Using Chat Interface for Mental Health Interventions

Conversations via instant messaging interfaces (also known as “chats”) are generating more and more interest in mental health care. A study even shows that it has equivalent results to “treatment as usual”, so they might be added to mental health doctors’ toolkit.  Indeed, this approach offers two significant benefits over face-to-face interventions (although obviously, the latter have other qualities), that I will explain to you right now.

First, it allows the patient, in the event that he or she feels more comfortable with a machine than with a human interlocutor, to exchange serenely, at his or her own pace and in the desired way, with the practitioner. This may be appropriate in cases where the person being treated has been abused by other human beings and is therefore reluctant to trust another human being, especially if it is to treat the abuse in question, or in cases where oral communication with the person is limited or impossible, and an introduction to written communication, via computer, has been made (or the person naturally goes towards machine environments to exchange). In some cases of unusual mental organization, a person’s logical intelligence may be more developed than their social intelligence, which will lead them to prefer the procedural way in which interactions take place through a computer.

Secondly, it provides support for people who do not dare to go to see a doctor, people who are concerned about mental health problems and who do not have a follow-up, for whatever reason. Indeed, at a time when technological culture is so developed and human beings sometimes find it easier to look at their phones than their peers (and therefore have difficulty seeking help for their potential mental health problems), it is a relevant approach to providing help to those in need. Moreover, it also allows people to go out on their own to seek support, rather than having it imposed on them. This is, therefore, part of a preventive, rather than curative, approach. For example, before a mental breakdown, which would lead to support being imposed. It is also a constructive response to the medical deserts that may appear in some rural areas.

A final reason to focus on instant messaging as an interface for mental health care is that it is also accessible to a large majority of the population. Thus, it is possible that many people may try an exchange in this context, especially if the exchange is automated using a chatbot (rather than having an army of practitioners behind the machine, while we still lack doctors for face-to-face exchanges). This would, therefore, contribute to raising awareness about mental health, for ourselves and for the people around us, to whom we could talk about this tool, again as a measure raising awareness or as a first step towards support.

In an analysis of 24 different types of intervention experiences (most using a conversation bot) to treat various mental health disorders (e. g. anxiety, distress, depression, eating disorders and addictions), it is shown that the immediate outcomes of these treatments are positive and even significant. The improvements observed are of equal quality to ordinary treatments, with an equivalent but not superior post-treatment improvement than the latter. This study (Application of Synchronous Text-Based Dialogue Systems in Mental Health Interventions: Systematic Review), published by the Journal of Medical Internet Research, makes a state-of-the-art of automated conversation techniques and opens the debate on the technologies that will best allow a conversation to be conducted with the patient. In the case of a chatbot, natural language processing (NLP) technology is very interesting because it allows the conversation to be adapted to the patient’s intentions and emotions. However, given the limited number of experiments, it is still too early to establish in any certainty and detail the most effective approaches when using chatbot technology.

Instant messaging exchanges, whether automated (with a chatbot) or not (with a human), have therefore shown that they are at least as effective as face-to-face exchanges in the context of mental health treatment. They are obviously not intended to replace them, but they are another tool in doctors’ toolbox to reach a wider audience. Moreover, the application of a scientific research method in the study of approaches to mental health care, and the fact that chatbots can be modified at will make it possible to move towards greater efficiency of this tool.

No Comment.