Health economic evaluation of Autism Adapted Safety Plans: findings on feasibility of tools from a pilot randomised controlled trial

Background

Autism Adapted Safety Plans (AASP) have been proposed to help prevent self-harm and suicidality among autistic adults. The introduction of such plans not only needs to be clinically effective but also cost-effective. The aim of this work was to establish how the cost-effectiveness of AASP could be assessed. Specifically, whether tools and techniques used to collect data for health economic evaluation of the intervention are feasible and acceptable to autistic people.

Methods

A feasibility and external pilot randomised controlled trial of the AASP intervention was conducted. Autistic adults recruited from diverse locations in England and Wales were randomised to either: AASP and usual care, or usual care only. Health economics tools (bespoke and adapted) were developed and focus groups were undertaken with participants, including autistic adults (n = 15), their family members/carers (n = 5), and service providers (n = 10), to determine their acceptability and feasibility. Tools considered worth further exploration were interviewer administered to participants during the pilot trial at baseline and at 6 months. Interviewer notes were used to record any issues reported while completing the tools. Response rates on the questions and completeness of the tools, along with participant feedback in the interviewer notes was assessed.

Results

Standard Gamble and Time-Trade Off approaches to measure health status were judged inappropriate to measure health outcomes with autistic adults experiencing suicidal ideation and with a history of self-harm. Contingent valuation and discrete choice experiments were also considered inappropriate, due to the heavy cognitive burden on respondents. The EQ-5D-5L/VAS, resource utilisation questionnaire and time-travel questionnaire were considered acceptable by participants. Response and completion rates (as a percentage of all returned questionnaires) for resource utilisation questionnaire (> 85%), time-travel questionnaire (> 79%), EQ-5D-5L (> 96%) and EQ-5D-VAS (> 87%) were good in general. Participants needed clear guidance and interviewer support to enable questionnaire completion.

Conclusions

It is feasible and acceptable to collect relevant data on resource utilisation, and costs of accessing care and the EQ-5D-5L in a future definitive trial. Clear guidance and interviewer support on how to complete the questionnaires and explanations of the importance of questions to the research would help autistic participants completing the health economic tools.

Trial registration

ISRCTN70594445; Trial Registration Date: 06/07/2020.