**2.4 Statistical analysis**

All continuous variables represented in this study were expressed as the means ± the standard deviation and all categorical variables were expressed as patient number and percentage (%). The differences between improvement and no improvement groups were analyzed by the independent-samples T test for all continuous variables if the distribution was normal. If the distribution was abnormal, the differences between groups were analyzed by the Mann-Whitney U test. In addition, categorical variables were analyzed using a Chi-square test or Fisher's exact test. These statistical analyses were conducted using SPSS 17.0 statistical software for windows.

We also analyzed our data using the feature selection model to classify the important variables. Then, based on these important variables, we analyzed our data using a decision tree model to identify and rank prognostic factors affecting the improvement of ISSHL by AT. These calculations were performed using SPSS Clementine version 12.0 statistical software for windows.

Probability values of 0.05 or less were considered statistically significant.

Effects and Prognostic Factors of

Gender

Age

Location of lesion

Vertigo

Hypertension

Diabetes Mellitus

Acupuncture Treatment for Idiopathic Sudden Sensorineural Hearing Loss 397

Yes No Total

Male 20 (27.8%) 20 (27.8%) 40 (55.6%)

Total 36 (50.0%) 36 (50.0%) 72 (100%)

≤60 27 (37.5%) 31 (43.1%) 58 (80.6%)

Total 36 (50.0%) 36 (50.0%) 72 (100%)

Left 22 (30.6%) 24 (33.3%) 46 (63.9%)

Total 36 (50.0%) 36 (50.0%) 72 (100%)

Yes 9 (12.5%) 7 (9.7%) 16 (22.2%)

Total 36 (50.0%) 36 (50.0%) 72 (100%)

Yes 11 (15.3%) 10 (13.9%) 21 (29.2%)

Total 36 (50.0%) 36 (50.0%) 72 (100%)

Yes 0 (0%) 1 (1.4%) 1 (1.4%)

Total 36 (50.0%) 36 (50.0%) 72 (100%)

**3.5 Time interval from the onset of hearing loss to the start of acupuncture treatment**  The time intervals from the onset of hearing loss to the start of AT for improvement and no improvement groups were 51.06 days (±69.39 SD) and 167.22 days (±265.01 SD),

loss to the start of acupuncture treatment 51.06 ± 69.39 167.22 ± 265.01 0.013\*

Table 2. Differences in the Time Interval from the Onset of Hearing Loss to the Start of

Acupuncture Treatment between Improvement and No Improvement Groups

\* Statistically significant difference, P<0.05 using χ2 test or Fisher's Exact test Table 1. Relationship of Character of Patients and Improvement of ISSHL

respectively, which differed significantly (P=0.013) (Table 2).

\* Statistically significant difference, P<0.05 using independent-samples T test

Time interval from the onset of hearing

*P*\*

Female 16 (22.2%) 16 (22.2%) 32 (44.4%) 1.000

>60 9 (12.5%) 5 (6.9%) 14 (19.4%) 0.234

Right 14 (19.4%) 12 (16.7%) 27 (36.1%) 0.624

No 27 (37.5%) 29 (40.3%) 56 (77.8%) 0.571

No 25 (34.7%) 26 (36.1%) 51 (70.8%) 0.795

No 36 (50.0%) 35 (48.6%) 73(98.6%) 1.000

Improvement

Yes No

*P*\*

improvement
