An exploratory look at pediatric cochlear implantation: is earliest always best?

Abstract

Objectives: Since the advent of cochlear implants, age at implantation has declined as investigators report greater benefit the younger a child is implanted. Infants younger than 12 mos currently are excluded from Food and Drug Administration clinical trials, but have been implanted with Food and Drug Administration-approved devices. With a chance that an infant without profound hearing loss could be implanted because of the limitations of the diagnostic measures used with this population and the potential for additional anesthetic risks to infants younger than 1-yr-old, it is prudent to evaluate benefit in the youngest cochlear implant recipients. The goals of this research were to investigate whether significant gains are made by children implanted before 1-yr-old relative to those implanted at later ages, while controlling for potential covariates, and whether there is behavioral evidence for sensitive periods in spoken language development. It was expected that children implanted before age 1 yr would have more advanced spoken language skills than children implanted at later ages; there would be a negative relationship between age at implantation and rate of spoken language development, allowing for an examination of the effects of sensitive periods in spoken language development; and these trends would remain despite accounting for participant characteristics and experiences that might influence spoken language outcomes.

Design: Ninety-six children with congenital profound sensorineural hearing loss bilaterally and no additional identified disabilities who were implanted before the age of 4 yrs were stratified into four groups based on age at implantation. Children's spoken language development was followed for at least 2 yrs after device activation. Spoken language scores and rate of development were evaluated along with four covariates (unaided pure-tone average, communication mode, gender, and estimated family income) as a function of age at implantation.

Results: In general, the developmental trajectories of children implanted earlier were significantly better than those of children implanted later. However, the advantage of implanting children before 1-yr old versus waiting until the child was between 1 and 2 yrs was small and only was evident in receptive language development, not expressive language or word recognition development. Age at implantation did not significantly influence the rate of the word recognition development, but did influence the rate of both receptive and expressive language acquisition: children implanted earlier in life had faster rates of spoken language acquisition than children implanted later in life.

Conclusions: Although in general earlier cochlear implantation led to better outcomes, there were few differences in outcome between the small sample of six children implanted before 12 mos of age and those implanted at 13 to 24 mos. Significant performance differences remained among the other age groups despite accounting for potential confounds. Further, oral language development progressed faster in children implanted earlier rather than later in of life (up to age 4 yrs), whereas the rate of open-set speech recognition development was similar. Together, the results suggest that there is a sensitive period for spoken language during the first 4 yrs of life, but not necessarily for word recognition development during the same period.

Figures

Fig. 1

Scatterplot displaying age at implantation for each individual participant.

Fig. 2

Left panel, Hypothetical curves for an outcome measure as a function of time in a group of three children implanted at 6 to 12 mos of age. Center panel, Similar to left panel, for a group of four children implanted at 13 to 24 mos. Right panel, Determination of the “developmental difference” between two children, one implanted at 11 mos and last tested at age 84 mos, and the other one implanted at 23 mos with a last testing age of 89 mos. The age range T, depicted by the horizontal arrow that extends from Tmin to Tmax (which are 11 and 84 mos in this example), is the age range over which comparison is done. The vertical arrows between the two curves represent the difference in outcome measure values for the two children at several ages between Tmin and Tmax. The average value of these arrows is the developmental difference between the two children and is depicted by the vertical arrow with the D label. With this method, differences between two developmental curves are boiled down to a single number. For a precise description of “developmental difference,” see Eq. (1).

Fig. 3

Calculation of developmental differences for the example data shown in the left and center panels of Figure 2. The left column of panels in this figure shows the average developmental differences between the first member of group 1 (in blue) and each member of group 2. The center and right columns of panels show the same for the second (in red) and third members (in green) of group 1, respectively. The arrows in the top four rows of panels indicate the developmental difference between the two curves in the corresponding panel, which are calculated with Eq. (1) and explained at a conceptual level in Figure 2. Arrows point up or down depending on whether the member of groups 1 or 2 has higher outcome measure values through the age comparison range. The bottom left panel shows the comparison between subject 1 and group 2, obtained by averaging the developmental differences between subject 1 and each individual member of group 2 [see Eq. (2) for a precise description]. The other bottom panels show the same type of comparison for subjects 2 and 3.

Fig. 4

Panel A shows average word recognition scores on the Mr. Potato Head task for each age-at-implant group, as well as scores for typically developing normal-hearing children, as a function of age. Arrows indicate mean age at implantation for each group. Panels B to E show individual and group-averaged scores for each age-at-implant group.

Fig. 5

Panel A shows average age equivalent scores on the Receptive Language section of the RDLS for each age-at-implant group. Some of these scores are predicted values based on the MCDI parent questionnaire. The thick diagonal line shows average scores for the children used to obtain the test’s norm. Two lines underneath the diagonal indicate 1 and 2 SDs below the mean for the same normative sample. Panels B to E show individual and group averaged scores for each age-at-implant group.

Fig. 6

Expressive language scores, displayed following the same format as Figure 5.