CINGLE-trial: cochlear implantation for siNGLE-sided deafness, a randomised controlled trial and economic evaluation
Study objectives
The main objective of our study is to compare CROS, BCD and CI in patients with SSD
evaluating speech perception in noise, sound localization, quality of life and tinnitus.
The second objective is to perform an economic evaluation and to evaluate adverse
events in all groups.
Study design
In this RCT patients are randomised in three groups. For a schematic overview of the
study, see Fig. 1.
Fig. 1. Flow diagram of CINGLE-trial. Abbreviations: BCD?=?Bone Conduction Device (* indicates
trial period on headband), CI?=?Cochlear Implant, CROS?=?Contralateral Routing of
Sound system, mo?=?months, wk?=?weeks
We will now discuss all consecutive steps in the study design as included in this
study protocol. This protocol is reported according to the SPIRIT Statement, an international
guideline on the reporting of study protocols 12].
Study population
The study population consists of patients who present with SSD at the outpatient department
of our tertiary referral center, the University Medical Center Utrecht, Utrecht, The
Netherlands. They must meet the following criteria to be eligible for the study.
Inclusion criteria
Age 18 years or older
Audiometry (Pure Tone Average [PTA] at 0.5, 1, 2, 4 kHz)
? ? Deaf ear: threshold of ?70 dB
? ? Better ear: threshold of ?30 dB
? ? Air bone gap ?10 dB (to ensure normal middle ear function)
Duration of deafness 3 months and ?10 years
Health status allows general anaesthesia and surgery for the potential implantation
of BCD or CI
Dutch language proficiency
Coverage of Dutch health insurance
Willingness and ability to participate in all scheduled procedures outlined in the
protocol
The minimum duration of deafness is 3 months, since that is the time in clinical practice
to await the natural course of sudden deafness. The maximum duration of deafness is
up to 10 years, since degeneration of the auditory nerve may occur.
Exclusion criteria
Previous experience with implanted BCD or CI
Retrocochlear pathology
Abnormal cochlear anatomy (i.e. ossification)
Comorbidity
? ? which could interfere with the completion of the tests or questionnaires (e.g. psychiatric)
? ? with an expected survival of less than five years
If eligible for inclusion, an Informed Consent (IC) form will be signed by patient
and researcher. Only after IC, a CT-scan of the mastoid will be made, if none is available
yet, to assess cochlear anatomy and check if no contraindications to cochlear implantation
(e.g. ossification) exist. Since the anatomical situation must allow cochlear implantation
(and thus randomisation), CT-scans will be performed prior to randomisation.
Randomisation and interventions
A web-based randomisation tool (Julius Center, University Medical Center Utrecht,
Utrecht, The Netherlands) will be accessed via a computer by one of the members of
the research team. Patients will be randomised into groups A, B and C (see Fig. 1), using a block size of 8 and a ratio of 2:3:3 for groups A:B:C and stratified for
age (45 years, ?45 years). Blinding and concealed treatment allocation is not possible
in our study design.
In group A, patients will be implanted with a CI from Cochlear Ltd., (type CI422).
A retro-auricular incision is made to expose the mastoid. The electrode is inserted
via a posterior tympanotomy and round window implantation. Intraoperatively, normal
functioning of the device is checked by measurement of impedance and neural response
telemetry. Four weeks after implantation, the CI will be activated by an experienced
audiologist. In the rehabilitation phase, patients will be encouraged to use the CI
each day and will be trained by experienced speech and language therapists. Patients
are instructed to train the cochlear implant ear using an International Speech Test
Signal (ISTS) noise 13] via an insert earphone to mask the better ear.
In the Netherlands, a trial period with both CROS and BCD is standard clinical care.
Patients in groups B and C try both devices for 6 weeks: group B starts with the BCD
(type: BP110, Cochlear Ltd.), then CROS (Phonak Audeo Q50-312T and CROS H2O), whereas patients in group C start with the CROS and then try the BCD. The reversed
order is implemented to correct for the order effect: patients judge their second
hearing aid based on experiences with the first hearing aid. After these two trial
periods, patients may, according to clinical practice, choose which of both treatments
they like best. When they choose a BCD, the implant and abutment will be surgically
implanted, and after six weeks mounted with a BCD (type: BAHA 4 system, Cochlear Ltd.).
When they prefer CROS, the patient is referred to standard clinical care where the
CROS will be adjusted. Patients can also opt for no treatment if none is preferred,
according to standard clinical health care.
Sample size
To detect a clinically relevant difference of 5 dB signal-to-noise ratio (SNR) (standard
deviation 5 dB) between the groups on the primary outcome (see Outcomes), with an alpha of 0.05 and a power of 95%, 27 subjects per group are needed. To
compensate for potential dropouts, a 10% margin is implemented, resulting in 30 patients
in group A. Initial group distribution will be n?=?45 for groups B and C. These groups
are bigger, since previous studies describe that only 45% of BCD-on-headband users
are satisfied after their trial period and opt for a BCD implantation (n?~?45) 14]. Therefore, we will include more patients in groups B and C than in group A. Patients
not choosing a BCD can opt for a CROS (approximately 60% of remaining patients, n?~?30).
The rest of the patients (n?~?20) will probably prefer no treatment; they will be
followed up to assess the natural course of single-sided deafness.
Approximately 25–30 patients per year present with SSD at the otorhinolaryngological
outpatient department of our tertiary referral center. We will actively invite audiologic
centers in the neighborhood to refer patients to our clinic. Therefore, we expect
the inclusion period to last for ~3 years.
Outcomes
The following outcome measurements will be recorded during baseline visit (one condition,
i.e. ‘no device’) and follow up visits at 6, 12, 18, 24, 36, 48 and 60 months ‘device
on’. During all audiometric tests, patients will be instructed not to move their heads
to improve speech perception or sound localisation. Head movements will be checked
by the researcher conducting the experiments. Patients will not receive feedback on
their performance on audiometric tests. All experiments will be performed by researchers
following the same protocol procedures.
Primary outcome measure
Our primary outcome is the performance on speech perception in noise, measured with
the Utrecht Sentence Test with Adaptive Randomized Roving Levels (U-STARR) 15]. In short, the U-STARR is designed to determine a patient’s ability to understand
speech in a noisy environment (signal from front, noise from front: S0N0. See Fig. 2, test set-up York Crescent of Sound 16]). A sentence is considered to be understood correctly when ?2 words are repeated
incorrectly. The noise level starts at +20 dB (roving 65–75 dB SPL). When the sentence
is repeated correctly, the level of the noise increases for the next sentence. Noise
is presented 500 ms before the start of the sentence and ends 500 ms after the sentence.
Sentences used are traditional Dutch sentences from everyday life 17]. The test provides a critical SNR at which 50% of sentences is understood correctly
(in dB).
Fig. 2. Set-up York Crescent of Sound. The patient is positioned in the center of an arch
of loud speakers (-90 to 90°) at head level with a radius of 1.45 m 16]
Secondary outcome measures
Speech perception in noise
Speech perception in noise is measured in two more configurations: S?60N+60 and S+60N-60 (Fig. 2). Again, Dutch sentences are used in the same audiometric set-up as the U-STARR 15], 17] leading to a SNR (in dB).
Sound localisation
The ability to localise sounds is also measured with the York Crescent of Sound set-up
16]. The stimulus (sentence by female speaker: ‘Hello, what’s this?’) is presented in
quiet at a roving level of 55–65 dB SPL in three configurations: 5 boxes separated
by an angle of 15° (box -30°, box -15°, box 0°, box +15° and box +30°), 5 boxes separated
by an angle of 30° (box -60°, box -30°, box 0°, box +30° and box +60°) and 3 boxes
separated by an angle of 60° (box -60°, box 0° and box +60°) (Fig. 2). The patient must indicate from which box the stimulus came. The test outcome is
a percent correct score.
Tinnitus
Tinnitus burden will be assessed using three questionnaires:
Tinnitus Handicap Inventory (THI): a 25-item questionnaire with statements/questions
about tinnitus burden 18]. Possible answers are ‘Yes’ (4 points), ‘Sometimes’ (2 points) and ‘No’ (0 points),
resulting in a maximum score of 100, representing a maximum burden of tinnitus. The
inventory is divided in a functional, emotional and catastrophic subscale.
Tinnitus Questionnaire (TQ): a 52-item questionnaire consisting of 5 subscales: emotional
and cognitive distress, intrusiveness, auditory perceptual difficulties, sleep disturbances
and somatic complaints 19], 20]. Possible answers are ‘Yes’ (2 points), ‘Sometimes’ (1 point) and ‘No’ (0 points).
Of the 52 questions, 38 constitute a final TQ-score on the validated Dutch version
of the TQ 21]. Tinnitus can be graded mild (TQ 0–16), moderate (TQ 18–34), severe (TQ 34–56) and
catastrophic (TQ 58).
Tinnitus Burden Questionnaire (TBQ): this is a self-developed questionnaire assessing
various aspects of tinnitus burden. It consists of 12 visual analogue scales (VAS),
ranging from ‘0’ (no tinnitus burden) to ‘10’ (maximum tinnitus burden).
Quality of life
Participants will be asked to fill in several questionnaires, each assessing different
parts of QoL.
Speech, Spatial and Qualities of hearing scale (SSQ): this questionnaire assesses
three domains of hearing: 1) Speech: consists of 15 questions about the ability to
separate speech from competing noise in a wide range of listening contexts; 2) Spatial:
consists of 17 questions to assess the ability to locate sound sources and their direction
of movement; 3) Quality: consists of 19 questions that assess naturalness and clarity
of sounds. The responses are given on a VAS ranging from 0 (not able to) to 100 (perfectly
able to) 22].
Abbreviated Profile for Hearing Aid Benefit (APHAB): this 24-item questionnaire documents
the outcome of a hearing aid 23]. The questionnaire yields scores on subscales for ease of communication, listening
under reverberant conditions, listening in background noise and aversiveness of sound.
Glasgow Benefit Inventory (GBI): a measure of patient benefit developed especially
for otorhinolaryngological interventions 24]. The inventory is validated to measure outcomes on health status after otorhinolaryngological
procedures. It measures QoL in three domains: social, general and physical. The domains
score on a scale of -100 to 100 (minimum versus maximum benefit, respectively).
Hospital Anxiety Depression Scale (HADS): a screening tool for anxiety and depression
in non-psychiatric clinical populations 25]. We use the HADS to measure baseline depression symptoms, which may confound/bias
the results of, for instance, tinnitus burden.
A VAS consisting of two questions (quality of life, quality of hearing).
Time Trade Off (TTO): comprises one question about how many years of their lives
patients would sacrifice for living with perfect hearing for the rest of their lives.
TTO (%)?=?((life expectancy – number of years to give up for perfect hearing) / life
expectancy) * 100. This question is generally considered a difficult question, so
it will not be presented on paper, but asked during the baseline and follow up visits.
EuroQoL5D (EQ5D): is a measure of general health status 26]. It contains 5 questions on mobility, self-care, daily activities, pain/complaints,
anxiety/depression and a scale to denote general quality of life (VAS 1–10). We will
use the Dutch EQ5D tariff 27].
Health Utilities Index 3 (HUI3): this is a generic quality of life questionnaire
consisting of 8 domains: vision, hearing, speech, ambulation, dexterity, cognition,
emotion and pain 28].
Economic evaluation
The latter three questionnaires can be used to calculate utility. Utility reflects
the value that is attached to health status. Utility values are important for the
calculation of Quality Adjusted Life Years (QALYs), which serve as the denominator
for the Incremental Cost Utility Ratio (ICUR). The ICUR is calculated as the incremental
costs of cochlear implantation as compared to current treatments (numerator) divided
by the incremental effects in terms of QALYs. Costs will be measured from a societal
and health care perspective. Both direct health care costs and indirect non-health
care costs will be incorporated in the analyses. Both categories of costs will be
quantified using a cost diary. This diary is completed on a monthly basis the first
2 years and on a quarterly basis the last three years. This diary assesses costs related
to hospitalisation, surgery, blood tests, complications (direct health care costs)
and sick leave, time and travel costs (indirect health care costs). Unit prices for
volumes of resources use will be taken from the Dutch guidelines for costing research
in health economic evaluations, as issued by the National Healthcare Institute 29].
Incremental Cost Utility Ratios, comparing CI with CROS and BCD, will be estimated
using bootstrapping. Cost-effectiveness planes and a cost-effectiveness acceptability
curve will be plotted to visually represent the results of the economic evaluation.
In addition to these outcome measurements at the previously specified baseline and
follow up visits, we will also objectify the experiences of the patients in Groups
B and C in the trial periods with CROS and BCD (see Fig. 1). To minimize patient burden, only the APHAB, GBI and SSQ questionnaires will be
administered to evaluate these trial periods.
Statistical analysis
Baseline characteristics per group will be described as means and standard deviations.
Differences between the three groups will be analysed using the Kruskal Wallis test.
The data of our primary outcome are quantitative and will be presented as continuous
variables. Between-group mean differences, rate differences and rate ratios with 95%
confidence intervals will be calculated. Again, the Kruskal Wallis test will be used
to analyse differences between the groups.
The secondary outcomes contain both categorical and continuous outcomes. Analyses
of between-group differences will be performed with Chi-square-tests for categorical
outcomes and Kruskal Wallis tests for continuous outcomes. Within-subject comparisons
will entail differences of mean values. These will be analysed using paired t-tests
for continuous measures.
Major test intervals are the same in all study groups (baseline and 6, 12, 18, 24,
36, 48, 60 months follow up; 6 and 12 weeks for group B and C). Missing values will
be imputed using multiple imputation. All analyses will be performed on an intention-to-treat
basis. A significant result is defined as a p-value??0.05. Statistical package SPSS will be used for statistical analyses of the
data.
Data will be presented according to the Consolidated Standards of Reporting Trials
(CONSORT) Statement, an international guideline on adequate reporting RCTs 30], 31].
Safety
This study will be conducted in accordance with the most recent version of the Declaration
of Helsinki (Fortaleza, 2013), good clinical practice guidelines and the Medical Research
Involving Human Subjects Act of the Dutch government. The research protocol was approved
by the Institutional Review Board (IRB) of the University Medical Center Utrecht (NL45288.041.13;
version 3, April 2nd, 2014).
All cases of serious adverse events will be reported to the local IRB and adequately
followed up. An independent monitor (Trial Form Support BV, Zaltbommel, The Netherlands)
is appointed to check trial quality (completeness of IC, validity of data etc.) twice
a year. All patient data will be stored on a password protected computer in a lockable
room. In the same room the signed Informed Consent forms will be kept in a locked
cabin.
Trial status
The trial is currently in recruitment phase.