Analysis of pharmacotherapy in patients with atrial fibrillation and stage 4 chronic kidney disease in real-world clinical practice: application of STOPP/START criteria

Introduction

In clinical practice, cardiovascular diseases, particularly atrial fibrillation (AF), are often combined with chronic kidney disease (CKD) [1]. This comorbidity is largely due to the presence of common risk factors such as arterial hypertension (AH), diabetes mellitus (DM), and coronary artery disease (CAD) [2].

The relationship between AF and CKD is bidirectional and extremely complex, forming the so-called “cardiovascular‑renal continuum”. According to the literature, CKD stage 3–5 is detected in 30% of patients with AF [3]. It should be noted that CKD has a significant impact on the hemostatic system, increasing the risk of both thromboembolic events and bleeding [4]. For example, according to the study by Arnson Y et al. [5], the risk of bleeding in patients with AF increases as CKD progresses: from 0.89 events per 100 patient‑years in stage C1 CKD to 4.91 in stages C4‑C5.

The combination of AF and CKD, especially with a decline in glomerular filtration rate (GFR), poses a significant clinical problem for the choice of rational pharmacotherapy. Particular difficulty arises in prescribing antithrombotic therapy. According to clinical guidelines, oral anticoagulants (OACs) are indicated for the prevention of cardioembolic events in AF [6]. At the same time, issues related to the prescription of anticoagulant therapy for stroke prevention in patients with AF and advanced CKD remain insufficiently studied [7]: the vast majority of clinical trials included patients with AF and concomitant CKD only up to stage 3 in severity, whereas the efficacy and safety of anticoagulant therapy in patients with stage 4 CKD have been poorly studied [8].

A characteristic feature of older patients is multimorbidity, i.e., the simultaneous presence of two or more chronic diseases [9]. As life expectancy increases, so does the number of elderly polymorbid patients with polypharmacy (prescription of 5 or more drugs): according to studies, polypharmacy occurs in more than 50% of older age groups [10]. Prescribing a large number of drugs leads to an increase in drug interactions and the risk of adverse drug reactions, which in turn negatively affects the efficacy and safety of pharmacotherapy [11]. It should also be noted that older patients have a number of age‑related changes in various organs and systems that affect drug pharmacokinetics and pharmacodynamics (slowed metabolism, inadequate nutrition, impaired blood‑brain barrier permeability, etc.), which must be taken into account when choosing a therapeutic strategy [12].

The problem of optimizing pharmacotherapy and minimizing inappropriate drug prescribing in persons aged 65 years and older has remained relevant for decades. The first tool for assessing the quality of drug therapy in patients aged 65 and older appeared in 1991 [13]. It was developed by a group of scientists led by geriatrician Mark H. Beers, initiating many years of practice in using various criteria to optimize pharmacotherapy [13]. Every year, there is modernization and development of more modern tools to assist practitioners in selecting therapeutic strategies to avoid adverse drug reactions. Various questionnaires have emerged, such as the MAI (Medication Appropriateness Index) for assessing the quality of drug prescribing, as well as other tools needed to study the adequacy of pharmacotherapy, such as FORTA (Fit fOR The Aged) and STOPP (Screening Tool of Older Person’s Prescriptions) / START (Screening Tool to Alert doctors to Right Treatment) criteria [14–16].

The first modified STOPP/START criteria were published in 2008 [16]. The convenience of this tool lies in the fact that the criteria themselves are divided by physiological organ systems, which facilitates searching and increases ease of use [17]. The second version from 2015 includes 80 STOPP criteria and 34 START criteria [18].

In view of the above, the aim of this study was to assess the compliance of prescribed pharmacotherapy with the STOPP/START criteria in hospitalized patients aged 65 years and older with a combination of AF and stage 4 CKD.

Materials and methods

Study design: retrospective, cohort. A retrospective analysis of medical records of patients from therapeutic departments (cardiology, internal medicine, neurology) of a multidisciplinary hospital in Moscow was performed.

We selected medical records of 86 patients aged 65 years and older with a combination of AF and stage 4 CKD who underwent inpatient treatment from July 1, 2018, to June 30, 2019; they formed Group 1 (27 patients). Group 2 (59 patients) consisted of patients who underwent inpatient treatment in the same multidisciplinary hospital in the same departments from July 1, 2022, to June 30, 2023.

Inclusion criteria: 1) presence of any form of AF, including patients with valvular pathology and reversible causes of AF; 2) presence of stage 4 CKD; 3) patient age at admission ≥65 years.

Exclusion criteria: age <65 years.

The complete clinical characteristics of the patients included in the study are presented in Table 1.

The median age of the patients corresponded to advanced age. Women predominated in both groups (81.5% and 78% women in Groups 1 and 2, respectively, p = 0.93). The most common form of AF in both Group 1 and Group 2 was permanent AF (40.7% and 57.6%, respectively, differences between groups not statistically significant). Notably, in Group 2 compared to Group 1, the median final HAS‑BLED score was significantly higher (p = 0.046, Table 1). In addition, blood pressure (BP) values were significantly higher in Group 1: both systolic (p < 0.001) and diastolic (p = 0.001). Gastric and/or duodenal ulcer (DU) was more common in Group 1 than in Group 2, but the difference did not reach statistical significance (25.9% vs. 5%, respectively, p = 0.05). Musculoskeletal diseases were also significantly more common in Group 1 compared to Group 2 (77.8% vs. 35.6%, respectively, p < 0.001).

Table 1. Clinical characteristics of patients

Notes: AH — arterial hypertension; CABG — coronary artery bypass grafting; DBP — diastolic blood pressure; DU — duodenal ulcer; CAD — coronary artery disease; PICS — post‑infarction cardiosclerosis; SBP — systolic blood pressure; GFR — glomerular filtration rate; FC — functional class; AF — atrial fibrillation; CKD — chronic kidney disease; HR — heart rate; * — high risk of thromboembolic complications — CHA₂DS₂‑VASc score ≥3 points for women and ≥2 points for men; * — differences between groups are statistically significant.

All patients included in the study underwent an analysis of their drug prescription structure, including the use of the STOPP/START criteria (version 2.0 from 2015) [18].

Statistical analysis was performed using IBM SPSS Statistics 27. Normality of distribution was assessed using the Shapiro–Wilk test. For non‑normally distributed variables, data are presented as median (Me) and interquartile range (25th and 75th percentiles); for normally distributed variables, as mean (M) and standard deviation (SD). Categorical variables were compared using Fisher’s exact test. Differences between non‑normally distributed variables were assessed using the Mann–Whitney U test; differences between normally distributed variables were analyzed using Student’s t‑test. A p value <0.05 was considered statistically significant.

Results

When analyzing drug prescriptions for compliance with the STOPP/START criteria, we identified a total of 178 START criteria (75 in Group 1 and 103 in Group 2) and 113 STOPP criteria (30 in Group 1 and 84 in Group 2). STOPP criteria were statistically significantly more common in Group 2 compared to Group 1 (64.4% vs. 51.9%, respectively, p = 0.03, Table 2). It should also be noted that the frequency of START criteria was lower in Group 2 than in Group 1, although the difference was not statistically significant (76.3% vs. 96.3%, respectively, p = 0.39, Table 2).

Table 2. STOPP/START criteria identified in patients with atrial fibrillation and stage 4 chronic kidney disease hospitalized in 2018–2019 and 2022–2023

Notes: AF — atrial fibrillation; CKD — chronic kidney disease; * — differences between groups are statistically significant (p < 0.05).

The most common START criterion in both groups was the absence of statin therapy in patients with a documented history of coronary, cerebral, or peripheral vascular disease. This criterion was statistically significantly more frequent in Group 1 compared to Group 2 (85.2% vs. 23.7%, respectively, p < 0.001) (Table 3).

The second most common START criterion was the absence of clopidogrel prescription in patients with a history of ischemic stroke or peripheral vascular disease (Group 1: 37.0%, Group 2: 23.7%, p = 0.39).

Statistically significant intergroup differences were found for the criterion “Warfarin for AF”: its frequency was 40.7% in Group 1 compared to 11.9% in Group 2 (p = 0.006) (it was the third most common in Group 1). These data indicate a significant improvement in thromboprophylaxis approaches: whereas in the 2018–2019 cohort it was often not performed, in the 2022–2023 cohort OACs were predominantly used for this purpose.

Table 3. The most common identified START criteria in patients with atrial fibrillation and stage 4 chronic kidney disease

Notes: ACE inhibitors — angiotensin‑converting enzyme inhibitors; GFR — glomerular filtration rate; AF — atrial fibrillation; CKD — chronic kidney disease; COPD — chronic obstructive pulmonary disease; CHF — chronic heart failure; * — differences between groups are statistically significant (p < 0.05).

The most common STOPP criterion was the prescription of drugs with anticholinergic activity for chronic constipation (18.5% and 30.5% of patients in Groups 1 and 2, respectively, p = 0.37). Prescription of drugs with anticholinergic activity for chronic glaucoma was also frequent (14.8% and 6.8% of patients in Groups 1 and 2, respectively, p = 0.25) (Table 4).

In Group 2, the STOPP criterion “Angiotensin II receptor blockers (ARBs) in hyperkalemia” ranked second in frequency, being identified in 23.7% of cases. Its prevalence was statistically significantly different from that in Group 1 (3.4%), p = 0.03 between groups (Table 4).

Table 4. The most common STOPP criteria identified in patients with atrial fibrillation and stage 4 chronic kidney disease

Notes: AH — arterial hypertension; ACE inhibitors — angiotensin‑converting enzyme inhibitors; ARBs — angiotensin receptor blockers; NOACs — novel oral anticoagulants; NSAIDs — non‑steroidal anti‑inflammatory drugs; DM — diabetes mellitus; GFR — glomerular filtration rate; AF — atrial fibrillation; CKD — chronic kidney disease; COPD — chronic obstructive pulmonary disease; CHF — chronic heart failure; * — differences between groups are statistically significant (p < 0.05).

Discussion

This study examined two groups of patients who underwent inpatient treatment (in the same therapeutic departments of one multidisciplinary hospital in Moscow) during different time periods: 2018–2019 and 2022–2023. In our study, the median age of patients corresponded to the advanced age group (75–89 years) and centenarians (90 years and older), respectively: median age in Group 1 was 87 [82;89] years, in Group 2 — 91 [87;93] years (p = 0.31 between groups). Analysis of comorbidities revealed that all patients included in the study had at least four diseases/conditions simultaneously (AF, stage 4 CKD, AH, and CHF). According to various literature data, up to 99% of patients with AF are polymorbid [19]. The prevalence of multimorbidity among patients with AF increases statistically significantly with age [20].

Pharmacotherapy in elderly patients is particularly complex and requires careful monitoring to ensure a favorable benefit‑risk balance, especially in polymorbid patients [21]. Elderly patients (≥65 years) are most vulnerable to inappropriate pharmacotherapy due to age‑related changes in drug metabolism [22]. Prescribing a second drug in such patients already increases the risk of adverse drug reactions by 10%, and when taking more than 10 drugs simultaneously, drug interactions occur in 100% of cases [23]. At the same time, optimization of pharmacotherapy can reduce the rate of preventable hospitalizations associated with adverse drug reactions, which according to several studies ranges from 30% to 50% [24–25]. This approach is key not only to reducing economic costs but also to improving prognosis and increasing patient life expectancy [26].

The problem of polypharmacy in elderly patients led to the development of drug prescribing screening tools designed to optimize pharmacotherapy. Currently, there is a transition from the previously dominant Beers criteria to the newer, more relevant STOPP/START criteria [16, 27]. In the patient groups we studied, the majority of patients had drug prescriptions that did not comply with the STOPP/START criteria. Our data analysis revealed that in most cases, patients were not prescribed potentially recommended drugs (START criteria), whereas potentially dangerous drugs (STOPP criteria) were prescribed to a lesser extent.

At least one START criterion was present in 26 (96.3%) patients in Group 1 and in 45 (76.3%) patients in Group 2. The most common START criterion was the absence of statin prescriptions in patients with coronary, cerebral events and/or peripheral artery disease (85.2% and 23.7% of patients in Groups 1 and 2, respectively, p = 0.31). It is well known that lipid‑lowering therapy in such patients reduces the risk of stroke by 60% and coronary artery disease (CAD) by 17% [27]. According to the Russian Ministry of Health clinical guidelines for stable CAD (2024) [28], lipid‑lowering therapy (statins) is indicated for all patients with stable CAD (level of evidence I A). It is worth noting that in this study, this criterion was significantly more common in the 2018–2019 patient group, indicating a prioritization of prescribing life‑saving drugs to improve the prognosis of patients with AF and stage 4 CKD in subsequent years.

Another common START criterion was the non‑prescription of clopidogrel in patients with a history of ischemic stroke or peripheral vascular disease (37.0% and 23.7% of patients in Groups 1 and 2, respectively, p = 0.39). However, according to current clinical guidelines in the Russian Federation, the use of this criterion is limited. This is because in elderly patients with AF in this clinical situation, the standard of therapy is anticoagulant monotherapy [6, 29].

Warfarin for AF was statistically significantly more often not prescribed to patients in Group 1 (40.7% and 11.9% of patients in Groups 1 and 2, respectively, p = 0.006). According to the Russian Ministry of Health clinical guidelines for atrial fibrillation and flutter in force at that time (2020) [6], the decision to prescribe anticoagulant therapy is based on the CHA₂DS₂‑VASc score. Long‑term oral anticoagulation for thromboembolism prophylaxis is indicated for a score of ≥2 points in men and ≥3 points in women [6]. In our study, the median final CHA₂DS₂‑VASc score in both groups was 6 points, which mandates anticoagulant therapy. The positive trend observed — a statistically significant improvement in adherence to clinical guidelines (Group 2 compared to Group 1) — is likely due to the expanded clinical use and increased availability of OACs.

Analysis of STOPP criteria revealed that more than half of the patients in each group had at least one such criterion. Statistically significant differences between groups were found in the overall frequency of STOPP criteria: they were significantly more common in Group 2 (64.4% (n=38) vs. 51.9% (n=14) in Group 1, p = 0.03).

The most common criterion in both cohorts was the use of drugs with anticholinergic activity in patients with chronic constipation: its frequency was 30.5% (n=18) in Group 2 and 18.5% (n=5) in Group 1, but this difference did not reach statistical significance (p = 0.37).

Key diagnostic criteria for chronic constipation include symptoms of obstipation lasting at least six months, including a reduction in stool frequency to less than 3 times per week in the last 3 months [30]. Chronic constipation occurs 2–3 times more often in women, and with age, according to studies, the prevalence of constipation increases to 50% among persons aged over 65 years and up to 74% by the end of life [31]. The development of constipation in elderly patients is due to a combination of factors: changes in diet and eating habits, significant reduction in physical activity, age‑related decline in intestinal motility, and others [32]. Drugs with anticholinergic activity play a special role, as they exacerbate existing constipation symptoms: their main mechanism is blockade of muscarinic receptors, which prevents the effects of acetylcholine, one consequence of which is inhibition of intestinal motility, leading to worsening constipation. In addition to selective M‑anticholinergics, other drugs (e.g., amitriptyline, diphenhydramine) also possess anticholinergic activity [33]. Therefore, the use of drugs with anticholinergic activity should be avoided in patients suffering from chronic constipation [34].

On the other hand, angiotensin receptor blockers (ARBs) were statistically significantly more often prescribed in Group 2 in the presence of hyperkalemia. Currently, this class of drugs is primarily recommended for the treatment of AH, CHF, DM, and renal dysfunction [35]. In patients with progressive CKD, a tendency to hyperkalemia is exacerbated by the use of ARBs and other drugs affecting potassium metabolism [36]. However, a study by Leon S. et al. [37] showed that discontinuation of renin‑angiotensin‑aldosterone system (RAAS) inhibitors was associated with higher mortality and cardiovascular event rates compared to continuation of this class of drugs. Given the above, there is a need to clarify this criterion; perhaps only a certain level of hyperkalemia (not just above formal normal values) should be a contraindication to the prescription of RAAS blockers in CKD patients. The current Russian Ministry of Health clinical guidelines for CKD (2024) [38] state that a transient decrease in GFR and the development of hyperkalemia during treatment with RAAS inhibitors are most pronounced in CKD stages 4–5, but these effects are not predictors of adverse renal outcomes in the general population. On the other hand, in clinical practice there is a cohort of patients in whom these effects become clinically significant, necessitating prevention, active monitoring and, if necessary, specific therapy or adjustment of the treatment regimen [38, 39].

Conclusion

Thus, the results of the study demonstrate that in real‑world clinical practice, pharmacotherapy in polymorbid patients over 65 years of age with AF and stage 4 CKD requires further optimization.

On average, half of the patients in both groups had STOPP criteria (drugs with anticholinergic activity for chronic constipation were the most common inappropriate prescriptions). At the same time, cases corresponding to START criteria predominated, reflecting the problem of underuse of therapy needed for secondary prevention and improvement of the duration and quality of life of older patients.

The obtained results allow us to conclude that there is a need for broader and more regular use of the STOPP/START criteria in multidisciplinary hospital settings to optimize pharmacotherapy in older polymorbid patients.