Adverse Outcomes of Antimuscarinics in Elderly Dementia

Date

2017-08

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Background- Dementia and Overactive bladder (OAB) are one of the most common comorbid conditions affecting older patients. Pharmacological management of OAB involves use of antimuscarinic drugs. Although antimuscarinics are effective, they vary in their safety and tolerability profiles due to differing selectivity for muscarinic receptor subtypes. Some of the antimuscarinics like oxybutynin, tolterodine, trospium, fesoterodine are non-selective as they have affinity for all muscarinic receptors, and others like darifenacin and solifenacin are selective due to their high affinity for M2/M3 receptors. Central adverse effects of antimuscarinics are a significant concern in dementia patients as these patients suffer from progressive cognitive decline due to damage to the cholinergic neurons system, and antimuscarinics can significantly affect patient outcomes. In addition, studies have consistently found strong evidence of increased medication-related morbidity and mortality due to anticholinergic use. However, very limited comparative safety data exists regarding antimuscarinics in the older patients, and even little is known about the safety of OAB medications in dementia patients.

Objectives - The specific objectives of the research were to: (1) To examine the prevalence and predictors of [A] antimuscarinic use among elderly dementia patients with OAB, and [B] non-selective antimuscarinic use among elderly dementia patients with OAB; (2) To evaluate the risk of falls/fractures due to antimuscarinic use among elderly dementia patients; and (3) To assess the risk of all-cause hospitalization due to antimuscarinic use among elderly dementia patients.

Methods - A new user retrospective cohort study design was conducted using Inovalon’s Medical Outcomes Research for Effectiveness and Economics Registry (MORE2 Registry®) from 2011-2015. The base cohort for objective 1 involved elderly patients with a diagnosis of dementia and OAB. New-users were identified with index dates from July 1, 2011 to December 31, 2015 with a 6 month washout period; non-users were identified with randomly assigned index dates from July 1, 2011 to December 31, 2015. For part A - new users and non-users were one-to-one matched without replacement, on age at index date (± 1 year), gender, dementia diagnosis and time lapse between dementia diagnosis and index date. For part B – all new users identified (categorize as non-selective vs selective medication users) were used as the study sample for further analysis. Descriptive statistics were used to determine the prevalence of antimuscarinic medication utilization; conditional logistic regression model and logistic regression model were used to identify the predictors associated with new use of antimuscarinic medication and non-selective antimuscarinic medication use in the study cohort.

For objective 2 and 3, the base cohort involved elderly patients with a diagnosis of dementia. New-users were identified with index dates from July 1, 2011 to December 31, 2015 with a 6 month washout period. The index medication categorized the patients as non-selective and selective medication users. Patients taking non-selective antimuscarinics (oxybutynin, tolterodine, trospium, fesoterodine) were matched with the patients taking selective antimuscarinics (solifenacin, darifenacin) using the Greedy 5 - 1 matching technique. The study evaluated the relative risk of falls/fractures and all-cause hospitalizations in a 6-month follow-up period using the Cox proportional-hazards regression model stratified on matched pairs.

Results - For the first objective, the initial cohort included 75,498 patients diagnosed with dementia from January 2011 to December 2015; 71,222 patients were diagnosed with OAB. Out of these, 4,837 new users and 21,493 non-users of antimuscarinic medications were identified. The 4,837 patients found to be new users of antimuscarinic medications, resulted in an incidence of 6.8%. For part A, after applying the 1:1 matching criteria, the study cohort reduced to 8,160 patients with 4080 patients in each group. Multivariable analysis revealed that patients with ‘other health plans’ that included [private fee for service (PFFS) plans, point of service (POS) plans, high-deductible health plans (HDHPs), exclusive provider organization (EPOs) plans] and myocardial infarction had decreased likelihood of receiving a new antimuscarinic medication. Conversely, patients with health maintenance organization (HMO) health plan, preferred provider organization (PPO) health plan, anxiety, gastroesophageal reflux disease (GERD), muscle spasms/low back pain, benign prostatic hyperplasia (BPH), urinary tract infection (UTI), vulvovaginitis and baseline anticholinergic load had increased likelihood of receiving a new antimuscarinic medication. For part B, the 4,837 patients found to be new users of antimuscarinic medications included non-selective antimuscarinic medication users 3,635 [75.2%] and selective antimuscarinic medication users 1,202 [24.8%]. Multivariable analysis revealed that patients with 75-85 years of age, muscle spasms/low back pain and Parkinson’s disease had a decreased likelihood of receiving non-selective antimuscarinic medication. However, patients with preferred provider organization (PPO) as their health plan had an increased likelihood of receiving a non-selective antimuscarinic medication.

For objective 2, the initial cohort included 75,498 elderly dementia patients from July 2011 to December 2015. A total of 4,633 new users of antimuscarinic medications was obtained after applying inclusion and exclusion criteria; 3,425 [73.93%] were non-selective antimuscarinic medication users and 1,208 [26.07%] were selective antimuscarinic medication users. The study sample consisted of 2,402 (1,201 non-selective and 1,201 selective) new users of antimuscarinic medications after matching on the propensity scores. Stratified cox regression model analysis revealed statistically significant difference between non-selective and selective users with respect to risk of falls/fractures [Hazard Ratio (HR), 1.34; 95% confidence interval (CI), 1.06–1.70]. Significant increase in the risk of falls/fractures was found among elderly dementia patients who initiated treatment with oxybutynin [HR 1.32; 95% CI 1.03-1.71) when compared with patients who initiated treatment with solifenacin.

For objective 3, a total of 4,633 new users of antimuscarinic medications as obtained from the objective 2 were used for analysis, with 3,425 [73.93%] non-selective antimuscarinic medication users and 1,208 [26.07%] selective antimuscarinic medication users. The study sample consisted of 2,398 (1,199 non-selective and 1,199 selective) new users of antimuscarinic medications after matching on propensity scores. Stratified cox regression model analysis revealed no significant difference between selective and non-selective users with respect to risk of all-cause hospitalizations [Hazard ratio (HR) 1.04; 95% confidence interval (CI) 0.87-1.23]. All the sensitivity analyses were found to be consistent with the main findings.

Conclusions - The study found that approximately 7% elderly dementia patients with OAB were incident antimuscarinic medication users with most (75%) of them using non-selective agents. Several predisposing, enabling and need factors associated with the use of antimuscarinic medications in general and non-selective in specific. The prescribing patterns of antimuscarinic medications in elderly dementia patients with OAB raised a significant quality of care concern. Given the high prevalence of OAB among dementia patients and the high utilization of non-selective medications relative to the selective ones, it is imperative to monitor the use of antimuscarinics, specifically the non-selective medications.

From objective 2, the study found significantly increased risk of falls/fractures with use of non-selective antimuscarinic medications when compared to selective medications among the elderly dementia patients. However, no significant difference in the risk of all-cause hospitalizations between users of selective and non-selective antimuscarinics among elderly dementia patients was identified in objective 3. Given the high prevalence of OAB among dementia patients, there is a significant need to be cautious while prescribing non-specific antimuscarinic medications among elderly dementia patients. There is a need for further evaluation of other adverse effects of non-selective antimuscarinics due to their muscarinic receptor affinity in the geriatric population.

Description

Keywords

Dementia, Overactive Bladder, Non-selective antimuscarinics, Antimuscarinics, Muscarinic receptor selectivity

Citation