EURURO Vol. 72 No. 3

390 476

1. Introduction
Overactive bladder (OAB) is a complex of lower urinary tract
storage symptoms characterised by urgency, with or
without urgency incontinence, and often accompanied by
frequency and nocturia in the absence of proven infection or
other pathology . The condition is common, affecting an
estimated 12% of adults
In the UK, pharmacotherapy is used to manage OAB
symptoms if lifestyle, behavioural, and conservativemeasures
fail . Antimuscarinics are the mainstay of pharmacother-
apy for OAB. However, systemic blockade of muscarinic
receptors leads to common, bothersome class-related adverse
events such as dry mouth, constipation, and headache
Mirabegron, a selective b3-adrenoceptor agonist, is an
alternative treatment option with established efficacy in
patients with OAB Overall rates of treatment-emergent
adverse events for mirabegron are similar to those observed
for antimuscarinics but the risk of dry mouth and
constipation is significantly lower for mirabegron .
Persistence (time from initiation to discontinuation of
therapy ) and adherence (extent to which a patient acts
in accordance with the prescribed interval and dose of a
dosing regimen ) with oral antimuscarinics are recog-
nised as among the lowest for medications used for
common chronic conditions and fall rapidly after
treatment initiation Depending on the antimuscarinic
prescribed, 65–86% of patients discontinue therapy after
1 yr . Real-world data, mainly from retrospective
medical claims databases, suggest that treatment persis-
tence with mirabegron may be greater than for antimus-
carinics in OAB
The objective of this study was to compare treatment
persistence and adherence for mirabegron compared to
tolterodine extended-release (ER) over a 12-mo period,
consistent with a previous long-term comparative study
, and with other antimuscarinics in UK clinical practice.
Our study was based on prescription records from a large
primary and secondary care database in the UK, Clinical
Practice Research Datalink (CPRD). One other UK-specific
comparative study of persistence for mirabegron versus
antimuscarinics is ongoing, but has only been presented in
preliminary form to date . As well as evaluating
persistence in the total study population, we also looked at
specific predefined subgroups of interest (ie, treatment-
naı¨ve, treatment experienced and elderly patients) and
applied matching to control for key baseline characteristics.
2. Patients and methods
2.1. Study design and data source
This was a retrospective, longitudinal, observational, cohort study of
patients who received a prescription for a target OAB medication in UK
clinical practice. The primary objective was to compare persistence on
treatment betweenmirabegron and tolterodine ER. Secondary objectives
included comparing persistence and adherence between mirabegron
and other antimuscarinics, and to describe patient characteristics that
affected persistence and adherence.
Anonymised data were taken from the CPRD GOLD, a large, nationally
representative, primary care research database that collates medical
records from 674 general practices across the UK. It includes data for
approximately 4.4 million active patients (ie, alive, currently registered)
who meet database quality criteria .
Approval for the study protocol was obtained from the CPRD
Independent Scientific Advisory Committee (protocol 16_097R; approv-
al date July 6, 2016). The study was conducted in compliance with
national and European Union requirements ensuring the rights of
participants in noninterventional studies.
2.2. Study population
Adults aged 18 yr with at least one prescription for a target drug issued
between May 1, 2013 and June 29, 2014 (selection period) were eligible
( Supplementary Fig. 1). The selection period was based on the
availability of mirabegron (approved in Europe in January 2013) and
to allow at least 12-mo patient follow-up before analysis of the database.
The target drugs were mirabegron, darifenacin, fesoterodine, flavoxate,
oxybutynin ER or immediate-release (IR), propiverine, solifenacin,
tolterodine ER or IR, and trospium chloride. The first prescription date
for a target drug initiated during the selection period was defined as the
index date, and the drug was designated the index drug. Only patients
with a new prescription were included; patients with a prescription for
the index drug in the 12 mo before the index date were excluded.
Patients were required to have at least 12 mo of continuous enrolment in
CPRD before (pre-index period) and after (post-index period) the index
date. Prescriptions for combination therapy with two target drugs at the
index date or a 5a-reductase inhibitor (suggestive of benign prostatic
hyperplasia due to an enlarged prostate) during follow-up were not
permitted.
2.3. Endpoints
Persistence on treatment was assessed using two endpoints: median
time to discontinuation (primary endpoint) and the persistence rate at
12 mo. Adherence was also assessed using two endpoints: medication
possession ratio (MPR) at 12mo and adherence rate at 12mo. Definitions
of the endpoints are provided in Supplementary Table 1.
2.4. Statistical analysis
Analyses included all patients meeting the inclusion criteria. Analyses
were also stratified according to the following prespecified subcohorts at
index date: treatment-naı¨ve (no prescription for any target drug during
the pre-index period); treatment-experienced (prescription for at least
one non-index target drug during the pre-index period); and elderly
patients ( 65 yr).
Time to discontinuation was calculated using Kaplan-Meier survival
analysis, and differences between cohorts were assessed via log-rank
test; hazard ratios (HRs) are reported for comparisons between
mirabegron (reference) and antimuscarinics. Patients were censored if
they reached the end of follow-up without discontinuation. Cox
proportional-hazards regression models, adjusted for potential con-
founding factors (gender, age, Charlson index, treatment status,
hypertension, polypharmacy), were used to compare cohorts (expressed
as HR and 95% confidence interval [CI]). For the primary analysis, OAB
treatment was defined as discontinued if the maximum allowable gap
duration (MAGD) was at least 1.5 times the intended duration of the
most recent prescription. Sensitivity analyses were conducted using a
MAGD ratio of 2 and fixed definitions of 30, 45, 60, and 90 d. Logistic
regression models, adjusted for potential confounding factors, were used
to compare 12-mo persistence rates between cohorts (expressed as odds
ratio [OR] and 95% CI). Proportions of persistent and adherent patients
E U R O P E A N U R O L O G Y 7 2 ( 2 0 1 7 ) 3 8 9 – 3 9 9 390