Not obtaining a medication the first time it is prescribed: primary non-adherence to cardiovascular pharmacotherapy

Primary medication non-adherence describes the situation when a first prescription for a new medication is never filled. Primary non-adherence is an important, yet understudied aspect of reduced effectiveness of pharmacotherapy. This review summarizes the frequency, impact, reasons, predictors, and interventions regarding primary non-adherence to cardiovascular/cardiometabolic drugs. The current literature reveals a high prevalence of primary non-adherence. The individual risk of primary non-adherence is determined on multiple factors, e.g., primary non-adherence of lipid-lowering drugs is higher compared to antihypertensive medications. However, the overall rate of primary non-adherence is > 10%. Additionally, this review identifies specific areas for research to better understand why patients forgo evidence-based beneficial pharmacotherapy and to explore targeted interventions. At the same time, measures to reduce primary non-adherence—once proven to be effective—may represent an important new opportunity to reduce cardiovascular diseases.


Introduction
Cardiovascular diseases (CVD) remain the leading cause of death globally [1].Pharmacological treatments can substantially reduce CVD morbidity and mortality.However, the effectiveness of these interventions is limited in cases of medication non-adherence and early discontinuation (nonpersistence).Moreover in an aging population, polypharmacy is a growing problem in the elderly with 50% of adults above the age of 79 years using ≥ 5 chronic medications and nearly 20% taking ≥ 10 medications, which increases the risk of adverse drug effects (ADE) and drug-drug interactions (DDI), and decreases medication adherence [2][3][4].
The vast majority of medication adherence research focuses on identifying factors and outcomes associated with either issues of implementation (secondary) adherence, i.e., whether or not patients refill their prescriptions after the initial fill and takes the medication as agreed and prescribed, exploring the quality of the execution and/or non-persistence [5,6], i.e., early discontinuation not intended by the prescriber [7,8].
Primary medication non-adherence describes the situation when a provider prescribes (or orders) a new medication for a patient and the order is never filled or dispensed and the patient does not obtain an appropriate alternative within an acceptable time.In this scenario, the patient either never takes the prescription to the pharmacy or does take or sends the prescription to the pharmacy, but never comes back to pick up the medication.Furthermore, initiation can be divided into patients having a new prescription that is not followed by dispensing versus prescribed and dispensed medication that is not followed by commencement of treatment (Fig. 1) [8][9][10][11].
Advances in technology, including electronic prescribing and the development of electronic medical records, have facilitated more rigorous population-based studies of primary non-adherence [12,13].As an important limitation, these methodologies are not able to measure whether a new medication is handed out by a pharmacist to the patient, but the drug is not taken by the patient.
Specific measures to reduce primary non-adherence may have a potential impact on atherosclerotic cardiovascular diseases (ASCVD).A comprehensive review of the frequency, impact, reasons, predictors, and interventions regarding primary non-adherence to cardiovascular/cardiometabolic drugs is absent.Therefore, this paper aims to close this gap.
The search identified 204 articles, 49 in MEDLINE and 155 in EMBASE, and 430 registered studies in Clinical-Trials.gov.Titles and abstracts were screened for suitability regarding all types of cardiovascular/cardiometabolic drugs including lipid-lowering medication, anticoagulants, and hypoglycemics.Descriptions of the methodology used to measure primary non-adherence, articles on study design, short summaries, editorials, conference abstracts, and letters addressed to the editor were excluded.The search was supplemented by a hand search of the reference lists of all selected full-text articles.After removing duplicates, 39 primary studies and two meta-analyses [14,15] were identified.
Descriptive statistical analysis and graphical illustration of the location and spread of estimates of primary nonadherence using boxplots were performed using IBM SPSS 28.

Results
The overall prevalence of primary medication non-adherence is high.By conservative estimation, the rate is > 10% [6,.
The overall rate of primary non-adherence among Danish residents in a GP setting was 9.3% with 4.7% for drugs addressing cardiovascular diseases.Polypharmacy (≥ 5 chronic drugs) and a diagnosis of ischemic heart disease were associated with higher rates of primary non-adherence.Older age and higher income were significantly associated with lower rates of primary non-adherence [19].
A recent study in the Netherlands included 65,877 subjects who received 181,939 new drug prescriptions issued by general practitioners.Overall, primary non-adherence was 11.5%.Within drug classes that are frequently used in primary care, primary non-adherence was 9.9%.Primary non-adherence was lowest for thyroid hormones (5.5%), 8.3% for cardiovascular drugs, and highest for nonsteroidal anti-inflammatory drugs (11.8%) and 12.8% for proton pump inhibitors.Several factors were associated with primary non-adherence, such as having comorbidities (OR 1.46, 95% CI 1.37-1.56for > 3 active diagnoses, compared to no active diagnoses) or reimbursement status (OR 2.78, 95% CI 2.65-2.92for not reimbursed drugs compared to fully reimbursed drugs) [18].
A study in Canada aimed to determine adherence to thienopyridine therapy after stent implantation, factors associated with suboptimal adherence, and association of suboptimal adherence with mortality.They evaluated 5263 older patients (> 65 years) who received drug-eluting stents (DES) and 6081 older patients who received bare metal stents (BMS).Primary non-adherence was observed among 6.9% in the DES group and 7.1% in the BMS group that did not fill a single prescription of a thienopyridine within 1 year of stent implantation.Premature discontinuation occurred in a progressive manner, with 28% in the DES group and 34% in the BMS group discontinuing therapy by 6 months.For DES patients, primary non-adherence (hazard ratio [HR] 2.68, 95% CI 1.77-4.07),12 months proportion of days covered < 80% (HR 2.39, 95% CI 1.67-3.43),and prematurely discontinuing therapy within 6 months (HR 2.64, 95% CI 1.60-4.35)were associated with an increased risk of death [42].
A recent study followed 12,257 DOAC patients.Of these, 10.4% were initially non-adherent; 12.8% for apixaban, 8.6% for dabigatran, and 10.8% for rivaroxaban.Patients aged ≥ 80 years showed lower odds of primary non-adherence compared to those aged < 65 years.A history of diabetes, hypertension, or stroke/transient ichemic attack was inversely associated with primary non-adherence [44].
Despite emphasis on efforts to prevent cardiovascular disease (CVD), 13-39% of people never fill a prescribed statin [45][46][47][48].In a study, patients with primary non-adherence to statin medications (n = 173) completed a self-administered cross-sectional survey that assessed their attitudes and beliefs related to primary non-adherence and to potential motivators for statin use [46].Ninety-nine patients (57.2%) never filled their prescription, and 74 (42.8%) filled but never took any statin.Over half failed to initially inform their prescriber they might not take the statin.Patients expressed desire for alternate treatment plans such as diet and/or exercise (77.4%) or natural remedies/dietary supplements (72.3%).56.6% of the patients worried about the possibility of statin dependence or addiction.15.6% of the patients noted that they would not take a statin based solely on CVD risk estimates, and hence for primary prevention [46].
Although proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i) were approved in 2015, their high cost has led to prescription restrictions or even prior authorization practices and use of PCSK9i in clinical practice has been low [49,50].Of patients receiving a prescription in the USA 2015/16, 47.2% ever received approval.Of those approved, 65.3% filled the prescription, resulting in 30.9% of those prescribed PCSK9i ever receiving therapy [49].
Table 1 summarizes the identified studies regarding primary non-adherence to cardiovascular/cardiometabolic drugs.
Considering the variability of studies (and their results) regarding different healthcare systems (e.g., countries or insurance plans with high vs. minor out-of-pocket drug costs), data sources, length of follow-up, and the impact of duplicate prescriptions, one can debate the precision of the estimates presented and their generalizability across therapeutic classes and patient populations.

Consequences of primary non-adherence
We were unable to identify studies that quantitate the effect of primary non-adherence for long-term cardiovascular/ cardiometabolic medication intake and for studies that correlate primary non-adherence with clinical events.The lack of published data on the clinical consequences of primary non-adherence identifies an urgent need for future research.

Interventions to reduce primary non-adherence
An RCT tested an automated outreach program to increase primary non-adherence to statins.Adult members of a health plan with no history of statin use within the past year who did not fill a statin prescription after 1-2 weeks were passively enrolled.The intervention group (n = 2606) received automated telephone calls, followed 1 week later by letters for continued non-adherence.The control group (n = 2610) received no outreach.Statins were dispensed to 42.3% of intervention and 26.0% of control participants (p < 0.001).Differences in the frequency of statin dispensations persisted up to 1 year (p < 0.001) [52].
In contrast, in a population of patients who had not picked up new prescriptions (asthma, hypertension, diabetes, or hyperlipidemia) after three calls from the pharmacy, additional nurse-directed outreach did not improve primary non-adherence [54].In another US-health system study, automated reminder calls had no effect on primary nonadherence.However, live calls from pharmacists decreased antihypertensive primary non-adherence significantly; the difference in abandoned prescriptions for antihypertensives was 6.9% (p < 0.0001), but only 1.4% for antihyperlipidemics (p > 0.1) [86].
Addressing long-term adherence rather than primary non-adherence only, a study found that the length of initial prescription for elderly patients after coronary angiography at hospital discharge was associated with long-term adherence.The strength of the association was consistent for each cardiovascular secondary prevention medication, i.e., angiotensin-converting enzyme inhibitor (ACEi)/angiotensin receptor blocker (ARB), beta blocker, and statin, even after adjusting for relevant clinical and sociodemographic factors.Most prescriptions at discharge covered less than 1 month.This may be based on a clinical assumption that short prescriptions encourage patients to attend early outpatient follow-up.Although early follow-up is essential to assess the patient and to address medication side effects, the vast majority of patients did have follow-up within 1 month, regardless of prescription length [87].
Observational studies suggest that post-myocardial infarction (MI) patients receiving longer initial prescriptions have greater long-term adherence to cardiovascular medications [88].The first, non-randomized interventional study to evaluate an intervention to standardize discharge prescriptions to prolonged duration to improve long-term medication adherence among post-MI patients, however, failed to meet its primary outcome.This was probably due to limited intervention fidelity, particularly at the pharmacy level [89].Although a 5.4% absolute increase in long-term medication adherence after 12 months post-implementation in the standardized prolonged discharge prescription forms plus education group was observed, this absolute difference was not statistically significant.The authors recommended that a similar intervention be implemented and evaluated in a larger population with fully electronic medical records coupled with policies to support the long-term dispensation of medications at the community pharmacy level [89].
Randomized studies testing the effects of more comprehensive interventions to reduce primary non-adherence are scarce.A rare example is the study of a complex multidisciplinary intervention involving patients in the decision-making process to improve primary non-adherence to cardiovascular disease and diabetes treatment in primary care patients (IMA-cRCT) [90,91].Results of the pilot phase identified opportunities for refining the intervention and improving the design of a planned cluster RCT to evaluate the effectiveness and cost-effectiveness of the intervention [92].

Discussion
Many patients do not fill their initial prescription for a cardiovascular/cardiometabolic medicine.At least one in ten newly prescribed medications do not make their way to the hands of the patients.Since 'drugs can't work if patients don't receive them' [42,55,60], primary non-adherence is a major healthcare problem.In addition to the negative impact on individual health, primary non-adherence is associated with a higher economic burden to the system in the short term, mainly due to higher productivity losses [93].
Most studies (62%) assessed primary non-adherence by following up (electronic) prescriptions for a 3-month duration or less.A sensitivity analysis of different durations (90 and 365 days) for defining primary non-adherence showed similar results to those when 30 days were applied [18].Pottegard et al. found that 65.2% of the patients redeemed their prescription for a new drug on the same date that the prescription was issued and that 89.3% of the patients redeemed their prescription by day 30.Most of the patients filled their prescriptions within the 1st week.In another study analyzing new medication orders, 75% of new prescriptions were filled within 7 days and 81% within 30 days [33].
Prescriptions at the expense of the German Statutory Health Insurance (SHI) system (insuring approximately 88% of the total population-that is, 73.3 million people) are valid for 28 days after the date of issue.Private prescriptions are usually valid for 3 months after the date of issue.
96% of SHI prescriptions issued at hospital discharge in Germany between 2018 and 2021 were redeemed within three working days.However, these were mainly drugs indicated for a short period after discharge, e.g., analgesics/ antiphlogistics, heparins, proton pump inhibitors, or antiinfectives.Long-term medications such as ramipril, bisoprolol, apixaban, or atorvastatin were prescribed less frequently at discharge (own unpublished data).
The review of the literature reveals a very heterogeneous set of data.For example, some of the observed primary non-adherence may reflect a wait-and-see approach [5].A patient with a first prescription for a statin and borderline LDL-cholesterol level and atherosclerotic risk may first seek to control LDL-cholesterol by dietary changes and increased exercise leading to weight loss, eventually.Another reason for primary non-adherence to chronic medications is starting a new treatment ('another pill') is a long-term, sometimes lifelong commitment.Patients need understanding of the nature of the medical problem as well as the pros and cons of (not) treating it with the most effective and safe medication selected.When out-ofpocket costs are considerable or health insurance is absent at all, financial consequences play a major role in primary non-adherence [5,14,15,58].
Primary non-adherence varied across therapeutic areas, drug classes, treatment duration, and individual drugs.One may expect lower primary non-adherence rates for conditions that present symptomatically (e.g., pain, anxiety, depression) or require short-term treatment (e.g., bacterial infections) compared to asymptomatic conditions with lifelong treatment indication (e.g., primary prevention of CVD).
Fig. 2 Boxplots of estimates of primary non-adherence to cardiovascular/cardiometabolic drugs.Outliers are not plotted [for citations of studies included, see Table 2] One may conclude that not all new medications for symptomatic conditions are comparably acceptable to patients.Compared to anti-infectives, new prescriptions for pain medications (analgesics, NSAIDs) may be less likely to be filled.
Clinicians may have good reason to assume that after a confirmed diagnosis (e.g., for heart failure, hypertension, ASCVD), primary intervention (e.g., percutaneous coronary intervention (PCI)), or clinical event (MI, stroke) in a shared decision environment, the patient follows the agreed treatment plan.Often, the treatment starts with the initial prescription and obtaining the new medication at the pharmacy.
We are unaware of prospective studies comparing primary non-adherence in primary vs. secondary prevention of CVD.Moreover, the precise indication is rarely available in electronic prescription/dispensing records and, hence, typically not reported in the publications included in our review.Lipid-lowering medications (mainly statins) when compared to other cardiovascular drug classes had, however, the highest rate of primary non-adherence.A plausible explanation is that statins are often used for primary prevention.Studies show that primary prevention, among others, is a risk factor for secondary non-adherence (implementation and discontinuation, Fig. 1) to statins (rate ratio 1.52, 95% CI: 1.50-1.53)[94].
Similarly, the reported primary non-adherence to PCSK9i prescribed for patients with high risk for ASCVD is high [49].In addition, approximately 7% primary non-adherence to thienopyridine therapy after stent implantation is concerning [42].
One problem is that primary non-adherence frequently goes undetected.The International Classification of Diseases (ICD) code Z91.1: "Patient's noncompliance with medical treatment and regimen" is rarely used.
A potential reason is that clinicians are frequently unaware of the extent or consequences of non-adherence.In addition, complex factors surrounding medication nonadherence make it a challenge for the healthcare professional to identify and confirm non-adherence in general and primary non-adherence in particular [95].Robust predictors for non-adherence to prescribed medication are absent and clinicians usually fail or significantly overestimate adherence behavior of 'their' patients ('no better than a coin toss') [96].Providers recognized non-adherence for less than half of patients whose pharmacy records indicated significant refill gaps, and often intensified blood pressure medications even when suspected serious non-adherence [96,97].
Hence, unidentified non-adherence may lead to unwarranted intensification of pharmacotherapy [95].Medical education and training for screening, diagnosing, or treating non-adherence is rare [95,98], but may represent an important opportunity to improve patient care.
Without access to electronic patient records including prescription and dispensing data, detection of primary non-adherence remains subject to trustful patient-provider communication [97].Occasionally, screening of blood plasma, urine, or saliva for antihypertensives is needed to confirm the diagnosis '(primary) non-adherence' (and, for example, 'resistant' hypertension) and to justify an intervention such as irreversible renal denervation [99][100][101][102].
Future research in different healthcare systems is needed to better understand why patients forgo evidence-based beneficial pharmacotherapy from the start and the clinical consequences of these decisions.We also need to improve our knowledge of detecting primary non-adherence and the efficacy and effectiveness of timely follow-ups and (digital) reminders [5,103] or community pharmacy-based or pharmacist-led interventions [54,61,[104][105][106].To our knowledge, a digital (eHealth) application addressing primary non-adherence is not available.
Of note, apart from poor medication adherence, therapeutic/guideline inertia [107,108] is independently associated with persistent poor blood pressure, LDL-cholesterol, or HbA1c control [109].Therapeutic inertia may have a higher impact on disease control compared to medication adherence [109].

Conclusions
Many patients fail to fill an initial prescription for a medicine.The rate of non-initiation of chronic cardiovascular/ cardiometabolic medication is estimated to be > 10%.The true prevalence of single dispensation of a new medication is not extensively studied, but is most likely higher.Research is needed to better understand why patients forgo evidencebased beneficial pharmacotherapy and to explore targeted interventions.At the same time, measures to reduce primary medication non-adherence-once proven to be effectivemay represent an important new opportunity to reduce cardiovascular diseases.
Novartis, Pfizer, and Sanofi, all outside the submitted work.Ulrich Laufs has received speaker honoraria and/or consulting fees from Amgen, Astra Zeneca, Berlin-Chemie, Daiichi Sankyo, Novartis, Pfizer, and Sanofi, all outside the submitted work.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http:// creat iveco mmons.org/ licen ses/ by/4.0/.