may be applicable in research or hospital settings, but their high cost and complexity make them less feasible for routine clinical practice. The present review shows that the main statins for which these methods were developed include atorvastatin, rosuvastatin, and simvastatin [62, 90, 92, 101].

Regarding persistence, it is generally measured through refill records, and although PDC may give an idea of adherence, it is not the ideal method for measuring persistence. PDC evaluates days covered by the medication but does not ensure uninterrupted treatment continuity, which is essential for accurate persistence measurement. Persistence is better assessed by analyzing periods without refill or long intervals without dispensing, providing a more realistic picture of patient behavior over the long term.

This review excluded studies in languages not using the Latin alphabet. However, this decision likely did not have a significant impact, as most reviewed studies were in English. Additionally, the search was conducted using only three databases, without accounting for gray literature or articles in other databases.

This study highlights the scarcity of validated adherence measurement methods for statins in secondary cardiovascular prevention, underscoring the need to develop a method applicable in clinical practice for this purpose, with consideration for gender perspectives. Most studies do not consider gender disparities in medication adherence measurement in cardiovascular diseases, despite evidence that gender may influence adherence behaviors and that being female is an independent predictor of non-adherence to certain medications, including lipid-lowering agents post-myocardial infarction [122,123,124]. Considering this factor would enable more personalized, gender-specific interventions and adapted clinical approaches, as biological and perceptual differences may influence adherence and persistence behaviors in statin treatment. Integrating a gender perspective could provide more comprehensive results aligned with each population group’s needs. Healthcare professionals must be familiar with tools to measure adherence to statins, given the severe implications of poor adherence in chronic conditions like CVDs. Failure to identify poor adherence as the underlying cause of inadequate disease control can lead to medications being incorrectly deemed ineffective, unnecessary treatment intensification, avoidable diagnostic testing, and even the misinterpretation of clinical trial results when adherence is not properly accounted for [28].

Consequently, we consider that direct methods, such as the detection of statins or their metabolites in blood or urine, are currently the most accurate tools available for measuring adherence to statins. However, their application in clinical practice is limited by cost and complexity. On the other hand, indirect methods such as prescription refill records and indices like PDC or MPR are practical and widely used, but do not guarantee that the medication has been taken. Notably, no indirect method has demonstrated sufficient validation metrics specific to statin adherence in secondary cardiovascular prevention, with the exception of the statin-adapted MAT, which showed low internal consistency. Thus, we highlight the need to develop and validate new tools that combine refill records and self-report. And we recommend using direct methods as the gold standard in research and validation studies to ensure reliable measurement of adherence. These tools should also incorporate a gender perspective, as gender differences can significantly influence adherence behaviors.

Conclusions

The methods used to measure adherence to statins in secondary cardiovascular prevention were mainly indirect, based on the review of prescription and supply records and self-report methods. Pill counting, electronic monitoring, and direct measurement through detection of statins and/or metabolites in blood or urine using the LC–MS/MS technique were used to a lesser extent. Regarding persistence, measurement methods were based on prescription refill records. None of the indirect methods identified was validated specifically for statin use in this population, and therefore, so their use to measure adherence to taking statins is not recommended. Based on current evidence, we consider that direct methods are the most accurate for measuring adherence and should serve as the gold standard in validation studies. In clinical settings, there is an urgent need to validate existing tools, originally developed for other conditions, and to develop new, mixed-method approaches that integrate refill data and self-report. We encourage future research and clinical efforts to prioritize the validation and implementation of reliable adherence measurement tools, as accurate assessment is essential for improving outcomes in cardiovascular disease prevention.