Automated clinical conversations across the cataract pathway with an artificial intelligence (AI) conversation agent: a UK regional service evaluation protocol

Abstract

AbstractIntroductionDigital technologies have the potential to support clinical pathways. This study aims to evaluate the impact of using an artificial intelligence-based conversational assistant known as Dora in the cataract pathway. Dora conducts clinical conversations with patients over a telephone call both before and after cataract surgery. Through automation of routine activity, the aim is to increase efficiency of the cataract pathway and the capacity of organisations.Method and AnalysisWe will use a mixed-methods cohort-based approach across all sites using Dora in South East England. The study has 3 key objectives, to: 1) Report site-specific variation on the implementation and impact of using Dora 2) assess the impact on the triple bottom line (financial, social and environmental performance) through implementation of Dora 3) understand the real-world patient outcomes of using Dora in clinical pathways. The Dora platform prospectively records symptom and outcome information from each call. We will retrospectively collect data from the hospital record and also collect qualitative data regarding the ease of implementation and patient acceptability of the technology.Ethics and disseminationThis will be registered as a service evaluation at each of the participating clinical sites. Research ethics is not needed as per Health Research Authority guidelines. Site-specific reports will be provided to each participant site as well as an overall report to be disseminated through NHS-England. Results will be published in a formal project report endorsed by stakeholders, and in peer-reviewed scientific reports.Article SummaryStrengths and limitations of this studyStrengths-This is the largest study on the use of an AI based natural language clinical assistant across multiple different hospital sites, with varied geographical locations, demographics and baseline clinical pathways.-This is a large-scale evaluation with input from multiple independent clinicians, patients, evaluators, economists and strategists.-Standardised data collection from autonomous clinical assistantLimitations-Retrospective collection of hospital level follow up data