INTRODUCTION
Overactive bladder (OAB) syndrome is defined as urinary urgency, usually accompanied by frequency and nocturia, with or without urgency urinary incontinence in the absence of a proven infection or other obvious pathology [1]. The overall prevalence of OAB in East Asia is estimated to be between 10% and 21% of adults [2-4], with data from South Korea showing a prevalence of 10% in men and 9% in women [2]. Prevalence of OAB increases with age, particularly after 40 years [2].
Antimuscarinic agents have long been at the center of OAB treatment [5]. However, these agents alter binding of acetylcholine to muscarinic receptors, which are widely distributed throughout the body; they are therefore associated with a number of unwanted adverse events such as dry mouth and constipation [6, 7]. Some anticholinergics may also cross the blood-brain barrier, particularly among older patients with comorbidities and polypharmacy [8, 9]. Thus, some of these agents could increase the risk of delirium and dementia in older patients [10-12]. Many older patients may be taking multiple medications that have anticholinergic effects (e.g., antimuscarinics, antihistamines, antidepressants), resulting in a cumulative effect known as anticholinergic burden [11, 13]. Oral β3-adrenoreceptor agonists, such as mirabegron, provide another treatment option for the management of OAB without contributing to anticholinergic burden [14, 15]; a recent meta-analysis demonstrated that mirabegron has similar efficacy to most antimuscarinics, but is associated with a lower incidence of dry mouth, constipation, and urinary retention [16].
Little is known about the contribution of OAB treatments to total anticholinergic burden in older patients in South Korea. Understanding the significance of this burden in older patients with OAB could raise awareness of the challenges associated with anticholinergic drugs in this population regarding adverse effects and long-term use. The objectives of this study were to measure the cumulative anticholinergic burden in older outpatients in South Korea with and without newly started OAB medications, and to assess the contribution of antimuscarinics used to treat OAB to overall anticholinergic exposure.
MATERIALS AND METHODS
Study Design
This was a retrospective cohort study that utilized data from the National Health Insurance Service (NHIS) database in South Korea (ClinicalTrials.gov: NCT03903094). The study cohort included patients ≥65 years of age who had at least one outpatient visit (any cause) captured in the NHIS database between January 1, 2016 and June 30, 2016. The overall cohort was categorized into 2 exposure groups: patients with OAB and patients without OAB. The following agents were considered OAB medications: fesoterodine, imidafenacin, mirabegron, oxybutynin, propiverine, solifenacin, tolterodine, and trospium. The patient-specific index date for OAB patients (i.e., the start of study observation) was defined as the date of the first OAB medication dispensing record captured in the database. For non-OAB patients, the same index date of his/her matched OAB patient was used. Study outcomes were assessed over a 100-day follow-up period after the index date.
Patient Groups
The OAB group included patients who had at least one captured OAB medication dispensing record. The non-OAB group comprised 5 matched “controls” for each patient in the OAB group. Patients in the non-OAB group had the same index date assigned, birth year, and sex as the matched OAB counterpart, and a Charlson Comorbidity Index (CCI) within one unit above or below the baseline CCI at the index date for the patient with OAB. A random selection was applied if there were more than 5 equivalent matches, and matching was performed without replacement. If a control patient was prescribed an OAB medication after matching, their data were censored after the OAB medication prescription date.
Between January 1, 2016 and June 30, 2016, all patients had at least one dispensing record of any medication, and patients in the OAB group required a diagnosis of OAB (International Classification of Diseases, Tenth Revision [ICD-10] codes N32.8, N31.0, N31.1, N31.2, N31.8, N31.9, N32.9, N39.8, and N39.40) and received a first dispensation of an OAB medication. All patients were aged ≥65 on the index date and had no dispensing record for an OAB medication during the 6 months before the index date.
In both groups, patients who were hospitalized on the index date or who did not have an active insurance subscription on the index date or during the 6 months before were excluded. Patients who received onabotulinumtoxinA and/or surgical intervention as part of OAB treatment and patients diagnosed with stress incontinence or a urinary tract infection during the 6 months prior to or on the index date were excluded. Full inclusion and exclusion criteria are shown in Supplementary Table 1.
Endpoints
The study’s primary endpoints included 100-day cumulative anticholinergic cognitive burden (ACB) score during the follow-up period, measured using the ACB scale [17, 18], a validated scale for assessment of ACB [19]. Herein, the abbreviation ACB is used only to refer to the ACB score and is not used more generally to refer to anticholinergic burden. Primary endpoints also included prevalence of any listed anticholinergic use, prevalence of strong anticholinergic use (defined as ACB score ≥2), and number of anticholinergics dispensed per patient.
In the OAB group, the proportion of the cumulative ACB score attributable to OAB medication was assessed as a secondary endpoint. Anticholinergic burden was evaluated in the OAB and non-OAB groups, with the OAB group further divided into patients who received mirabegron only, patients who received an antimuscarinic with mirabegron, and patients who received antimuscarinics only.
Statistical Analyses
The 100-day cumulative ACB score (cACB100) was calculated as:
ACB, ACB score of anticholinergic medication; D, number of days supplied; i, patient i in the dataset; k, anticholinergic medication k dispensed to each patient; l, number of anticholinergic medications dispensed to each patient n sample size for each subgroup; TPD, total patient days of each group.
Per patient, the prevalence of anticholinergic use, prevalence of strong anticholinergic use, and number of anticholinergics dispensed during the study period were calculated.
The attribution proportion of OAB medication to total anticholinergic burden for the OAB group only was calculated as:
ACB, ACB score of anticholinergic medication; AOA, attribution proportion of OAB medication to total anticholinergic burden; D, number of days supplied; i, patient i in the subgroup; k, all anticholinergic medication k dispensed to each patient; l, number of anticholinergic medications dispensed to each patient; n, sample size for each subgroup; y, OAB medication with anticholinergic property y dispensed to each patient; z, number of OAB medications with anticholinergic properties dispensed to each patient.
Data for missing medication supply days were imputed using mode imputation. No other missing data were imputed.
RESULTS
Patient Population
In total, 2,796,947 eligible patients had at least one outpatient dispensing record captured in the NHIS database during the study period, of whom 14,036 were included in the final study cohort (Fig. 1). The OAB group included 2,360 patients with at least one OAB medication dispensing record during the study period. The non-OAB group included 11,676 matched patients.
The mean age (standard deviation) of the overall cohort was 74.2±5.9 years, and the average age of the patients who had received an OAB prescription was 74.9±6.7 years compared with 74.0±5.7 years for the non-OAB group (Table 1). The mean ages (standard deviation) of mirabegron-only and antimuscarinic-only users were 75.3±6.4 and 74.9±6.8 years, respectively. A greater proportion of the mirabegron-only group was aged >70 years (184 patients [79.0%]) compared with the other OAB treatment groups (antimuscarinic with mirabegron: 69 patients [76.7%]; antimuscarinics only: 1,539 patients [75.6%]). Overall, 8,734 patients (62.2%) were male, and the same proportions were noted for the OAB (1,469 patients [62.2%]) and non-OAB (7,265 patients [62.2%]) groups. For the OAB treatment groups, the highest proportion of males was seen in the mirabegron-only group (158 patients [67.8%]) compared with the antimuscarinics-with-mirabegron group (51 patients [56.7%]) and the antimuscarinics-only group (1,260 patients [61.9%]). The OAB group had a similar comorbidity profile to the overall cohort, with 52%–55% of patients in each group having <3 recorded comorbidities (Supplementary Table 2). Notable differences between exposure groups were observed in the 1,852 patients with dementia (13.2%; OAB group: 549 patients [23.3%]; non-OAB group: 1,303 patients [11.2%]) and the 3,926 patients who had a chronic pulmonary disease (28.0%; OAB group: 908 patients [38.5%]; non-OAB group: 3,018 patients [25.8%]). Further baseline characteristics are shown in Table 1.
Primary and Secondary Endpoints
The mean 100-day cumulative ACB score was 15.2 times higher among the OAB group than the non-OAB group (320.1 vs. 21.0, respectively; Table 2). Anticholinergics were widely used in the entire study population, particularly in the OAB group (2,287 patients [96.9%] vs. 3,921 patients [33.6%] for the non-OAB group). The prevalence of strong anticholinergic use was almost 4 times higher in the OAB group (2,234 patients [94.7%]) compared with the non-OAB group (2,817 patients [24.1%]). On average, 0.9 anticholinergics were dispensed per patient (Table 3). Of the 6,208 patients who received ≥1 anticholinergic, only 1,524 (24.5%) received more than 2 anticholinergic drugs.
The proportion of the ACB score attributable to OAB medications was highest in the antimuscarinic-only group (468,369/699,902 [66.9%]) compared with the total OAB group (485,472/755,320 [64.3%]) and the antimuscarinics-with-mirabegron group (17,103/28,110 [60.8%]). In the mirabegron-only group, 0% of the ACB score was attributable to OAB medications (Table 4 and Fig. 2).
DISCUSSION
It is important to accurately establish the contribution that antimuscarinic agents can have to the anticholinergic burden experienced by older patients with OAB. By quantifying this, it may be possible to indicate whether the burden experienced by patients will have a substantial effect on their cognition and thus their quality of life. In this database study, we assessed anticholinergic burden and the contribution of antimuscarinic agents to the overall burden in patients aged ≥65 years in South Korea.
In our study, the 100-day cumulative ACB score was 15 times higher in the OAB group than the non-OAB group. A total of 64.3% of the ACB score in the OAB group was attributable to OAB medications. As expected, patients taking mirabegron only had an ACB score attributable to OAB medications of 0. In contrast, the highest percentage ACB score attributable to OAB medication was observed in the group of patients who took antimuscarinics only (66.9%). These findings align with the results of a previous Japanese pharmacy claims database analysis, which found that both the prevalence of anticholinergic burden scale–listed anticholinergic use and mean ACB scores were higher in patients with OAB versus patients without OAB [20]. Similarly, the patients who received antimuscarinics had higher ACB scores than patients who received mirabegron only. Interestingly, this previous study also showed that similar ACB scores were noted for patients with OAB who had received mirabegron only and patients without OAB. The authors concluded that these results showed that mirabegron could be used as an alternative to antimuscarinics to potentially reduce anticholinergic burden.
The increased ACB score in the OAB group is a troublesome finding given that previous investigations have noted specific associations between higher ACB scores and negative outcomes. For example, a one-point increase in total ACB score was associated with greater healthcare utilization (number of inpatient admissions, outpatient visits, and emergency department visits) and a 13% increase in the risk of cognitive impairment in a large community-based study [21]. Furthermore, a 2-year UK study found that the use of anticholinergics that were included in the ACB scale was associated with cognitive impairment and increased mortality [22]. Given these findings, clinicians and policy stakeholders should consider an approach that minimizes anticholinergic burden for patients with OAB.
The average age of patients who had received an OAB prescription was marginally higher than patients who had not (74.9 years vs. 74.0 years). International geriatric guidelines recommend caution in prescribing antimuscarinic drugs to older patients owing to the risk of anticholinergic side effects in these individuals [23, 24]. Despite these recommendations, a previous Japanese pharmacy claims database investigation of patients with and without OAB found that prevalence of such prescriptions increased with age [20]. In our study, the highest proportion of patients aged ≥70 years was in the mirabegrononly group (79.0%) compared with the antimuscarinics-withmirabegron group (76.7%) and antimuscarinics-only group (75.6%). The higher age for patients who receive mirabegron has been typically observed in previous database studies [25, 26], which may indicate that patients receive antimuscarinics before mirabegron in their treatment pathway or that mirabegron is being preferentially used in older patients, possibly due to concerns over anticholinergic burden or the risk of urinary retention [16, 20, 27].
Males constituted almost two-thirds of the patient cohort in both the OAB and non-OAB groups. This is higher than the proportions noted in other real-world studies of OAB, which generally report proportions of 50% or less [2, 3, 20, 25, 26, 28, 29]. In terms of the OAB cohort, the highest proportion of males was seen in the mirabegron-only group (67.8%) compared with the antimuscarinics-only group (61.9%) and antimuscarinics-with-mirabegron group (56.7%). This finding may be due to a preference for mirabegron over antimuscarinics in older men, who commonly have benign prostatic hyperplasia. As stated above, antimuscarinics are associated with an increased risk of urinary retention [16, 27], and may therefore be unsuitable for men with benign prostatic hyperplasia due to the association of the condition with acute urinary retention [30]. Treatment guidelines recommend extreme caution when prescribing antimuscarinics to patients with a history of urinary retention [31].
The prevalence of dementia as a baseline comorbidity was ~10% higher in the OAB group compared with the non-OAB group and overall study population. The cause of this difference is unclear, but it may relate to the slightly higher age of the patients in the OAB group or the presence of comorbidities that required treatment with drugs that are associated with cognitive impairment, which may have included the use of antimuscarinics more than 6 months before the start of the study. Importantly, anticholinergic medication may increase the risk of cognitive impairment, dementia, and mortality in older patients [10, 32]. A real-world study of patients with OAB in Canada found that use of anticholinergic medications was associated with increased risk of new-onset dementia when compared with mirabegron [29], and an analysis from a phase 4, randomized, placebo-controlled study showed that 12 weeks of mirabegron treatment was not associated with any increase in cognitive side effects in patients aged ≥65 years [33]. Therefore, mirabegron could be considered a viable treatment option in patients with OAB and dementia.
The prevalence of chronic pulmonary disease at baseline was also approximately 10% higher in the OAB group compared with the non-OAB group. Previous studies have suggested an association between urinary incontinence and chronic pulmonary disease [34]. It has been hypothesized that age-related changes in smooth muscle, increased abdominal pressure while coughing, and altered micturition control may explain the association between these 2 conditions, although the exact physiological mechanisms involved in this association remain unclear [35]. In contrast, some clinical investigations have not observed a significant association between chronic pulmonary disease and OAB specifically [28].
This study is subject to limitations inherent to its nature. First, the definition used for OAB prescriptions during the 6 months leading to the index date and a 100-day follow-up period to measure cumulative ACB score may have restricted the number of potential patients with OAB who could be included in the study. However, our study still likely reflects the vast majority of the OAB population in South Korea. Secondly, despite the strengths of the ACB scale, it was developed in the United States and may therefore include anticholinergics that are not approved in South Korea or vice versa, meaning the anticholinergic burden highlighted in this study may be a conservative estimate. Thirdly, this study did not examine the consequences of anticholinergic use on patient outcomes. Further research should investigate the side effect profile experienced by patients with OAB in South Korea following treatment with antimuscarinics or mirabegron.
In conclusion, in patients aged ≥65 years in South Korea, the 100-day cumulative ACB score was 15 times higher in patients with OAB than in patients without OAB. The main source of anticholinergic burden was antimuscarinic OAB medications, many of which have strong anticholinergic properties; in this study, the β3-adrenoreceptor agonist mirabegron did not contribute to this burden.
Anticholinergic burden is associated with unwanted clinical effects, such as cognitive impairment and increased risk of dementia. As anticholinergic drugs are often used to treat conditions such as OAB in older individuals, it is important to understand this burden in patients with OAB compared with the general population. Clinicians should consider the anticholinergic burden of each individual patient in the use of pharmacotherapy to treat OAB.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement1
Print
