The Speech Intelligibility Index (SII) is a quantitative measure that predicts speech understanding by assessing how much of the speech signal is audible to a patient based on his or her audiogram. The SII is typically a value between 0 and 100% and can be a valuable tool for counseling and is used during hearing aid fittings. This blog from AudiologyOnline does a nice job of explaining the basics.
Calculation of the SII is defined in the ANSI s3.5-1997 standard. In the standard, there are 4 methods for this calculation. Each method uses a different number of data points to calculate the SII “score” for each audiogram. Because there are multiple methods for calculating the SII, the same patient may appear to have a different SII based on the equipment being used. SII isn’t necessarily about the number of thresholds you measure, it’s more about how many data points a piece of equipment uses to calculate.
GSI uses method 4 to calculate the SII. The figure below is taken from the ANSI s3.5-1997 standard and describes how many data points are required for the calculation of each method. The table from the ANSI s3.5-1997 standard defines how the 4th method is calculated. GSI uses patient audiologic threshold data points for the calculation; therefore, the 4th calculation option is most appropriate for GSI. There is not a user setting to change the way that GSI calculates the SII.
Other manufacturers may use different methods for SII calculation based on their purpose and capabilities. For example, AudioScan uses the 1/3 octave band method, or method 2, described in the ANSI s3.5-1997 standard. Method 2 uses 18 data points which reflects the measurement methods of their verification system and offers a high level of accuracy for SII calculation. In other words, AudioScan has access to additional data points when the properties of the ear canal are measured. More information on their measurement and calculation can be found in the user guides.
SII values vary between different diagnostic equipment due to calibration differences, varying frequency weightings, and measurement conditions like background noise levels. Understanding why these values differ from one diagnostic equipment to another helps you make an informed decision about which SII score to use, ensuring you choose the most accurate and relevant measurement.
Laura Prigge, AuD is the Manager of Clinical Education and Training at GSI. Laura received her Doctorate of Audiology degree from AT Still University, a Master's of Arts in Audiology from UCONN and a Bachelor's of Science degree in Communication Disorders from Western Illinois University. Laura’s 20+ years of experience includes providing manufacturing support for a leading hearing aid manufacturer as well as technical audiology training and support for an international audiologic equipment company. Prior to that, she managed education and training at another hearing aid manufacturer and conducted audiologic evaluations on children, adults, and geriatric patients at a retail hearing center.
Sources:
Audioscan
ANSI s3.5-1997 Standard
AudiologyOnline
E3 Diagnostics