Assessment of speech recognition abilities in quiet and in noise: a comparison between self-administered home testing and testing in the clinic for adult cochlear implant users

Figures & data

Table 1. Demographic characteristics of the participants.

Participant	Gender	Age (years)	Duration of severe hearing impairment (years)^a	CI experience (years)^b	Implant type^c
1	M	74	7	2	CI24RECA
2	F	81	73^d	5	CI512
3	M	58	16	2	CI24RECA
4	F	58	36	6	CI24REHybrid
5	M	51	43^d	6	CI512
6	M	67	53	6	CI24RECA
7	F	68	28	7	CI24RECA
8	M	67	47	6	CI24RECA
9	M	48	40^d	17	CI24M
10	F	81	23	6	CI24REHybrid
11	M	63	33	5	CI512
12	M	83	34	2	CI24RECA
13	F	61	21	6	CI512
14	F	44	27	2	CI24RECA
15	F	65	8	7	CI24RECA
16	F	53	8	3	CI24RECA

F: female; M: male.

^a Years between self-reported onset of (severe) hearing impairment and implantation;

^b Years of CI experience;

^c Type of implant from Cochlear Limited;

^d Estimation of the duration of severe hearing impairment.

Table 2. Summary of test procedures and standard errors of measurements (SEM) for speech recognition in quiet (65 and 55 dB) and noise (adaptive and fixed procedures).

Session	Location	Presentation mode	Tester	Test	Presentation level	CVC Lists	SEM
1	Clinic	LS	Clinician	DIN adaptive	65 dBA		1.2 dB
				DIN fixed	65 dBA		11.2%
				CVC	65 dB	1–3	7.7%
				CVC	55 dB	4–6	6.5%
		AC	Clinician	DIN adaptive	65 dBA		1.7 dB
				DIN fixed	65 dBA		8.3%
				CVC	65 dB	7–9	8.3%
				CVC	55 dB	10–12	7.2%
2	Home	AC	Subject	DIN adaptive	65 dBA		1.2 dB
				DIN fixed	65 dBA		6.3%
				CVC	65 dB	13–15	7.1%
				CVC	55 dB	16–18	6.3%
3	Clinic	LS	Clinician	CVC	65 dB	19–21	6.9%
				CVC	55 dB	22–24	6.6%

AC: Audio cable; CVC: Consonant-vowel-consonant; DIN: Digits-in-noise; LS: Loudspeaker; SEM: Standard error of measurement.

Figure 1. Speech recognition in quiet measured in four different conditions across the three test sessions. In each test condition, CVC words were presented at 65 dB (left panel) and 55 dB (right panel), respectively. The symbols represent individual scores and the horizontal lines represent mean and ±1 standard deviations.

Table 3. Means and standard deviations (SD), and summary of paired samples t test results for speech recognition in quiet and in noise.

Clinic vs home testing		Mean	SD	Mean	SD	t	df	Sig. (2-tailed)
Speech recognition in quiet		Clinic + loudspeaker		Home + audio cable
65 dB		80.9%	7.6%	77.6%	9.9%	1.623	15	.125
55 dB		82.3%	9.3%	85.6%	10.5%	2.373	15	.031*
Speech recognition in noise
Adaptive		−3.4 dB SNR	2.2 dB SNR	−4.8 dB SNR	2.2 dB SNR	3.182	15	.006*
Fixed		57.3%	14.3%	67.1%	17.6%	−1.998	15	.064
Presentation mode		Clinic + loudspeaker		Clinic + audio cable
Speech recognition in quiet
55 dB		82.3%	9.3%	81.1%	11.8%	0.322	15	.752
Speech recognition in noise
Adaptive		−3.4 dB SNR	2.2 dB SNR	−5.0 dB SNR	2.1 dB SNR	3.218	15	.006*
Test format		Clinic + audio cable		Home + audio cable
Speech recognition in quiet
55 dB		81.1%	11.8%	85.6%	10.5%	−4.343	15	.001*
Speech recognition in noise
Adaptive		−5.0 dB SNR	2.1 dB SNR	−4.8 dB SNR	2.2 dB SNR	−0.571	15	.576
Learning effect		Clinic + loudspeaker (session 1)		Clinic + loudspeaker (session 3)
Speech recognition in quiet
65 dB		80.9%	7.6%	84.6%	8.4%	−2.186	15	.045*
55 dB		82.3%	9.3%	85.4%	8.8%	−2.952	15	.01*

* Significant p value (paired samples t test).

Figure 2. Speech recognition in noise measured in three different conditions using the adaptive (left panel) and fixed (right panel) procedures. The digits-in-noise test was used to obtain speech reception thresholds in dB SNR from the adaptive procedure and % triplets correct scores from the fixed procedure. The symbols represent individual scores and the horizontal lines represent mean and ±1 standard deviations.