Comparing the Effectiveness of AI Chatbot with That of Telephone Hotline (AI chatbot)

The University of Hong Kong (HKU)

Status: Completed

Conditions: Telephone Hotlines; AI Chatbot

Treatments: Other: Telephone hotline; Other: AI Chatbot

Study type: Observational

Funder types: Other

Identifiers

NCT06621134

Collaborative Research Fund (Other Grant/Funding Number)

UW21-344

Details and patient eligibility

About

The COVID-19 pandemic has significantly impacted the wellbeing of people in Hong Kong, leading to social distancing policies and changes in healthcare service utilization. School closures and remote work have increased stress levels for parents and children. Vulnerable populations, such as low-income families and children with special needs, are at higher risk of maltreatment and mental health issues. Parental burnout has become a concern as parents juggle work, childcare, and education responsibilities. There is a need for research on the physical and mental health effects of COVID-19 on families and the potential role of AI in addressing these challenges. AI, particularly chatbots, can provide accessible healthcare information and support, aiding in early diagnosis and treatment. AI chatbots offer timely responses, accurate information, and continuous availability, making them valuable tools for remote health assistance. While AI chatbots are not without limitations, further research can help integrate them more effectively into healthcare services.

Full description

The COVID-19 pandemic has had an unprecedented impact on the wellbeing of people in Hong Kong since the outbreak in December 2019. The Government has adopted social distancing policies to minimise the risk of infection. These include but are not limited to; school closure, remote working, and the prohibition of group-gatherings. These anti-infection measures have led to a change in pattern in the use of healthcare services and help-seeking activities. Studies have also shown that a dearth of socialisation leads to higher stress levels for both parents and children.

As school closure and remote work measures continue, both children and parents are under great pressure. UNESCO (2020) reported that over 1.58 billion children and youth in 200 countries were affected by school closure, as of mid-April 2020. Although the long-term effect of COVID-19 on children's and parents' mental health is unknown, cases of child abuse, neglect and exploitation have increased in the face of such unprecedented times. Low-income families or families with children with special education needs (SEN) are prone to children being maltreated and/or having mental health crisis . Parents who work from home are facing challenges of fulfilling a triple role: work, childcare and homecare. Worse still, children's lack of learning interests and motivation adds extra burden on parents as they take up the role of teachers. Parents are inclined to experience parental burnout, which is characterised by mental and physical exhaustion, with a feeling of hopelessness. Therefore it is clear there are strong societal needs for COVID-19 physical and mental health research. It is imperative to prevent potential and mitigate existing problems regarding parent-child relationship, parental stress and family functioning caused by COVID-19.

Consequently, exploring more easily accessible and efficient ways of dealing with potential and existing health problems (both physically and mentally) should be a priority. Artificial Intelligence (AI) in healthcare services has the potential to reduce the workload of healthcare workers by answering frequently asked questions through the AI system all from the comfort of the subject's home. Considering the potentially detrimental effect of COVID-19 on both children and parents it is important to fill the research gap as to how AI may serve as a platform for help-seeking, particularly during times of social distancing.

AI has been widely adopted in healthcare services in the past decade. The use of chatbots, in particular, has enhanced public engagement in health service all from the comfort of the subject's home. AI chatbots utilised natural language processing (NLP) to facilitate interaction with users in conversations, making appropriate medical advice accessible to the public. Intelligent algorithms in AI enables early diagnosis of disease and offers treatment techniques to those who may otherwise have been diagnosed too late. For instance, the U.S. Centres for Disease Control and Prevention (CDC) has launched a chatbot named Clara to help users access information on potential symptoms of coronavirus and help enable them to make decisions about the need to seek medical care). This is especially useful as it identifies high-risks groups in need of medical attention by triaging patients according to their symptoms, therefore reducing hospital visits for minor cases. It also provides support to family members of high-risk groups as to what measures can be taken to prevent infection and ways to relieve pressure in taking care of patients within their family.

AI chatbots merit attention in its prompt response to users' questions as it provides a service around the clock. In addition, answers provided by AI are considered more accurate than that of search engines, subject to the proficiency of data mining methods. These features are of significance as users are able to seek psycho-medical advice while practising social distancing, without face-to-face appointments with clinicians.

AI chatbots may serve as a self-help tool for gaining insights in dealing with both mental and physical conditions but it is far from perfection. The hope is that this study can contribute to making AI chatbots an integrated part of the health care service.

Enrollment: 48 patients

Sex: All

Volunteers: No Healthy Volunteers

Inclusion criteria

• Subjects who give consent to participate in the study.

Exclusion criteria

• Subjects who do not give consent to participate in the study.

Trial design

48 participants in 2 patient groups

Control Group

Description:

Participants will be asked to consent to randomization on their first access to our system. Users ask questions covered by the question bank and specific questions not covered by the question through a telephone hotline.

Treatment:

Other: Telephone hotline

Intervention Group

Description:

Participants will be required to provide consent for randomization when they first access our system. Users can ask questions covered by the question bank, as well as specific questions not covered by the bank, through an AI chatbox.

Treatment:

Other: AI Chatbot