effectiveness of multi-channel unilateral cochlear implants for profoundly deaf children: a...

Effectiveness of multi-channel unilateral cochlear implantsfor profoundly deaf children: a systematic review

Bond, M.,* Elston, J.,* Mealing, S.,* Anderson, R.,* Weiner, G.,† Taylor, R.S.,* Liu, Z.* &

Stein, K.*

*Peninsula Medical School, University of Exeter, Plymouth, UK, and �Royal Devon and Exeter Foundation Trust,

Exeter, UK

Accepted for publication 18 February 2009

Clin. Otolaryngol. 2009, 34, 199–211

Background: Annually an estimated 223 children in the

UK are born with or acquire permanent profound bilat-

eral deafness (PBHL ‡ 95 dB). These children may gain

little or no benefit from acoustic hearing aids. However,

cochlear implants might enable them to hear.

Objectives of the review: To bring together the diverse

research in this area under the rigor of a systematic

review to discover the strength of evidence when compar-

ing the effectiveness of unilateral cochlear implants with

non-technological support or acoustic hearing aids in

children with PBHL.

Type of review: Systematic review.

Search strategy: This examined 16 electronic data bases,

plus bibliographies and references for published and

unpublished studies.

Evaluation method: Abstracts were independently

assessed against inclusion criteria by two researchers,

results were compared and disagreements resolved.

Included papers were then retrieved and further indepen-

dently assessed in a similar way. Remaining studies had

their data independently extracted by one of five

reviewers and checked by another reviewer.

Results: From 1,580 abstracts and titles 15 studies were

included. These were of moderate to poor quality. The

large amount of heterogeneity in design and outcomes

precluded meta-analysis. However, all studies reported

that unilateral cochlear implants improved scores on all

outcome measures. Additionally five economic evaluations

found unilateral cochlear implants to be cost-effective for

profoundly deaf children at UK implant centres.

Conclusions: The robustness of systematic review meth-

ods gives weight to the positive findings of 15 papers

reporting on this subject that they individually lack; while

an RCT to show this would be unethical.

Background

Annually an estimated 223 children in the UK are born

with or acquire permanent profound deafness (‡95 dB

HL), and the estimated prevalence of profound deafness

is about 31 cases per 100 000 children.1 An epidemiologi-

cal study that examined birth cohorts of those born in

the UK between 1980 and 1995 with permanent bilateral

hearing impairment found that 4262 (25% of the cohort)

had profound hearing loss.2 Some of these children would

gain little or no benefit from acoustic hearing aids. How-

ever, cochlear implants may enable them to interpret

sounds, understand other people, be better understood

and function more safely in their environment.

Previous reviews of paediatric cochlear implantation

include Cheng et al.3 who conducted a meta-analysis

comparing the results from published and unpublished

literature on the effects of age at implantation and age of

onset of deafness for speech perception benefits from

cochlear implantation. More recently Thoutenhoofd et al.4

published a comprehensive literature review evaluating the

different outcome measures used in paediatric cochlear

implantation. Similarly, in 2006 Lin and Niparko5 con-

ducted a systematic review of methods used to measure

Health Related Quality of Life (HRQOL) in children with

cochlear implants. However, their study was not about the

actual effects of cochlear implants on HRQOL. This

systematic review fills this evidence gap by reporting the

clinical effectiveness of unilateral cochlear implants in chil-

dren. Thus, providing a methodologically robust design

to support a literature base of variable design strength.6

Objective of review

The main objective of this paper is to present a system-

atic review of the clinical effectiveness of unilateral multi-

channel cochlear implants for children with profound

Correspondence: Mary Bond, Peninsula Medical School, University

of Exeter, The John Bull Building, Tamar Science Park, Research Way,

Plymouth PL6 8BU, UK. Tel.: +44 1392 403968; fax: +44 1392 406401;

e-mail: [email protected]

RE

VI

EW

� 2009 Crown copyright • Clinical Otolaryngology 34, 199–211 199

bilateral hearing loss (PBHL), when compared with non-

technological support (e.g. lip-reading or sign language)

or acoustic hearing aids. Secondly, it summarises a sys-

tematic review of economic evaluations of unilateral

cochlear implantation. In particular it addresses the fol-

lowing questions for profoundly deaf children:

1 Are unilateral implants more effective than non-tech-

nological support or acoustic hearing aids; in particular

what is their magnitude of benefit and improvement in

quality of life?

2 What is the published economic evidence to support the

use of unilateral cochlear implants in children in the UK?

Methods

Search strategy

The following databases were searched: MEDLINE (Ovid),

EMBASE (Ovid), Ovid MEDLINE(R) In-Process & Other

Non-Indexed Citations, ISI Science Citation Index, Coch-

rane Database of Systematic Reviews, CENTRAL, NHS

EED, DARE, HTA (NHS-CRD); EconLit, Biosis Previews,

ISI Proceedings, Current Controlled Trials, National

Research Register and Clinical Trials.gov. Additionally

bibliographies and reference lists for published and

unpublished studies of cochlear implants were explored.

Searches were carried out from inception to July 2007 and

were limited to the English language. Searches were made

using MeSH heading and text; words included: hearing

loss, unilateral, cochlear implants, cochlear implantation

and severe to profound deafness. The full search strategy

is available from the authors on request.

Study selection

Abstracts were independently assessed against inclusion

criteria by two researchers (MB and JE); results were

compared and disagreements resolved. Included papers

were then retrieved and further independently assessed in

a similar way.

As this paper is part of a larger systematic review, stud-

ies of children and adults with severe to profound deafness

who used one or two multi-channel cochlear implants and

whole speech coding strategies were searched for together.

Severe and profound deafness are defined for this whole

review as the inability to detect tones below 70 dB HL and

95 dB HL respectively in their better hearing ear. Studies

in this part of the systematic review were included if they

compared unilateral implanted devices with either non-

technological support or acoustic hearing aids.

All previous systematic reviews and randomised con-

trolled trials (RCTs) were accepted, including those with

waiting list controls. Due to the methodologically highly

variable evidence base, other types of controlled studies

(i.e. non-randomised controls, cross-sectional studies and

pre ⁄ post-studies with participants acting as their own

controls) were also included. The outcomes of interest

were; hearing thresholds, speech perception, speech pro-

duction, adverse events, health-related quality of life and

education. Here, we only report on unilateral implanta-

tion in children with PBHL.

Studies of single channel cochlear implants or those

that used feature extraction coding strategies were

excluded as they are qualitatively different from the cur-

rent generation of cochlear implants. In cases where the

coding strategy was not disclosed in the research paper,

attempts were made to contact authors for this infor-

mation. Where there was no response it was assumed

that studies that collected data after 1995 used whole

speech processing and that those before did not. Studies

that compared cochlear implants with normal hearing

controls or which had no control group were also

excluded.

Data extraction and quality assessment

Data were independently extracted by one of five

researchers (MB, SM, JE, ZL and CM). Each data extrac-

tion form was checked by another researcher. Disagree-

ments were resolved by discussion.

Consideration of internal validity included the selection

of an appropriate population, identification of sources of

possible confounders and their effect on analyses, whether

the study was prospective, blinding of assessors and data

analysts, the validity and reliability of outcome measures,

reporting of attrition and appropriateness of data analy-

sis. External validity was judged according to the ability

of a reader to consider the applicability of findings to a

patient group in practice.

Results

Search results for the whole systematic review

The systematic search of electronic databases for clinical

effectiveness produced 1580 paper titles and abstracts.

From the search results 1435 items did not meet the

inclusion criteria. One meta-analysis and 144 other pri-

mary research papers were obtained for further examina-

tion. Assessment of these articles led to the exclusion of

97 papers, leaving 47 studies. This included the meta-

analysis by Cheng et al.3 because their paper compares

the differences in published and unpublished literature of

the effects of age at implantation and age of onset of

200 Bond et al.

� 2009 Crown copyright • Clinical Otolaryngology 34, 199–211

deafness rather than the direct effectiveness of implanta-

tion. No other admissible systematic reviews were found.

Further papers (n = 27) were obtained from references

of the included papers. When these had been assessed

four papers were added to the review, giving 51 primary

research papers in the review of clinical effectiveness.

Twenty five of the 51 studies were in adults and 26 in

children (19 unilateral and 7 bilateral), 15 of these in

children with PBHL (n = 1058). The flow of studies

through the assessment process can be seen in Fig. S1.

Data presentation and synthesis

The high degree of heterogeneity of the studies (e.g. design,

outcome measures, comparators), combined with limita-

tions in the reporting of methods – plus a preponderance

of non-randomised studies – meant that meta-analysis

was inappropriate. Therefore, the study characteristics,

methodological quality and results of included studies are

presented using narrative synthesis with tables.

The studies used 51 outcome measures in total, with 27

in this systematic review of studies in profoundly deaf chil-

dren (details of all these can be found in Tables S1–S4).

Results of studies of unilateral cochlear implantation

in profoundly deaf children

Study characteristics and quality

The quality assessment of the studies was undertaken fol-

lowing the general principles published by the NHS Cen-

tre for Reviews and Dissemination.7 Fifteen studies were

found, of which none were randomised controlled trials,

10 had pre ⁄ post-designs, three were cross-sectional

designs and two were prospective cohort studies. Table 1

shows a summary of study characteristics.

Table 1. Summary characteristics of studies of unilateral cochlear implantation for children

Reference Design

No.

Participants Age at assessment

Degree of

deafness Outcomes

Length of

follow-up

Comparator: non-technological support

Nikolopoulos et al.12 Pre ⁄ post-prospective 82 <7 years Profound Speech perception 5 years

Manrique et al.8 Pre ⁄ post-prospective 182 0–14 years Profound Hearing 12 years

Staller et al.9 Pre ⁄ post-prospective 78 1–17 years Profound Hearing Speech

perception

6 months

MED-EL10 Pre ⁄ post-prospective 82 18 months – 17 years Profound Hearing Speech

perception

6 months

Bollard et al.13 Pre ⁄ post-prospective 10 37 months Profound Speech perception

& production

18 months

Cohen et al.11 Pre ⁄ post-prospective 19 20 months – 15 years Profound Speech perception 6 months

Comparator: acoustic hearing aids

Mildner et al.17 Cross-sectional 49 CI mean 11.6 (7–15)

AHA mean 12.9 (7–15)

Profound Speech perception NA

Horga & Liker22 Cross-sectional 20 Range: 7.4–15.2 Profound Speech production NA

Tomblin et al.23 Pre ⁄ post-prospective 58 CI mean (sd): 10 (2.9)

AHA mean (sd):

9 (3.65)

Profound Speech production 5 years

Osberger et al.18 Pre ⁄ post-prospective 58 5.4 years Profound Speech perception 1.5 years

Svirsky et al.19 Non-randomised

controlled trial

297 CI mean: 4.2

AHA mean: 8.4

Profound Speech perception 1.5 years

Brown & McDowall24 Pre ⁄ post-prospective 24 3.6 years Profound Speech production 6 months

Osberger et al.20 Pre ⁄ post-prospective 30 Mean 9.0 years Profound Speech perception 6 months

van den Borne

et al.16

Non-randomised

controlled trial

43 CI mean: 5 years,

9 months

AHA: 1 year,

4 months

(mean)

Profound Hearing Speech

perception

3 years

Truy et al.21 Cross-sectional 26 CI mean: 6.9

AHA mean: 6.2

Profound Speech perception NA

CI, cochlear implant; AHA, acoustic hearing aid.

Visual summary results tables: unilateral cochlear implants versus non-technological support

Effectiveness of multi-channel unilateral cochlear implants 201


There was a large degree of heterogeneity in the

included studies; in design, methods of reporting and

outcomes. Overall the studies were judged to be of

moderate to poor quality with weaknesses in design

(affecting internal validity). Many of the studies were

poorly reported: results were not always described in

the text but had to be interpreted from figures; the

methods of participant selection were not well

documented; accounting for all participants did not

always occur and it is not known whether those who

assessed or analysed the outcomes were blinded to the

condition of the participants.

Assessment of potential confounding factors showed

that few studies reported or allowed exploration of how

outcomes varied with: age at implantation; duration of

deafness; or levels of hearing loss. However, the small size

of many of the studies would have precluded statistically

appropriate analysis of such factors. No effectiveness

studies separately reported outcomes for subgroups of

deaf children with different levels of functional hearing

(informal hearing tests) or for children with other sensory

impairments or other complex co-morbidities. In some

cases those with other disabilities or who performed less

well in screening tests were excluded. There was a lack of

power calculation in all cases. None of the included stud-

ies reported quality of life, educational or adverse event

outcomes. Figure 1 shows the percentage of quality indi-

cators present.

Unilateral implantation versus non-technological

support

Six studies, all prospective with pre ⁄ post-designs, made

this effectiveness comparison. The variety of outcome

measures used, range of methods of data analysis and

limited reporting meant that pooling of data was not pos-

sible and drawing firm conclusions difficult. However,

weight should be given to the large total number of par-

ticipants (n = 453) and the prospective design of all of

these studies.

All studies reported gains on all reported outcome

measures, some demonstrating greater gain from earlier

implantation. Due to the large number of outcome

measures we have compiled the results into a series of

visual summary tables (Tables 2–4). Boxes are shaded to

show whether the results were significantly positive and

whether significance was either not recorded or not

present.

The measurement of hearing provided the strongest

evidence to support the use of cochlear implants. Clear

gains were made from 6 months post-activation onwards,

with pure tone average thresholds before implantation at

>91 dB HL increasing to between 32 and 44 dB HL post-

implantation.8 This reflected a significant improvement in

aided threshold scores (P < 0.05) at 12 months post-acti-

vation compared with pre-implantation (pre-implant =

115.8, sd 3.25 to 12 months post-implant = 34.3, sd

8.25), indicating that a fundamental change in theFig. 1. Summary of combined study quality indicators for uni-

lateral implants in children.

Table 2. Hearing outcomes – cochlear implants versus non technological support

202 Bond et al.


children’s ability to detect sound had occurred. These

gains remained stable during the 8-year follow-up.

The studies’ results that measured speech perception

and production have almost certainly been biased by con-

founding from maturation: as children grow older their

ability to understand and produce language may have

improved independently. However, the degree of

improvement in ability to understand the speech of

others and to produce intelligible speech is likely to be

greater than that due to aging alone. This is evidenced by

a 50% improvement in understanding speech in noise

found by Staller et al.,9 for children, age 3–17 years, over

6 months, (HINT-C: before implantation 11 ± 21%, after

6 months 61 ± 37%).

Further evidence for the benefits of cochlear implants

came from the MED-EL report 2001.10 The scores for

younger children ranged from 50% difference on the ESP

spondee ID (two long syllables) test to a 70% difference

on the Early Speak Perception battery (ESP – pattern

perception test). Older children’s scores ranged from

53% difference with the Bamford–Kowel–Bench test

(BKB-simple sentences) to a 79% difference with the ESP

Table 3. Speech perception outcomes – cochlear implants versus non-technological support

Table 4. Speech production outcomes



spondee ID test and the Glendonald Auditory Screening

Procedure (GASP). However, not all children were

entered for all tests.

In the same year Cohen et al.11 found that all their

measures showed improvements at 6-month follow-up

with mean percentage improvements ranging from 26%

for the Phonetically Balanced Kindergarten test (PB-K)

words to 50% for the Common Phrases test.

Additionally, the MED-EL, Staller and Kessler studies

reported parental ratings of listening behaviours (e.g.

responding to a door bell), using the Meaningful Audi-

tory Integration Scale (MAIS). Although significance was

not reported, all found a positive trend in the increase of

scores (MED-EL = 38%, Staller = 16%, Kessler = 20%).

Age at implantation

A key question to be answered is what effect a child’s age at

implantation has (if any) on the benefit gained from coch-

lear implantation? We found two studies that reported on

this aspect, they showed a positive link between earlier

implantation and improved functional ability.

First, an association between age at implantation and

positive outcomes was found by Nikolopoulos et al.12

who found a link between earlier age at implantation and

greater understanding of the construction of English

grammar. The proportion of those with understanding

comparable to normal hearing peers rose from 2% pre-

implantation to a remarkable 67% after 5 years when

measured with the Test for the Reception of Grammar

(TROG).

In a much smaller study by Bollard et al.13 (n = 10),

speech production was measured by the Mean Length of

Utterance (MLU) before and after implantation. Results

showed that MLU scores had improved from 1.8 words

pre-implant to 4.8 words. Similarly on the Peabody Picture

Vocabulary Test14 their mean scores had improved from a

language age of 12.4 months, prior to implant to 55

months and their Reynell Development Language Scale15

scores had also improved from 20.4 to 40.8 months.

Although benefit was clearly gained from cochlear

implants, given such a disparate range of studies it is dif-

ficult to quantify the magnitude of benefit gained by

cochlear implantation, and it is lamentable that none of

the studies included direct measures of quality of life or

educational measures as these are arguably where the

overriding expected benefits would be seen.

Unilateral implantation versus acoustic hearing aids

Nine studies made this effectiveness comparison, using

a range of study designs. Again heterogeneity and lim-

ited reporting precluded meta-analysis. However, the

results on a variety of outcomes for 605 profoundly

sensorineurally deaf children indicate that greater gains

in hearing, speech perception and speech production

can be made with cochlear implants than acoustic hear-

ing aids.

Visual summaries of the results can be seen in Tables 5,

6 and 7.

Forty-three children had hearing outcomes measured

by van den Borne et al.16 The ability to detect everyday

sounds was measured on a scale of 1 to 4; both groups

were measured with acoustic hearing aids (AHA) prior to

implant. Post-implant the cochlear implant group was

measured with implants alone. Measures were taken at

6-month intervals up to 24 months post-implant. At the

final follow-up the cochlear implant group had improved

their score by 3.5 points, whilst the acoustic hearing aid

group only improved by 1.9 points.

Six studies measured the ability to understand speech.

Most recently Mildner et al.17 used a cross-sectional study

design to compare children with cochlear implants or

acoustic hearing aids. They found a mean percentage gain

in understanding visually and orally presented words for

the cochlear implant group, with an overall difference in

word scores of 22.4% (P < 0.01) (CI group = 82.8%,

AHA group = 60.4%).

An earlier study by Osberger et al.18 measured speech

perception using five tests in a pre ⁄ post implantation

Table 5. Hearing outcomes – cochlear implants versus acoustic hearing aids

204 Bond et al.


study. Their results showed improvements on all mea-

sures over 18-month follow-up ranging from a mean

score difference between groups of 19.9 on the common

phrases test to 56.5 on the ESP. All measures showed a

significant difference in favour of cochlear implants

(P < 0.0001).

Table 6. Speech perception outcomes – cochlear implants versus acoustic hearing aids

Table 7. Speech production outcomes – cochlear implants versus acoustic hearing aids



A much larger study by Svirsky et al.19 (n = 297)

compared the difference between actual PB-K words

scores for implanted children, with predicted PB-K

scores for children using acoustic hearing aids.

However, the two groups had outcomes measured at

different time intervals and numbers of times and the

hearing aid group were on average older than the

cochlear implant group. The cochlear implant group

mean scores improved by 6.3% for those aged <6 years

over 18 months and 6.5% for those aged between 6

and 12 years over 12 months.

A small study by Osberger et al.20 (n = 30) measured

speech perception using three instruments. Measures were

taken before implantation with acoustic hearing aids and

6 months post-implantation with cochlear implants. The

results showed improvements on all measures over

6 months for the cochlear implant group. The difference

in scores between the groups ranged from a mean per-

centage score difference of 33.3% on PB-K phonemes to

49.6% on PB-K words. However, statistical significance

was not reported. These participants may be a sub-set of

Osberger et al.18

van den Borne et al.16 (n = 43) also reported on speech

perception. They found a small ‘relative to baseline’

improvement in verbal receptive skills for children with

cochlear implants compared to those with acoustic hear-

ing aids of 0.1 over 24 months, using the Scales of Early

Communication Skills for Hearing Impaired Children.

However, the actual scores at 24 months were better for

acoustic hearing aids (CI = 50, AHA = 54). It should be

noted that the baseline scores were lower for the cochlear

implant group (CI = 43, AHA = 47.5). As both groups

made gains from their baseline scores (CI = +7.0,

AHA = +6.9) and the cochlear implant group were on

average 4.5 years older than the hearing aid group, it

would appear that maturation effects may have contrib-

uted to improvements in receptive language.

Further results from Truy et al.21 support the addi-

tional gain from cochlear implants in this small study

(n = 26). They compared matched pairs of children with

either cochlear implants or bilateral hearing aids. The

cochlear implant groups’ mean PTA scores were greater

than those of the control group (PTA threshold,

CI = 113 dB HL, hearing aid = 104 dB HL P < 0.05).

Only the implanted children’s speech perception was

measured before surgery, but both groups were measured

at follow-up. It is not reported how long after implanta-

tion the follow-up measures were taken. Age equivalent

scores were obtained using three French language tests,

and linear regression used to produce the following equa-

tions: Preoperative cochlear implant group: equi = 0.14

chrono + 0.79 (R2 = 0.45, P < 0.05). Postoperative cochl-

ear implant group: equi = 0.55 chrono + 1.41 (R2 = 0.90,

P < 0.001). Hearing aid group: equi = 0.49 chrono +

0.25 (R2 = 0.58, P < 0.01). The equations indicate that

receptive language scores significantly increase over time

after surgery in children with cochlear implants compared

to pair-matched non-cochlear implanted children with

hearing aids.

Speech production

Three studies examined speech production in 102 chil-

dren. The most recent study by Horga & Liker22 (n = 20)

measured speech production by asking participants to

copy spoken words and sentences. They found that the

cochlear implant group had better vowel differentiation,

pronunciation, more intelligible vowels except for ⁄ a ⁄( ⁄ i ⁄ : CI = 80% AHA = 63%, ⁄ a ⁄ : CI = 58% AHA =

64%, ⁄ u ⁄ : CI = 88% AHA = 43%), and a better ability to

distinguish between accents than the hearing aid users.

However, voice onset time and closure duration showed

no difference between the groups.

Some years earlier Tomblin et al.23 (n = 58) reported

speech production measures, using the Index of Produc-

tive Syntax (IPSyn) to analyse transcripts of children

retelling stories. Their results showed a mean change in

5 year post-implant total scores of 19.6 in favour of

cochlear implants. However, these results may be suscep-

tible to bias as the cochlear implant group had the advan-

tage of repeated exposure to the test whilst the acoustic

hearing aid group had only one exposure. Regression

analysis showed that when age was included, length of

experience of use of cochlear implants was the main fac-

tor in IPSyn scores.

In the same year a small study by Brown & McDo-

wall24 (n = 24) reported results favouring cochlear

implants using the Identifying Early Phonological Needs

in Children with Hearing Impairment Scale25 (IEPN) to

assess phonological processes. Measures were taken prior

to implantation with acoustic hearing aids and 6 months

after implantation with cochlear implants alone. Signifi-

cant improvement was found in eight of the nine charac-

teristics examined; syllable production, +28% P < 0.01;

stress feature, +34% P < 0.05; initial consonants, +19%

P < 05; vowel production, = +27% P < 0.01; dipthong

production, +15% P < 0.05; manner, +13% P < 0.001;

consonant place, +16%; consonant voicing, +13%

P < 0.01.

Summary

Taken together these results confirm clinical experience

and the findings of individual studies. The additional

206 Bond et al.


value of this study is that it brings the strength of system-

atic review methodology to bear on the weak methodolo-

gies of the included studies, so that although the

individual studies lack robustness the fact that there are

so many children involved (n = 1058) and that the results

are all going in the same direction gives a collective

weight to their findings that individually they lack.

An overview of the results from all the studies included

in the clinical effectiveness systematic review can be

found in Table 8. This shows outcomes reported by type

and the strength of the result.

Systematic review of economic evaluations

We also conducted a systematic review of published eco-

nomic evaluations using the quality assessment criteria

developed by Evers et al.26 Of the 24 studies identified as

reporting cost-effectiveness or cost-benefit ratios, 20 were

full economic evaluations and five were in children and

from a UK NHS perspective.27–31 All of these UK-based

studies were cost-utility analyses, included some analysis

of education cost savings, one also included the cost of

support services at home,29 and another study costs to

the family.28 However, only the study by Barton et al.28

based their utility estimates on the actual (although par-

ent-reported) quality of life of hearing-impaired children,

and using a well-established generic instrument for mea-

suring quality of life (the HUI-3). Although the HUI-3 is

the only generic health-related quality of life instrument

which assesses the impact of hearing impairment, it has

limitations. The questions are predicated on different lev-

els of hearing with ⁄ without a hearing aid (not a cochlear

implant) and are a measure of functioning, rather than

capturing how those different levels of hearing impair-

ment impinge on normal daily activities or a person’s

social life.

Three of the studies involved either a mixture of pre-

and post-lingually deafened children, or failed to specify

their age of onset of deafness,28–31 with the other study

reporting that only pre-lingually deafened children were

included.27 The most rigorous study, published in 2006

by Barton et al.,28 used regression analysis of a survey of

over 2000 deaf children – including 403 implant recipi-

ents – to estimate the gain in health utility associated

with the implant, and combined these estimates with

comprehensive NHS costs (mainly from Barton et al.32)

to produce a range of incremental cost-utility estimates

for children of two different ages (3 and 6 years old),

three different levels of pre-implantation hearing loss, and

according to three analytical perspectives. The results of

the evaluations can be seen in Table 9.

These economic evaluations of paediatric unilateral

implantation have assessed it to be cost-effective for pro-

foundly deaf children who have been clinically selected

for implantation at UK centres. As with clinical

effectiveness, the particular characteristics of children for

whom unilateral cochlear implants would be the most

cost-effective are uncertain.

Discussion

Key findings

A total of 15 studies with 1058 profoundly deaf children

met the inclusion criteria for the systematic review of the

effectiveness of unilateral cochlear implantation and five

relevant economic evaluation studies from the UK were

found.

We found that the clinical research evidence base that

supports the use of unilateral cochlear implants in chil-

dren is methodologically weak; this is due to poor study

design and inadequate reporting. There are good clinical

Table 8. Summary of outcomes reported by type and strength of result

Comparison

Total number

of outcomes

(No. reporting

significance)

Positive

significant

N outcomes (%)

P £ 0.05

Positive trend

NS ⁄ NR

N outcomes (%)

Negative

significant

N outcomes (%)

P £ 0.05

Negative

trend NS ⁄ NR

N outcomes

Cochlear implant versus non-technological support

Hearing outcomes 1 (1) 1 (100%)

Speech perception 27 (6) 6 (22%) 21 (78%)

Speech production 3 (0) 0 (0%) 3 (100%)

Cochlear implant versus acoustic hearing aid

Hearing outcomes 1 (0) 1 (100%)

Speech perception 10 (9) 9 (90%) 1 (10%)

Speech production 3 (1) 0 (0%) 3 (100%)

NS, Not statistically significant (at P £ 0.05 level); NR, statistical significance of difference not reported.



reasons for the lack of RCTs; an RCT in 1993 established

the additional gains of multi-channel over single channel

implants33 and the research agenda changed to refining

processors. Unilateral multi-channel implants have estab-

lished effectiveness making future RCTs in most pro-

foundly deaf people illogical and unethical.

Table 9. Results of full economic evaluations analysed from a UK NHS perspective in children

Study

Analysis

year Setting

Effectiveness

data source Comparison

Reported ICERs

(per QALY)

*Barton et al.28

From an NHS

perspective

2001 ⁄ 2 UK The survey

presented

in the study

Unilateral

implantation

No cochlear

implantation

Implanted at

age 3 year:

£17 521�£11 645�£10 006§

Implanted at

age 6 year:

£20 932�£15 042�£13 225§

*Barton et al.28

From a societal

perspective

2001 ⁄ 2 UK The survey

presented

in the study

Unilateral

implantation

No cochlear

implantation

Implanted at

age 3 year:

£15 868�£9029�£7012§

Implanted at

age 6 year:

£19 062�£12 532�£10 331§

*O’Neill et al.31 1997 ⁄ 8 UK, the Nottingham

Paediatric Cochlear

Implant Programme

Study by

Summerfield

& Marshall34

Unilateral

implantation

No cochlear

implantation

£2532 ⁄ QALY

*O’Neill et al.30 1997 ⁄ 8 UK, the Nottingham

Paediatric Cochlear

Implant Programme

Study by

O’Neill et al.31

and study by

Summerfield

& Marshall34

Unilateral

implantation

No cochlear

implantation

Results stratified

by education

authority–:

County: £8,310

London: £12 282

Metropolitan:

£11 177

Unitary: £10 360

*Summerfield

et al.27

1996 The Nottingham

Paediatric Programme

with 3 educational

settings: school for deaf

children, special unit

attached to mainstream

school, and main stream

school with support

Cost data derived

from Summerfield

& Marshall

Unilateral

implantation

No cochlear

implantation

£15 600 ⁄ QALY**

£12 100 ⁄ QALY**,

taking into account

saved costs in

education

£10 000

⁄ QALY**, taking into

account saving in cost

in special equipment

for daily living in

adulthood

*Hutton et al.29 1994 UK Assumption Unilateral

implantation

No cochlear

implantation

£16 214

208 Bond et al.


However, the robustness of systematic review method-

ology that has been brought to bear on this methodologi-

cally weak research base, gives it a weight it does not

otherwise have; so that the strength of the whole is more

that of the parts, confirming clinical experience in a

methodologically vigorous way.

Unfortunately the heterogeneity of the studies pre-

cluded assessment of the magnitude of benefit from

cochlear implants and the lack of reporting of quality of

life or educational outcomes meant that these important

indicators of gain could not be addressed. Measures of

hearing provided the most objective evidence to support

the use of cochlear implants. Clear gains were made from

6 months post-activation onwards, with hearing thresh-

olds ranging from 32 to 44 dB HL post-implantation.

Where significance tests were reported they showed

improvement from pre-implant of P < 0.05 at 1 year and

P < 0.01 at 2 years. The results relating to speech percep-

tion and production show a 50% improvement in scores

for understanding speech in noise compared to before

implantation (HINT-C: before implantation 11 ± 21%,

after 6 months 61 ± 37%).

Previous economic evaluations indicate that, in the UK

NHS setting, unilateral cochlear implantation in children

is cost-effective and especially so when educational cost

savings are included.

Strengths of the systematic review

The strengths of this review are that it is systematic, up-

to-date and conducted by an independent research team

to address explicit research questions.

Limitations of the systematic review

There are a number of limitations of the systematic

review. First, the reviewed studies were of moderate to

poor quality; this reflects the standard of reporting more

than the choice of design. The strength of conclusions

that can be drawn is limited by the quality of the evi-

dence reviewed.

The heterogeneity of the studies meant that meta-anal-

ysis was not possible. Although there are ethical reasons

for the lack of randomisation of participants, this leads to

the possibility of selection bias, and possible confounding.

Further bias may occur from patient attrition, reporting

methods, observation and measurement.

Our assessment of potential confounding factors showed

that few studies reported or allowed exploration of how

outcomes varied with different: age at implantation, age of

onset of deafness, duration of deafness or level of audiolog-

ically measured hearing impairment.

A further limitation is through possible bias intro-

duced when studies did not differentiate in their speech

development results between those who were congenitally

deaf and those with acquired deafness.

None of the studies that met our inclusion criteria

reported quality of life or educational outcomes. These

are arguably two of the most important aspects of cochl-

ear implantation in children; it would be valuable to see

effectiveness studies that included these measures.

Five studies excluded children with other physical or

learning disabilities.8,12,18,20,23 We did not include foreign

language studies, which also may have biased the overall

implications of the review.

Comparison with other evidence

There have been no previous systematic reviews that

address these research questions with which to compare

our findings.

Conclusions

This systematic review confirms that, universally, research

papers show benefit in many aspects of unilateral cochlear

implantation in children; while an RCT to show this

would be ethically unacceptable.

Table 9. (Continued)

*Number of implants used not stated in any of the studies, but known to be exclusively unilateral.

�Corresponds to a subgroup of children with preoperative AHL 105 dB.

�Corresponds to a subgroup of children with preoperative AHL 115 dB.

§Corresponds to a subgroup of children with preoperative AHL 125 dB, ICERs for Barton et al.28 converted from Euros, at

£1 = €1.54, as reported in the paper.

–Original paper presented results in dollars; Converted by the authors of this review on an exchange rate of £1 = $1.45 (as per

August 2000).

**Results were derived using a range of scenario analyses using a range of assumptions about costs, cost savings and utility gains

associated with paediatric cochlear implantation. Values quoted are therefore speculative rather than based on data.

ICER, Incremental Cost Effectiveness Ratio [the difference in mean costs divided by the differences in the mean outcomes]; QALY,

Quality Adjusted Life Year [an index of survival, adjusted to account for quality of life during this time]; AHL, Average Hearing

Level.



Implications for practice

Reservations about generalisability remain due to the lack

of representativeness of some of the study populations

and the size and quality of the studies themselves. Never-

theless the positive direction of change in all cases indi-

cates that unilateral cochlear implantation is likely to be

an effective treatment for profoundly deaf children.

Implications for research

Further research is needed to determine other factors that

influence effectiveness i.e. age at implantation, duration

of deafness and age of onset of deafness. Comparative

effectiveness studies are also needed that include quality

of life and educational outcomes with long-term follow-

up to find the long-term impact of interventions on edu-

cation and employment.

Keypoints

• The robustness of systematic review methodology

brought to bear on methodologically weak research

gives a collective weight to the findings of those

studies that they individually lack.

• Annually over 200 children in the UK are born with

or acquire profound deafness and are potential can-

didates for cochlear implants.

• The evidence found was of moderate to poor quality

with great heterogeneity of design, reporting and

outcomes between the 15 included studies.

• All the studies found that cochlear implants improve

hearing, speech production and speech perception.

• Due to the heterogeneity of the studies it is not pos-

sible to quantify the overall magnitude of benefit.

• The review of economic evaluations found unilateral

implantation of children to be cost-effective in those

profoundly deaf children selected by implant centres.

• The quality of the studies means that their individ-

ual results must be viewed with caution.

Funding sources

This project was funded by the NIHR Health Technology

Assessment Programme (project number 06 ⁄ 59 ⁄ 01). This

will be published in full in Health Technology Assessment.

See the HTA Programme website (http://www.hta.ac.uk)

for further project information.

The views and opinions expressed therein are those of

the authors and do not necessarily reflect those of the

Department of Health.

Conflict of interest

None to declare.

References

1 Davis A., Bamford J., Ramkalawan T. et al. (1997) A critical

review of the role of neonatal screening in the detection of hear-

ing impairment. Health Technol. Assess. 1, 1–176

2 Fortnum H.M., Marshall D.H. & Summerfield A.Q. (2002) Epi-

demiology of the UK population of hearing-impaired children,

including characteristics of those with and without cochlear

implants – audiology, aetiology, comorbidity and affluence. Int.

J. Audiol. 41, 170–179

3 Cheng AK, Grant GD & Niparko JK. (1999) Meta-analysis of

pediatric cochlear implant literature. Ann. Otol. Rhinol, Laryngol.

Suppl. 177, 124–128

4 Thoutenhoofd E., Archbold S., Gregory S. et al. (2005) Paediat-

ric Cochlear Implantation: Evaluating Outcomes. Wiley Black-

well, London

5 Lin F.R. & Niparko J.K. (2006) Measuring health-related quality

of life after pediatric cochlear implantation: a systematic review.

Int. J. Pediatr. Otorhinolaryngol. 70, 1695–1706

6 Flood P.G. (2008) The systematic review in otolaryngology: a

guide to preparation. Clin. Otolaryngol. 33, 531–535

7 NHS Centre for Reviews and Dissemination. (2001) Undertaking

Systematic Reviews of Research on Effectiveness, 2nd edn. NHS

Centre for Reviews and Dissemination, York

8 Manrique M., Cervera-Paz F.J., Huarte A. et al. (2004) Prospec-

tive long-term auditory results of cochlear implantation in pre-

linguistically deafened children: the importance of early

implantation. Acta Otolaryngol. Suppl. 552, 55–63

9 Staller S., Parkinson A., Arcaroli J. et al. (2002) Pediatric out-

comes with the nucleus 24 contour: North American clinical

trial. Ann. Otol. Rhinol. Laryngol. Suppl. 189, 56–61

10 MED-EL. (2001) Summary of safety and effectiveness data.

FDA, Silver Spring. P000025

11 Cohen N.L., Waltzman S.B., Roland J.T. Jr et al. (1999) Early

results using the nucleus CI24M in children. Am. J. Otol. 20,

198–204

12 Nikolopoulos T.P., Dyar D., Archbold S. et al. (2004) Develop-

ment of spoken language grammar following cochlear implanta-

tion in prelingually deaf children. Arch. Otolaryngol. Head Neck

Surg. 130, 629–633

13 Bollard P.M., Chute P.M. & Parisier S.C. (1999) Specific

language growth in young children using the CLARION

(TM) Cochlear Implant. Ann. Otol. Rhinol. Laryngol. 108,

119–123

14 Dunn L. & Dunn L. (1981) Peabody Picture Vocabulary Test

(PPVT). American Guidance Corp., Circle Pines, Minn

15 Reynell J.K. & Huntley M. (1985) Reynell Developmental Lan-

guage Scales – Revised, 2nd edn. NFER Publishing, Windsor,

England

16 van den Borne S., Snik A.F., Hoekstra C.C. et al. (1998)

Assessment of basal sound identification skills and communi-

cation abilities in profoundly deaf children fitted with hearing

aids or a cochlear implant. Clin. Otolaryngol. Allied Sci. 23,

455–461

210 Bond et al.


17 Mildner V., Sindija B. & Zrinski K.V. (2006) Speech perception

of children with cochlear implants and children with traditional

hearing aids. Clin. Linguist. Phon. 20, 219–229

18 Osberger M.J., Zimmerman-Phillips S., Barker M. et al. (1999)

Clinical trial of the CLARION cochlear implant in children.

Ann. Otol. Rhinol. Laryngol. Suppl. 177, 88–92

19 Svirsky M.A. & Meyer T.A. (1999) Comparison of speech per-

ception in pediatric CLARION cochlear implant and hearing aid

users. Ann. Otol. Rhinol. Laryngol. Suppl. 177, 104–109

20 Osberger M.J., Fisher L., Zimmerman-Phillips S. et al. (1998)

Speech recognition performance of older children with cochlear

implants. Am. J. Otol. 19, 152–157

21 Truy E., Lina-Granade G., Jonas A.M. et al. (1998) Comprehen-

sion of language in congenitally deaf children with and without

cochlear implants. Int. J. Pediatr. Otorhinolaryngol. 45, 83–89

22 Horga D. & Liker M. (2006) Voice and pronunciation of cochl-

ear implant speakers. Clinical Linguistics & Phonetics 20, 211–

217

23 Tomblin J.B., Spencer L., Flock S. et al. (1999) A comparison of

language achievement in children with cochlear implants and

children using hearing aids. J. Speech Lang. Hear. Res. 42, 497–

511

24 Brown C. & McDowall D.W. (1999) Speech production results

in children implanted with the CLARION implant. Ann. Otol.

Rhinol. Laryngol. Suppl. 177, 110–112

25 Paden E.P. & Brown C.J. (1990) Indentifying Early Phonological

Needs in Children with Hearing Impairment. AG Bell Associa-

tion, Washington, DC

26 Evers S., Goossens M., de Vet H. et al. (2005) Criteria list for

assessment of methodological quality of economic evaluations:

Consensus on Health Economic Criteria. Int. J. Technol. Assess.

Health Care 21, 240–245

27 Summerfield A.Q., Marshall D.H. & Archbold S. (1997) Cost-

effectiveness considerations in pediatric cochlear implantation.

Am. J. Otol. 18 (Suppl 6), S166–S168

28 Barton G.R., Stacey P.C., Fortnum H.M. et al. (2006) Hearing-

impaired children in the United Kingdom, IV: Cost-effectiveness

of pediatric cochlear implantation. Ear Hear. 27, 575–588

29 Hutton J., Politi C. & Seeger T. (1995) Cost-effectiveness of

cochlear implantation of children. A preliminary model for the

UK. Adv. Otorhinolaryngol. 50, 201–206

30 O’Neill C., Archbold S.M., O’Donoghue G.M. et al. (2001) Indi-

rect costs, cost-utility variations and the funding of paediatric

cochlear implantation. Int. J. Pediatr. Otorhinolaryngol. 58, 53–

57

31 O’Neill C., O’Donoghue G.M., Archbold S.M. et al. (2000) A

cost-utility analysis of pediatric cochlear implantation.[see com-

ment]. Laryngoscope 110, 156–160

32 Barton G.R., Bloor K.E., Marshall D.H. et al. (2003) Health-ser-

vice costs of pediatric cochlear implantation: multi-center analy-

sis. Int. J. Pediatr. Otorhinolaryngol. 67, 141–149

33 Cohen N.L., Waltzman S.B. & Fisher S.G. (1993) A prospective,

randomized study of cochlear implants. The Department of Vet-

erans Affairs Cochlear Implant Study Group. N. Engl. J. Med.

328, 233–237

34 Summerfield A.Q. & Marshall D. (1995) Cochlear Implantation

in the UK: 1990–1994. HMSO, London

Supporting information

Additional supporting information may be found in the

online version of this article:

Figure S1. Flow of studies through review process.

Table S1. Hearing (sensitivity to sound) measures.

Table S2. Speech perception measures.

Table S3. Speech production and vocabulary develop-

ment measures.

Table S4. Quality of life measures.

Please note: Wiley-Blackwell are not responsible for the

content or functionality of any supporting materials sup-

plied by the authors. Any queries (other than missing

material) should be directed to the corresponding author

for the article.



effectiveness of multi-channel unilateral cochlear implants for profoundly deaf children: a...

Documents