Closing the accessibility gap to mental health treatment with a personalized self-referral chatbot

She has counseled hundreds of patients facing issues from pregnancy-related problems and infertility, and has been in charge of over 2,000 deliveries, striving always to achieve a normal delivery rather than operative. Physicians must also be kept in the loop about the possible uncertainties of the chatbot and its diagnoses, such that they can avoid worrying about potential inaccuracies in the outcomes and predictions of the algorithm. And while these tools’ rise in popularity can be accredited to the very nature of the COVID-19 pandemic, AI’s role in healthcare has been growing steadily on its own for years — and that’s anticipated to continue. Ideally, the difference between pre-intervention and post-intervention data for each group should be used in a meta-analysis [47].

The Physician Compensation Report states that, on average, doctors have to dedicate 15.5 hours weekly to paperwork and administrative tasks. With this in mind, customized AI chatbots are becoming a necessity for today’s healthcare businesses. The technology takes on the routine work, allowing physicians to focus more on severe medical cases. This way, clinical chatbots help medical workers allocate more time to focus on patient care and more important tasks. However, for this vision to become a reality, successful integration and widespread adoption of these AI-powered systems will necessitate collaborative efforts from various stakeholders. Key players such as healthcare providers, technology vendors and regulatory authorities must come together to facilitate the seamless implementation of conversational AI in the healthcare ecosystem.

An Overview of Chatbot Technology

Chatbots with access to medical databases retrieve information on doctors, available slots, doctor schedules, etc. Patients can manage appointments, find healthcare providers, and get reminders through mobile calendars. This way, appointment-scheduling chatbots in the healthcare industry streamline communication and scheduling processes. As computerised chatbots are characterised by a lack of human presence, which is the reverse of traditional face-to-face interactions with HCPs, they may increase distrust in healthcare services. HCPs and patients lack trust in the ability of chatbots, which may lead to concerns about their clinical care risks, accountability and an increase in the clinical workload rather than a reduction.

• The outcome in the 3 studies was measured using the Positive and Negative Affect Schedule.
• Regarding the study design, we included only randomized controlled trials (RCTs) and quasiexperiments.
• This free AI-enabled medical chatbot offers patients the most likely diagnoses based on evidence.
• For example, in 2020 WhatsApp collaborated with the World Health Organization (WHO) to make a chatbot service that answers users’ questions on COVID-19.
• Chatbots collect patient information, name, birthday, contact information, current doctor, last visit to the clinic, and prescription information.

QliqSOFT’s Quincy chatbot solution, which is powered by an AI engine and driven by natural-language processing, enables real-time, patient-centered collaboration through text messaging. The tool helps patients with everything from finding a doctor and scheduling appointments to outpatient monitoring and much more. Now more than ever, patients find themselves relying on a digital-first approach to healthcare — an arrangement that, at first, might not involve a human on the other end of the exchange.

Chatbot Keeps Your Patients Satisfied

Chatbot algorithms are trained on massive healthcare data, including disease symptoms, diagnostics, markers, and available treatments. Public datasets are used to continuously train chatbots, such as COVIDx for COVID-19 diagnosis, and Wisconsin Breast Cancer Diagnosis (WBCD). Chatbots must be regularly updated and maintained to ensure their accuracy and reliability. Healthcare providers can overcome this challenge by investing in a dedicated team to manage bots and ensure they are up-to-date with the latest healthcare information. It’s essential to understand the pros and cons of chatbots in the healthcare industry.

Physicians’ autonomy to diagnose diseases is no end in itself, but patients’ trust in a chatbot about the nature of their disease can impair professionals in their ability to provide appropriate care for patients if they disregard a doctor’s view. Chatbots are now able to provide patients with treatment and medication information after diagnosis without having to directly contact a physician. Such a system was proposed by Mathew et al [30] that identifies the symptoms, predicts the disease using a symptom–disease data set, and recommends a suitable treatment. Although this may seem as an attractive option for patients looking for a fast solution, computers are still prone to errors, and bypassing professional inspection may be an area of concern. Chatbots may also be an effective resource for patients who want to learn why a certain treatment is necessary. Madhu et al [31] proposed an interactive chatbot app that provides a list of available treatments for various diseases, including cancer.

Issues to consider are privacy or confidentiality, informed consent, and fairness. Although efforts have been made to address these concerns, current guidelines and policies are still far behind the rapid technological advances [94]. Although there are a variety of techniques for the development of chatbots, the general layout is relatively straightforward. As a computer application that uses ML to mimic human conversation, the underlying concept is similar for all types with 4 essential stages (input processing, input understanding, response generation, and response selection) [14]. First, the user makes a request, in text or speech format, which is received and interpreted by the chatbot.

Conversely, health consultation chatbots are partially automated proactive decision-making agents that guide the actions of healthcare personnel. Dennis et al. (2020) examined ability, integrity and benevolence as potential factors driving trust in COVID-19 screening chatbots, subsequently influencing patients’ intentions to use chatbots and comply with their recommendations. They concluded that high-quality service provided by COVID-19 screening chatbots was critical but not sufficient for widespread adoption. The key was to emphasise the chatbot’s ability chatbot in healthcare and assure users that it delivers the same quality of service as human agents (Dennis et al. 2020, p. 1727). Their results suggest that the primary factor driving patient response to COVID-19 screening hotlines (human or chatbot) were users’ perceptions of the agent’s ability (Dennis et al. 2020, p. 1730). A secondary factor in persuasiveness, satisfaction, likelihood of following the agent’s advice and likelihood of use was the type of agent, with participants reporting that they viewed chatbots more positively in comparison with human agents.

The underlying technology that supports such healthbots may include a set of rule-based algorithms, or employ machine learning techniques such as natural language processing (NLP) to automate some portions of the conversation. To reduce selection bias, two reviewers independently selected studies, extracted data, and assessed the risk of bias in the included studies and quality of the evidence. Agreement between reviewers was very good, except for the assessment of the risk of bias (which was good).

In the last decade, medical ethicists have attempted to outline principles and frameworks for the ethical deployment of emerging technologies, especially AI, in health care (Beil et al. 2019; Mittelstadt 2019; Rigby 2019). As conversational agents have gained popularity during the COVID-19 pandemic, medical experts have been required to respond more quickly to the legal and ethical aspects of chatbots. In September 2020, the THL released the mobile contact tracing app Koronavilkku,1 which can collaborate with Omaolo by sharing information and informing the app of positive test cases (THL 2020, p. 14). Chatbots have the potential to address many of the current concerns regarding cancer care mentioned above. This includes the triple aim of health care that encompasses improving the experience of care, improving the health of populations, and reducing per capita costs [21].

search

The first was an RCT conducted in Sweden [33], and the second was a quasiexperimental study conducted in China [37]. A meta-analysis was not carried out for this outcome as 1 study [37] did not report data required for the analysis. Forest plot of the 2 studies assessing the effect of using chatbots on the severity of anxiety. Risk of bias graph for quasiexperiements, showing the review authors’ judgments about each risk of bias item.

Most would assume that survivors of cancer would be more inclined to practice health protection behaviors with extra guidance from health professionals; however, the results have been surprising. Smoking accounts for at least 30% of all cancer deaths; however, up to 50% of survivors continue to smoke [88]. The benefit of using chatbots for smoking cessation across various age groups has been highlighted in numerous studies showing improved motivation, accessibility, and adherence to treatment, which have led to increased smoking abstinence [89-91].

The risk of bias due to missing outcome data was judged as low in 3 studies while it was rated as moderate in the remaining 3 studies due to availability of less than 95% of the participants’ data. The risk of bias in the measurement of the outcomes was serious in all studies (Figure 3); assessors of the outcome were aware of the intervention received by study participants, and this could affect the assessment of outcomes. In 5 studies, there was moderate risk of bias in the selection of the reported results (Figure 3); this is because there were insufficient details about the analyses used in the study. While the overall risk of bias was rated as critical in 1 study, it was judged as moderate and serious in 3 and 2 studies, respectively. Multimedia Appendix 8 shows the reviewers’ judgments about each “risk of bias” domain for each included quasiexperiment. Of 1048 citations retrieved, we identified 12 studies examining the effect of using chatbots on 8 outcomes.