THE AAC RATE INDEX IN CLINICAL PRACTICE

Katya J. Hill
Edinboro University of Pennsylvania
Barry A. Romich
Prentke Romich Company

ABSTRACT
AAC evidence-based clinical practice requires data collection and outcomes measurement. Recent advances in methods and tools to support evidence-based practice facilitate the generation of a communication performance summary measure report. One of the identified summary measures possible using automated data logging is the rate index. The rate index has been defined to separate the selection rate from other factors influencing communication rate. This study reports preliminary results from collecting language samples from five individuals who rely on AAC systems and reporting the rate index.

BACKGROUND
For people who rely on AAC, speech-language pathologists (SLPs) are now expected to provide services in accordance with the principles of evidence-based practice (1). Automated tools and methods are available to support data collection and outcomes measurement (2). Summary measures of communication performance include communication rate in words per minute (3). Factors that influence communication rate include the selection rate (bits per second) (4), use of language representation methods, and errors (5).

The rate index is a communication performance summary measure that is calculated by dividing the communication rate in words per minute by the selection rate in bits per second and dividing the result by 60 seconds per minute (6). Thus the unit of measure for the rate index is words per bit. The rate index then is a measure of communication performance that is independent of the selection rate. Rate index provides for the comparison of communication rates adjusted for differences in selection rates. Rate index comparisons can be made between individuals using similar or different systems or for one individual under different conditions.

A first step in the evidence-based therapy process is characterizing the individual and determining the level of performance that would constitute the desired communication performance for that individual. Next the collection and analysis of language samples from the individual form the basis of current and past performance. Finally the current performance is compared to desired performance and to past performance to drive the therapy program. Rate index may serve as a summary to facilitate clinical decision making and make comparisons between different user profiles.

RESEARCH QUESTIONS
Can the AAC rate index be a clinically useful summary measure? What factors influence the calculation of the rate index? Is rate index consistent across language sampling contexts?

METHOD
Using automated language activity monitoring, language samples were collected from five subjects who participated in a controlled study. All subjects were individuals who rely on AAC systems. The five individuals ranged in age from 18-48, had cerebral palsy, and used an AAC device with synthetic speech output that allowed for the use of the three-language representation methods: single meaning pictures, alphabet based methods, and semantic compaction. Four subjects accessed the AAC keyboard using unassisted direct selection. One subject used optical headpointing. All subjects reported that they considered themselves competent communicators. Table 1 shows background information on the five subjects. Language samples were collected in two contexts, an interview and a picture description task.

Table1. Background information on augmented communicators

Subject
Age
AAC System
Array size
Selection Method
1
48
Words Strategy/Liberator

128

Unassisted Direct
2
21
Words Strategy/Liberator
128
Optical Head Pointing
3
18
Unity/Pathfinder
128
Unassisted Direct
4
45
Unity/Deltatalker
128
Unassisted Direct
5
36
Custom/Vanguard
45
Unassisted Direct


RESULTS
The analysis of the collected language samples included the calculation of communication rate, selection rate, and rate index. Table 2 shows these results for the five subjects in the study for each of the two language sampling contexts.

Table 2: Communication rate, selection rate, and rate index for five subjects.

Interview
Picture Description
Subject
Comm. Rate
Sel. Rate
Rate Index
Comm. Rate
Sel. Rate
Rate Index
(words/min.)
(bits/sec.)
(words/bit)
(words/min.)
(bits/sec.)
(words/bit)
1
14.8
12.39
0.0199
13.0
12.04
0.0180
2
6.5
7.00
0.0155
5.3
2.60
0.0340
3
10.9
24.61
0.0074
10.3
22.14
0.0078
4
11.4
7.61
0.0250
11.8
21.00
0.0094
5
16.6
8.75
0.0316
14.0
8.40
0.0278


DISCUSSION
While selection rate can be addressed in AAC therapy, most therapy time is used to work on other issues. Without consideration for differences in selection rate, communication rate comparison between individuals can be meaningless. What was needed was a communication performance summary measure that normalizes the impact of selection rate.

Selection rate for an individual can change from time to time. In the short term, selection rate can be a function of general energy level, positioning, fatigue, medication, and other like factors. In the longer term, selection rate can change as a result of efforts to optimize physical access, progression of a condition, etc. However, selection rates that are calculated from language samples taken in close proximity could be expected to be comparable.

For all subjects, the two language samples were collected in close proximity. For both Subject 2 and Subject 4, the reported selection rates for the two contexts were significantly different. A review of the data used to calculate these revealed that spelled words (the basis of selection rate calculation) that are short can produce outliers in the data. This is because the resolution of the time stamp is only one second. A suggested solution to this is to base the selection rate calculation on only those spelled words that are longer than the mean length of the spelled words that meet the criteria for consideration. For these subjects, that would bring the selection rate difference for the two sampling contexts to 17% and 4% respectively.

Likewise, communication rate can change from time to time and may be a function of the language sampling context. Since rate index is based on both selection rate and communication rate, these factors need to be considered in use of the rate index.

For people who are using similar AAC systems, rate index comparison can draw attention to opportunities for improvement in communication rate, even though the selection rates of the individuals may be vastly different. A relatively low rate index would prompt a closer review of the various other factors that contribute to communication rate: use of language representation methods, selection errors, spelling errors, etc. Comparison of these summary measures can be made without concern for selection rate differences.

Rate index can be a useful clinical tool in identifying opportunities to improve communication rate and hence communication effectiveness. The end result can be higher personal achievement for the individual who relies on AAC.

REFERENCES
1. American Speech Language Hearing Association (ASHA) (2001). Scope of Practice. Rockville, Maryland.

2. Romich, B.A. and Hill, K.J. (1999). A language activity monitor for AAC and writing systems: Clinical intervention, outcomes measurement, and research. Proceedings of the RESNA ’99 Annual Conference. Long Beach, CA. pp 19-21.

3. Romich, B.A. and Hill, K.J. (2000). AAC communication rate measurement: tools and methods for clinical use. Proceedings of the RESNA ’00 Annual Conference, Arlington, VA: RESNA Press. 58-60.

4. Romich, B.A., Hill, K.J., and Spaeth, D.M. (2001). AAC selection rate measurement: a method for clinical use based on spelling. Proceedings of the RESNA ’01 Annual Conference, Arlington, VA: RESNA Press. 52-54.

5. Hill, K.J. and Romich, B.A. (2001). A summary measure clinical report for characterizing AAC performance. Proceedings of the RESNA ’01 Annual Conference, Arlington, VA: RESNA Press. 55-57.

6. Hill, KJ and Romich, BA (2002). A Rate Index for Augmentative and Alternative Communication. International Journal of Speech Technology. 5(1), 57-64.

ACKNOWLEDGEMENT
The development of AAC performance measurement methods and tools has been supported in part by grants from the National Institute for Deafness and other Communication Disorders of NIH awarded to Prentke Romich Company.

Katya Hill, Ph.D., CCC-SLP
Compton 102
Edinboro University of Pennsylvania
Edinboro, PA 16444
Tel: 814-732-2431
Fax: 814-732-1580
Email: katyaaac@aol.com