Appendix 4 An Appraisal of two Economic Evaluation studies related to AAC
Hass, U., Andersson, A., Brodin, H., Persson, J.(1997) Assessment of Computer - aided assistive technology: Analysis of Outcomes and Costs, Augmentative and Alternative Communication, 13, 125-135
Tolley, K., Leese, B., Wright, K., Hennessy, S., Rowley, C., Stowe, J., Chamberlain, A. (1995), Communication aids for the speech impaired: cost and quality of life outcomes of assessment programs provided by specialist communication aid centers in the United Kingdom, International Journal of Technology Assessment in Health Care, 11:2, 196-213
In the Hass et al (1997) study a cost and outcome analysis of computer-aided assistive technology was explored. Crucially the study included people with sensory impairment who had difficulties only with written communication as well as individuals with speech difficulties. In addition, participants in the study were provided with standard PC based systems. There was no provision of dedicated voice output communication aids for people with speech difficulties reported. The results of the study indicate reasonable marginal costs but limited utility rating, particularly for people with speech impairment. Only direct costs such as costs of assessment and training were included but not indirect costs such as carer support, travel costs, etc. No summary benefit measure was used and outcome measures were analysed. Costs for 'selection process' that included assessment, trials and training were reported as 30% of the total first year costs.
The UK study by Tolley et al (1995) compared the costs and outcomes of AAC assessment programmes by specialist communication aids centres for the speech impaired with areas where there are no specialist centres. It is a significant study relating to this guidance and is discussed in detail. The study included 6 regional communication aid centres and 4 districts with no regional communication aid centres in England. The comparator sites with no communication aid centres had recognised AAC specialist staff. Four of the communication aid centres had mixed adult and child caseloads while the remaining two had exclusively children and adult caseloads respectively. The child only centre was located within a special health authority that provided a service to children nationally. Both direct and indirect costs were used including costs such as trial and review costs and time spent on travel and attendance at assessment.
The low numbers of participants in the comparator group possibly indicated that non-specialist speech and language therapists were managing AAC needs within their respective districts rather than transferring clients to a colleague with a special interest in AAC, or that a level of unmet need was not being identified within non- communication aid centre district. Analysis of the clients referred to the communication aids centres and to the 4 non-communication aid districts included in the study demonstrated that just 25% of referrals were under 18 years. The evidence that most referrals, 75%, to communication aid centres were adults may also suggest that the AAC needs of children are poorly recognised or that these needs are catered for within existing non-specialist services.
Two outcome measures were used in the study. Neither of them validated a modified standard quality of life measure and a self-assessed perception rating scale. The outcome measures were not felt to be suitable for use with children therefore; final analysis did not include children. The total number of clients completing both measures was 148. No sensitivity rating or detailed statistical analysis is reported.
The outcome of the study demonstrated increased costs where a communication aid centre provided the assessment. These costs were modest. For clients receiving input from communication aid centres, modest gain in outcomes is reported. These reduced where loan equipment provided and input was protracted. In non-communication aid centre districts improvement in outcomes were lower than in communication aid centres. These outcomes are reported within the limitations of the study, described above, and should be interpreted cautiously.
The lack of summary measures in the above studies does not support the calculation of cost-effectiveness ratios.
Email: Peter Kelly